My 1100D modding thread *IR Cut mod finished* ETA CARINAE images Stock VS Modded

[2stroke]

Heya guys I will be using this thread to post pics and progress as I turn one of my 1100D into a AP cam. I have being reading through Gary Honis's work and Gina's of the stargazers lounge, along with a large number of others. Thanks to Gina's experiments and findings it helped me alot with my modding plans, it looks like a cold finger won't be needed thank god :)

I grabbed this camera for $398 at harvey normans with a non IS 18mm-55 lens, so if it breaks i won't be left in tears. My first mod will be to remove the canon I/R and place back the AA/dust filter, I don't need AF so I won't be putting a replacement filter in. My next mod will be a cooler box with a twist. I plan to use ducting into the the camera along with air filters to kill of any dust, i will also draw heat of the toshiba IC with a heat sink. I Plan to duct air between the sensor and the pcb in place of using a cold finger, this is because Gina's findings found you don't need sub zero temps and at 20 cel noise is pretty sweet.

This won't happen overnight, I have yet to order some tools of ebay and make a dust free chamber for filter removal. I might Just get slack and use the bathroom and some steam, along with some talc free latex gloves yet. Any input is welcome and here is a pic of the cam before things get under way :)

[rcheshire]

Looking forward to seeing the results Jay. The cooling mod sounds interesting. I will look that up.

[2stroke]

Cheers Rowland, have got a few ideas with the cooling if the ducting isn't going to be workable. The main backup plan is using a non-conductive material which is great thermal transfere such as this http://www.pccasegear.com/index.php?main_page=product_info&cPath=207_335&products_id=19703 to pack the rear of the sensor and turn the shim into heat transfere surface :) I've used this stuff quite alot on pc vga cards for power mosfets, though i wouldn't mind hunting around for the very best.

This camera will never be used again for normal use, which lets me sort of goto town modding it lol. I'am even planing to lose the lcd&mic and any un-needed parts, I may even fully re-house to make life easier. I'll see how things go, just getting all the prep work done atm with screw sheet and getting screws drivers ordered ect :)

[rcheshire]

Jay. I did the filter mod in the bathroom, then stored the sensor assembly in a glad bag with desiccant for a few days to remove any moisture before sealing the entire assembly with Silicon. I've had ice on the face of the low pass filter - on clearing it, the inside surfaces are free of moisture.

The limitation with the cooling mod is dew point.

[2stroke]

Cheers for the heads up man, lol would love to go to the extream and vacum seal camera unit or something like argon or co2. Maybe you could isolate the lp filter by haveing a clear filter infront of the sesnor like those clip in ones though sealed from outside atmosphere. Someone here posted a dew heater eliment just infont of there dsi II which solved the prob for them. Will have to expirement when i cross that bridge.

[2stroke]

Little update have started by removing the first few screws only to find you really do need a #000 and #00. I will have to get back to it friday after i have a chance to get to altronics :) So if your doing yours don't be tempted to skip and try cheap jewlers drivers.

[2stroke]

Got the #000 and #00 from jaycar today, along with a full set of tweezer and hobby knife. I also got a few pairs of latex non tac gloves and zip baggys. Starting to look like a drug lab crossed with a bomb making den hahah. Have Just popped in half and have gotten the worest job over with not tearing the lcd cable :) Step 12 of the guide atm :) http://ghonis2.ho8.com/rebelmod1100Dstep12.html

[2stroke]

Well the sensor is out w00t, zero damage to cables and all is good. 1 tip you don't need to fully removing housing to get to the top sensor screw, just gently pry up the top plastic casing with your finger to gain access to the screw. Heres tonight's pics guys i will hopefully get to the IR cut removal tomorrow as well as the LP filter replacement :) It's really quite simple to do so don't let the manual frown you off it, just double check before doing each step.

[2stroke]

LP filter is out without a scratch or crack nor chip w00t. Just have to remove the IR cut now which will be put in the bin, may even keep to make a cool pair of shades if i do a second haha. 2 pics showing naked cmos sensor and LP #1 removed but IR cut still in the housing to be removed :) Just worried about the dust bunnies now :/ but a clean will sort that out.

[2stroke]

IR Cut removed and LP replaced and glued. Bah what a mess and I have yet to put the whole thing back together. I did run into trouble with the LP filter and a residue from the cling wrap "Don't use cling wrap, Use baggy's". I got this off with some non treated toilet tissue that was moistened by 30 secs in a microwave , how ever this then left dust. I busted these dust bunnies with a fine sponge makeup applicator, these things are better then those pricey lens cleaner swabs :) I can only hope i used enough glue as i didn't want to go overboard and risk LP damage :O Anyhow 20hrs should tell and hopefully my next post will be it re-assembled. I will then be run some test shots to post and after that will be phase II Cooling mahaha

BTW My IR Cut got broken no ozzy osbourne glass's for me :( but i just pushed it out so it has no chance anyhow hahah

[rcheshire]

Isopropyl alcohol works well with the applicators. Don't forget the dessicant in the bag with the assembled sensor.

[2stroke]

cheers Rowland, gotta chase down some baggys from a shoe store lol. Might just add a little more to the top of the LP wells just to be on the safe side :)

hmm better check it didn't off gas :/

[2stroke]

WooT WooT , shes all go guys. On first go it wasn't though, i already thought cable H was being a little ***** but after quickly re-inserting with force it was all systems go. On first fire up it asked for date time ect once set it was alive again :) Now the Fiber optic cable is a bit fiddle i made a tool to help guide this into home base, don't use tweezers because you will break this. Once my wife gets back from the MPAS public viewing ill pump some test shots out, i have run a few on the modded cam but will setup so you can see between the 2. I wont get a chance tonight to do m42 :( its pouring rain and windy at cranbourne. Any how some more pics which may help out other users doing this to a 1100D.

[rcheshire]

Good one Jay. Soon to see some images I hope.

[2stroke]

Now for the pics you have being waiting for, well its no sky but show results :) Quality is piss poor due to IIS having a 200kb file limit, the raws show the god like power :) Cheers Rowland, can't wait to start cooling mate. I'll get some darks tomorrow for reference and some flats to find the dust bunnies. Also planing some m42 shoots to show the diff of modded vs stock , thats if its good weather:). The first shot is downlights warm halogen bulbs and the hall cool white cfl, don't have color temps sorry :/ Lol looks like it will have to be eta carinae since m42 is to low after dusk :(

[rcheshire]

Looks good Jay. Have you taken any flats to see how clean the modification is?

[2stroke]

Not yet Rowland, might get motivated today and gets some out. Have being more in wait for clear sky's lol, every night as being rain and cloud since i did the mod. Will be starting cooling once i get some time to plan it out, pretty settled on a full re-housing though the on/off will need relocation. Getting a enclosure to suit is going to be the hard part, something air tight and light. Really looking at a rip off of the Koren's, though I might go for multi TEC's to control temps.

[skysurfer]

Another advantage is that the green tinge in photos taken with CFL (even the warm white ones) is compensated....

[2stroke]

T-shirt flats here ISO 800 and 1600 / AV MODE, lol sorry about the others someone should have said WTF are you doing with 15secs exposure. Anyhow not bad, there's maybe 3 or 4 small dust bunnies but thats what flats are for. Really happy in away as i've seen way worse around the web. Also this could be the shirt i used or dust on the ED100. Will have to get some light box's built for the scopes stat :) , hmm guess I could just use a 18-55mm and homebrew lightbox it? Not a expert here guys so input welcome :)
http://www.users.on.net/~paulafl/IMG_0021.CR2 (http://www.users.on.net/%7Epaulafl/IMG_0020.CR2) ISO 800
http://www.users.on.net/~paulafl/IMG_0027.CR2 (http://www.users.on.net/%7Epaulafl/IMG_0020.CR2) ISO 1600

[2stroke]

Just did a reference shot with the 18-55mm on a plaster wall done in flat white paint. This makes very happy as this is near perfect for my first attempt and working with what i had anti dust wise IMO.

[rcheshire]

If it's helpful, a laptop/TFT screen, with a white background, illuminated so that 1/10 sec @ iso800 puts the camera histogram around 75% is OK for flats - if you don't have a light box. The other benefit is that there is no appreciable dark noise to warrant taking dark flats. Manual focus on infinity of course.

Try equalizing the frames in PS or GIMP, which accentuates detail and may provide a better view. I'm not sure, but you may have focused on the wall. Else you need to remove the gradient. I suspect it's OK, because you have been careful with the replacement.

Not sure if that helps?

[2stroke]

Cheers for the info Rowland , yer focused on the wall for that shot above, lol can see the texture, the raws are with a shirt though. Will try the notebook screen method thanks man, trying to get the screen clean is harder then the dam camera hahah. Will wack up a light box this week though as i wanna give it a bash before the cooling mod starts :) If you have any links to enclosures send them my way, jaycar and altronics don't really have anything light weight and strong. Really casing a ally case that has a seal though is not overboard on strength. It needs to also fit the camera chassis with a inch of space at the rear extra.

[2stroke]

Heres what everyone's wanted to see, what really makes modding a dslr worth while. These are 10 secs at eta carinae both 1600 and 3200 ISO non modded vs stock 1100D. These are raws converted to psd then re-sized to 1366x911 and saved in level 8 of jpeg compression. No flats No darks :welcome:

My wife has done a quick 5 x10 with dark here http://www.iceinspace.com.au/forum/showthread.php?t=89361

[2stroke]

Small update:
Still playing around with the cooling ideas guys, so far the winner is TEC cold finger with water cooling. The main choice is do i drop the LCD and mount that way, do i go the normal route of side housing or do i go for a complete rehouse. I'am in the middle of shopping around for the TEC and sourcing some cheap water cooling gear from ocau atm. I have also being looking at potting which is filling the internals with a highly thermal conductive silicon which shields from condensation like this http://www.insulcast.com/technical/Silicone_condcure/sil_condcure.html#Anchor-LOW-363 . Though there's problems with moving parts and them getting filled as well lol, though i don'[t want to use thermal grease or other grease's which may get into the sensor.

[rcheshire]

Jay. Sounds like fun. The limitation will be condensation on the face of the low pass filter at and below dew point temperature. I have designed a circuit to warm the filter glass, but it is difficult to fit because of space restraints.

I used BluTak pressed into the components of the sensor electronics PCB to exclude condensation, and, of course the sensor to cold finger assembly is sealed with silicon. Providing the cooling is operated just above dew point, condensation won't be a problem.

If on the other hand, you want to go below dewpoint, sealing components is important. Having said that, I have had the sensor temp to -13C iced up for extended periods with no harm to the camera. At other times below dew point and high humidity with water flowing off the cold finger with no adverse effects. A cold box might be different because the cold is not localised.

[2stroke]

Cheers Rowland, yer was thinking the same with the LP filter. I thought of maybe using a heater close to it like a element inside the camera but going in on the lens interface side. One extreme option would be to make a demister element on the LP filter it self lol but i think thats a bit out of my skill. I might google it up and could try an experiment with a demister repair kit though the shutter may get in the way :/

http://www.youtube.com/watch?v=jQ7nGrXehJs&feature=autoplay&list=UUwV3e2fXrm4wRJMgwyWOVhg&playnext=2

lol love watching that

[rcheshire]

I have a spare sensor assembly and given the time will investigate further. The sensor frame chassis fit is very tight, with the shutter right in the way. Alternatively, directing filtered warm air from the lens side of the shutter on to the filter glass, might do the job.

[2stroke]

Cheers Rowland here another idea i could try, I was thinking maybe use a heating element on the ground strap. If i heated the ground strap would it be enough to keep the condensation off the LP? Heres 2 pics of the 1100D, it doesn't have the sonic lens cleaner, which lol it not very useful anyhow, though you can see what i mean :)

[rcheshire]

The ground strap looks isolated from the glass, but then again you only want to keep the environment warm.

The red line is what I had in mind. Heating direct to the glass. But that might have its own problems.

I'm looking at a spare 1000D sensor assembly and thinking about how best to solve this:question:

[2stroke]

Prevent condensation: All contact surfaces between heatsink and body housing, body mount and clear filter, all the screws are applied silicon O-ring for perfect seal. In the inner space we installed morecular sieve desiccant to minimize the absolute water vapor of airtight chamber.
The clear filter is wormed by the heat from heatsink and maintained the temperature over dew point.
Enhanced noise immunity: The housing material of EOS 1100D, 600D are plastic and this is easy to be affected by electric noise comming from outside.The sealed metal housing of CDS-1100D,CDS-600D protecting the inner electric circuit and added the ground connection from main board, power board, storage board to base housing made more stronger anti-noise system. This aluminum alloy housing of CDS-1100D, CDS-600D is more durable and resistant to shock than Astro50D, Astro60D.

lol they give all there trade secrets away at http://www.centralds.net I think ill have to find a clear filter now and do a rehouse :)

[2stroke]

Just some copper sheet i sourced for the cold finger, slowly getting there. Its more now getting the time to pull down the camera now as its the only astro modded one i have :( It's .025" or .635mm thick

[rcheshire]

That's a narrow gap. The 1000D is just over 1.2 mm. Looking forward to the result.

[2stroke]

The .635mm is the metal gauge man, i think the gaps about 1mm on the 1100D. Its going to be a tight squeeze for sure though, thinking about epoxy to bed against the pcb and copper with some good quality pc grade tim between the cmos and copper cold finger. May even just use epoxy as a protective coating on the pcb exposed side, will have to have a play once i get the camera stripped down again :( . I bet we have clear skys as soon as i start haha always the way :/

[2stroke]

Since starting the 20D i have also decided to return to TEC cool the 1100D and rehouse. I just ordered up one of these sweet baby's for rehouse http://www.ebay.co.uk/itm/HAMMOND-1590WE-IP65-ALUMINIUM-STANDARD-WATERTIGHTLID-ENCLOSURES-/200851116438?pt=UK_BOI_Electrical_C omponents_Supplies_ET&hash=item2ec3a8d196 Hopefully this rocks up soon and i can get this done by xmas :) I will only strip flash and LCD though along with plastic housing and rehouse inside. Once again its going to be a mc clear T-thread job. The case will act as a heatsink which will stop frosting and dewing of the mc clear and allow sub zero temps off the sensor, i may even fill it with argon since its water and hopefully air tight.

[2stroke]

Got the next piece of the puzzle the TEC



Measures: 40 x 40 mm with lead wires.
Temperature Differential (DT) (Th=27 degrees C) : >=68 degrees C
Optimum Input Voltage (Vmax) (Th=27 degrees C): 15V
Optimum Input Current (Imax) (Th=27 degrees C): 6A
Maximum Cooling Power (Qmax): 51.4W

Just waiting on the encloser to rock-up and then i can get this baby on the go. I need to gather up some case usb and power fitting for the tec and camera, from my collection of junk lol. I also need to investigate EMF shielding to bring the noise down even more and its placement, obviously the usb, tec and power will be shielded but what else can i shield. I may cascade the TEC later on for more extream sub zero temps though i don't think it will benefit a cmos sensor. I also need to make up the plates for the tec to optimize transfer of temperature difference.

[rcheshire]

Hi Jay. Looks like it's all coming together.

If this helps. I found that bonding the camera chassis to GND reduced noise significantly, while the TEC was not operating. There was noise in the line plugged into the mains, through the 12V PSU. Noise was also a problem when the MOSFET was switching on and off maintaining temperature during normal operation. Not what I wanted.

The crude solution - I placed a 22uf capacitor across the Gate and Source of the MOSFET, effectively removing voltage spikes caused by the switching. The only effect on the system was to slow down the rate of cooling marginally. Still as quick as an SBIG from what I have seen.

Every now and then a line would appear across an image, not optimal, but stacking software rejected these pixels as outliers - so no problem.

Hope that's useful - depends on your system I guess.

[2stroke]

Cheers for that info Rowland, its being years since i was into doing a lot of electronics. I plan to use a computer 400watt PSU and just run it at 12volt without any PWM control, there shouldn't be any spikes i'am guessing?

After a quick test last night though i don't think the encloser is able to move the heat off the TEC and may need to go for the 32watt model TEC. Did you find out what sort of heat load the cmos sensor seem to put out in your experiences and what tec was best off?

[rcheshire]

Hi Jay. I found the Tellurex FAQ's valuable. Long read but great info.

Keep in mind that I am self taught and tend to over-engineer - just in case.

This is my take on things TEC.

TEC devices are not very efficient. The heat sink must remove energy from the object being cooled and dissipate the energy required to drive the TEC. For example, if you need to move 20 watts to cool a component you need a heat sink/fan capable of handling ~40 watts - in rough figures - otherwise the TEC becomes a heater and reduces the effectiveness of cooling.

The other issue is the mass that the TEC is expected to cool overall. Is the structure in which the TEC is operating sucking the life out of it. Isolating the TEC from sources that don't need cooling, such as the box, adds to overall efficiency. I don't think the CMOS is the issue.The TEC is cooling a whole lot of additional mass as well.

Looking at cooling boxes, I would keep them as tight as possible, insulate the inside. Not have the TEC in contact at all and make a cooling path from the cold side of the TEC directly to the camera chassis, to localize cooling as much as possible. But that's my take on things.

If you can get 0C, dark noise is reduced substantially.

[2stroke]

Just grabbed a 33watt model and will do some experiments with it, might have to go the water cooling route :( The 33watt tec is running great just sitting on the p4 cooler with some tim and no pressure. Its icing up nicely and running around 30 at the heatsink i guess. I ran this for 10mins straight and it seems i should be able to pull this off with the p4 cooler and encloser. This will be great as the encloser will have the mc clear t-thread thermally connected to it and should sit above ambient easy which will mean frost free :)

Going to use a cold finger Rowland, kindia doing a quick half arsed rip of the centrals 1100d tec cooled camera http://www.centralds.net/cam/?p=1687 . Lol would just try and go phase change cooling if it was a box only style, if i didn't it would have to be some massive TEC's and series and parallel lol. I think the cooler boxes aren't bad but unless you have got airflow to the cmos sensor then there wasted alot. Really hoping to get sub z's with just this TEC and some cleaver installation of the cold finger to prevent cooling loss while on its way from the tec to the cmos sensor :)

[rcheshire]

Jay. As long as you can keep it cool and prevent condensation, it should be fine.

[2stroke]

Well just grabbed a 240x120mm rad, a nice waterblock and pump/res combo. If this doesn't keep the TEC cooler nothing will lol. Should rock up next week but still waiting on the enclouser which i had to get from the uk :(

[2stroke]

The encloser got here today and all i can say is wow. It has a very nice rubber gasket for a seal and also rubbers for the screws. Unlike most alloy cases which way a ton this thing is super light and really feels like aluminum. The size is excellent and can fit my 1100D in it by just removing the eye veiwfinder rubber on the top of the body, this leaves me lots of room to play with :) Anyhow will get started this weekend fitting the 1100D into the encloser, need to get a astronomic mc clear once this is done to finish it off :)

[2stroke]

240x120 xspc rx240 rad rocked up and the pump and block for the water cooling of the TEC today w00t. The blocks all copper and weighs a bit so i might grab a cheap alloy block instead, got all this second hand for $70 to which was sweet.

[2stroke]

I think i killed my 1100D guys :( I removed flash and tryed to strip down the top housing, then seen it was to complex. I stuck the sensor back in after loosing to may shims lol and the mainboard. I then stuck a battery in it only to find there was no power :( I will get some pics up on the weekend, all i can put it down to now is a bit of metal that came loose from the top housing and its involvement in the no power issue. I will now strip it down with out fear as to find the course. At worst case i guess i have some great spare parts and will just have to grab another one. If i can't get it going then i can't pull down another one for TEC cooling and will just have to use a cooler box.

[2stroke]

OM*G!!! So i went to get a 3rd camera to mod it, after the mod dead again. I was thinking it was the top fiber ribbon cable H that had some damage (after 4 tear downs and re-assembles) and told Forgey she can pull down the camera for parts and ill use them to fix the 1100D i broke. After she stripped it down i used the top section of the camera replace the one on the modded, and no go again. So then i also re-soldered the flash wires on and no go again BAH!!!! I was about to give up as everything was connected fine and grounded out ect (4hrs later from modding a new 1100D) then i though might as well just stick the battery door on as well and try for laughs. Well powered on and bam going, cut flash wires and bam again going still. What cannon have done is added a switch to the battery door to stop power on unless shut, thanks cannon for $300 of parts :) But now the extreme mod lives on and more to come :)

[steve000]

Very cool, also love the Bill Hicks sig. not a fan of tool are ya?

[2stroke]

Haha yer man, mad tool fan was them that got me onto bill hicks.

[2stroke]

So after some testing for cooling i've ditch the water as its really not needed for the 33w TEC i'am using. I've just done some testing with a few heatsink fan units to find the best to used. After testing some heavy CPU coolers i came across a dead (well has BSOD's with media playback) ATI/AMD HIS 5750. The unit weighs next to nothing maybe less then 100grams, and boy does it do a nice job and will suit my project perfectly. I havent tested with the fan at 5volt yet but i know its going to be fine already. Here is some images, remember warm air is being forced onto the tec and still look at that ice wow :)

[2stroke]

Hmm whats in that bag your thinking, well a full 1100D in parts hahaha. As you can see i have to keep the top section due to the rotary switch for manaul selection and on/off, why because i just don't have the gear to work with such fine pcb tracks to bypass it. Though as you can see on the spare top i have stripped the shot button and flash off to make the camera more low profile. What i will do is get a clip-in clear and seal the sensor chamber and see how that goes first, its not ideal because air can get by the EF-S mount clips, maybe i could seal them with a fine film tape externally though. I Really want to keep that EF-S mount system at this stage and well i have a spare everything and nothing to loose lol. I next plan to cut the whole for the EF-S mount into the case and then make mounting studs.:thumbsup:

[2stroke]

Camera EF-S whole is cut, also made up the heatsink unit for the TEC. I have tapped the heatsink with 6mm thread, then drilled through the housing which will also act as a heatsink and anti dew forming unit. Next goes the tec then the cold plate. The cold plate will have cork bushes and top washers inside the 8mm holes, this keeps the heat off the cold plate from the 6mm hex heads. Next goes a brass washer then spring then brass washers, followed lastly by the 6mm thread hex head bolts. This puts force down on the tec to gain maximum heat transfere without the risk of cracking or breaking the TEC. Heres some pics which might put the story together better then words.
This is an on going project and don't expect completion straight away, i just do this when i have spare time, which isn't often enough lol.

[Chris85]

Hi Jay in your original post you mentioned you had to order special tools off ebay to do the job. What were they? I have a astrodon filter ready to go in my 600D, but I don't wanna tear into it until I have the right tools
Cheers
Chris

[2stroke]

The main tools are the 00 and 000 phillips, get them from jaycar for $3.50 http://www.jaycar.com.au/productView.asp?ID=TD2065 . I used the biggest 1 of the 2 on all the screws mostly, also get a cheap jewelers screw driver kit from your $2 that has a pointed piece as its great for prying open the locking cams, I used it as well for the cables though tooth picks are a safer option. Grab some markable zip baggies and place you parts in them to aid dust, damage and mark them on your first run with the step number so you know what goes where in case.

GL with it, though its pretty hard to screw up and post a thread so other can use it for help in the future to :)

[Chris85]

Excellent, thanks for that! I think i'll swing by Jaycar this week now.
Also I've read on some pages that they use rubber tipped tweasers? Think it's necessary?

[2stroke]

Not worth the hassle, i found a good small pair of pointy nose pliers best for the power connectors, though they maybe useful for the fiber optic cable from the flash where it goes into the motherboards optic connector. A pc repair kit with a grappler for small screws can help out alot with non magnetic tip drivers. It's all quite easy to do its just the initial fear the puts so many people off, just take your time and don't go tearing things off before making sure ribbon cables are free, you don't want to break them.:thumbsup:

[rcheshire]

Using the case as part of the heat sink and as a condensation collector is a clever idea, Jay. Sounds like you have given this mod a lot of thought. It seems odd to attract moisture, but it actually keeps it away from sensitive electronics.

I've been working away on the electronics side and have what I think is a 'quiet' PWM set up. There is no such thing as quiet PWM , but working through the problem with the Arduino guys I think I'm close to reducing interference in the TEC leads and smoothing out inductance and capacitance issues with the MOSFET switching off. Just for interest I've posted a screenshot of the project so far.

Note: The TEC is the 17.9R (R3) resistor.

[2stroke]

Can't wait to get it finished, just haven't had the time though :( The mods more a cds rip off now haha, so i can't take any glory for it besides figuring out how to strip un-needed camera parts off ect. I did notice that cds are now using a demister on there filters in the style of that used on a common car. By running a resistive film around the filters edges it can keep dew at bay with out the need for a sealed enclosure atleast from the sensor out. This also beats the hell out of using nic wire and risistors due to the wire melting plastic housing, the size of common used resistors and there ineffectivness of not making optimal contact and heat transfer with the filter. I would really love to see what i can pull off with some conductive paint and insulation like epoxy paint or liquid elec tape over it. ;) Maybe ill get around to it one day when i have the time, lol sucks when there just do much to do and not enough time.

That ooks pretty sweet, so osc's are your mosfets, what about using some more for smoothing out the ripple even more? That looks like its doing a great job :) This was a great google http://www.proaxis.com/~wagnerj/PWMfil/PWM%20Filters.pdf may help out but no doubt your problem read books the size of bibles on it now haha.

[rcheshire]

Thanks Jay. That's a great resource. Will read from top to bottom. The resistive film sounds just the ticket. Must look it up.

The problem with switching mosfets is avoiding linear inputs which cause overheating, so your stuck with the pulses. Although within limits that's not been my experience. Linear drives are less efficient but more desirable. Working on one right now.

[wasyoungonce]

TEC temperature control is a vexed issue.

Really the best method is proportional voltage control (http://ghonis2.ho8.com/rebelmod450d16g.html) ...that is apply a linear voltage proportional to the temperature (cooling) your require and vary this to keep at this temperature. A PID system.

However this is not easy to do and it's much easier to use PWM. TETech (http://www.tetech.com/) has some primers and info on TEC cooling and calculators.

Another simple method is to use a simple fleabay thermostat controller (http://www.ebay.com.au/sch/i.html?_odkw=WH7016R&_osacat=0&_from=R40&_trksid=p2045573.m570.l1313&_nkw=WH7016C&_sacat=0) for ~$20. Yes...crude but it managed to keep my DSLR cooler box and DSLR sensor temp to within +/- 1 degree C. Even I was surprised at this.

[2stroke]

Cheers for the info brendan, just going all out to start with unless dew and condensation become an issue, though using a sealed unit it shouldn't. Have read a stack of tec papers :) and have found the overclocking community has more then a wealth of information :) Thats a great link and thanks, will have to chase through my books marks even have one to work out specs from an unknown TEC :)

[rcheshire]

Thinking outside the box, proportional voltage to a TEC or heater, could be controlled by attaching a stepper motor to a potentiometer shaft, controlling with a microprocessor. A little calibration may be required, but once the correct position is set, fine control could be achieved by dithering either side of set point. Changes in energy demands, due to ambient temperature could be programmed as well. I dont think this would be too difficult to do with an Arduino, for instance.

[wasyoungonce]

Problem would be power dropped over potentiometer. I draw Peltier ~ 2.5amp I (nominal), your lower Peltier voltage from 12V to lets say to 5V. of course you lower Peltier voltage it's I draw will change as well.

Thus that this stage we have 7V dropped across pot. Power across pot P = EI, P = 7 x 2.5 = 17.5 watts energy dissipated thru pot...a lot.

This is why proportional voltage control is difficult.....if you used a transistor to lower the voltage you would have similar power drop across it.

This is probably why devices control is done with PWM...changing mark space ratio, driving a FET and Peltier.

[rcheshire]

It's not such an easy thing to do. Maybe for a dew heater drawing a few watts. Tellurex say they have developed a linear style using MOSFETS with reduced efficiency. No details that I can find.

[2stroke]

So the work goes on with little time i have lol, i striped some un-needed parts of the top section to make it fit inside the enclouser and also free it of non needed fuctions to save weight and hopefully EMF. I still need the top section as it has the dam switch and also mode dial bah. Heres a few pic's hopefully next post i can start the dam coldfinger itself lol.

[rcheshire]

You're braver than I Jay. I've been playing around with a 5mm Neoprene jacket on my dslr. Cooling is to the plate on the bottom of the chassis - removed the plastic cover to get to it. Best reduction measured at the bottom plate is 20 below ambient, but I dont think this is conducted to the sensor. Have a recirc fan blowing air within the camera body flowing over silica gel. Dark noise improvement is markedly better. If I insulate the cold finger at the bottom plate, I get another 5C below. I've toyed with the idea of using reflective foil under the neoprene. Sealing around the lens seems OK and so far no condensation, even at -5C, with a dewpoint of +5C. Except for one or two controls the Neoprene is flexible enough to push buttons without cutting holes. For the other controls and screen I've cut flaps that fold back in place. See how it goes...

[2stroke]

Thats a whole new approach you have gone with there, that's really thinking outside the box haha. Good to hear you getting great temps with it, the only problem would be getting the temp down to start with which could be solved in a fridge :)

Sort of going right out with this mod, have a whole spare camera in parts so i can afford to experiment a bit. Will do a full pc run with the final strip down, no back lcd or buttons / plastic so only the top is left.

[rcheshire]

Yeah. It's takes about 10 minutes to get down to max diff, but I think a while longer to chill the chassis and anything connected. The internal recirc fan should even out the air temp inside the camera, I hope. Keep up the good work.

[2stroke]

Quick shots of the 1100D stripped down with no back or front working in BYEOS. The flash has also being stripped out along with some of the functions on the top, this is also the final product in terms of strip down as there inst much else to take off it lol. Starting the coldfinger after this post XD

[2stroke]

Well its on with the cold finger, i have just mod the cmos housing so no cutting is required for the studs, also i will use longer screws to replace toughs that hold the cmos in place. Now why longer screws you ask? because they will also force the cold finger down firmly against the cmos back maximizing transfere of cooling and also hold the cmos. No more cold finger sliding all over the place either and also no need to worry about contact with the pcb and those prom pins sticking down w00t. Now for anyone else out there these nubs are only used to guide the cmos bracket to the housing and can be removed with a twist action with a pair of pointy nose pliers :) Kitty not recommend in work space :rofl: