DIY ICX453AQ, Sony sensor CCD camera

[wasyoungonce]

Following from this thread (http://www.iceinspace.com.au/forum/showthread.php?t=43574), time to form its own thread.

Following on from CN discussions, here (http://www.cloudynights.com/topic/497530-diy-astro-ccd-16-bit-color-6mpx-camera/), as lead by the great work of Hippieua and Vakulenko [/URL](Sergiy), also see[URL="http://astroccd.org/2015/04/cam84/"] here (http://www.cloudynights.com/user/243447-vakulenko-sergiy/), the object is to build an astro camera, ASCOM compliant using the Sony ICX453AQ sensor (as used in QHY8 and Nikon camera.

Initially this started a a great project but got a little bogged in changes and lack of cooling and the need for users to adapt and machine housings with UV/IR filter etc. ATM most bugs are ironed out for the "square PCB ver" but the "round PCB ver" still needs work in curing noise etc.

A few IIS'ers have expressed interest and this is a great project. I will be making one soon and can offer some assistance mainly in the SMD soldering area. Obviously I cannot help if dozrns of users want to build...but can help a few.

I'm still waiting for cooling cold finger to be adapted to this project but are building a BOM for parts for the square PCB board that can be sourced in Australia. This said...one of 2 parts are now not available or not sold here but I am chasing alternates. I'll be doing this for next 2 weeks as it takes a long time to cross check specs and read document datasheets. But I'll post it soon.

The other thing is the U1 dual rail buck converter, -8V/+15V out in QFN package. I'm not happy with this and have started some searching for alternates but this will require PCB changes, something I'd rather not do as the implication might degraded performance. Still thinking about this. I can do QFN packages with my hot air gun...but...I'm not overly joyed by this prospect.

I won't repost PCB gerbers etc as they are up to date and best held by the designers. We will also need to put these PCBs out for manufacture from a PCB fab house. The more the cheaper. Also buying parts in bulk can lower costs.

Anyway, lets get the discussion ball rolling.

Brendan

[luka]

Hi Brendan

I was also doing parts list and preparing to start a new thread :-)

As you said the round board is not as good as the square one (higher noise levels). Also the round board has 10cm diameter which I thought was quite large.

Here are my findings so far:

All parts are available at RS components apart from the sensor (obviously) and:
CXD1267AN-N can be sourced from ebay, $25
AT93C46-10S (discontinued???)
AD9826JRSZ - RS has AD9826KRSZ, I have not looked yet to see what the difference is, it may be suitable
BC847B.235 - it is a transistor, should be easy enough to find a replacement.
Ordering all from RS comes to $86, not including the capacitors, resistors and inductors which should not cost too much. RS also has free shipping so it is a good source for small orders :)

I have attached the Excel sheet showing the RS pricing for all parts, rounded up to a nearest $ (some parts were only few c but I rounded them to 1$).

As you said soldering should not be too difficult apart from U1. I don't have reflow oven or anything similar but this video (https://www.youtube.com/watch?v=c_Qt5CtUlqY) shows how to do it by hand. Not for faint-hearted but I am happy to try. I am also happy to help few others with SMD components.

Also I had a quick quote for the manufacture of the square board from PCBway. I have to double-check but I think it was about $25 for 10 boards from China + $30 postage. The price per board drops significantly on larger quantities of course as shipping is the killer. I used them in the past and the boards were quite good. Ah, and they have few $ paypal fee... squeezing every cent.

The biggest issue I have is actually making the box and especially the cooling. I do not have any machining capabilities. One option would be to use the recommended Gainta G107 box (which probably can be found in the local shops, it looks like a standard size). But again we need to solve the cooling. It looks like most of people used copper cylinder and a large heatsink with a fan at the back of the box.

Luka

edit: added the Excel sheet

[glend]

I am pretty sure someone on IIS was going to get some composite boxes fabricated for this camera. Brendan might remember that?

[wasyoungonce]

Hi folks. Composite box sounds nice but I guess we also should look at sealing and ability to shim the PCB so the sensor is orthogonal.

Luka I'll download your spread sheet and compare to mine. I was up to just starting the resistors identification (Cam84 ver 4.1).

CXD1267AN (not -N ver but this should be ok) ....UTS (https://www.utsource.net/search.aspx?keyWords=CXD1267AN-N&page=1) for $3.33. UTS are pretty good.

AT93C46-10S ...an alternate...RS, AT93C46DN-SH-B, RS number, 696-2743, a 1Kb, 64 x 16 or 128 x 8 bit eeprom. I think this is suitable but need to check a little more. Princely sum of $.35 ea.

AD9826JRSZ - RS has AD9826KRSZ,...I think this is fine, the nomenclature on the end is just minor revisions like temp rating colour package etc. I am looking at data pdf's to ensure comp ability. $17 ea.

The rest easy to find or alternates. I'll do my own sheet and compare with yours then we might be best to post it in the thread so that everyone can acces but only a few can change!

I've quite a bit of work comparing data pdf's on alternates....this takes some time! It should take me a full week or a little less. trying to type when I've just had a right elbow ulnar nerve de-compress operation a day ago. Quite funny really....I look like an idiot with one bad arm!:lol:

I'll get back soon.

edit:
forgot to mention a lot of parts are bulk order, ie: 74HC595D in pkg of 10 we need 1 ea. Transistor T1, BC857B.215 comes in pkg of 100, at $.053 ea. Breaking them up for more users is a cheaper way, $.053 ea (doubt we will have 100 users)! And so on...

[RickS]

Cool project, Brendan & Luka. I wonder if I could sneak a few into our next prototyping run at work :lol:

[wasyoungonce]

My quick XLS cross check comes up with ~$75 AUD for each PCB in parts (PCB not included) and a few small parts missing from my sheet atm.

But the real cost is buying some parts in bulk packages which atm is around $168 AUD all up. Obviously some items come in packages of 50 etc and we need one or 2 so this is where the cost comes down by buying and split share these packages.

I'm now back to cross checking the main IC's etc.....this'll take time to triple check.

[wasyoungonce]

Here is first iteration of Parts in Australia. I've re-checked Main IC's but are missing some small parts like some Tantals. I've included RS & element 14 stk numbers, costs packages etc

I haven't included case, header for sensor etc. I zipped it using winrar but re-named it to a ".zip" ...hope it unzips ok.

Any mistakes pls say so.

[wasyoungonce]

Hi Rick...of course..if you feel the need to slip one or 2 past at work....who are we to complain!:lol:

[luka]

Things are heating up :)

Since I saw this project a few weeks ago I started looking for a good deal on D40, D50 or D70 on ebay. This morning I got a D40 for $65 shipped. Partially broken (flash + cracked housing) but it takes photos, i.e. the sensor is good. I feel bad for having to throw away the rest of the camera :(

Anybody else looking for a sensor, buried deeply inside the thread on Cloudy Nights is this:
"D40 and D50 are on the flex board. This certainly makes it easier to remove. You can remove the chip a pin at a time as opposed to having all the pins released at the same time."
For D70 I would recommend using something like ChipQuick to remove the sensor.

Brendan, did you see the "upgraded" parts that were used to improve the round version? I saw a post (http://www.cloudynights.com/topic/497530-diy-astro-ccd-16-bit-color-6mpx-camera/page-37#entry7263088) on CN and I am trying to see if there is anything we can upgrade. Especially the lower noise voltage regulators may be easy to swap if the footprint is the same.

And thank you for doing the list. Note that minimal order from Element14 is $45 to avoid paying for postage.

Regarding the sockets for the sensor, this (http://www.cloudynights.com/topic/497530-diy-astro-ccd-16-bit-color-6mpx-camera/page-32#entry7170123) is probably the most relevant post at CN with DigiKey part numbers. Also see the post on CN, just after the one I linked, regarding using beryllium-copper. I have found a few sockets at RS/Element14 but a much cheaper source would be this (http://www.altronics.com.au/p/p0536-oupiin-18-pin-0.3-inch-dil-machined-pin-ic-socket/). It may need some trimming.

[luka]

Brendan, I had a look at upgrading components as in the CN post I mentioned before. Not sure if it is really worth doing it.

We could swap LP2985 with LP3985 versions. The LP3985 has the same noise levels but lower dropout voltage = less heat generation
5V replacement (http://au.rs-online.com/web/p/low-dropout-voltage-regulators/7616706/) (U4 and U5)
3.3V replacement (http://au.rs-online.com/web/p/low-dropout-voltage-regulators/5360868/)(U6 and U7)
(note the extra A in the part number. I could not decipher the meaning from the datasheet)

Unfortunately LM2937IMP-5.0 and LD1117S50CTR have the same package but different pin orientations. The latter has MUCH lower noise levels.


Also an interesting note from CN:
"I supply my cam84 with only 6.5Volts, That is enough for the U2 to regulate the 6Volt and also enough for all other VRs. By doing so the heat generated by the VRs decreased a lot.On U4 I measured a temperature of 70°C when operated with 12 volts and with 6.5volts its is only 50°C. The 12volt I use now only for my TEC."


And finally, do you know if the PCB design files were ever released? I can't find them in the files I got from CN.

[wasyoungonce]

Luka the PCB are in gerber format. You can get the PCBs made with these files. I think they were using steed studios.

There are some pdf's of the schematics...easy to convert to your own PCB cad...well not quite....takes some time. I was thinking of doing this but as said you could muck up noise figures by fiddling with the schematics, track size etc. Vakulenko and Hippieua will not release PCB cad drawings apart from those gerber files.

I concur with the regulator changes...but...I was reading a slight difference in bypass and output capacitor selection with the LP3985 vers. Also reading the bypass caps should be ceramic COG or NPO's.

More reading and cross checking required on this. Yes I saw the comment of the supply voltage. I think most ICs run from 5V or 3V so no great need to feed it 12V...just creates more heat than necessary.

I'm not fussed on PCB type, round or square but think the logistics of a round PCB is more difficult. I have not done any BOM for the round PCBs.

On another note...I ordered some parts from UTS as they were cheaper:

CXD1267ANT: SSOP; SONY; qty 5@$3.33 ea, = $16.65 (USD)
AD9826KR: SSOP; AD; qty 3@$4.00 ea = $12.00 (USD)
They are cheap enough to to grab some. Should take ~ 2 weeks to arrive. Already waiting for some parts from them to fix a 13.8 PSU I managed to destroy!:lol:

[luka]

I saw the Gerbers but I was hoping they had released the original drawings. Gerbers are too difficult to change and redrawing would take a long time and possibly introduce errors that need to be fixed...

Regarding the comment about the input voltage, the ADP3336 has maximum 400mV dropout voltage so in theory 6.4V at the input should be enough. I had a quick look and there are heaps of (simple 3-pin) DC-DC converters to get 12V down to 6.5V. For example this (http://au.rs-online.com/web/p/switching-regulators/6664372/) ($10) or this (http://au.rs-online.com/web/p/switching-regulators/8179404/) ($15). The first one is only 0.5A. Of course this needs to be separate from this board.

The round board had worse noise performance and would need a custom made housing. Square may be the one to go for, at least until they fix the round one.

[luka]

Can I ask everybody else who is reading this and is interested in building this camera to reply. It would be good to know the rough number of people because bulk ordering parts can significantly drop the price.

At the moment it looks like there is Brendan and me.

[wasyoungonce]

The ADP3336 has very low noise at rated output current...typically 27~45µ volts. These other units as small switchmode's with ~30~50MV P/P noise. Which is 100 times more noise. Not good for signal noise in this system.

Yes I agree that some of the power regulators are very close to their limits and probably need an upgrade.

I can do eagle PCB CAD but this is a big job and placement, track size etc will effect noise etc...a lot. This is the problem with the round PCBs.

on another note...I'll need to hunt around for another sensor ICX453AQ. Should have purchased Glens.:doh:

[wasyoungonce]

Meant to say....I'll be building 2~3 boards for myself. One colour one mono one spare. Or at worst case...at least one will work:lol:

[glend]

I hope the purchaser of my sensor is following this thread.

[luka]

Brendan, you probably already know but I thought I should mention it anyway if anybody else is reading, ebay is the source for the old cameras (sensors). New sensors seem to cost more than the old broken cameras.

And I think we should forget about redesigning the board. It is too much work and basically we will be going back to prototyping stage. Other people had good results with this design. I have never used Eagle but I can work my way around Altium Designer.

And I was also thinking about building 2... one colour and one mono... We may need more than 2 sensors though, the failure rate of debayering seems to be very high. Glen debayered a Canon 450D sensor but not without casualties.

Also I have access to a 3D printer... possibly for enclosure??? But the plastic box would need EM shielding then (copper tape inside?).

By the way, I have posted a link to this thread in the old thread. Just in case anyone was following that and did not see this thread.

[luka]

Do you think this can be used for L1 and L2?
http://au.rs-online.com/web/p/wire-wound-surface-mount-inductors/4153338/

It may need some scraping of the solder mask.

[wasyoungonce]

Luka that's not a bad spec, comparasion wise, but only one thing.

The Sumida unit has current rating at saturation of 1.2A (2.4A max) where the Wurth unit just lists 1.2A max....no idea if this is saturation or max current! The Wurth unit is a little larger but will probably make it.

So atm, it's worthy but I need to find out the saturation/max current specs of the Wurth unit. Arrrrhhh data sheets...don't you love/hate them!:sadeyes:

[luka]

Well spotted, it was getting late last night.

Another potential candidate here
http://au.element14.com/bourns/sru6025a-6r8y/inductor-power-6-8uh-1-3a-30/dp/2374259
Saturation current is 1.3A and the resistance is 0.042 Ohm instead of 0.048. Not sure how much difference it would make???

Another way would be to get it from DigiKey Au and get more parts from them. Free shipping on orders over $200... otherwise $34. That's why I never order from them.

[wasyoungonce]

Yeah that unit look ok although a tad large on the current PCB pad size. Probably why we might need to draw up our own PCB cad. To give a little fat on some pads.

I've asked on CN...they reckon farnell (now element14) has it. I did a search of Element14 UK...nil stocks. Anyway karlfoxman is going to share his BOM from element14 and Mouser. I have a mouser account. Not sure of their shipping nowdays though.

Edit:
Mouser shipping free if over $60. You know this might be easier than RS. I'll do a comparasion and get back

Brendan

[luka]

We only need to get enough parts from Mouser to get to $60. The rest can come from RS. But the part at Mouser (http://au.mouser.com/search/ProductDetail.aspx?R=0virtualkey0vi rtualkeyCDRH4D28CLDNP-6R8PC) shows "not in stock, call to quote"... it does not sound too good.

Now I have been thinking about making the PCB again. We do have the gerbers and images of the PCB... retracing the tracks may not be too time consuming...

edit: clicked post instead of preview

If we would to design our own (and it looks like it is only two of us and you can do Eagle and I can do Altium), did you have a look at cam85? I did not bother reading the details as it is work in progress but it fixes few issues that cam84 has. So partially incorporating those fixes could be beneficial. And we could make the housing smaller, cam84 has lots of "wasted" space as it was made to fit a pre-selected enclosure.

[wasyoungonce]

Damn...I was going off the mrn: CDRH4D28NP-6R8NC, (http://au.mouser.com/ProductDetail/Sumida/CDRH4D28NP-6R8NC/?qs=%2fha2pyFaduhsaTzWuq1zDTzdasUkQ Ox8P5cJcoia7tZwOQlM0OJnwQ%3d%3d) as given in the CN forum. Looking at the different data sheets...damn...it's a no go on using this one.

Looks like I made a mistake...the correct mrn is:CDRH4D28CLDNP-6R8PC (https://www.digikey.fr/product-search/fr?keywords=308-2039-1-ND) . This is the digikey mrn previously given.

Mouser has one..guess which one...the wrong one! The Digikey info is the correct mrn.

Damn....Damn....

I haven't looked at the Cam85 and indeed was kinda waiting for this...but like many thing the originators have a lot on their plates so progress is slow. Not that I care for the delay, I'm happy to wait or indeed do our own CAD. That said, trace impedance; noise; PSU ripple...are all things that can turn a simple design into a nightmare.

Which leaves me to say either use CAM84 sq board, maybe some small minor changes ( like up-rated voltage regulators, +5V; 3.3V, with our own CAD) but keeping the basic layout the same.

I think a Frankenstein approach between CAm85 and CAm84 is asking for trouble. Like as seen in the CAM84 round boards.

FWIW...only musing ...chewing the cud atm.:shrug:

[luka]

I was not thinking to do any major changes, that would be a nightmare. I was thinking along the same lines as you, copy the signal paths of the cam84 exactly as they are so that when overlayed it looks exactly the same. And then to upgrade the transformers or whatever else looks bad. And then possibly move the non-signal lines to save space... there is lots of empty space on cam84 board.

But this was the section I liked in cam85m (google translation, not sure what some of the parts mean):
I was just looking into it and it is not a simple change. We do have circuit diagrams from cam85 and cam85m which show the change. This part of circuit in cam85 looks almost identical to our cam84.

And as a side note regarding the choice of inductors, this is google translation from cam85 indicating that we need shielded inductor:

[Ausrock]

I had considered doing it for the round pcb as we are a bespoke resin formulating company (at Cardiff) but real life dictated other priorities so I won't be proceeding.

[wasyoungonce]

Thanks Chris.

posting an new updated parts spreadsheet...but...since RS doesn't have all parts it'll have to be an RS/Element14 buy. Which brings me to the possibility buying from Digikey or Mouser might be cheaper as they both free ship if over $60 aust order.

Now going to have a look at cam85. I was busy doing a cad drawing for a powertech PSU I destroyed.

Now I'm free...

edit:

forgot to add the sheet...doh! Again this was done with winrar and re-named to "xx.zip"...so hope it works

[luka]

Beware that there is cam85 and cam85m. The latter is the improved version.
As far as I can see cam85 has gerbers (with a mistake) and schematics while cam85m has only schematics.

[wasyoungonce]

Yeah saw this....8eV noise level...now that is quite low indeed. Again...it might be best to wait. Vakulenko certainly doing a damn fine job

[luka]

It may be worth waiting. I was reading the translation of the camera forums (https://translate.googleusercontent.com/translate_c?depth=1&hl=en&ie=UTF8&prev=_t&rurl=translate.google.com&sl=auto&tl=en&u=http://www.astroclub.kiev.ua/forum/index.php%3FPHPSESSID%3Dovbvtf7g0t0 s4e9l5krn05brg0%26topic%3D28929.144 0&usg=ALkJrhgp7-4SokvtLvxiZjd5mnN1_V2egg) and came across this (post 1456, from 31st of May 2016):
Interestingly, on the same thread, post 1470, Hutch84 posted their Altium designer files of the cam85m. The design seems to work although it got criticised for running the USB data lines all over the board.

[Ausrock]

Brendan,

I'd love to be doing this having DIY'd guitar amps and other stuff since the early-mid 70's but it just can't happen.

Disregarding (much to the annoyance of SWMBO) home projects, approx., five years ago for workish reasons I shelved two personal projects, a still (yes, I like my liquor) and building an arch-top guitar......a few months ago, in an attempt to regain something for me, I returned to the still and when finished I'll resume the guitar........logic said that I can enjoy my liquor while building the git BUT I can't enjoy playing the git while building the still ;). Add to that the fact that my scope hasn't seen light in over a year........say no more.

[wasyoungonce]

My you have been doing some reading...I do need to catch up! That site is a gem. I think I remember reading some threads on this years back when I had a Gotostar unit. They were trying to hack the firmware so you could set your own gear ratio dividers...to use it on other mounts.

Oh...those Russians!

Chris....moonshine?...I love it...:lol:

[wasyoungonce]

Just downloaded Sprint layout free viewer (http://www.abacom-online.de/uk/html/sprint-layout.html)to view some of the "xxx.lay" PCB cad files used.

It's god awful but you get the idea...so I can re-draw into eagle PCB. I'll do the TEC cooler as an exercise. Also been looking at Cam85m in the Russian forum. mighty impressive. Also saw some discussion on sealing cases with HTS2000. I've used similar Aluminium brazing. It works ok. It's not TIG but it's easy to use especially if you use a stainless brush or wire scraper to clean the surfaces when the metal joint in molten (clears oxidisation as aluminium alloys form oxide layers almost immediately).

[luka]

Brendan, which version of the cooler are you making? There is an Arduino Nano version on Vakulenko's github (https://github.com/vakulenko/CAM-Cooler-Control) which is much simpler.

I have been following the thread about the Cam85m and I would say that the design has been finalised. Unless I am misunderstanding the google translation, it does not look like there will be any more changes. This will probably be the version to make and not Cam84. There are already gerbers for it from "faddy" and quite a few images of his build.

Regarding Cam86, this is what will change:
For me this is not important and worth waiting for. What do you think?

By the way, my D40 arrived. I feel bad opening it just to get the sensor.

[wasyoungonce]

There have been some minor improvements and some personal changes of cam85m by "Faddy" and they look great. I also think he use G106 case, which was slightly smaller. His set-up is pretty damn good indeed.

Looks as if cam86 might be final instalment with the Atmega328 to provide flash firmware updates and a USB hub. The thing I saw with Faddy's unit was some of the parts used are a little harder to get but so far a quick check shows them available!

I am doing an eaglecad drawing of his unit at the moment but it'll take some time as I just went from ver 5.6 to 7.5. Glad I did 5.6 was too old and lacked parts and footprint data.

This said.....I'm happy to use his gerbers and order PCBs now if you wish! The only downfall is I have little mechanical engineering equipment but can do basics. We will need a nose piece for filter and to seal against the case.

Not fused on cooling, yeas saw the other versions but since cooling is controlled but out of case electronics I'm not fussed.

edit:
doing a full parts list/cost and availability now.

edit edit:
Errrrgggghhh been at it for hrs...lots of cross checking. There are quite a few changes made by Faddy.....Might have to ask for parts list!

[wasyoungonce]

Luka...the case for Cam85m "Faddy" is the G106. From the case machine drawings, it has 25mm inside height clearance from inside base to inside top of lid. The PCB sits on a base land, 10mm up from base. Guessing a PCB thickness of 2mm, this leaves 13mm for component height (maximum)?

Do you agree? I need to have this in agreement as I am finding it difficult to find capacitors of suitable dia and height to those specified. I can substitute longer caps if the clearance mentioned is correct.

Specifically:
I can find C5; C6; C9; C10 (6mm diax 7mm H).
I Cannot find height suitable for C11 (330uf, 10V), looks to be 6 dia x ~9 h (I can only find 11mm h caps).
I Cannot find suitable for C8, 220uf 25V, 6 dia x 9 h (can only find 6.3 d x 11 h).
I Cannot find suitable for C47, 47uf 25V, 5 dia x 8 h (can only find 6 d x 7 h).

Lastly it is difficult to find any ICs: CXD1267AN apart from fleabay. I have some coming from UTSource, but given the parts I just got from them...I have doubts on them.

[luka]

Brendan, where did you find the information about the case (G106)? Also where did you find the BOM with Faddy's modifications? I could only find his/her gerbers. Or did you get the BOM from the original Cam85 and then added the modifications yourself?

I do not have any real mechanical workshop capabilities here but, depending what we need, I could as a friend to help. Their resources are limited as well though (no mill for example).

I do have access to a 3D printer so at least that is sorted. Hope you don't mind red plastic :-)

[wasyoungonce]

I got the BOM from Cam85 and went thru the related posts in the forum (http://www.microsofttranslator.com/bv.aspx?from=&to=en&a=http%3A%2F%2Fwww.astroclub.kiev.u a%2Fforum%2Findex.php%3Ftopic%3D289 29.1540) as well as reviewing the PCB image he posted. I tracked his changes about the only questionable part I have identifying is Dr4, an 0805 inductor, I think it's ~1.5uh. There are minor errors on his PCB. He has 2 "C11's". One an electro one an 0805 cap. These are minor errors though. But to fully check I need to complete my own PCB CAD of Cam85m.

I have the rest of the Cam85m BOM pretty much done (see attached) except the small 0805 sized capacitor and resistors, they are common items no need to chase after atm.

The case G106 is a Gainta case. All info on it is here. (http://www.gainta.com/en/g106.html) Because he used the 106 case, height and size of component is critical. That said I calculated that we have an available clearance of 13mm (absolute max) so 11mm height electrolytic caps would be ok.

About the only "machining" part needed is the nose piece adaptor. Funny you should mention "printed parts"...he use some in his case parts design.

[luka]

Geez, going through all posts and tracking all changes must have taken some time. Thank you.

I was actually wondering how did you know that faddy's case is G106 as I did not see G106 mentioned anywhere... until this morning. I looked specifically for this last night but could not find it. This morning I started the computer and the browser was still opened on the post I was looking at last night and G106 was there. Tired evening eyes...

To confirm the G106 dimensions, the inside clearance is 25mm and the PCB ledge is 10mm. PCBs are usually 1.6mm but if we round it to 2mm we have 25 - 10 - 2mm = 13mm clearance, just as you said.

Can we get the 106 case here? I found a case with the same outside dimensions here (http://www.jaycar.com.au/sealed-diecast-aluminum-enclosure-115-x-65-x-30mm/p/HB5036)but there are no drawings to confirm the inside dimensions. Alternatively if we run out of space Jaycar has the same boxes but with 55mm height. Not the best but...

If we get too desperate with the too long capacitors, depending on space it may be possible to solder them with "long legs" and then bend the legs 90 degrees so that the capacitor is on its side instead of upright. Then a bit of hot glue tends to hold them in place quite well.

I saw the printed parts in faddy's design, that's why I mention the 3D printer :-)

And I may be able to source some of the small components that come in bulk packaging from work. I also have stacks of SMD resistors/capacitors at home but have to check sizes.

Regarding mechanincal parts, what is the plan to cut the openings in the box? Get it milled somewhere or drill-and-file method :-) And what about the copper finger?

Do we have a proper drawing of the nose piece adaptor? I may ask around how much it would cost to get it made. I also remember someone getting a cheap DSLR extension ring from ebay and gluing it to the front of the box :-)

[wasyoungonce]

Urrrggghh don't ask It was dizzying to say the least.



I've seen the 106 case at element14 (http://au.element14.com/webapp/wcs/stores/servlet/Search?pageSize=25&st=G106&catalogId=15001&categoryId=800000002596&langId=43&storeId=10184) and fleabay. Element14 sells the 106 ver for $26 but he 106mf(basically a flanged version, we can cut this off) for $12.25. Also fleabay. best to stay with these cases as the PCB was designed specifically for this case.



Yep exactly what I was thinking but looks to be 11mm height caps are ok.



Excellent. Would need to get a copy of his specs, we might need to ask on their forum. I think they would be fine with this, they have done a good job.

.

Talking dirty...I like it. Some of the parts I found can be substituted and indeed I have done this, like the 6V sot89 regulator, but some also come in big packs which we don't need. Element14 is a possibility to source these. The only hard to get part is the CXD1267AN. It's an old IC, no longer made. I have qty 5 coming from UTSource but are a little unsure of them being
"original". There are other sources...fleabay and some parts houses, but suffice to say...if it's from China..."its probably fake!" We shall see!



Looks to be we can hand do most except the nose piece. This looks to be internal threaded for probably 2" adaptors, filter...not sure. Yes we could get it made ...not sure atm. As for the cooling finger, look any cold finger extension to the sensor, copper would be best thermally.

Probably we need to contact the guys on the Cam camera thread, Faddy and grim for parts check, use his gerbers and other parts specs. I mean they may have a reason not to go this path....like cam86 around the corner.

I'm keen to make a jump. I'll do the cam85m parts into a "cost" structure. I'll probably just leave the 0805 sized parts out as they are easily available. I'll Just look at prime costs items. I'll get back on this.

[luka]

Brendan, I just saw cheap CXD1267AN (http://www.ebay.com.au/itm/CXD1267AN-Encapsulation-TSSOP-CCD-Vertical-Clock-Driver-/370729978091)s (or here (http://www.ebay.com.au/itm/SONY-CXD1267AN-TSSOP-Vertical-Clock-Driver-For-CCD-Image-/221585600591)) on ebay. Somehow I remembered them to be $25 each (on ebay) and not $5.

I will order 3 and see what I get...

Can you know if they are fake without soldering them first?

[wasyoungonce]

Yeah pretty sure I saw those as well...was "littlediode"?. Just be careful, look at sellers history many sell fakes! That link was to UTSource, the units I'm getting.

Note most sellers say these are TSSOP package, where the sony datasheet states they were made in SSOP package. Probably same but nomenclature is important sometimes.

[luka]

I just noticed that the link was to UTS, sorry. But I am looking at these (http://www.ebay.com.au/itm/SONY-CXD1267AN-TSSOP-Vertical-Clock-Driver-For-CCD-Image-/221585600591) ones now...

[wasyoungonce]

ok price currently is $122 but this includes a lot of packages at inflated prices when we just need one or 2. Price if we can obtain individual is $64.

I need to do more work sourcing ea parts instead of some packages but all parts are available....except the CXD1267AN...which we have on the way and have spoken about.

[luka]

Good news, I just saved us $5.30 :cool:
I found a large stash of BC857 (RS part number 761-3587). Part VT3 on Cam85m sheet. No need to order it.

I also have a box with various 0805 resistors and chip capacitors. It probably has all parts we need in this size.

[wasyoungonce]

:thumbsup:Welcome news indeed. As said we both have parts on the way and can share. I would like to have enough parts to build 2 PCBs each (at least for myself), though depends upon expensive parts.

I'm trying to register on the Ukrainian forum to ask questions...server is down. :shrug:

I have doe a few more parts and identified cheaper options in lieu of packages. That said, the cost doesn't faze me. I'm more concerned with the nose piece collar getting done...also cooling. We can do this on the fly though, no need to fuss too much atm as they are "addons" to the main camera.

[luka]

The cost of bulk parts saving is really only few $, as you said not really important.

I am looking for the drawings of the nose piece... I have seen them somewhere but the forum threads are very long...

I am also looking to build 2 PCBs. Happy to share everything.

I don't think there is anybody else building this.

[wasyoungonce]

Exactly what I was thinking on some of the bulk parts. Hoping "Faddy's" nose piece adaptor brings backfocus of sensor to 55mm and allows filter fittment! That would be handy.

The only issue I see is case sealing.

[luka]

Case sealing = silicone :-)

Also have a look at this (http://www.cloudynights.com/topic/497530-diy-astro-ccd-16-bit-color-6mpx-camera/page-26#entry7051267) for the front adapter. The black part is few $ on ebay and it connects to the scope via DSLR t-adapter.

I am still looking for drawings for faddy's adapter. I am sure I have seen them somewhere. But it is just a ring with a thread inside and a few tapped holes. I am guessing that the thread is the same as the common 2" filters???

[luka]

Flange drawings are here (https://translate.googleusercontent.com/translate_c?depth=1&hl=en&ie=UTF8&prev=_t&rurl=translate.google.com&sl=auto&tl=en&u=http://www.astroclub.kiev.ua/forum/index.php%3FPHPSESSID%3Dovbvtf7g0t0 s4e9l5krn05brg0%26topic%3D28929.msg 468215&usg=ALkJrhjr3SRJrpyJuU633iVRxk3ayDk ntg#msg468215) (scroll down). To be interpreted tomorrow once I get some sleep :zzz2:

Also check out the peltier here (https://translate.googleusercontent.com/translate_c?depth=1&hl=en&ie=UTF8&prev=_t&rurl=translate.google.com&sl=auto&tl=en&u=http://www.astroclub.kiev.ua/forum/index.php%3FPHPSESSID%3Dovbvtf7g0t0 s4e9l5krn05brg0%26topic%3D28929.msg 515365&usg=ALkJrhjgMIyYnMatO4I9zLlmxZFxLny 9_g#msg515365) (scroll down):eyepop:

[wasyoungonce]

Damn you're up late! Appears Astro Club Kiev is off line atm! :shrug:

Edit:
Are we socketing the Sensor? The only SIL I know of is this (http://www.digikey.com.au/product-search/en?keywords=317-93-121-41-005000)from digikey. But, I don't see why we couldn't use one of these. (http://au.element14.com/mill-max/117-43-642-41-005000/dip-socket-42-position-through/dp/1621393) Would need to cut it down and separate the halves so they were SIL headers. Not cheap though. I know the pitch, 1.78mm is fine but the socket depth is ~3.7mm. The sensor legs are 5mm long. Also the socket pins accept .015" to .025" leads. The Sensor leads are .46mm/.018". It is a tad small but will do. I think the digikey socket pin depth is the same.

[wasyoungonce]

Still plugging away on parts list. Also doing a digikey basket. This is actually an easy way to do it. I've found some errors on parts, like qty 2 C20's on PCB...one is supposed to be R20 (the one top of DD9). Cannot find R9 atm.

Still doing CAD PCB but had to make a few packages like the sensor ICX413AQ. Getting this footprint etc wrong will be a major catastrophe lots of time doing that one. Just checked resistor now doing caps then I add them to the drawing and connect.;)

[luka]

Regarding the sockets, let me summarise:
1. Digikey ones are recommended for Cam84. The CCD should (hopefully) fit in them.
2. The element14 ones are identical to these (http://www.cloudynights.com/topic/497530-diy-astro-ccd-16-bit-color-6mpx-camera/page-32#entry7170123) suggested on CN, apart from being lead free. They have identical dimensions as the digikey ones but would need cutting.
3. Have you seen these (http://www.altronics.com.au/p/p0536-oupiin-18-pin-0.3-inch-dil-machined-pin-ic-socket/)? Local, cheap, need cutting. Identical dimensions. Beryllium copper as well.

Now, looking at Faddy's design, I cannot see if he/she socketed the sensor. But we should beware of the height of the sensor as you mentioned that some capacitors were at the max height. Socketing will increase the height by at least 4.2mm. But I suppose we have 10mm height to play with.

Do you have a datasheet with dimensions of the sensor? I could not find it anywhere.

[wasyoungonce]

Hi luka, sorry long rant,

I have a pdf on the sensor (attached) it's not a full data sheet but it does show the package dimensions. The problem with the sensor is that the pins are 1.778mm (.07") pitch on centre (¢), not the std 2.54mm, the pins are 25.4mm ¢ (1.0") across. Not much uses this spacing and I suspect the 1.778mm is a standard of 1.78mm, used in some SIL and DIL packages (not many though). There is an Element14 package here (http://au.element14.com/harwin/d8864-42/socket-ic-shrink-dip-64way/dp/1023069) (that is 1.778mm pitch) but it needs to be cut to make into qty 2 SILs (that should be achievable) as it's row pitch is too narrow. RS has nothing in this pin pitch, digikey does (http://www.digikey.com.au/product-search/en?keywords=ED90488-ND) (again this needs trimming as its 21 pins long, we need qty 2, 16pin SILs.

The other issue is most SIL/DIL sockets allow for a pin insertion of 3.683mm (.145") maximum depth. The ICX413 sensor pins protrude 5mm below the package body. Thus the IC will insert and bottom the pins in the SIL socket with the body proud of this (not by a lot).

I have found some "single pins" that are soldered into the PCB and many allow for longer IC leg insertion, like this (http://au.element14.com/mill-max/8579-0-15-15-11-27-10-0/pin-receptacle-pc-board-solder/dp/1441581)at element14. But again, we are thwarted. The Pin head dia is .055" (1.397mm) which means the sensor pin pitch is too close, they will overlap! There are no single PCB pins that are small enough head diameter to fit, so, back to the SIL.

Its not a big issue atm but it's something we need to be aware of, especially for cooling and sensor orthogonality. I have the sensor package done...just hope its right. I like to print the PCB at original size and check parts fit before committing to the PCB being sent for production but FADDY's gerbers are obviously correct.

[luka]

Hi Brendan, thank you for clearing that up. I checked every other dimension several times but completely missed the pitch :doh:
So much for staying up so late...

Cutting the Element14 DILs should be manageable. I have done similar things in the past (never with 1.778mm pitch though). Luckily there is enough spare length for a few practice cuts. Possibly even use the soldering iron to get a pin very hot, melt the plastic around it and then pull the pin out. Cutting on that spot should be easier then or even plastic can be melted instead of cut. Actually a hot knife may be the best option depending how brittle the plastic is.

Also, it was probably just a typo but better safe then sorry, you said that we need 2x 16pin SILs. The sensor is actually 2x 17 pins, not 16.



On a different matter, here is a quick calculation of the distance between the sensor and flange:
back of PCB to box = 10mm
box thickness = 3mm
Total = 13mm

We need to subtract the distance between the back of the PCB and the sensor surface:
socket height = 4.2mm (7.2mm but the legs partially go through the PCB)

sensor legs = 5mm
insertion in socket = 3.683mm
difference = 1.317mm, say1.3mm

sensor height (from legs) = 2.6mm

Total back of PCB to sensor surface = 4.2 + 1.3 + 2.6 = 8.1mm

So we have sensor to outside-box-wall-distance of about 13mm - 8.1mm = 5mm roughly.



To this we have to add the flange thickness. The bad news is that Faddy's flange is different than the one I mentioned before (https://translate.googleusercontent.com/translate_c?depth=1&hl=en&ie=UTF8&prev=_t&rurl=translate.google.com&sl=auto&tl=en&u=http://www.astroclub.kiev.ua/forum/index.php%3FPHPSESSID%3Dovbvtf7g0t0 s4e9l5krn05brg0%26topic%3D28929.msg 468215&usg=ALkJrhjr3SRJrpyJuU633iVRxk3ayDk ntg#msg468215) (scroll down), the one with the drawings. But I believe Faddy tried to reproduce this (https://www.myastroshop.com.au/products/details.asp?id=MAS-005K4) in one way or another.

If we guess Faddy's flange of 15mm we get sensor to flange distance of 20mm. A big guess???

I am not sure about your setup but the short distance is good for me as it this gives an option to fit a filter wheel or something else in between my flattener and camera. With my DSLR there is not much space for anything really. Flattener needs 55mm distance to sensor while Canon has 44mm sensor-to-flange distance. The T-ring adapter takes care of the short difference.

With this setup I will have at least 55-20mm = 35mm room to play with. And will likely need extensions...


Now to do some research on the flange design...

[wasyoungonce]

Hi Lucas, that's some good work and timely on spacing. I was just starting to think about this.;) That tilt-able flange is is a very good idea.

I've finished the digikey parts list...found quite a few changes between the Cam85m and FADDYs version. This will make it a little difficult to PCB CAD draw, I'll have to use FADDYs gerbers to understand the differences. This'll add more time to completing this. I only have one more package to draw, the shielded inductors, 2525 package. As usual, inductor pad sizes on parts like this are not a common sizes across mfgrs.:mad2:

Yes I made an error with the sensor 32 pins (it's 34)...I even made this error when drawing the CAD package and had to fix it. Damn 32 pins on the brain!:lol:

[wasyoungonce]

Damn, there is more than one version of Cam85m out in pdf format. They both neglect to state the +5V generation, except the latest ver appears to have it coming off board. But Faddys has a +5V Dpack regulator on PCB (not on schematic).

Sigh....I'm doing CAD drawing now, I'll make if from the schematic then check against Faddys PCB. I can ten make changes and re-name the schematic to a different version.

[luka]

I have seen that. Have you seen post 1556 (by Faddy) in the thread? It shows external 5V input.

[wasyoungonce]

On Faddys post 1541 you can see the PCB, side 2 he has a +5V DPack fixed regulator on the bottom of the board (not in any schematic). I think the sub-board shown in post 1556 shows 2 wires down the bottom the the board.

Urrrrrgggh....Just noticed my 7806A voltage regulator package has pins backwards compared to what it should on PCB...grrrr. I have to fix this.

Referring to the PCB image these bottom leads are +12V and ground. +12V feeds Vin DD10, the 6V regulator we can see this on the PCB image, yet the schematic (dated 14.06.2016) shows 15V2 going to DD10 as Vin. So Faddy decided to simplify and feed it with +12V.

Also noted on the schematic that the PCB is feed with +6V externally which is dropped by a diode to Vusb (+5Vusb). This is not on Faddys PCB. Still have no reference to value of DR4 on his PCB as well. There is a bit of lack of info.

I think it best I draw the schematic as given then I can make changes to suit Faddys PCB. This'll muck me around a bit but its might be a useful exercise in understanding the changes.

Of course the easy way is just use his gerbers...but...that would be too easy.;)

Edit:

Ahhh just found a newer schematic in post 1446, by grim. I did note is has the same date as the other one that showed external +6V. This schematic looks better, shows the +5V regulator (DD15) and +12V route to DD10. All good then I'll use this. Sanity at last.

[wasyoungonce]

More disaster from UTSource

IC's arrived....nice box, good, antistatic bags, good,.....shame about the rest.:mad2:

Its a pity. I've had good stuff from them before but lately they appear to have gone to the dogs.

[luka]

Is that sticky tape?
Think positively, they could have superglued the chips to the cardboard.

By the way, for one of the orders (at work) we have received a fully populated PCBs packed in... ziplock bags. Just good old ziplock bags that you put your sandwiches in. Miraculously everything survived.

[wasyoungonce]

Although I've seen and used conductive sticky tape that is safe for ESD sensitive packing, suffice to say, this is plain old sticky tape! Which probably produces a couple of kilovolts of static charge when pulled from reels!

ESD precautions from a component supplier...."a dirty word". I hope they are ok, cheap enough, but the issue is using them and they may be degraded or worse.

Might look for some more.

edit:
meant to say, this IC was used in quite a few webcams...and I think playstation (don't know which version). I might even think of ratting them if I can get my hands on some webcams. I can remove them quickly with a SOP adaptor to my Hakko. I would have more confidence in a part sourced this method than the UTSource ones. I have 2 Philips SPC800NC webcams (one mono sensor, one colour)...but...I'm not ratting them!:lol:

[wasyoungonce]

Just found the mfr for sensor socket, Sony and Kodak etc, Andon Electronics (http://www.andonelect.com/). I've downloaded their catalogue and while they do not have a carrier for the ICX453AQ, they have one for another sensor that is 34 pins, 1" row width, 1.78 pin pitch ...all other specs look same as our sensor.

Mrn is:IS230-1034-75S-R29-L14. An alternate DIP package is this:

Mrn: IS230-1034-75S-R29-L14-A.

Also they have SIP packages in 1.78 pin pitch, mrn:
416-17-75S-R29-L14, 1.78 pin pitch, 17 pin length, (again with longer throat for longer pins. Looks to me they use a DIP carrier to hold SIP arrays then remove the carrier after soldering the array in to the PCB. Similar method to solder individual pins. This holds the SIP or pins square and flat to the PCB.

I'm trawling the web for this but so far none. I did note they appear to make and sell individual pins that are made for sensors...aka smaller head dia, long throat for longer sensor pins etc.

Looks to me so far...false alarm. It looks as if they are specialist items only available from Framos (https://www.framos.com/products/en/sensors/sensor-sockets/is416-17-75s-r29-l14-10395.html) or Eureca (http://www.eureca.de/english/imagesensors_index.html). Both of these places are not cheap!

[wasyoungonce]

Damn...I used ICX413AQ pin/package info to draw my CAD in my schematic part. Problem is, It's not the same as ICX450AQ. Added another 3 hrs to sort it out. Gawd I was really confused for awhile.:shrug:

Starting to get there now, just about complete schematic but haven't started PCB routing yet. There will be difference between the schematic and PCB because each user (FADDY) put in some differences, but I can easily accommodate this once I have the basic schematic done.

The PCB routing will show these differences.

[wasyoungonce]

Ok, I've completed the schematic, and are laying out the PCB. This could take some time. FADDY did some changes as he was previously building the CAM85 then went to CAM85m.

About the only issue I have seen is that he left on R13, a 3.9K between VT2 drain to base of VT3. Since Vt3 is now not a FET (it's now a NPN bipolar BC857B) it looks to require inclusion. But its simple enough to leave out if required.

I pretty much have all component values sewn up and cross checked parts to actual PCB. But transposing his PCB to mine.... placing the PCB components on my PCB requires high accuracy for: holes; via's; traces; everything... etc. FADDY has already completed this. I don't mind doing this butit will take some time.

So, the question is:
We can use "FADDY's" gerbers right now as I'm confident they are ok and I have a handle on his components; or
Wait till I complete my PCB and use this to cross check each trace etc.

So...what's it to be?:shrug:

Decisions Decisions...:question:

[luka]

Brendan, again thank you for doing all the hard work.

If you are confident that Faddy's design is fine, the gerbers would most likely be fine as well. I would say let's go with his/her gerbers. It would be a waste of your time to duplicate his/her work... unless you really enjoy designing PCBs ;)

[wasyoungonce]

Yes I'm confident with his gerbers,but this commits us to the G106 case.

Oh... I made an error...Vt3 was a transistor on cam85 but later changed to FET on Cam85m. No Idea why FADDY left the resistor R13 on but suspect he's used the Cam85 transistor on the cam85m in lieu. Easy enough to leave this item on the PCB and bridge but we need to clarify this part VT3. VT3 on Cam85m is IRLML2803 N channel FET.

As said, I'm confident on his gerbers. It was an exercise in understanding his changes and components.

I'll do some track laying and get back.

[luka]

If you remade the gerbers, can we go to a smaller case or to a different format of the case, keeping in mind that moving some of the components can affect the noise levels?

[wasyoungonce]

Inducing noise or thermal issues is a real problem I am thinking of. Also there are some minor changes in capacitor values for Cam85m around the IC's forming timing pulses, DD8, DD9, DD11. I think I have the correct values (I do have alternate values noted but are unable to determine which cam has what atm!) ...gawd it hard to decipher.

[wasyoungonce]

Urrrgggh...doing my head in after about 10 PCB traces. :lol:

VD5 anode and DD7 (pins 1 & 2) connect to 15V, as per schematic. 15V and 15V2 are separated by DR3 (inductor) and 15V2 has some extra capacitors, like C47. Yet FADDY pulled a swifty and also put VD5 on this 15V line, right at top of PCB images, top side.

Cannot believe it...only 13 traces laid and it's all pear shaped...Hmmm unlucky "13" anyone!

In reality this probably means little but it's little details like this that can mean a lot, especially with this. Grrr I'll figure out what to do...looks as if I'll do a correct schematic and PCB and another, with the differences included.

[luka]

Brendan, you will love this... NOT...
Have you seen the page 83 (last page ATM) of the Ukrainian forum? Gerbers, schematic and BOM posted by grim, the person who started the thread and is developing the cam86. Also few questions of relevance by Faddy.

[wasyoungonce]

yep...well that kills off Cam85m!

Looks to have major improvements over Cam85m. Major changes are: PCB layout much much better (edit: is smaller as well, 85mm x 54mm);
ATmega328; control of clks; TEC cooling (VT1) and DSB18B20;
More, improved regulators and these moved to bottom of PCB;
Changed AD converter to AD9826;
And more...

Suffice to say, Cam85m is dead, long live Cam86.:lol:. Beginning to gather info for all the changes...Including a much better Schematic and re-order listing of parts on Schematic.

Brendan

[jscott]

I'm interested, if I'm not too late.

John

[wasyoungonce]

No, its not to late...pretty much we haven't done anything apart from talk and point fingers :lol:

We were doing PCB cad for Cam85m but Cam86 is released which is has better layout and is flash programmable, meaning I've seen a bigger bone and I'm going for it.:D

What we will do is ensure we have a path ahead with PCBs, parts, and programming before we commit. Pretty much we have PCBs and parts already, just need info and microcode for ATmega328....and grim's bets wishes

Brendan

[luka]

New gerbers with silk screen and programmer for AVR are out on the Ukrainian forum. Just waiting for the firmware now...

And John, welcome :-)

[wasyoungonce]

Looks as if he's tweeking the clk signals. We gotta thank them. I'm away atm for a few days...might be worth while to join their forum to say hello and thank them.

Brendan

[luka]

John, just wondering if you have machining skills with a workshop access? In particular to make the nose-pieces for the cameras. Brendan and I can handle the electronics but the noise piece is still the missing part of the equation.

I found two solutions but they are not great. We could, for example, use an off-the-shelf camera tilting unit (http://www.gerdneumann.net/english/astrofotografie-parts-astrophotography/ctu-camera-tilting-unit.html) and screw it to the camera box or we could go without tilt adjustment and use something simpler like this (http://www.teleskop-express.de/shop/product_info.php/language/en/info/p6556_TS-M68-System---Adapter-from-M68-to-M48x0-75.html), screwed to the box. Then we would need extension rings which can be purchased separately. Neither is perfect though.

[luka]

Brendan, have you seen this (http://www.jaycar.com.au/sealed-diecast-aluminum-enclosure-115-x-65-x-55/p/HB5040) case? Same dimensions as G106 but deeper, if we need more depth. I am not sure about the internal dimensions for mounting the board. If of interest I could go to Jaycar and measure the insides.

Jaycar also has a diecast case (http://www.jaycar.com.au/sealed-diecast-aluminum-enclosure-115-x-65-x-30mm/p/HB5036) with identical outside dimensions as G106.

[wasyoungonce]

Actually I was going to say...they don't have a seal....but they do! I'll go look at them, this could be a cheap alternate.

Brendan

[wasyoungonce]

Started Cam86 PCB cad drawing, a few errors found but minor. Will take a day more to do then check out PCB placement.

[luka]

John, hopefully I did not scare you with my question ;) Don't worry about it.

Brendan, what is your plan regarding colour or mono (de-Bayered sensor)? This is all new for me and I am still thinking what to do. Do you know what size filters we would need for the sensor size?

[luka]

I was looking for ways to control moisture inside the enclosure and came across these (http://au.rs-online.com/web/p/humidity-indicating-desiccators/0838726/). There are 3 different sizes.

[wasyoungonce]

IHi Luka

I haven't done any measurements so far but the sensor is 28.4mm diagonal, however, the pixel array is smaller than the package. I think Gary mentioned 26mm filters. I'm staying with colour bayer atm, but, If I get 2 running, one will go under the knife for some de-bayer action!

That's a dam good idea for silica gel. I think sealing the case might be a real issue. We only need to seal the sensor area, which is also not an easy thing. I suspect we will not get it fully sealed. I have read (an IIS user) who had sensor dew issues and he used a small piece of tube to direct some of the TEC cooler hot air over the sensor. Apparently it worked.

Finished drawing schematic of cam86 with eagle PCB. I'll now cross check parts on the schematic to the PCB images.

[luka]

I have spent last two years battling humidity, salt and water in devices (at work)... it is much more difficult problem due to salt...

Anyway, we should try to seal the sensor area as good as we can and then also seal the whole camera. A thin layer of silicone around the joints will ensure no air gaps. If you pick the right silicone it can be removed without much drama when camera needs opening.

Then there are few ways to prolong the operation time. You want to close the box with as little humidity inside as possible and then there is a possibility of a leak.
Silca gel bags (in little cages) will work fine but I like the ones from RS as they screw into the housing and have an outside indicator. In other words you know when the humidity is too high inside and you can take them out for baking without taking apart the housing. I just found them and have not used them yet. May be too bulky though.

If you look at some of the 450D cold finger builds (I just finished mine), there is lots of discussion about this. The other idea is to fill the camera with dry gas, for example with argon (easy to get from Bunnings). This will help a bit as well.

Finally, we could put a humidity sensor inside and measure humidity this way. If it goes too high, time to look for leaks, replace the silca gel, and refill with argon :)

[luka]

I forgot to say, my CDX1267ANs arrived today. No stickytape anywhere and properly packed :)

Got them from here (http://www.ebay.com.au/itm/221585600591?_trksid=p2060353.m2749 .l2649&ssPageName=STRK%3AMEBIDX%3AIT).

[wasyoungonce]

Hi luka

I would have though that the the electronic humidity sensors are too big to fit in the case, near the sensor area?

I ran a cooled 450D, al la Gary Honis style (camera in a TEC coolbox) and it worked well, took lots of time to cool down, but would cool to below 0º C. I kept the sensor free of moisture by wrapping a dew heater on the camera/drawtube nosepiece. I used to find ice in the case area after use. Mostly the electronics doesn't ice up as they are slightly warm from passing electrons.

We can conformal seal the PCBs to protect them but I guess no matter what we do, sealing the case is a priority. Gary has some ideas on this, like "FADDYs" camera build. I see the cover filter as the most difficult area to seal, "sealing the UV/IR filter". We absolutely need to backfill the case with an inert gas, probably Argon. I guess we should pressure check them first. This is probably best with the sensor out as we don't want to damage it. I don't know what pressure differential the sensor coverglass can be subjected to.

I looked at the RS silica Gel Desiccators and my guess we need the small 10gram version (http://au.rs-online.com/web/p/humidity-indicating-desiccators/0838798/), mainly due to the mounting hole size (17mm Ø). I saw some IP67/IP68 sealed "D" connectors, like this (http://www.digikey.com.au/product-detail/en/conec/6STD15PCM99B30X/626-1165-ND/1017401). I'll do a search for more or cheaper units but these look as good as we can get to seal the housing electrical connector. Which means we might need a small breakout box (like "FADDs") on the back of the camera to convert this "D" sub out to power, USB etc.

Can I ask where you got the CXD1267?

I'll get back on connectors and do some sealing case/UVIR filter research.

[luka]

See the ebay link in my post above :)
Interestingly the ebay feedback was in German but the CXD1267 were apparently sent from Hong Kong. And I paid in Euros.
The postage is a bit steep, ordering some spares is advised. I have three, i.e. one spare assuming I make two cameras.

Regarding the waterproof electrical connectors, I should have offered my help before, I can easily find some. Somehow my brain didn't register that you were looking for them. Sorry. Too many things to think about...

Do we know what connectors we will have coming out of the box?
1. USB (I actually bought one of these (http://www.altronics.com.au/p/p9719-amphenol-ltw-usb-type-a-female-chassis-ip67-waterproof/) yesterday, or this (http://au.rs-online.com/web/p/usb-cable-assemblies/6120740/)? A bit bulky but we could probably find space for it).
2. Power (what kind, i.e. how many connectors/voltages)?
3. Anything else?
Will the Peltier sit on the outside of the box, i.e. will it be powered from the outside? I assume it will be PWM controlled and that can interfere with the other communications (USB for example). It may be the best to keep them on separate connectors.


The electonic humidity sensors are bulky (this (http://www.jaycar.com.au/humidity-temperature-sensor-module-for-arduino/p/XC4246) for example). You are right, they would not fit in the small space where the sensor is but could be used to measure humidity inside the case. We would also need a small controller... You went through all the Cam86 circuitry, does the build have any provisions for connecting temperature sensors? I don't believe they will release the source code for the firmware, otherwise it would be trivial to modify the ATmega code to include temperature/humidity sensor or whatever else we want.

If you are worried about the sealing the front window, here is a different idea... When I mentioned the 450D cooling mod, I meant the cold finger mods. Two different approaches were used to combat the fogging up of the sensor:
1. SMD resistor array to heat up the front of the sensor
2. Astronomik MC-clear glass in front of the sensor. It created a sealed chamber that was filled with argon.
We could 3D print a small plastic spacer, glue the MC-clear glass to it and then place it on the sensor (under argon atmosphere). This will create another sealed chamber providing extra thermal insulation.
It increases the cost significantly as the 450D glass was about $100 when I bought it about a month ago.

[wasyoungonce]

Hi luka

I was looking more at the D sub sealed connector as it is not just moisture sealed it has glassified epoxy on the backshell around the pins. Thus its environmentally sealed, well, in a manner, and a flange seal, which will seal air inside, hopefully. Funny enough the RS USB sealed connector looks to be the goods! My guess you're right up this field already.

The peltier can sit outside the box and a cold finger (Slug) arrangement to the sensor but mostly I have seen them directly on the sensor. Since he's put peltier control on the PCB I suspect they are looking at locating the TEC in the case against the sensor.

I'm sure we can seal the front glass, depends on the mating arrangement. RTV3145 or such like.

Cam86 uses a DS18B20 sensor which can operate more than one on a one wire bus, though I suspect they are just using one. It connects to pin 14 or ATMEGA328. I don't think grim will release the source code. Not that he really cares, it's more someone using it to sell lookalikes.

Sorry have to dash...le tour highlights are on.

[luka]

Just realised RS USB connector is not good. Explain later, don't waste time looking into it.

edit: I am back, sorry about the cryptic message, I just didn't want you to waste any time on those connectors.
The RS connector has two parts. The actual plug is removable from the round housing - it is just a push-fit and it is not sealed. The waterproof rating applies only if a cap is used, or a matching connector is connected into the USB plug. The matching connector has a housing which screws onto this connector's housing. So, unless we want to keep the matching connector connected at all times, i.e. have a long cable going from the camera at all times, the plug is not good.

The Altronics connector (http://www.altronics.com.au/p/p9719-amphenol-ltw-usb-type-a-female-chassis-ip67-waterproof/) is filled with epoxy and has pins coming out on the other side. A bit more soldering but it should be OK to use.

[wasyoungonce]

Looking at some hybrid D sub connectors (http://au.rs-online.com/web/p/solder-d-sub-connectors/0484919/) at RS atm. They would be good to carry power and signals. I don't think they have any sealing rating though as you select an insert the power pins you want. I'll look at digikey some more.

edit:
also looking at sealed (IP68) USB connectors element14 (http://au.element14.com/bulgin/px0845-a/sealed-usb-2-0-type-a-receptacle/dp/1166684)...but they all appear to have rather large mounting diameters, like the Altronics version. We would have trouble with the case size and the connector size.

[luka]

I think that the RS connector had similar size as the element14 and Altronics. I may even have them all here.

Question, how many connectors do we actually need to go into the box? USB + just one voltage (and GND)? Is everything powered from this voltage internally?

Edit: What I meant is that if PWM for the Peltier is generated inside then I don't see any issue passing USB + voltage through the same plug.
Something like this (http://www.altronics.com.au/p/p9850-amphenol-ltw-5-pin-4a-m12-screw-on-chassis-male-ip68-waterproof/) or this (http://www.altronics.com.au/p/p9617-amphenol-ltw-7-pin-20a-locking-male-chassis-ip66-waterproof-plug/) may be suitable.
(http://www.altronics.com.au/p/p9617-amphenol-ltw-7-pin-20a-locking-male-chassis-ip66-waterproof-plug/)

[luka]

Or we just get any connector, solder wires to it and mount it and then put a big blob of silicone on it (from inside) to seal it :lol:

[wasyoungonce]

We need a separate +12V for the peltier. PWM puts a lot of noise on the power line and grim has allowed for this (separate +ve). Reality...it should also be separate gnd as well.

The Altronics IP68 connector looks good. Yep of stuff I'm looking at and also in D sub. I looked for sealed USB....nothing available. USB was never intended for this. Probably why FADDY made a breakout for all this. Actually his last images, page 84 look, (https://translate.google.com.au/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.astroclub.kiev.u a%2Fforum%2Findex.php%3Ftopic%3D289 29.1660&edit-text=) good. Hi Cam85m is obviously working well.

[luka]

Brendan, I was just looking at the circuit diagrams and the Gerber files for Cam86 and, unless I am mistaken, 12V and Peltier+ are joined by a track in the copper top layer. Can you please check this.

Also, (I know we are not using it but...) the element14 USB connector is not sealed unless you put a cap on. They are the same as the RS ones. This is not obvious from the datasheet but the seal is only achieved by the outside o-ring. There is a sentence there saying "dust and waterproof sealing when mated". We use them at work. The connector itself can be popped out from the housing with a bit of pressure from the back. It makes it easy to replace when they get full of dirt.

I did lots of searching for a sealed USB connector that is smaller than the Altronics one over the past few years and didn't manage to find anything. If you do please let me know.

By the way, Altronics has lots of sealed D-sub connectors. Just search for "waterproof".


Back to Faddy's design, did he just poke the wires through a hole in a camera housing into a breakout box? I suppose that could be a way to go :)

[wasyoungonce]

Just noticed I have the Peltier power rail wrong. Yes on the PCB image they are indeed joined. Sorry missed by ...that much!;)

Faddy made a hole in his case rear and sealed this with something that looked like silastic. These wires then ran to his breakout/peltier control box. It depends upon the material used. Something like a metal set like JB weld/Devcon would be ok but I suspect they wouldn't stick to some wire insulation.

The Altronics sealed D subs look just like the digikey versions or copies there of. I'd like to see their data sheet on them.

[wasyoungonce]

I meant to link this before, Martin Raabe.

A German guy who built his own ICX453 camera (https://translate.google.com.au/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.maksunaut.bplace d.net%2Fccd453.htm&edit-text=), quite awhile ago, ~2010. Obviously is very talented, doing his own firmware, machining, design even its own inbuilt control and CF card etc. He obviously also paid attention to the sensor sealing.

He also built his own scope, autoguiders, stand alone autoguider, telescope mount and goto controller.

Pity he hasn't released any details.

Also of note is the way he had a secondary chamber around the sensor and desiccant system. Which I have been thinking about. From the RS images and data sheets, it's not entirely obvious that we can access the desiccant or remove the desiccant plug externally, without camera disassembly. It looks as if the plug seals with a nut and O ring but its not obvious if we can remove the desiccant without undoing it's housing. Thus we may not be able to purge our camera. Am I missing something here?

Brendan

[glend]

Brendan why not put the dessicant on a plug that inserts into the housing, screws in and seals with the O ring. Several commerical cameras use this method. You can always charge the interior with Argon before you insert the plug, assuming you have it properly sealed. The new ASI1600 cooled cameras, have a top sensor cover section that unscrews from the rest of the case, exposing the desicant pill holders (4) underneath the cap; simple to change. I am sure you guys can come up with a solution.:thumbsup:

[luka]

Brendan, if you open the Altronics page with a D-sub connector of interest, say for example this one (http://www.altronics.com.au/p/p9654-amphenol-ltw-de9-female-pcb-ip67-rated/), you can click on specifications to get to a mechanical drawings. Also, to the left just above the image of the connectors it says "Amphenol LTW SDB-09PFFP-SL8001". That is the original manufacturer and the part number. This is not what Altronics usually does, only for some more industrial parts.

Glen, we are looking at commercial screw-in silica gel holders. However, the one we found here (http://au.rs-online.com/web/p/humidity-indicating-desiccators/0838726/) is VERY bulky. If you know of a smaller one it would be wonderful!!!

Brendan, the RS silica gel holder has a nut and o-ring. The way to do the mounting would be to screw it in properly and then use epoxy or some other strong glue to glue the nut to the box from the inside. Just have to be careful not to get the glue on the thread :)

However, the size of it is still a problem for the sealed chamber around the sensor. We could drop the idea of having a separate chamber for the sensor only and dehumidify the whole box. Or we could redesign the 3D printed part that the sensor sits on (see Faddy's images) and enlarge the chamber so it can have silica gel insert on a side... somehow...
As Glen suggested, we may have to make our own smaller insert, the RS one is too large.

[wasyoungonce]

Yep spot on. The desiccant carriers we were looking at have an "O" ring and nut that would sit inside the housing, thus when undone the nut comes loose nor can it be taken out to "charge".

Which means its not suitable and yes I've been looking at other camera set-ups just like you mention. You can actually buy some of these plugs from camera mfgrs.

We would need to ensure the housing is threaded for this...but the housing walls are a little "thin". Maybe glue in place a doubler?

[wasyoungonce]

Thanks for the info Luka. They did look the same as the Amphenol. IMHO the DSUB 9 pin might be best connector. Put it down the base of the case. Looking at the DB9 drawing and G106 drawing, there should be just enough room (edit...maybe not the G106 will need serious machining). Probably more if we use the Jaycar case. (Edit: might be better on the case lid at bottom. This way we could use a female to plug in a small power/gender box to hold a small PCB for USB/Power plugs)

The way the PCB is laid out should allow room for this as well even enough to put the connector on the case rear down the bottom.

Some desiccant plugs: here (http://www.teleskop-express.de/shop/product_info.php/language/en/info/p6074_Farpoint-SBIG-Replacement-Plug---converts-the-original-to-a-field-serviceable-tool.html); here (http://www.highpointscientific.com/farpoint-sbig-ccd-camera-replacement-desiccant-plug-system-fp329); here (http://www.darksights.com/SRP.htm); here; (http://www.modernastronomy.com/shop/accessories/qhy-accessories/desiccant-tube-for-qhy-cameras/)

Gary is asking how many camera cases we want. Send him an email.

edit:
oh it would be easier if the D sub was male pins with solder cups...just easier to solder wires into. Some of the D subs have long pins coming out the rear which may interfere with the PCB, or we can cut them down, but soldering to a pin is more problematic than a cup.

[luka]

Agree, solder cups at the back would definitely be easier to solder to than pins.

(If you were looking at the connector I sent a link to before, I just picked a random one to show you the part number on the top. I did not mean to imply that we should use either male or female connector)
Normally when building stuff like this I would use a female connector on the cable as it carries power and it is more difficult to short compared to pin connector.

Off-topic question, have you "freed" your sensor from the aluminium holder yet? I just noticed on Martin Raabe's page (your link few posts up) that he machined the aluminium to get the sensor out.

[luka]

By the way, I made a quick draft of the front 3D-printed plastic part, based on Faddy's images.

Dimensions are just a guess as nothing is know yet but I have to print it first, figure out the printed dimension for the sensor step and adjust the model for the plastic shrinkage.

I will need to get my sensor out of my D40 soon...

[wasyoungonce]

Yep totally agree. If we wanted we could use a breakout sub-PCB like Faddy. This would allow us to use a DB9 on the rear of the camera that mates with another DB9 (opposite gender) outside the case, connected to USB/Power connectors etc. Mount it all in a small plastic case. This will allow use to seal the case yet route proper USB/Power to the camera with proper connectors.



I have another D70 coming, I'll do them both at once. I'll have to get some DMSO. I did laugh at Martin Raabes method of getting the alloy holder off. I was thinking...no way...but...he's obviously not only practical he's damn 1st class at everything he touches, so the laugh is definitely on me. I also noted he included a sensor PCB jackscrew arrangement for sensor orthogonality, but I also guess he had micrometer dial gauges to set this all up.

Gary asked ..."why Faddy used the plastic sensor cover"...really I had to say...I'm not sure!:shrug:

[luka]

Are you still following the CN thread? Filip Lolic is another person of great skill, see his build here (http://www.cloudynights.com/topic/497530-diy-astro-ccd-16-bit-color-6mpx-camera/page-40#entry7316889). No plastic in front of the sensor.

[luka]

One use of the plastic in front of the sensor would be to locate the sensor in the centre of the window. But the PCB should already be doing that.
Another use I can think of is to keep it parallel with the window. The sensor is socketed so keeping it "flat" by pushing it all the way in may not work too well.

[wasyoungonce]

Sure am...gazed upon his case with envy eyes. Which I had lather and mill...errr and skills to operate them!:lol:



Yes this is what I was thinking. The PCB keeps the sensor at centre. Would be nice to have PCB metal standoffs near this to keep the sensor absolutely still and square.

[luka]

Have you seen this (http://www.cloudynights.com/topic/497530-diy-astro-ccd-16-bit-color-6mpx-camera/page-41#entry7330965) already?

[wasyoungonce]

Unfortunately yes... with desire.

Flolics machining is excellent. Doesn't look like he sealed the sensor area...probably impossible with the sensor pins and socket. His case is designed to accommodate connectors, silica gel plug etc.

edit:
looks like he has some youtube vids (https://www.youtube.com/watch?v=wNoOK8Nkvp8) of making his enclosure.

[luka]

Lesson of the day: don't type a reply in a window that you left open the whole night... you will get "token expired" and have to retype everything.

Anyway, the main reason I mentioned Filip's latest images is that I believe that his sensor holder will be spring loaded and not for us to start drooling over it ;)

Faddy's design relies on accurate distance between the PCB and the 3D printed piece, also including how deep the sensor goes into the socket.

Spring loaded sounds like a better idea and that piece should be easily printable. Not sure how Filip is dealing with the sensor alignment though...

[wasyoungonce]

I think he spring loaded the sensor cap to allow contraction, expansion. He probably should have use qty 3 or 4 spring mounts. Maybe even and adjustable spring load...that also adj's camera tilt? Like a spring loaded jackscrew

[luka]

You probably already knew but I just found out that the source code (https://github.com/vakulenko/CAM8_software) of the Ascom drivers for the Cam8x builds is actually available. Not sure if we want to play with that...
The microcontroller firmware does not seem to be available.

[wasyoungonce]

Hi Luka

ASCOM driver (object interface) is all open development ...to make it compatible. Gotta love how ASCOM has swept to popular use since inception.

I noted how grim has his cam86 (https://translate.googleusercontent.com/translate_c?depth=1&hl=en&ie=UTF8&prev=_t&rurl=translate.google.com.au&sl=auto&tl=en&u=http://www.astroclub.kiev.ua/forum/index.php%3FPHPSESSID%3Dt9a0hhkae8i 0u8p4ui1f8vrq03%26topic%3D28929.168 0&usg=ALkJrhgxTZRogWrEjj3wfFW-BEhp6vSYGg)up and running. He reversed the PCB case/lid so the sensor faces the lid. No images of the TEC or TEC HSF but you can see he routed wires to the TEC. Also out the crystal on the bottom PCB layer...so he is still "fiddling".

It appears he used a microphone type connector on the camera base, with what looks like a plate and seal.

I'm still doing the Cam86 PCB. Had an error on my schematic that wouldn't let me open it today...I managed to open one of the backups, thankfully. Only lost 2 hrs work. The Schematic and PCB do have differences. No idea how it got this way, the PCB appears to be correct, not the schematic.

I'll be around another week on this at least before it's close.

Oh received my 2nd D70 so I'll be pulling them apart for the sensors soon. I have a D70s and D70.

[luka]

Altronics/Jaycar have a whole bunch of waterproof connectors that would fit inside the case on the side, like Grim did.
Something like this (http://www.altronics.com.au/p/p9366-amphenol-ltw-6-pin-2a-locking-male-chassis-ip66-waterproof-plug/)? 2A per pin only though.

edit:
Probably even something like this (http://www.altronics.com.au/p/p9416-amphenol-ltw-6-pin-5a-screw-on-male-chassis-ip67-waterproof-plug/) would fit. 5A per pin, 6 pins. Quite big.

On the other hand, this (http://www.altronics.com.au/p/p9650-amphenol-ltw-de9-female-solder-cup-ip67-rated/) D-sub connector has 5A rating, according to the data sheet. I assume per pin???

[wasyoungonce]

Yes that current rating will be per pin. Probably over optimistic though. The only issue is we need to take this plug to a breakout board like Faddys as there is no such thing as sealed USB sockets, as we know.

Interesting in that Flolic on CN didn't seal his sensor area more like left vents. I wonder what's his thinking on this.

The Cam86 board is coming along slowly. There are a few errors or parts not numbered or miss-numbered. Still its a much better iteration of the original.

[luka]

We can always use multiple pins to get more current into the box if needed. D sub connector will have quite a few spare pins.

There exists one completely sealed USB socket (http://www.altronics.com.au/p/p9719-amphenol-ltw-usb-type-a-female-chassis-ip67-waterproof/) at Altronics that I know of but it is quite bulky. It would probably fit... just... with the latest inverse PCB orientation.

I can see a possible problem with the new inverted PCB design and the alignment of the sensor. Unless the lid is tightened metal-to-metal the alignment will change every time the camera is opened. I don't have any diecast boxes here to check, do you know if the lids get tightened all the way?

I keep thinking if it would be possible to use the 4 screws which hold the PCB to change the PCB and hence the sensor alignment. Drill the holes through, and then spring-load the PCB somehow... not sure how... but the idea was to turn the screws from the outside to change the alignment of the PCB and hence the sensor.

Looking at Filip's design, I think he will seal the sensor area. The red plate that the peltier sits on has only cutouts for the posts for the feet from the sensor to go through. A bit of silastic around the posts will seal this nicely. I think I saw somewhere that he is using multiple stacked posts to gain height (and reduce heat transfer?). This way he will even be able to leave one row of posts glued in while being able to separate the PCB from them.
Then there is something that looks like an orange seal around the Peltier.

Here is a question. If the whole box is sealed and moisture inside controlled carefully (dry gas and desiccant and possibly a humidity sensor), is there any need to keep the sensor in a separate dry chamber?

This is an interesting post about the sensor temperatures (http://www.iceinspace.com.au/forum/showthread.php?t=81592), in particular see Richard's comments and PDF. HQY8 has the same sensor as we are using.

[flolic]

Hi all, Filip from "the other" forum here :)
I just want to clear few things regarding design of my camera housing. First, sensor will be sealed in it's own chamber. This is not the easiest thing to do, but it is certainly the best thing to do if we think about moisture control and possible condensation problems. Air volume in that chamber is quite small (comparing to other design possibility to seal complete electronics). Plus I will fill the chamber with dry Argon gas and include some small piece of desiccant sheet.
Connection between CCD and PCB is via few stacked sockets, with one of them glued to aluminium bracket that will provide mechanical support and sealing (with additional O-ring).

Ok, let me just include few images for you to see what I am talking about :D

CCD socket (with sealing)
https://dl.dropboxusercontent.com/u/21387397/Astroforum/Cam84/cam84_ccd_socket.jpg

https://dl.dropboxusercontent.com/u/21387397/Astroforum/Cam84/cam84_ccd_socket1.jpg

This is how CCD will be connected, with individual pins to reduce thermal transfer. Plastic cup over the sensor will keep it pressed onto the TEC module and centered in the chamber.
(please ignore the greasy fingerprints :lol:)
https://dl.dropboxusercontent.com/u/21387397/Astroforum/Cam84/cam84_ccd_socket2.jpg

[bojan]

Hi Filip, welcome to "this" forum :welcome:
I hope we will see more of your posts in the future

[wasyoungonce]

Hi Filip,

a warm welcome to "downunder". We have been admiring your work but couldn't completely understand "sealing" until you posted the above images. Now we get it "a picture tells all"! Good idea, pot SIL connectors in a sensor mount housing. Double stack them for TEC clearance! Wish I could think of things like this.:sadeyes:

You then only have to worry about moisture in the sensor chamber! This is the problem with us atm. Using the enclosure the grim and others used. Its really not the best solution.

I suspect we will stay with this till we get some cameras "running" then push ahead to a different housing. I say this with eyes of envy towards your set-up!

Keep up the good work.:thumbsup:

Luka, I can add stand-offs to my PCB board drawing, easy as. But I think it would be difficult to hve this as a jackscrew arrangement to adjust from outside the box. Impossible to seal this.

If the box is sealed and flushed with gas then this enough for the sensor. No separate chamber needed. This is the beauty of Filips' enclosure. A small sealed chamber is easier to control than a bigger volume. I have some doubts the case we are using is fit for this (sealing). Sorry to say.:shrug:

On another note....my eagle PCB files got corrupted late last night. Wouldn't open, gave errors ..."not consistent" and "wrong technology... at line xxx" blah blah. After hrs and hrs...and a sleep..found Eagle has an editor I can use to goto the line, delete the "technology" field save...volià. It lives and re-linked back to the board!:D

[luka]

Hi Filip :welcome:

I have seen your name crop up on several forums and I was always impressed with your work :bowdown:
Thank you for the explanation, ideas and photos :thumbsup:


Brendan, as I said, I keep thinking about the jackscrew arrangement but cannot figure out how to do it. Anyway, good work with Eagle, but don't forget to keep making backups.

How will we actually mount our Peltier? One side to sensor and the other to enclosure (with copper inbetween to fill the space)?

When you say the box will not seal, are you talking about the seal between the lid and the base? We could run a thin layer of silastic on the outside after everything is screwed down. Not the best look but it will do the job :D
Or even squeeze some into the gap where the rubber seal is. It will look nicer but it will be more difficult to take the camera apart.


Regarding the moisture control, what if we keep everything (sensor + electronics) in one chamber, fill it with Argon and silica gel packs. The moisture levels will be quite low. Then we could add a humidity sensor inside the box with one wire going outside to a small display which shows the humidity inside the box. This will show when refills are needed.

Or we could copy Filip's approach, get a piece 3D printed (or milled) that screws into the existing PCB posts and has a large rubber o-ring which will seal the sensor enclosure. We could also glue the extension sensor posts, just like Filip did.

[wasyoungonce]

Gary has this covered, a slug of copper between the rear HSF and TEC. The rear HSF mounted on rear case and sealed to it. Of course the distances to Sensor etc have to be reasonably precise.



Just a little dubious of sealing capabilities of the box. Its not exactly machined quality! I've seen air leaks thru pins of connectors ...tiny very slow bubbles but more than enough wreak havoc. A perfect seal is probably not achievable with the case IMHO. We can probably get it well sealed but intrinsically its not something ever designed for sealing like this. This doesn't mean it won't work...obviously it does, proof seen in other builds.



Yes some silastics seal ok when run in thin beads, especially between mating surfaces screwed down. The case comes with a seal, neoprene or such like.



Yes this is the plan, seal it all, as discussed we might run into issues as we need a purge port (which we looked at with silica gel), which is probably right on the limit of fitting the case.



I think the original designers looked at this and decided its much cheaper and easier to try a "sealed case". Which is what we need, a cheap easy approach. Thus sticking with a case approach, with a purge port and sealed connector is out best bet. Seal the rear HSF and front nose piece to the case with a thin silastic bead. The connector is sealed, the purge port can be the silica gel ports we looked at.

We assemble the unit, leave open the purge port, soak it in Argon in a plastic bag, open bag seal the port...done. When you open the plastic bag keep its opening upright (Argon being denser than air) we can keep the item reasonably filled with Argon as we screw in the port. Of course we can leave the port in the bag and probably start screwing it in wile still sealed.

[wasyoungonce]

Slowly getting the board done...one bugger up at a time. Finally now doing trace laying...the difficult part! Manual routing FTW!:lol:

[bojan]

as RF electronics engineer, only thing I can say is that manual routing is the ONLY way to properly route the layout ;)

[wasyoungonce]

I guess I could lay signals I want laid then rely on autoroute...nah! I don't fancy RF...damn backplane on RFELATS was a rats next of solid shielded coax. I still shudder thinking about it.

1st Firmware hex (https://translate.google.com.au/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.astroclub.kiev.u a%2Fforum%2Findex.php%3Ftopic%3D289 29.1680&edit-text=) released, many thanks grim, ASCOM handler later

[luka]

Is there any other way :P

Seriously, good job and thanks for doing it.

[luka]

I could not stop thinking about using the 4 PCB mounting screws for the PCB/sensor alignment. I may have come with a solution.

Here is back-of-envelope drawing :D of the possible solution.
We would need to:
1. Drill a hole all the way through the box and make it large enough to remove the existing thread in the mounting holes
2. Bore a hole from outside with diameter larger than the existing hole and deep enough to fit a rubber o-ring in it
3. Put a screw into the new hole so it sits on the o-ring
4. PCB sits on a spring
5. The screw goes through the spring, through the PCB and is screwed into a nut that is fixed so it cannot rotate. For example, it could be glued to the PCB.

The screw will compress the o-ring when tightened and create a seal. Tightening/loosening the screw (from outside) will change the spacing of that corner of the PCB. We have 4 corners instead of 3 which makes it a bit trickier to do the alignment but it should work.
Longer screws will even allow adjustment of the sensor distance from the glass.

Sounds too simple...

[wasyoungonce]

The idea is sound but I think we cannot counter-bore the hole enough, the enclosure thickness is too thin. Wish I had answers sorry.

[wasyoungonce]

Slow progress on routing. I'll get there...someday. Found some differences and errors between schematic and board. So sometimes laying tracks...ripping them up, moving components points of a mm then re-doing it, more than once, is the work flow order. But, over all a much better layout than cam85m.

Also, so far it's easier to lay these tracks, compared to cam85m, changing this board to a different layout shouldn't be a major hurdle in the future. All that said....I'm not a fan of placing IC on the bottom layer...a few resistors and caps are ok but it gets complicated with ICs etc on base layer. I feel the PCB size is about on the limit...for my meagre skills. Also the board size limits adding things like sensor orthogonality jackscrews.

Having a ulnar nerve decompress surgery on Wednesday...might be not able to do do more on this for a week from then. lets hope I can get most done prior.

[luka]

As you said IC on the bottom is not the optimum. But I did not realise that the board is so crowded. I was thinking if we every build a round board, the way to go would be to stack two smaller PCBs. If possible to physically separate power and signal boards...

All the best with the surgery and my best wishes for a quick recovery!!!

[wasyoungonce]

Hi Luka

it's just minor surgery...already had right arm done and I lived...suspect I'll be back to havoc more destruction...:rofl:

The board is a little crowded in areas. I was also thinking of splitting the PCBs into 2. One power, the other signal. Double stack them. Not difficult to do, ideal for a round board design. Could even slip in an EMI shield (http://www.digikey.com.au/product-detail/en/3m/1170-7.7-X10-/3M1170E-ND/2213534) between them. :shrug:

Maybe next iteration?

[luka]

That's exactly what I was thinking. Build one as it is and then try to improve. We will also have something to compare to.

And if we manage to put all power circuitry on one side of one of the stacked boards we can use the backplane (copper pour) of the power PCB as the EMI shield. Just trying to be cheap :lol:

Managed to pick up a broken D70 on ebay for $12 (+ $13 shipping) :)
Hopefully the sensor is working. No way to test it as now I have D40 (working) and D70 (broken).
(And my daytime-use D90 + cold finger & IR modded Canon 450D + IR modded 1200D. Damn I got quite a collection :lol:)

Regarding the humidity absorption, have you seen these (http://www.harveynorman.com.au/cameras-printers-photocentre/camera-accessories/sony-anti-fog-sheet.html)? Usually retailing at about $25. Not that they are any better (or cheaper) than silica gel but it may be easier to mount an anti-fog sheet that a bag with small balls inside the camera :shrug:

[wasyoungonce]

Got some new 63/37 solder, .56mm from Mektronics, on sale. I'll need to do some solder practice boards to re-learn eutectic soldering again. I generally use 64/40 due to cost.

Also got some more equipment to get my Vacuum de-solder working 100% again. This will really help de-soldering the sensors. I did some trial vac de-soldering on an old cct board...around a second per leg "bingo" all solder gone. Sony does state (with their ICX413 sensor):
"To dismount an image sensor, do not use a solder suction equipment. When using an electric desoldering tool, use a thermal controller of the zero-cross On/Off type and connect it to ground."

Which I take to mean hand vacuum solder suckers are out proper vac de-solders are ok as long as zero point power switching and earthed tip. Gary asked me about this I said vac de-soldering should be ok. Mine is a proper vacuum de-solder (...a little old) but I'll interweave all the sensor pins with conductive wire (and probably connect to earth) before I attempt de-soldering.

Funny thing about the Sony anit-fog sheet...$5 HN...much more other places! Who'd have thunk it!

I was watching fleabay for Nikons, I saw one "damaged not working" they wanted ~ $60. I wasn't brave enough. Funny enough fleabay has Nikon sensors for sale reasonably cheap.

edit:
I'm also eyeing off some DMSO to "un-stick" the sensor from the carrier. Once used I should be able to send it on to others for them to use.:shrug:
Reading up Onsemi handling notes on CMOS sensor atm, AN56521. Nothing I'm not aware of though.

[luka]

I remember reading somewhere (I read a lot so I don't remember where) that D40 and D50 use flex boards for the sensor mounting, i.e. it is possible to unsolder it pin by pin. D70 is more difficult as it is a normal PCB.

That is one of the reasons I got D40 first and was planing to get another one. But the D70 for $12 + postage was way too cheap to miss, broken or not. As you said usually they go for around $60. My highest bid was actually $40 but nobody else went that high :lol:

I doubt that the Ukrainians + people on CN had access to proper vacuum de-solderers. I don't have one either (apart from a hand vacuum solder sucker). I may try Chipquik (the sensor may be too big for it) or copper braid :shrug:

[wasyoungonce]

If the sensor is on a flex then might be easiest to cut away extra flex material or use the flex to pry off the solder as you melt it. Chipquick would be ideal for this.

Either way it should be easy.

[wasyoungonce]

Ok rough completed PCB. Needs some clean up on layers "restrict" and "keepout"...well a lot, not just some! But finally getting there...where there is ...no idea, the damn image graphic hurts my eyes!:scared2:

And...no DRC errors! Been fixing them as I go.;)

Had a lot of warning as I used "0V" in lieu of "GND" so Eagle PCB has a hissy on this but it's not an issue. Guess I should use proper conventions.:shrug:

[luka]

Excellent work, let me know what I can do to help. Are we building your or Ukrainian Gerbers?

Now we just have to wait for drivers. Do we have the latest BOM for Cam86 yet?

Just a thought, making the PCB square (no extra "wings" on sides) will let us fit some connectors on side of the box if needed, like grim did in his latest "upside-down" version.

[wasyoungonce]

Thanks Luka..

I'd use our gerbers but it doesn't matter. There were some small issues I found with grims/faddy's layout, apart from small things like miss-naming or qty 2 R4's on base layer. The only major deviations were:

DS18B20 temp sensor is powered from 3.3V direct but the PCB had this powered from same 3.3V but after inductor DR6. Not a big issue really. This output from DR6 is also put to a pad "VDD", no reference about this on schematic, but I can do this.

VT1 schematic gate from pin 12 DD5, yet board shows this to be pin 13? Mine is put to pin 12.

DD5 pins 7 & 12 broken out to pads PB6 and PB0. Which I haven't yet done as this PB0 conflicts with VT1 gate pin from DD5. I need to look a little more closely at this.

Apart from these...just need to clean up orphans and layer restrictions. As you said, this also allows us some space for connectors etc.

edit:
Oh DD13 +5V supply was rotated wrong way..or something strange...on their board. Fixed on mine.
I can generate parts lists from Eagle PCB CAD....straight into places like digikey. As long as the parts were fully recognised.

[wasyoungonce]

Ok, grim posted up pictures of his PCB from board house (https://translate.google.com.au/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.astroclub.kiev.u a%2Fforum%2Findex.php%3Ftopic%3D289 29.1700&edit-text=). They fixed:

DD5 pin 15 now same as mine;

Still confused about DR6 3.3V feed to R10. This is not an issue really, its just they powered R10 after DR6 inductor, the schematic has it powered straight from 3.3V, not thru DR6 (I have it straight from 3.3V);

They corrected PB6/PB0 via. But these not on schematic nor mine but I will place them as per his board.

Still confused on DD13 +5V LDO. They have it rotated in silkscreen wrt to it's pins. No idea why. They have pins connected ok, as is mine.

They tended to use thinner traces...I tend to oversize most of mine, especially if I had room. The rest looks correct. I'm just laying extra 0V vias between layers and doing some layer restrictions to clean up orphan planes.

I won't be able to do this RCB for a few days from now. But I'll be thinking about it.;)

[luka]

I have started compiling a BOM. Looks quite different that Cam85m. It is partially finished but I wanted to order few more parts.

Best of luck tomorrow.

[luka]

By the way, do you know if we are using Atmega328 or Atmega328P? There is a small difference between the two.

[wasyoungonce]

Thanks Luka...I will export mine from eagle and compare. Sorry I assume its a Mega328!

Actually I'll email you my BOM exported from Eagle. Its in CSV format. Haven't checked it but it should be right. Looks terrible in CXv..sent as txt as well. Just open with excel. FWIW...it's not useable without a lot of work.

[luka]

Going through the components one by one.

By the way, your BOM shows mega328P (note the P at the end).
Grim's BOM shows only mega328 (no P).
The chips have the same footprint but the firmware may be different.

edit: Someone on the Ukrainian forum just posted a link (https://translate.googleusercontent.com/translate_c?depth=1&hl=en&ie=UTF8&prev=_t&rurl=translate.google.com.au&sl=auto&tl=en&u=http://ru.aliexpress.com/item/4-Pcs-M-F-1-778mm-Pitch-40P-Straight-IC-Socket-Connector-for-PCB/2049899422.html&usg=ALkJrhjqGNOMAKXhGIVP0zp36yAEUqg qOA) to the sensor sockets from China.

[wasyoungonce]

Hi Luka...I'm alive....!


any mistakes..it was the drugs ;) Anyway I'll go thru each you listed to clarify:



Schematic shows mega328P.

I used this device package (element14 eagle libraries (https://www.element14.com/community/community/cadsoft_eagle/eagle_cad_libraries)) as it's the only one I could get for TQFP32 for this device. I believe the pin out is the same as ATmega328. Thus it shows as 328P in my BOM. I've amended the schematic name to ATmega328.

This raises an issue with CAD drawings...who does the drawings for different PCB CADs? In this case Element14 has an agreement with many suppliers.



DD6 - your footprint is SO14 while Grim's BOM shows SOIC14

I looked at datasheets prior to putting it down, yes it says S014 but drawing showed part number is "D" variant which is a SOIC14. Just to be sure, I just measured IC on PCB (using measure tools) and it was 9.6mm x 3.9mm body dimensions, so, it has to be a SOIC.



DD9 has part number and value reversed

Not sure on this I removed part numbers from PCB as it was too crowded. I cannot see part number on board for DD9?



DD10 and DD11 are both 78L06 in Grim's BOM. However, your BOM shows 78L06 and 7806A.

Yep that;s a typo. I chose a SOT 89 voltage reg and mis typed it when putting down the 2nd one. Fixed now.



ZQ1 crystal, your Farnell part number is for 3.68MHz crystal, not 12MHz

Yep this is definitely wrong. I selected 49S standard small thru hole package from variants in parts but this listed it as a surface mount crystal in BOM. No idea how that happened, but it's only in BOM not on PCB or schematic. Just amend manually in BOM.



DR3 - Grim has RLB0712-101KL which I found at RS part number 811-8796. Your BOM shows WE-TI-7447462010_5075, Element14 part number 1848262. Both are inductors and more-or-less the same. Any reason you changed this or am I missing something?

Ahhh you have me... I cheated as I couldn't get the correct package and I was too lazy to make my own so I found one close enough but forgot to amend any reference to the wurth package. Yes it shouyld be as on grims list, RLB0712-101K/


DR4-DR6 show resistor in the description but they are inductors

Ahh caught me cheating again. Couldn't find 0805 inductor package (in std Eagle library) so I used a 0805 resistor. It's confusing but suffice to say the PCB is ok but BOM is in mucked up. I'll fix this.


C11 = 1uf (Grim) or 0.1uF (you)

Yep my error, I've fixed it now.




Missing are C22 + C25, C26 + C32 and C37 (aluminium capacitors). I was not sure of dimensions, they do not seem to be mentioned anywhere.



No idea how these escaped..here they are:
C22 , 100uF, 25V, CPOL-EUE2.5-6 , size: E2,5-6 , POLARIZED CAPACITOR, European symbol

C25, 100uF, 25V, CPOL-EUE2.5-6, size: E2,5-6, POLARIZED CAPACITOR, European symbol

C26, 220uF 10V, CPOL-EUE2.5-6, size: E2,5-6, POLARIZED CAPACITOR, European symbol

C32, 220uF, 25V, CPOL-EUE2.5-6, size: E2,5-6, POLARIZED CAPACITOR, European symbol

C37, 220uF, 10v, CPOL-EUE2.5-6, size: E2,5-6, POLARIZED CAPACITOR, European symbol

As you said, we are now starting to get closer and the more eyes that look at this then mistakes will be seen and fixed, thus confidence is high ;) :lol:

As for the SIL packages...I've never ordered from ALiExpress or Taobao...found any thing I needed wither RS/Element14/Digikey/Mouser or dare I say Fleabay. Guess I've lived a sheltered life! if your brave enough...then go for it. If you need funds for parts purchases, pls let me know. As for element14...yes I've seen them put up prices. I suspect it from "search demands" . I've always been sus about this. In fact I was ordering "H" bridge drivers for Losmandy Gemini awhile back and the price went up double o'night. I needed them RS had none so I complained as I said the part was xx to fix for the user now it's ZZ. They dropped the price back and gave me discount. Worked for me!

[wasyoungonce]

Oh meant to ask...do you prefer thermals on vias and pads? I think I should do them its easy to do, otherwise soldering to ground plans is a "inert expletive here"!

Also I have the sensor package pads/pins as 1.27mm dia with a drill of .6mm I looked at the SIL package measurements and it lists the SIL pins at .51 nom but I just saw that it can be up to .64. I might need to increase my hole size. Also my pads appear quite a bit larger than the Ukrainian forum images. I like bigger pads and it doesn't interfere...I hope.

I'll fix the drill to .7mm for the sensor....

edit:
Done sensor pin drill holes .7mm. Images loaded to show differences
Also got my 63/37 solder...very very nice. Doing a practice board, the joints are just so clean and shiny. Perfect filleting and much better amount with .56mm solder (could have gone smaller? I feel I've been far to lazy using 60/40 for too long.

[wasyoungonce]

Luka just going thru your parts BOM...noticed on DD10/11 you put in RS stock 661-4591 (http://au.rs-online.com/web/p/linear-voltage-regulators/6614591/), a 6.2V LDO. The LDO is supposed to be 6V version, but ok 6.2V Vs 6V isn't much should be ok though. Just sayin!

Ok found another...VT3, IRLML6302TRPBF, is an extended range min order qty 50 @$.28 ea (50 x $.28 = $14). Alternate at RS is: IRLML6302PBF, 610-6851 (http://au.rs-online.com/web/p/mosfet-transistors/6106851/), in Pk of 5 for $3.93 (@$.786 ea)...exactly same as other one as "TR" reference is just "tape Reel". So the alternate is cheaper.

Also its good to cross check everything! I'll do the whole BOM and check it against the parts on Schematic. I'll go thru all then do a summation post as posting everytime I find something is in dribs and drabs.

[luka]

Brendan, it is great to have you back!!!

Thank you for cross-checking the BOM. I did heaps more work on it last night before ordering lots of stuff (already got the VT3 that you mentioned).

Luckily I did not order DD10,DD11 yet. I totally did not see that the part I picked was 6.2V. Neither was I expecting 6.2V for a part labelled 78L06 when all other 78L06s are 6V exactly.
The alternatives do not look good at RS, they have only 25 packs which are quite more expensive :mad2: Must do a bit more searching.

I will finish the BOM later tonight, only 5 components are missing now.

[wasyoungonce]

Hi Luka...
I'm up to resistor now but are hurtin a bit so I'll stop for the night, but I have to give you what I found:

Caps C17/C30/C33/C38 are 100uF, 1206 package 6.3V....buuuut...they are on 6V rails...so the original part told to us is wrong, they should be 10V versions. RS, 788-3057 (http://au.rs-online.com/web/p/ceramic-multilayer-capacitors/7883057/?searchTerm=788-3057&relevancy-data=636F3D3126696E3D4931384E525353 746F636B4E756D6265724D504E266C753D6 56E266D6D3D6D61746368616C6C26706D3D 5E283F69292852537C5253207C52532D293 F5C647B337D285C73293F5B5C732D2F255C 2E2C5D285C73293F5C647B332C347D24267 06F3D313426736E3D592673743D52535F53 544F434B5F4E554D4245522677633D4E4F4 E45267573743D3738382D33303537267374 613D3738383330353726), on backorder atm till 15/08/2016. You might find an alternate but they should be 10V versions as these are not as cheap.

C32 should be 220uf, 25V. Some how this was entered as 10V, Its on 12V rail in so has to be at least 16V or above. I found RS 749-7177 (http://au.rs-online.com/web/p/aluminium-capacitors/7497177/?searchTerm=749-7177&relevancy-data=636F3D3126696E3D4931384E525353 746F636B4E756D6265724D504E266C753D6 56E266D6D3D6D61746368616C6C26706D3D 5E283F69292852537C5253207C52532D293 F5C647B337D285C73293F5B5C732D2F255C 2E2C5D285C73293F5C647B332C347D24267 06F3D313426736E3D592673743D52535F53 544F434B5F4E554D4245522677633D4E4F4 E45267573743D3734392D37313737267374 613D3734393731373726), 6.3mm x 11mm, 2.5mm pitch (my PCB is set to 2.5mm pitch for electro radials, but if you find one at 2mm or 3.5mm pitch I can change PCB), 105º, 7000h life, high ripple , low leakage.

Sorry don't know how these slipped thru!

This appears all atm, I'll email you what I have atm. I haven't been looking at E14 or digikey yet. If there are high count package buys we can break them up and sell to each other.

Brendan

[luka]

Hi Brendan

"the resistors are hurtin"... describes it quite accurately :rofl:
I actually did them all last night so the latest BOM will have them all... hopefully correctly. Email is sent.

Also Faddy commented about condensation problems but he said that his camera seal was working OK. Apparently the housing warmed up (from Peltier) and silica gel does not work above 35 degrees C, it actually releases moisture.

Also Grim said there was a typo in BOM and that R5 is 1K and not 10K. Already corrected in my BOM.

Luka

[wasyoungonce]

Thanks,,,on to it, Yes saw that about heating case. Their HSF is probably too small.

[luka]

Looking at AliExpress part here (http://www.aliexpress.com/item/4-Pcs-M-F-1-778mm-Pitch-40P-Straight-IC-Socket-Connector-for-PCB/2049899422.html) (you may need to clear all aliexpress cookies if you looked at the Russian translation before).

4 x 40 sockets... how many would we need? 40x should do one camera, i.e. one pack should do 4 cameras (if we don't stack pins). Should I get one pack first? It is over a month for the delivery.

It does not specify if it is beryllium copper.

[wasyoungonce]

Just saw some of these on fleabay, here (http://www.ebay.com.au/itm/Male-to-Male-1-778mm-Pitch-Distance-40P-Position-IC-Socket-Connector-4-Pcs-/391261772501?hash=item5b1904dad5:g: ESUAAOSwayZXjC6U) here (http://www.ebay.com.au/itm/4-Pcs-M-M-PCB-Board-1-778mm-Pitch-40Pin-Straight-IC-Socket-Connector-/121744690358?hash=item1c588c8cb6:g: KwcAAOSwqfNXjC~z).

edit:
Sorry they are straight thru pin to pin not sockets. Here is a 1.778mm pitch socket. (http://www.ebay.com.au/itm/64-WAY-SHRINK-DIP-IC-SOCKET-AJ-64-HZL-ASSMANN-MACHINED-TURNED-PINS-1-778mm-PITCH-/271423535242?hash=item3f321a788a:g: dDEAAOxyrM5TICSv) A little big but can be cut down.

[luka]

Ordered one set of sockets from AliExpress. It should be enough for our first round of cameras. It is free shipping anyway.

edit: Just saw your edit with the ebay link. Already ordered from AliExpress.

[wasyoungonce]

Noticed an error on grims PCB layout. He has C26, 220uf 25V electro (on -8V line) with polarity +ve to gnd (0V) on schematic but on etched PCB its reversed (and image). +ve is on R19 side which is the -8V rail. It's wrong, the +ve has to be towards 0V.

I have to flip it around. I suspect they noticed this and just turned it around when building the PCB.:shrug:

edit:
Done.

Ok Capacitors, C17, C30.C33,C38 are 1206 sized 100uf ceramic (a lot in a small package). RS has them on back order, but we can change size to 1210 RS has some in stk now, 766-1179 (http://au.rs-online.com/web/p/ceramic-multilayer-capacitors/7661179/), $5.95 for pk of 5. But if we changed I'd need to amend the schematic and drawings. Not difficult to do though. I've left it as is atm.

[luka]

Found some 1206s at E14 here (http://au.element14.com/tdk/c3216x5r1a107m160ac/cap-mlcc-x5r-100uf-10v-1206/dp/2211193). Would they be OK?

edit: SIL sockets I ordered from AliExpress have been shipped.

[wasyoungonce]

yes there fine...I'll leave those caps as 1206 size.

[flolic]

Hi guys, thanks for warm welcome, and sorry for not responding earlier.

About sockets, I bought these: http://www.ebay.com/itm/111186383448
Now I have stock for the rest of my life :lol: These must be cut to needed length, and quality is excellent.

Unfortunately, I have not finished my camera yet. Too much work and family related obligations...

OTOH, my camera PCB is obsolete, version cam84 R3.1. I must design new cam86 round PCB. Brendan, may I ask you for Eagle .sch file? :)

[wasyoungonce]

Thanks for the link Filip. Just PM me your email address I'll bang of the files to you...it's in eagle 7.5ver. Of course it'll be for a rectangular PCB....but at least it will cut down on some work you need to do. As you know the most difficult part is placement and routing.

[wasyoungonce]

Just testing outputting Gerber files using cam processor. Odd thing, drill file cannot be read (there are no drill plots) with Gerbv unless I use Excillion24 cam processing.

I think its an issue with eagle. all PCB mfgrs use gerber files I can usually upload them at a place like OSHpark to check them. On the subject of PCB mfgrs...PCBshopper (http://pcbshopper.com/)looks to be the place to get quotes. Of course depends on the mfgr on the quality of the made board, but, PCB way gives you quotes for all the big suppliers. I did a quick quote and prices range from (for 5 PCBs):
Elecrow: $13 + $40 (AUD) P&P EMS delivered in 13 days, at least I read it as that.
Seeed: $24+$51 (P&P) 12 days; and
It goes up exponentially.

I read the inputs as the PCB costs and P&P. 10PCBs wouldn't appreciably cost a lot more. You can select slower (cheaper) lead time. We also have to have screen printing on both sides and milling the sensor hole.

While I'm confident the PCB is ready....I need to re-check the locating holes and sensor is perfectly centred. They have to be spot on. Also I think I should delete the top/bottom PCB tabs to allow room for connector/dessicant.

Attached is a rectangular PCB showing this and an image of the Gerbv of the PCB.

[luka]

Excellent work.

I think we should get 10 PCBs, the extra cost is almost nothing. Also consider that our version 2 (round design) may not work well enough, for example due to the increased noise. In that case I would like to build a 2nd camera on this PCB. A spare would be good as well...

Last chance if anyone else wants to join in.

Also add few $ to those quotes. From my experience they will try to squeeze every cent out of you and we will probably encounter things like credit card fees etc.

By the way, my $12 broken D70 arrived today. No way to test the sensor as the camera is broken, but (in theory) I have two sensors now. Did you get DMSO yet and/or tried "freeing" your sensors?

[luka]

I did a bit more research on silica gel and it is not true that it is releasing humidity when heated above 35°C, as stated on the Ukrainian forums.

It only desorbs moisture when heated above 104°C, see for example here (http://www.desiccants.cn/news.htm) or here (http://www.giebel-adsorber.de/en/products/adsorbents-silicagel).

However, when heated above 25°C it will adsorb less moisture. There is a graph here (http://www.sisweb.com/art/pdf/desiccant-performance-information.pdf) showing the adsorption capacity vs temperature. The capacity drops from about 22% at 25C° to about 18% at 50°C.

So, heat in the case is not really a problem unless our case gets to 100°C :P

[flolic]

PM sent ;)

[flolic]

I am worried about signal integrity (placement of decoupling caps and ground fills and vias).
It's hard to see from your images, I'll wait until you send me project files...

Small changes here can make big difference regarding noise and interference.

[wasyoungonce]

Absolutely Filip...

What needs to be done for each PCB design is not only a base level noise level but also a readout noise signature. I guess grim would be best to talk to on this. IMHO...wee need to separate the Power supplies to a different sub PCB and shield this. Then make the main PCB. Of course this philosophy needs to be measured...again back to grim and measuring signal noise. Ensure de-coupling caps are as close to the ICs as possible. Look at ceramic caps noise values...especially high value caps...they have an effect depending upon their dialectic.

About the only figure I saw was Grim stating he got down to ~ 4.7eV dark noise value, which is about base value for this sensor and indeed very good for any CMOS sensor. Again readout is an issue.

As said, separating and shielding the power would help a lot but...this needs quantification to which only grim can supply. This is what he mentioned in his original round PCB. He basically said...the noise floor was to high...and did not elaborate. Could be a language issue at play here.

But Luka and I both agree....separate the power from main PCB, shield it, would go a long way to improving the system. We both agree this is an iteration of design (cam 86 rectangle PCB) that will not end here. Oh .....there are other small issues...like ceramic capacitor noise values (from pizeo effect) that can be effected by dielectric used... a whole can of worms...

Brendan

[wasyoungonce]

DMSO ordered, 99.9% Fleabay. I'll probably be removing the sensors and separating them from their carriers this week.

[wasyoungonce]

It's remove sensor time!:D. Only did one camera...the other later.

Vacuum de-solder made very short work of the pin joints. I wire wrapped the sensor pins to afford some ESD protection for it. This splayed the pins a bit, but they should return when removed.

Would be real nice to keep the sensor metal carrier and sensor together...maybe in the future.

[wasyoungonce]

Annnd so it won't be lonely here's the other ICX453AQ sensor removed.

This is where the Vacuum de-soldering comes into it's own...just seconds per joint. Pity I cannot use it on SMDs

Also had to fix the sensor socket pin row pitch..managed to stuff it up...fixed now. Remember...if a jobs worth doing ...do it twice!;)...errr in my case 3 or 4 times:lol:

edit:
Gawd I need to clean my ESD mat!

[luka]

Excellent work :thumbsup:
I have been sick in bed for a few days and haven't done anything at all.

Regarding reusing the metal frames for mounting, I am not sure if the frames are the same for D40, D50 and D70. I think you said you had 2x D70. I have D40 and D70, will try to disassemble them soon and compare the frames.

[wasyoungonce]

Hi Luka...hope your on the improve, get well soon.

Yes my cameras were a D70 and D70's. The sensor carrier is so damn well made...shame to toss it considering it would simplify mounting.

I noted even though the vac de-solder worked a treat...it did leave little dags of solder on some legs which need cleaning.

Interesting post on Ukrainian forum by "Wind" (https://translate.google.com.au/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.astroclub.kiev.u a%2Fforum%2Findex.php%3Ftopic%3D289 29.1760&edit-text=) regarding sealing and backfill etc. Well worth a look, download the image.rar file. Reading Lukas work on silica gel absorption. Good work, nice to see a critical eye and facts instead of conjecture. The Language translation is difficult enough as is. Despite this they have done some good work.

Meant to say if anyone wanted their sensor removed vac-desoldered...I can do it no problems, literally a 5 min job (apart from disassembly and this only took a few mins) ...only issue is postage.

Had to fix my errors on centreing the sensor and changing the measurements to layer 47 "measure".

[wasyoungonce]

Ok we have the PCB layout pretty much wrapped up and ready, looking at connectors.

We really looked at D sub DB9 mainly due to the fact we could get an IP67 sealed version, ideal for our needs. Pretty much all DB9, sealed D sub (IP67) have the seal on the front face. Given most if not all are around 39mm across the face width...this means we need a space of 39mm to mount this. The bottom of the G106 case is 30mm.

We could not fit it on the bottom of the case without major metal excavations. Nor does fitting it on the lid (or base of case) work given the dimension and where it would sit in relation to the PCB.

About the only connectors I see that will fit and are IP67 are Hirschmann CM series. Here (http://au.rs-online.com/web/p/mil-spec-circular-connectors/2377598/) is a 6 way that I think will fit. 19mm dia hole, 23mm centre for mount holes (this might be too big and interfere with the lid flange?), IP67, has a flange seal. We could fit it at the bottom base of the case, "I think" ?

Any ideas appreciated as we are getting closer to making the PCBs, thus we need a decision soon'ish. Jaycar has a similar but in round style, here (http://www.jaycar.com.au/chassis-plug-hirschmann-c16-6-pole-pe/p/PP1002) and here (http://www.jaycar.com.au/ip67-6-pin-panel-mount-plug/p/PP0546) not sure on dimensions though.

I'll upload the Hirschmann CM data sheets to the Cam86 google site.

Brendan

edit:
Actually there is a whole swag of IP67, 5 and 6 way mini sockets at the major suppliers, RS (http://au.rs-online.com/web/c/connectors/circular-connectors/industrial-automation-circular-connectors/#esid=4294960342&applied-dimensions=4293691580,4293691672,42 94878613,4294885220,4294885229,4294 885351) etc. I started to upload some datasheets...but there are too many types. Best to look at them and narrow ideas down. Many will fit out needs, size, IP rating, current and number of ways.

[luka]

Sorry, no time for long messages but I have used Hirose miniature connectors in the past. See here (http://au.rs-online.com/web/p/miniature-connectors/5365778). I may even have a some here.

[wasyoungonce]

Thanks Luka...I think those Hirose use a synthetic/rubber insert...which while might be IP67 would probably not be enough for hermetic seal, or a seal that lasts.

I'm mainly looking at Sockets that have the glass fibre type sealed inserts (that will fit, also have a good current rating ~3A or more), like:
Molex Ultra-Lock M12 (http://au.rs-online.com/web/p/industrial-automation-circular-connectors/7527505/).
Molex/Brad 120084 series. (http://au.rs-online.com/web/p/industrial-automation-circular-connectors/7527509/)

Yes some of these have "PCB" type pins one the rear, i'm sure there would be other terminations available...or I cam make a sub-PCB to solder wires and this connector to. "I'm sure these are more that I've missed.

edit:
We could resin pot the end of a connector ourselves....like the Hirose?

[luka]

How about IP68 connectors, here (http://www.altronics.com.au/p/p9845-amphenol-ltw-5-pin-5a-m12-screw-on-line-female-ip68-waterproof/) and here (http://www.altronics.com.au/p/p9850-amphenol-ltw-5-pin-4a-m12-screw-on-chassis-male-ip68-waterproof/).
4A per pin, 5 pins - did we need more than 5? 4 for USB + 1 for peltier power (shared GND).

edit: The real Amphenol part number is just above image, to the left.

[luka]

I finally removed the sensor assemblies from my cameras, D70 and D40. Few notes:


D70 was soooooo easy, it took less than 5 minutes to have the sensor out and the camera assembled again. Back off, sensor out, back on.
D40 has a PCB behind the sensor and this needs to be removed. There is even a wire that needs to be desoldered. The removal is bottom off, back off, PCB off, sensor out, PCB back on, back back on, bottom back on. Similar to Canon 450D.
The sensor metal frames are completely different for D70 and D40. So much for reusing the existing metal frames... it would limit the camera choice to one camera only.
D40 has a flex PCB. It will be easier to desolder the sensor from it.



Photos attached. Sorry about the quality, I can retake if needed.

[wasyoungonce]

Hi Luka good stuff.

On the D70 sensor...I was looking at the part numbers on the ICs...found EEPROMs, vertical driver IC, shift registers....all the ICs we are substituting. But reverse engineering that sub-PCB...would be a nightmare.

I found the D70 sensor so easy to remove as well. When de-soldering the carrier helps to sink off some of the heat...bonus!

[luka]

New corrections on the Ukrainian forum:


Also new BOM + Gerber is there.

And there is talk now about adding humidity sensor inside the box.

[wasyoungonce]

Thanks Luka I'll have a look.

I have done changes to my lists and Schematic/PCB. But he is talking about the ATMega328P...I think he's saying he programmed it as a ATMega328P with fuse bits as ATMega168P and it was fine. But he also mentions his IC designation "ATmega328" was rubbed off so I'm not sure if its ATMega328P or nor vanilla ATMega328?

I think he's saying its an ATMega328P yet parts list as ATMega328! Sigh! Oh the other attachment was a program update for the ATMega/ATMega328P.

edit:
I asked in CN forum about this and vacuum of sensor!

Received my DMSO this morning. My little friends, the sensors, won't have long to wait before being freed of their earthly carriers.:lol:

[wasyoungonce]

Tapped case holes with M4 x .7. Holes were exactly the correct size for M4 but the screw feels a little loose. Would have liked it with a tighter drill hole.

Had to angle grind one of my M4's to make it a bottoming tap but it all tapped ok right to bottom. I have just over 6mm thread travel. The material probably tapped too easily, it feels a little "weak". So I'll have to be careful tightening screws.

Just using M4 countersunk to feel the threads, have M4 cap head screws on the way.

[luka]

Perhaps using nylon M4 screws will be easier on the thread than the stainless ones.

[wasyoungonce]

Hi Luka..yes thats' a possibility

Just got an answer regarding ATMega328 from Sergiy on CN....it's supposed to be an ATMega328P. The reason is the 328P is more modern and code compiled for 328 can run on 328P but not backwards.

I'm changing schematic now and looking at pin differences. I already had 328P in schematic...probably didn't have footprint of 328!

Sergiy said he didn't know if the sensor could withstand vacuum but they had a sensor cover glass fail moisture got in a destroyed the sensor. They are having issues with dew on the sensors.

edit:
further to this..while I have the correct package there is an issue with the difference between packages pin changes:

Pin ATMega328 ATMega328P
Pin 3 GND PEO/ACO......... A problem with schematic
Pin 6 Vcc PE1................A problem with schematic
Pin 19 ADC6 ADC6/PE2
Pin 22 ADC7 ADC7/PE3

Pin 3 on our schematic/PCB is GND, "0V"
Pin 6 on our Schematic/PCB is VCC, "3.3V".

Yet the ATMega328P data sheets list these pins on the "P" as GND and VCC respectively. TQFP32 package...sigh!

[wasyoungonce]

A new version PCB and Gerbers has been released on the Ukrainian forum. (https://translate.google.com.au/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.astroclub.kiev.u a%2Fforum%2Findex.php%3Ftopic%3D289 29.1800&edit-text=)

Basically they:
Fixed some PCB polarity errors (C26);
Replaced C10 with .1uF;
Replaced R9 with inductor 4.7uF, now called DR7;
Added connection for DHT temp humidity sensor, pin 13 of ATMega328P. Connected to Vcc via new R9 (10k) on bottom of PCB;
Added suitable quick connect connectors for USB and +12V/Gnd inputs.

I'll go over these changes and change our PCB respectively. The SIL for the connector for the power could be an issue. Those 2.54mm connectors are only rated to ~2.5A ...I reckon even less. Jaycar has .156 (4mm) connectors rated to 7A...these might be a better bet. I'll see if I can get them to fit.

EDIT:
Ok got the USB 3 pin SIL on the PCB. I can get the .156" 2way SIL but I'll have some small movements, and I'll look on fleabay/RS to see if I can get a slightly small body version than Jaycar's. I have connectors similar to those used in their PCB but no 3way versions or .156" version.

[wasyoungonce]

Ok PCB/Schematic updates done uploaded to Google hangout.

Back at the ranch..dinner is served..."Sensor Soup"! Mmmm! Yum!

Soaking sensor in DMSO to free it from alloy carrier. Soak and forget come back in a few days. It apparently falls away.....I hope. I hope I don't come back and the sensors dissolved!:question:

[wasyoungonce]

The DMSO did it's job, this morning...the ICX453AQ sensor just fell away from the carrier. That was easy! Just some residual gunk to clean away.

[wasyoungonce]

Just have to post this...new toy to play with, 3D render of PCBs. Much better than the old eagle3D and POV Ray.

You can look between the layers, pads etc and imports BOM etc. Apparently the netlist file is IPC-D-356 format and it contains way more than just connections. Obviously some of my parts are not correctly rendered correctly. I have to work on this. Also my colour selection is.."I chose poorly"!

[luka]

Congratulations on successfully freeing the sensor!!! Excellent work :thumbsup:

So the black plastic frame and the glass stayed glued to the sensor. Interesting...
Rowland, in his Canon 450D cold finger mod, glued few SMD resistors around the front glass and was passing low current through them to keep the glass warm to prevent fogging. From memory it raised the sensor temperature slightly but it was nothing major.

[wasyoungonce]

Hi Luka

The IC package, sensor glass & glue, totally unaffected by the DMSO. Even the copper wire I used as shorting wire between the pins...everything unaffected except the glue.

I have no idea what's in DMSO apart from what GOOGLE (https://en.wikipedia.org/wiki/Dimethyl_sulfoxide)tells me. Suffice to say...the sensor fell off the carrier in ~3.5 days!

I agitated the sensor from time to time to pick off softened glue and also rotated the sensor in the solution face up/down. I put a saucer over the container to contain the solution. Quite a bit of solution evaporated and was on this saucer after 3 days.

Dew mitigation...hard to tell. One side we are cooling the other side heating....poor old sensor will be split for choice! :lol:We will probably have to do something like resistor dew control.

[wasyoungonce]

2nd sensor free from carrier. Took around 2.5 days for this. Just has some cleaning spray on it in pics.

In the mean time...kids have used all out internet data allowance so I'm throttled back to dialup speeds...Grrrrr. Painfully slow. Where is my NBN oh that's right I'm on pair gain so I won't be getting it until they remove my copper....in other words "never".

[luka]

Sorry I was slow to reply, had quite a few things to sort out. Roughly, what size Peltier are we after so I can start looking around. I assume we want to get it as cold as possible?

Also, the SIL sockets I ordered on Aliexpress just arrived. HUGE >10cm x 7cm x 7cm box with tons of bubblewrap with 4 tiny sockets inside. I first thought it was another DSLR with a new sensor that I forgot that I ordered :-)

The pitch looks OK (measured by ruler) but I don't have a free sensor to try. Things just got too busy recently...

[wasyoungonce]

Hi Luka...sorry I don't know the Peltier size. I would expect probably no bigger than the Mill cut-out for the cold finger. Also probably no more than 3A, 12V...probably best 2A @12V.

Sorry I cannot help till tomorrow...kids used all my internet data allowance and even loading this page took 5 mins....Grrrr. Data resets tomorrow.

[luka]

The Peltiers...
There are few cheaper ones at Jaycar around $20, all 40x40mm. The smallest one (http://www.jaycar.com.au/thermoelectric-peltier-modules/p/ZP9100) may be able to run at our power levels.
Then we move a step up to RS/E14 and similar shops and the prices start at $40. Large range of Peltiers to chose from, even stacked ones for over $100.

After lots of reading it looks like running the Peltier at about 70% of max power is the best for long life. This means that we should look for about 15V Peltier that runs at less than 4A.

For example, something like this (http://au.element14.com/european-thermodynamics/aphc-12704-s/peltier-cooler-module-14-7vdc/dp/2484934) at 12V will give us around 3A max current. The max temperature differential will be about 40deg.

Or there is a fancy dual-stage Peltier here (http://au.element14.com/european-thermodynamics/ap2-158-10-15/peltier-cooler-module-16-2vdc/dp/2519561). A bit pricey at $100 but it should be able to achieve temperature differential of 95C!!! However, the cooling power is 8W only... I suppose that will be the main question, how much cooling power do we really need.

[wasyoungonce]

Hi Luka...been absent for a few days. Enjoyed the break...fixing soldering stations. Now I want anything but soldering stations!:lol:

I agree on Peltiers...we don't ant to over kill...but ...seeing that beauty from E14.....my testosterone levels went up a notch!:lol: I would have thought a 40mm sq unit is a bit big for our needs. Looks like the Ukrainians (and others) (https://translate.google.com.au/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.astroclub.kiev.u a%2Fforum%2Findex.php%3Ftopic%3D289 29.1820&edit-text=)are looking at a much smaller 3A unit (https://translate.googleusercontent.com/translate_c?depth=1&hl=en&ie=UTF8&prev=_t&rurl=translate.google.com.au&sl=auto&tl=en&u=http://www.ebay.fr/itm/262174876804&usg=ALkJrhjg02ZI7ol4Tm6MLk5hH-RXM0cC-A). Notice its 8V? What's the deal here? Odd voltage rating? Or is this an avg voltage?

Those TEC units are a little pricey, nice, but pricey. Might have to revert to fleabay. From my understating of TEC units (I may be wrong) they:
Are best run at a slightly lower voltage than specified;
They don't like PWM but hey this is probably the easiest method of controlling average cooling capacity (depends on what PWM freq used in the firmware);
Current draw for our application is limited ~3A'ish;
The centre PCB hole for the cold finger is 20mm dia which might limit TEC sizes to around this dimension;
A dual unit wins hands down over a single unit;
What cooling capacity do we need? Hard to tell...I noticed my Canon 450D pretty much plateau'd off noise wise at around 5ºC and below. meaning the noise floor had been reached. That said...what to QHY run their units at?

I'm not fussed atm...I'll start looking now and get back on this.

Brendan

[luka]

Haha, I knew you would like the E14 dual stage Peltier. Interestingly it is only 7W so it may not have enough capacity to get the heat out. Also the dual stage Peltiers are extremely inefficient... single stage has about 10% efficiency and dual stage then 1% efficiency.

Regarding the temperature, the colder the better. QHY8L (same sensor as us) goes down to -40C while QHY8-Pro goes down to -45C.

I have cooled 450D as well and, as you said, the noise plateaus around 5C. However, this is only part of the story:


Canons do some heavy processing even on RAW files which probably explains why the noise plateaus. I believe Craig Stark had some excellent articles on this, I can find the links if you are interested.
The ~5C plateau applies only for relatively short exposures. I tested it for 3 min exposures and the dark frames were identical to the bias frames at 2C exactly (for my camera/cold finger build).
Going colder at 3 minutes exposure may not reduce the thermal noise but it definitely reduces the intensity of the hot/cold pixels.
I recently tried doing 10 min exposures and things are very different compared to the 3 min exposures. At 10 min the noise definitely gets less when colder than 5C. There is still a significant noise even at -3C when compared to the bias files (-3C is my limit due to my insufficient power supply). I suppose that Canon's manipulation of the RAW files works differently for such long exposures.

Also I was unable to find any 20x20mm 12V Peltiers on ebay. Only ~3V and ~5V versions exist.

Few of the larger and possibly suitable 30x30mm Peltiers are here (http://www.ebay.com.au/itm/TEC1-12707-3A-12V-35W-30x30x6-5mm-Thermoelectric-Cooler-Peltier-Plate-Module-/391530981365), here (http://www.ebay.com.au/itm/TEC2-19003-3A-12V-35W-30x30x6-5mm-Thermoelectric-Cooler-Peltier-Plate-Module-/331938293865) and here (http://www.ebay.com.au/itm/TES1-12702-Heatsink-Thermoelectric-Cooler-Peltier-Cooling-Plate-30x30mm-12V-2A-/262174876843).

[Gary47]

Hi,
The cameras I have looked at all have a larger TEC unit and the cold finger reduces in size to fit the sensor or through the PCB. The lump of copper I have for this project is about 35 mm diameter so we can use a sensor that will cover that. As well it would be no drama to add a larger flange to the cold finger to fit a larger TEC, I would only have to find a piece of copper plate, a bit of old bus bar or similar would do the trick and we could have full coverage for a wide range of TECs.
Does anyone know how much power the sensor outputs?
Need to look at the TEC specs because TECs only achieve their rated deltaT when they are operating at maximum efficiency which is not actually moving any heat. Two stage TECs are notoriously inefficient.
An "over size" TEC will operate at better efficiency and give a faster cool down time.
Gary

[luka]

Fully agree with you Gary, the larger the Peltier the better, if we can fit it. We are probably limited to about 30W max power due to 12V input with <3A max current due to our connector.

By the way, QHY8L (same sensor) uses 3.6W-30W, i.e. about 4A of current at max cooling. Advertised delta T is -35 degrees and it has a single stage Peltier.

[luka]

12V/3A/36W, dual stage, 30x30mm Peltier on ebay here (http://www.ebay.com.au/itm/TEC2-19003-12V-36W-2-stage-Multistage-Thermoelectric-Cooler-Peltier-Plate-Module-/391489710823). Max temp difference is 80 deg.

I don't like the ebay cheapies as they don't really have much details compared to, for example, the RS/E14 Peltiers.

[luka]

Did you see the Ukrainian forum. There is talk of a new camera with ICX613AQA sensor. Read noise is 4eV, half of our current build.

edit:
ICX493AQA, not ICX613AQA

[wasyoungonce]

Hi Gents

yes I saw the work done by Ruslan. Very nice. He's been doing this camera for awhile (https://translate.googleusercontent.com/translate_c?depth=1&hl=en&ie=UTF8&prev=_t&rurl=translate.google.com.au&sl=auto&tl=en&u=http://www.astroclub.kiev.ua/forum/index.php%3FPHPSESSID%3D67bju07oeq9 5fmkol9s3lu3g92%26topic%3D27737.msg 531605&usg=ALkJrhi0xwWLYt77b_FCUjk5TkMGz38 IYQ#msg531605)but obviously gained enthusiasm to continue from the CamXX project. he's done some fine work and indeed a whole new range of DIY may come from this.

Just think...all those 2nd hand cameras on ebay are about to get more expensive!:lol:

I'd be up to building one...but after we get this one running.;)

As for TECs..I'm going to buy some 3Amp vers and as Gary said around 40mm square. Actually 40mm sq is a little large for the case? We are cutting the case so the TEC clears it!

[luka]

I am in as well for building the next version :D

See if you can find any 12V 40mm square Peltiers on ebay. I retract my word when I said we can/should use >12V Peltier so it does not run at 100%. PWM will do this for us and we should not limit ourselves from the start to reduced performance.

Also if looking on ebay, the descriptions of Peltiers have lots of incorrect/misleading information. It looks like lots of descriptions were copied/pasted and some corrected and some not.

[wasyoungonce]

Ordered qty 2 30mm sq TECs, one 3 Amp one 2Amp (fleabay). They might take awhile to arrive.

AstroCCD.ord has posted ASCOM for Cam86. (http://astroccd.org/2016/09/cam86-software-pack-01/)

The Camera C45, 1uf ceramic, grim was replacing this with a higher capacity ~10uf or up to 22/33uf 20V in 1206 size. I'll leave ours atm but probably need change it.

Others are building their PCBs and noticed +5V/+3V regulators getting a little hot. They are within temp specs (but the 3.3 is hot) and the I draw is in specs but probably on the upper end of what they can supply. We probably need to maybe look at a slightly higher rating regulator, especially for 3.3V.

I'm keen to resolve this "regulator issue" and get on an build our PCBs...but its good we waited somewhat. But I feel we won't be far off.

Brendan

edit:
Ok we might get away wit han 3.3V LDO with a more narrow voltage input range than the current LDOs LP2950XX. These have a wide voltage input range (up to 29V in) but I think the trade off for this is a limit of power dissipation. Where as MCP1702XX is rated for 13.2V in and has 250mA I out in TO92 package. not so sure atm have to read the pdf on this device, but it appears to use a pass FET thus can handle I flow with less heat.

[flolic]

Brendan, no matter what kind of regulator you use, dissipated heat will always be the same because it is function of the input/output voltage difference and amount of current through regulator. That's how all linear regulators work.
What can be done is better thermal management, for SMD parts that means using larger area of PCB for heatsinking.
Off course, one should not exceed maximum allowed power dissipation of used voltage regulator.

[wasyoungonce]

Actually you're right Flip. I was quickly looking at MCP1702 specs and saw it using a pass FEt in lieu of more traditional transistor array for the device. I was thinking hey this will have less V drop across it.

When in fact its exactly the same as a traditional device P = (Vin -Vout) x Iout. Reading the TI tech paper on this, SBVA027, details lots of info on the SOA of these devices.

I have to now go back and look at the current thru the regulator. No use chasing tails if it's in the SOA of operations.

[luka]

Saw this (https://astronomy-imaging-camera.com/products/accessories/t2-tilter/) posted by Glen in this (http://www.iceinspace.com.au/forum/showpost.php?p=1267266&postcount=433) thread. May be useful for the camera/sensor alignment, it is way cheaper than any other devices I saw before.

[luka]

From the Ukrainian forums, it looks like the camera will work with voltages over 7V. Regulating 7V will definitely produce less heat than trying to drop down 12V.

Either we find TEC that works between 7V and 12V or we could have a separate power for TEC? 12V or more depending on the TEC. It would require an extra pin...

[wasyoungonce]

Ahhh got it ...I was confused about all this. Makes sense as it'll drop pdip across regulators as well.

Brendan

[wasyoungonce]

Ok a little of topic but something we may need for this project is the ability to solder and de-solder SOP/SIL/QFP smd ICs. Since these are quite small and real-estate for these is at a premium the pcb is packed closely...I picked up a used Military Royel RE900-4 station. (http://www.ebay.com.au/itm/Royel-RE-900-4-Soldering-Station-Main-Unit-/182266231172?_trksid=p2047675.l2557&ssPageName=STRK%3AMEBIDX%3AIT&nma=true&si=FtGCi2QKZyqbZjjYvxdIQ%252FjxkFk% 253D&orig_cvip=true&rt=nc)

Its a little rough but my older Royel irons (and de-solder iron) are exact matches for use on this unit and I know these units, I used to use them and repair them. I had my own de-solder system but this is far better as well as twin transformers/control PCBs for duel concurrent iron use. But it gets better I have fittings to plug straight into the vac and pressure ports and I know the rotary vane pump on this unit is a ripper.

But there's more...it has resistance soldering facility. I need a handpiece (can make my own) but basically this uses around 4V (variable) at high current (variable) to basically heat by current flow or spot weld. This is a very valuable feature not used much by techs...costs too much. Damn nice unit indeed and will come in handy.

Also looking at getting some Hakko 950 soldering tweezers for my Hakko 936, mainly for SMD IC removal. It appears the Ukrainians had to remove and change some ICs and a heat gun is a poor tool for such a tightly packed PCB. But, hakko 950 hand pieces are expensive...I found Aoyue T001 (http://www.ebay.com/sch/i.html?_from=R40&_trksid=p2047675.m570.l1313.TR0.TRC 0.H0.Xaoyue+t001.TRS0&_nkw=aoyue+t001&_sacat=0) knockoffs for around $35 which I can make fit my 936 station (actually my 936 station is a home made knockoff with exact copy PCBs and uprated 75Va transformer (http://www.eevblog.com/forum/blog/eevblog-596-world's-cheapest-soldering-station-yihua-936/msg708737/#msg708737)) I also have proper Hakko 907 irons as well as clones (on that note I found a place selling cheap clone 907 irons with exact specs as the hakko units). Also the Aoyue 950 handpiece has much cheaper tips for QFP, and other smd packages. Hakko tips cost ~ $100.

A lot of fiddling but it'll be worth it. I hope:lol:

[luka]

That is some nice gear. I am sure it will come handy. Nothing so advanced here.

I also have some news. I got inspired and tried debayering my D40 sensor. Well, didn't quite get to debayering yet. I only got as far as the glass removal and I already damaging something by accident. I bumped the camera/sensor while trying to put the sensor back in for testing and now the D40 does not even power up any more (I have tried reseating all the connectors multiple times). I am not even sure what is damaged...

Hopefully the experience will be very useful for everybody, here are my notes regarding the glass removal:


I have a feeling that glass removal and debayering may be easier while the sensor is still in the carrier.
I used a heatgun (858D clone from ebay) to heat the glass and separate it from the mounting.
The choice of the temperature and the heating time is critical. While places like this (http://www.jtwastronomy.com/tutorials/debayer.html) suggest heating only for a few seconds, I needed to heat for over 10s at the time to get any result. Probably even as close as 30s!!! It could be because they used a butane torch which is hotter?
Temperature... Eventually I settled to 450 degrees C which is the maximum my heater will do. After doing lots of short tests I believe that hotter is better. We want to quickly heat up the glass so it expands and the glue cracks before heating up the surroundings. The faster this happens the better (up to a limit of course).
I noticed that when using over 350 degrees C the glass gets cloudy but this just wipes off. I am not sure if some coating gets burned off at this temperature?
Heating works the best when started on corners. After a little while the glue will change colour indicating that it separated. It then becomes easy to "push" and expand the separated section with heat.
Working in small sections is the best and sensor MUST be let to cool down to ambient in between.
Patience, patience, patience. Some sensors may be easier to remove than the others.


MOST IMPORTANTLY:


The glass is surrounded by glue on all sides. This must be cut with a sharp blade. It will not be separated by heat. This glue will hold the sensor in the carrier even when all bond under the glass has been broken. I did not take any photos but the glue is clearly visible on sides.
When cut with a blade the glue on sides will make a white/yellowish powder. Blow the powder off and keep cutting. You can stop cutting when there is no more powder being created, i.e. when you went through all the glue.
Use a magnifying glass/lamp. Make sure all the glue on sides is gone before attempting to heat up the sensor.
The blade may not follow the sides of the glass. USE a magnifier to check!!!
I had a quick go at a corner from D70 glass and it seemed to unglue much, much quicker/easier than D40. Unfortunately my D70 body is broken so I decided to proceed with the functional D40 so I could monitor the debayering.


Anyway, that is the progress... or the lack of it.

I am not sure if I accidentally broke the body or the sensor electronics, the sensor is likely still OK... I hope. I will try debayering it once I get a bit more time but I cannot test it as the D40 does not power up any more. So I am scouting ebay for a cheap D40 body or D40 sensor.

[luka]

Quick update, I found a broken solder joint inside the D40, probably from when I "bumped" the camera. It is partially working now (the back LCD is dead) but it refuses to take photos.

[wasyoungonce]

Well done Luka....I only have a D70 and D70s you can scrounge parts from. If I had a D40 I'd send you the rear LCD. The LCD may just be a ribbon flex?

I'm guessing you removed the glass with the sensors still on their carriers?

You're braver than me removing the cover glass...yet when we get this camera running...that's what I will be doing!;)

[luka]

I am sure that all of us will have a go at debayering at some stage. I just thought to try it out while I had a working system (D40 + sensor) which would let me test the progress. And I also acquired a set of NB filters and I wanted to play with them and a mono D40 :D

But I did not have much success really, the D40 focuses (in viewfinder) but refuses to take photos which may indicate that the sensor, or more likely its supporting electronics or ribbon, is broken. Nothing obvious like the crack on the body that I fixed.

Removal of glass with sensors on carriers is much easier. I had to apply a fair amount of force to cut through the glue on the side. Having a sensor sitting on pins would not be the best idea for this. I suppose using less force would work as well but it would take longer.

I am not sure what to do next... I could debayer this D40 sensor but without any means to test it. I do not fancy doing it "blind". And my D70 does not work so there is no advantage trying to do it on the D70 sensor. Hmm :shrug:

[wasyoungonce]

Annnd just to finish off the post on the Royel soldering station......It works perfectly. Minor clean up, had to re-configure my iron pin config to suit the Royel (they are Royel irons but were on Adcola stations...I think this is how Royelston started), but that's a minor issue as I had changed them to 5 pin DIN......as everyone did with these systems.

Just some old caps to replace, but everything is "A OK". Funny thing about Royel...they were one of the few "MIL Spec" systems back in the mid 80's. Damn nice specs, way over engineered and a joy to use.

Mannn that Vac pump is way over spec'd as well as the resistance txfmr. The station is surprisingly goo condition. Just a few old electro's to replace....not that it needs it.

Yes my old "ad Hoc" Adcola with attached vac pump system worked but hey why dine on crab when you can have Lobster!

Keen to try the resistance tweezers!

[Gary47]

I got this link from Astronomik in reply to a query about filter availability.

http://www.astronomik.com/en/neue-produkte-new-products/luminance-filter-l-1-l-2-l-3.html.

Can you please advise which filter you prefer for your set up?

Regards,
Gary

[wasyoungonce]

Gary that really looks smick! I guess in my case I'd go to the L3, 400~690nm but I'm no expert in this...far from it in fact.

Kinda...I'll have what he's having...in this case what you are both having.

I got one of the TECs....40mm Sq. Was supposed to be 30mm Sq. Grrr

Back is playing up atm....will be down doe a day of so and no idea if what I'm typing is correct...on some nice meds!:confused2:

[Gary47]

Thanks Brendan,
I'll file that one away for when we order.
That's a formidable looking bit of gear, makes my little Royel soldering set look pretty humble. Do you have to open it and unscrew the glass jars to remove solder or is there an external solder trap?
Watch those meds, its all good fun at the time but the after effects, oh dear.
Regards,
Gary

[wasyoungonce]

Hi Gary...the iron has a solder trap, filter etc in the handle as well. A bit of overkill but I guess they don't want foreign muck in the pump. The Glass jars also have filters!

Meds don't work except make me:( ...not me!

[luka]

That looks really, really nice Gary :2thumbs:

Regarding the luminance filters, I am not sure. I have ED80 + flattener ATM and I used it with Baader UV/IR-cut filter in the past, until I dropped the filter and smashed it :mad2:. That worked well to remove bloated stars.
The Baader UV/IR-cut seems to have the same spectral response in IR as the Astronomik L2 but it is somewhere in between the L2 and L3 filters in the UV range.

I am tempted to say L2 but I will post a question on this forum regarding my setup. Hopefully one of the more knowledgable people will be able to give advice (and could probably also help you with your decision).

Brendan, I hope you get better soon. Back pain here as well but nothing very serious.

[wasyoungonce]

Sorry gents in hospital last few days, still here. Turns out back pain was a cyst on kidney that burst and had kidney bleeds. Was in bookooo abdominal pain thus then thought appendix (pain was on RH side only) Well funny thing....I said to my wife " it'll be ok I'll see how I am in the morning"....famous last words!

I hate hospital food!:help:

Should be out in a few days;)

[luka]

If I was any closer I would bring you a pizza. But you will probably be out before I traverse the Nullarbor. According to IIS you are only 2,869km away ;)

All my best wishes for a quick recovery!!!

[wasyoungonce]

Thanks Luka will be out tomorrow...gone as fast as it came on....:thumbsup:

[Gary47]

Hope you are out soon, you are right about hospital food, its a plot to deter you from using the health system.
I just had my gall bladder out because there were growing cysts, panic time and then after it was all over they told me that the cysts turned out to be cholesterol and were benign. I don't know whether to be happy or annoyed++.
Still tender and unsteady as a result.

There is no need to worry about filter Luca, I just thought that if we purchase the filters in one lot we will save on postage (about 11 euro quoted) but we don't need them until we are ready for final assembly. Astronomik advise that they have them in stock.

[wasyoungonce]

Thanks Gents

yes I am still...not 100% yet. How do I know....I looked at a half consumed cab sav bottle last night and felt sick to the stomach! Definitely not well indeed!:lol:

I'll be up an about soon. I know how you feel Gary!;)

Watching the Ukrainian site...they appear to have ironed out most issues and indeed most issues appear to be soldering! Also noted the issue with the differing ATMEGA328! As far as I'm concerned the PCB is pretty much done really the only change was increasing C45 to anything above 10uF ceramic. Thus this needs a 1206 package and PCB change. I'll do this now on our drawings.

I have one of the TECs...unfortunately I ordered one 40mm SQ and another 30mm SQ. The 40mm one arrived. II think its too big but will play with it soon. I really wanted to compared differing Power dissipation with the same size TECs.

[flolic]

So, this Friday my camera saw starlight for the first time :)
I am quite pleased with the final result, but I know it can be even better because subframes are not so clean. There are some noise and interference patterns visible in them, but thankfully they calibrate fine.

Cooling is working fine but I am not entirely happy about delta T, camera can achieve "only" 30C below ambient temperature. On the plus side is that it takes only 1.7A on 12V supply :)

Anyway, full resolution image of the M33 galaxy is here:
https://dl.dropboxusercontent.com/u/21387397/Astroforum/Cam84/M33_cam84.jpg

Image is processed in Pixinsight. There is no noise reduction applied to the image.

SW ED80 with SW 0.8 FF/FR
cam84 (round board)
Exposure 6x600s, 4x900s, sensor temperature -10C
Belt modified NEQ6 mount

There is a noticeable lack of Ha regions, it seems that cheap Chinese UVIR block from ebay that I used as a optical window is completely cutting that wavelengths...

[wasyoungonce]

Flip that's brilliant. These sensors really do have potential! Well done indeed.

Brendan

[bojan]

Filip,
Which UVIRCUT did you use?
I am using Rocolax (from ebay), according to graph (http://www.iceinspace.com.au/forum/showpost.php?p=1252973&postcount=4) supplied by manufacturer, it is supposed to pass Halpha.
Gary Honis also sort of claims (https://stargazerslounge.com/topic/168057-need-help-with-choosing-cooled-modded-dslr/) this filter is OK...
However, I am not totally convinced this is the case, I couldn't see much difference between photos taken with modified 450D+filter and my standard 60D without filter.

However, it seems Rocolax filter is beneficial when used in combination with 60D on my Canon 400 f2.8L at full aperture - stars are not bloated, but with 450D the effect is somewhat visible.

[bojan]

Since I have Hartridge spectrometer at hand, I measured (with help of red LED) the cut-off wavelength of my Rocolax UVIRCUT filter.. 656 nm is passing through without much attenuation.

[luka]

Bojan, can you measure the transmission of the filter in the far IR? A clear light globe with a glowing filament will be a good source of IR.

The 450D sensor will happily detect photons at wavelengths over 1100nm and your filter seems to start transmitting again around that range. See for example here (http://dslrmodifications.com/rebelmod450d14a.html).
On the other hand the Canon filter in your unmodified 60D may be blocking this far IR.

[bojan]

I will try.. later.
The setup will be a bit more difficult than this one, because Hartridge spectrometer is visual instrument, and was holding the whole thing in hand.

Previously taken spectra of Eta Carinae (attached) doesn't show a lot above 945nm.. it depends of how much IR is in eta spectrum for th estart [NB: my 450D is modified (filter was removed), but not de-bayered (yet)]

[luka]

Filip, congratulations, that is great news :thumbsup:

Which Peltier did you use? Was it single or double-stacked?

[luka]

Bojan, I did a bit more reading tonight and the maximum photon wavelength that will interact with the monocrystaline silicon is 1125nm (1.1eV photon energy). Photons with lower energy (higher wavelength) will go through it without any interaction. Now I am not sure what else Canon added to their sensors but this wavelength is probably the maximum wavelength you need to worry about.

Now, the transmission graph for the Rocolax filter stops at about 1090nm, i.e. short of our maximum of 1125nm. While not likely it is possible that the transmission goes up in this small but unknown range. For example, Baader BCF (http://www.holland-observatory.net/astrophotography/cameras/canon-xsi-450d-mod/) does exactly this, the transmission rises from 0% to 80% between 1100nm and 1140nm (it is about 40% at 1125nm).

Also to consider is that the sensor efficiency at this far IR end is quite low.

Just some food for thought :shrug:

[flolic]

Thanks guys!
I really don't know what's wrong with my Rocolax filter (and I don't think it's image processing error either). Ha regions are just not there...

Could it be that I turned filter the wrong way around? The red reflective layer should point to the telescope side, maybe I installed it backwards. Does it really matters? Hmm...


Luka, Peltier is double stacked DIY type :)

https://dl.dropboxusercontent.com/u/21387397/Astroforum/Cam84/2_stage_peltier.jpg

[wasyoungonce]

Hi Flip

I have a few UK admirers of your work asking questions. I said they should join and post. Excellent work!

bojan...thanks for your assistance and input...much valued indeed!:thumbsup:

[flolic]

Than they really should join ;)
I'm sorry that I can't participate more, I am stretched between my job and my family, with very little spare time dedicated to this project...
Progress on my cam86 board is exactly 0% :(

[bojan]

I will check spectrometer calibration with neon lamp....

[bojan]

Calibration looks OK.
According to Christian Buil's webpage spectrum (http://www.astrosurf.com/buil/us/spe2/calib2/osram_1.gif) of domestic energy saving lamp (https://sendyourlightbulbstowashington.file s.wordpress.com/2012/07/cfl_spectrum_500_250.png) (bright red line), my spectrometer is almost spot on (my reading of 6113A line is 6119A).
So, Rocolax filter does pass 6560A Halpha line.

For my measurement I used LED (which has a pretty broad spectrum around 656nm), and measured the wavelength of the (sharp) edge introduced by IR stop filter.

[flolic]

So, in the end, what is conclusion? My filter is somehow different or what?
I have Astronomik UVIR block but only in 1.25" size, on my next imaging session I will try with that.

[bojan]

Hard to believe..

[bojan]

I just repeated the (visual) measurement with Energy saving lamp as reference.
The line I was monitoring (650.5nm) is marked with ellipse on the attached annotated spectrum of the lamp (it is quite faint).

Summa sumarum: Without Rocolax filter it is visible, with filter it is not.
Conclusions;
- I have made a mistake with first measurement reported earlier (it is not easy to pinpoint the cutoff wavelength visually)
- Rocolax filter is NOT good for Halpha (contrary to manufacturer's claim and supplied transmission curve) (I contacted them and asked for advice)

[wasyoungonce]

Getting lots of interest requests from CN on cam86, Mainly referred them here to join for access to our GD files etc. Makes it much easier. I think people are having a difficult time following the Ukrainian forum thread or finding info.

They discuss more than one camera type and its difficult with the language. That said...we wouldn't be here without their initiatives! Again much praise for Grim and Vakulenko.

So welcome to all new readers...join up and become active. The more mouths the better the tune!:lol:...well maybe!;)

Brendan

[flolic]

:thumbsup:

[flolic]

About the Rocolax UVIR blocking filter, last night I again take few shots of M33 without it. And I can confirm Bojan's measurement, this filter is not good for astrophotography. Without filter Ha regions are clearly visible.

On the other side, I am totally surprised with the stuff I can pull out from my cam84 v3.1 camera. Cam86 should be even better, with lower noise.

M33 from the last night:
(8x900s, gain 63, offset 12, CCD temperature -10C)
https://dl.dropboxusercontent.com/u/21387397/Astroforum/Cam84/M33_crop_no_UVIR_reduced_stars.jpg

[wasyoungonce]

Damn you are getting good data. Damn that camera is performing well. Kinda like the Canon D20a of the modern era. Less mega pixels is more! Kinda hitting the sweet spot for many arc seconds per pixel wise!

Well done indeed.

Brendan

[gehelem]

yes, here i am
i'm Gilles, living in France. i completed Cam84, it took me 9 month
no picture under real sky, because my own challenge is to make it work with indlib under linux... No sucess for now .. No pictures

I'd like to try this Cam86, i thougt it could work easier with linux because of the Atmega, replacing FT2232 bitbang communication with a good old serial. But i might be wrong...
Doesn't matter, it'll be funny

[osult]

Thank you Brendan for the invite to Iceinspace. I'm Tom - a new arrival from the CN forum. I was about to begin construction of the CAM84 v4.1 when I saw the Ukrainians building CAM86. How do I go about gaining access your CAM86 pcb files on google drive? The gerbers from the Ukrainian AstroCCd site are difficult to use.

Thank You again.
:thanx:

Tom

[wasyoungonce]

Good morning Tom and Gilles....welcome to downunder. Remember we are upside down here so hang on!:P Oh did I mention the time zone differences.....yeah this can produce disjointed conversations so please be patient, we are 10hrs faster than UK but 10 years behind!:lol:.

I think the ATMEGA328 does the "bit banging" timing in lieu of hardware and ASCOM driven timings in cam84/85m. This takes some communication load off the connection. Also the ATMEGA328 is capable of doing more...like on-board TEC cooling control, etc etc.

I've sent invites to our GD with "read" permissions at the moment. Please by all means use what you want but I need to discuss with the other IIS gents on accessing "write permissions" for you.

I think I've already sent the GD invite...please tell me if it does not show up.

Really good to see new members interested. Kinda.."new blood to spill"!:lol:

Brendan

[wasyoungonce]

One bit of news I should say. I was talking to a gent on CN thru PM and he didn't have experience nor links to files....(like many others) but wanted to see if he can get a camera like cam86.

He had gone to PCBWay and they said they can produce the PCBs and populate them cheaply..here is an extract (no names given):

"I wanna built my own ccd camera, and already looking for a defect or second hand nikon. But I could't find the schematics or files related with the cam86 project hence I'm contacting you. I can also join google/yahoo groups if there are any. I'm not so handy with smd soldering but I think I can get it done in a company like PCBway. On their web site it says 10 boards costs 10USD and PCB assembly is 20USD so I think I can get it done under 100USD including shipping."

Which I found very interesting. He wasn't looking for a freebie build but at getting the PCBs made and populated. Indeed this looks cheap. Yeah I know...pay cheap end up with cheap! Maybe we should investigate this. Also saw this similar discussion in a French forum but I think the costs they were looking at was quite a lot more. Obviously they cannot populate the PCBs with sensors.

Anyway...worth a mention.:shrug:

Brendan

[luka]

Tom and Giles, welcome!!!

I have used PCBway few times in the past. We actually got a PCBs made there only about a month ago. The boards were always of decent quality. In one instance the silk screen was out a fraction of a mm and a bit blurry but that does not matter really.

I would estimate the cost of our PCB to about AU$25 for 10 + about AU$35 for shipping to AU + few $ for PayPal/CC fees. The price per PCB would be cheaper for larger orders. Easy to get quotes on the PCBway website. The biggest cost usually is the shipping.

I was never involved getting populated PCBs with PCBway. We recently ordered 10 prefabbed PCBs from another Chinese manufacturer for about $1200 (not including parts). Mainly through-hole components and not too many. The base US$20 will probably cover 50 solder joints and they charge per solder joint. AtMega already has 30 solder joints. >$100US probably may be reasonable cost to expect.

The quality of the hand solder joints we had on our prefabbed boards was quite good.

It may be worth getting proper quotes. Another option would be to selectively only get the difficult parts soldered. Note that it will all be hand-soldered unless we order over 100 boards. Testing etc all costs extra.

Brendan, do you trust someone else to do the soldering for you ;)

[wasyoungonce]

Got me by the short an curly's....Nooooossss...

Actually....yeah someone trained in HRHS and or MicroMin PCB repair. The guys doing this in our workshops were way better than me. Unfortunately we tended to leave them in these work areas as their experience was too valuable to keep changing workers.

Even though we were all trained to the same standards...some are better than others! Sadly I was not at the top end of the soldering gene pool ...but...still passable!;)

[luka]

As I said, the soldering on the boards we got made up looked really good. Better than mine :-)

I suppose those guys sit and solder the whole day and have more soldering experience than we ever will.

Another option is to get some boards prefabbed with the difficult parts and to also order some blank boards. That gives us an option to do them manually at almost no extra cost.

[luka]

Hi Giles, :welcome:

If you are planing open-sourcing the indi code I may be able to help with this, once my Cam86 is working and if I can find time (no promises). I don't have experience with indi but, among other things, I am a software developer with lots of experience with embedded devices.

[wasyoungonce]

That Idea has a lot of Merritt. I'd like to claim it as my own:D...but sadly it wasn't:sadeyes:.

I saw Grim was putting on the FT2232 and ATMEGA328 (and the appropriate power supply lines, hooking these up to a 9V battery then installing ATMEGA firmware. Damn fine idea.

There could be a number of ways we can get boards part done...but I'd like to at least try some. I've been reading the Ukrainian thread and noted they had some issues with soldering, bridging pads and degraded ICs. I'm ramping up my equipment now to be able to R&R these SMD ICs easily.

If I have to replace a bad IC I'll use my hot Air Yihua 828D. I'm getting some ChipQuik (and some more nozzles) so I can apply the chipquik across the pins/pads of a bad IC (with an iron). Isolate it off on the PCB with Kapton tape then use hot air to low temperature melt these bridged joints all at once. The ChipQuik will lower the melt point significantly thus lessen PCB/Pad over heat. Of course I don't care about he removed IC.

I put them in a Voodoo container I have. Kinda like keeping the bones of something I've killed. I gaze at this container of dead parts with satisfied joy on occasions. This would be a good study case for some Psychiatrist.:rofl:

This looks to be the easiest, safest method of IC removal! We certainly have attracted some outside interest and it's gaining.

Brendan

[luka]

I also like the idea of assembling the board in sections. This is a complicated build and needs assembly in stages and testing of each stage... not something I would normally do but in this case I would.

Imagine having the whole thing done (in China or here), powering it and it goes off in smoke because voltage regulator was faulty and was outputting 15V instead of 5V killing most of other components...

Also there are two aspects of the build for me:
1. Getting a working camera
2. A challenge of building it
Having it pre-built in China removes half the fun :D

[wasyoungonce]

Yes..its the trill of the hunt in completing something! The journey is half the fun!

I've just ordered a medium size pack of Chip Quik from Oritech. Gawd the cost of this stuff is high and varies greatly across Australia. Even shipping was high...$16.5 from Melbourne to Melbourne!

Sigh...I much prefer RS with free shipping...but they don't have this stuff. Anyway...this will make it easy to remove degraded ICs etc without damaging the PCBs. Adds Bismuth or such like to the solder alloy that is then used to bridge all the IC pins. Since its low melt point, 58ºC and remains in plastic stage for a long time. Its easy to heat all pins/pads to remove ICs.

I can use my hot air station and Kapton tape to heat and protect the rest of the PCB...easy peazy IC removal! Then clean up pads as normal!

I looked at many other options like the Hakko 950 SMD tweezers (or Ayuoe T001 knock off's) but the Hakko tips were exhorbitant and the Ayuoe tips range was small. Also the damn thing was ~25Watts per side. Not enough heat for larger SMD ICs!

The other way was using a Hakko modified barrel for my 907 irons and SMD tips to for to these. But each SMD tip is ~$40 AUS ea. I'd need 5 or 6 tips...and more!

In the end..chip quik won out! Much cheaper and less heat thus less propensity to damage the valuable PCBs. I'll need some extra hot air nozzles but fleabay has plenty cheaply!

Lest hope that..." I chose wisely"!

[pat30]

Hi,

I'm Patrice and I look like you Cam86, I have opened the post on the French forum, since the beginning of your subject I read you without speaking.

By the way, hello Gilles :) .

My PCB are achieving by PCBWAY they were costing € 20 for 5 pieces with delivery, I'm surprised for the price you speak above.

For soldering not to worry it is not hard, I started by voltage regulators to take action and be sure to have not a short circuit, after the rest, I has a small soldering station cheap.
I find it unfortunate not to weld ourselves it removes the fun of building and it's easy.

Sorry for my english

[wasyoungonce]

Hi Patrice

I have seen you post on the Ukrainian forum thread (https://translate.google.com.au/translate?sl=auto&tl=en&js=y&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.astroclub.kiev.u a%2Fforum%2Findex.php%3Ftopic%3D289 29.2120&edit-text=)quite a few times as pat30. Good to see you here and welcome to Ice In Space. This thread and camera is generating a lot of interest and its good to see people all over the world with the same excitement and interests.

Your English is better than my French!:thumbsup: Even my English is bad!:lol:

So welcome and if you have questions ask and we will help.

Brendan