Performance of FLI ProLine cooling

[strongmanmike]

I was just testing in the lounge to see if a new USB-serial adapter for my mount control was working and that some driver updating I had done recently wasn't interfering with my CCD systems and I decided to test the colling capabilities of my FLI ProLine 11002 CCD camera in the hot weather.

I have always used -35C as my chip temp and thought.. I wonder if I can reach that today? A digital thermometer on our wall aircon controller read 27degC (we run it modestly to conserve energy) so I set the chip temp in Astroart at -35 and sat back and watched to see if it could do it. Over the course of a couple'a min I watched as the cooling pured and slamed that temp down pretty fast to about -30. At this point it slowed and operating at 99-100% cooling capacity it struggled on but after about 5-6min eventually the chip temp stopped dropping and stabilised on an amazing -34.7C :eyepop:

This means the delta cooling was 61.7C from ambient or in other words I can realistically (running at just under 100% cooling capacity) image at -35C on any night where the temp isn't more than about 26C.

In this heat, that's pretty handy :thumbsup:

Just thought some might find this interesting :)

Mike

[strongmanmike]

I've posted this at the equipment forum but thought some on here may be interested too?

I was just testing in the lounge to see if a new USB-serial adapter for my mount control was working and that some driver updating I had done recently wasn't interfering with my CCD systems and I decided to test the cooling capabilities of my FLI ProLine 11002 CCD camera in the hot weather.

I have always used -35C as my chip temp and thought.. I wonder if I can reach that today? A digital thermometer on our wall aircon controller read 27degC (we run it modestly to conserve energy) so I set the chip temp in Astroart at -35 and sat back and watched to see if it could do it. Over the course of a couple'a min I watched as the cooling pured and slamed that temp down pretty fast to about -30. At this point it slowed and operating at 99-100% cooling capacity it struggled on but after about 5-6min eventually the chip temp stopped dropping and stabilised on an amazing -34.7C :eyepop:

This means the delta cooling was 61.7C from ambient or in other words I can realistically (running at just under 100% cooling capacity) image at -35C on any night where the temp isn't more than about 26C.

In this heat, that's pretty handy :thumbsup:

Just thought some might find this interesting :)

Mike

[spearo]

drool
Frank (who owns a DSLR that testing reveals operates nice and hot...)

[gregbradley]

That is amazing Mike.

How long did it take to get there?

The Microline is like that too. I can get -35C almost anytime but that's a much smaller chip. It takes about 2-3 minutes to get down that low.

Greg.

[strongmanmike]

It got to about -30C (ie delta 57C) in about 2-3min but then took another few minutes to slowly drop to the final resting point of -34.7 so all up perhaps 10min? But that includes waiting untill I was happy that it wasn't dropping any further so most likely it took about 6-7min or so..?

I also rested the camera body with its thinnest edge against the table to allow the body most contact with air for heat disipation but at the scope it would have been hanging in the air freely and maybe even in a slight breeze so perhaps even a few more decimal points of cooling may be possible..?

A delta cooling of some 62C is pretty good though and very handy.

Mike

[AlexN]

Daaamn.. :) thats hellishly impressive.. :)

Tell me though, are you sure that the camera is cooling to -35°C, or is it saying -35°C from ambient? meaning ~-8°c

Forgive me if thats a dumb question... I have no idea if it clearly states that its running at the -35 temp etc.. but thought it best to ask anyway... If you dont ask, you dont learn. :)

Impressive in any case thats for sure, if it is giving a -61° delta, thats incredible.. :)

[Gama]

Yep, thats quite good cooling.
Just be carefull Mike you dont crack the sensor. You should be looking at a 10 to 15 minute change to max cooling to prevent thermal stress.

With that sort of temp, you could dispose of the darks....

Once the camera starts to warm up, you may find it will drop a little, as the heat dissapation efficiency decreases as the body and heat sink absorb the heat.
Bit like running an engine for 5 minutes when cold and running the engine 5 minutes when hot. More heat and harder to dissipate when hotter.
But you now why the costs are so high for these cameras. Thats why you have one in your hands..

Theo

[gregbradley]

He means the final temperature is -35C.

Yes I think the FLI cameras are at the forefront with their cooling. And good cooling solves a lot of problems with CCD cameras and makes life a lot easier.

Greg.

[avandonk]

The delta T depends on the heat load and the heat pumping capabilty of the Peltier stack. You are cooling a single chip directly with no real heat load apart from the much higher internal heat generated by CCD's as compared to CMOS. The numbers are as impressive as is the condensation problems.

The only reason these CCD's work at all is their brutal cooling. I would like to see a Canon CMOS at the same conditions and without Bayer filters.

There is more but I can't be bothered.

bert

[Peter Ward]

Yep...very cool Mike.

But shouldn't you be ringing a cow bell with this (Toyota-Valvoline-Wild Turkey-FLI ) sponsored post ?) ;)

[Omaroo]

That's pretty cool Mike. (guffaw.... guffaw)

I was out last night and only made it down to -10C at 85% capacity with ambient temp sitting at 22. That's a delta T of 32 degrees from the SBIG which is what is advertised. I can't wait until winter in other words....

Mmm... STL/X or ProLine.... hmmm....

[strongmanmike]

No, what I wrote is exactly how it happend and since many on here have been frustrated with the hot weather lately with even their cooled cameras not being able to get much below zero on some nights, some may be in a position to upgrade and this is one hell of a good reason to if you were thinking about it.

Mike

[strongmanmike]

You are right Bert, some chips really benifit from super cooling others not so much. I notice with my camera KAI11002 chip that at lower temps things do indeed improve and beinga able to run at -35C all year also means much less bother with dark libraries etc...

I think the Sony chips are more tolerant of warmth though..?

I have no condensation problems at all that I can tell. The CCD chamber is purged of air and backfilled with Argon.

Mike

[Terry B]

My ST9E with water cooling and a secondary cooler will give me a 43deg C drop in temp. I can run at -20 almost always as we tend to cool down at night below 20degC.

[Peter Ward]

I beg to differ. (normally I'd say "Bollocks" :) )

The on-chip (vs off-chip with CCD) charge to voltage conversion of CMOS hamstrings its ability to produce extremely clean pixel to pixel signal read uniformity.

To date, all major science cameras (in Hubble, Keck etc.) remain CCD for this and other reasons

[Omaroo]

Water cooling coming up for my ST-8i soon Terry. Can't wait.

I have to say that my lights were pretty clean last night, even at -10.

[Peter Ward]

Gong! Tell 'em the price son :)

[strongmanmike]

Aparently Kodak claim the KAI11002 will not suffer thermal stress with rapid cooling and heating and I can even turn the camera off with the chip at -35C aparently? I don't do that though and usually warm it up to zero before tiurning off.

While the frames are indeed clean for a KAI11002 they still need darks at -35C, I haven't tried taking the chip below -35C really but on a cold night in Newcastle -55 or -60C would be quite acheivable, must try and see if it does anything?

Mike

[AlexN]

Terry, what secondary cooler are you using? I've devised my own little plan (being quite the ingenious fellow I am :P) Rather than just running a bucket of water under the mount with the pump feeding the water jacket of the ST9E, I'll be putting a dual core radiator in line with 2x120mm fans feeding air through the radiator... this will ensure the water is as close to ambient temperature as possible... every little bit counts :)

[strongmanmike]

In case it wasn't clear, these cooling figures were WITHOUT water cooling so no messy buckets, pumps and stiff water tubes (to pop off unexpectedly in the night)...water and electronics in the dark???:scared:

As a result of this impressive cooling the ProLine is a big camera with three cooling fans so when coupled to the filterwheel it is a big lump of a camera so your adapters and couplings better be robust:

http://www.pbase.com/strongmanmike2002/image/79551561/original

and

http://www.pbase.com/strongmanmike2002/image/80650047/original

Mike

[avandonk]

Oh yes you have the added advantage of one crook pixel stuffing up a whole vertical line! When I was at the Synchroton in Japan Professor Sukabi showed me his CCD detector for photons from xrays. For time dependant experiments he was only using half of the CCD to collect and then downloading the data to the unexposed part of the CCD. This data was then read by the much slower bottleneck before the next lot of data was collected.

Bert

[AlexN]

Watercooling hoses need not be stiff.. you're just using the wrong hosing if it is... Tygon make the best hosing around, its expensive at $25 per meter, but its what hospitals use for all their hosing needs.. so thats the stiffness solve... from there, norma hose clamps are $1 each at mitre 10... Theres that solved.. :) Watercooling of anything need not be considered dangerous, its as safe as houses provided you think before you execute the design.

I myself have watercooled many MANY computers in my time, with hoses going between the cpu/graphics card, ram and hard drives.. all up a $6000 computer with 1.2L of water inside it... never once had a single issue... :)

Alex.

[strongmanmike]

Monte Wilson of ASNSW water cooled his SBIG STL11000 by rigging an amazing set of gravity fed tubes from the water tank tower at the imagers house at Wiruna. He ran tubing under the ground through his camera and up to the tank, had to dig a trench and all....So extra cooling is deffinitely desirable given the lengths some will go to get it but not having to use water cooling in the first place and still having better cooling is..?..even better! :P.

Mike

[avandonk]

I have come to the conclusion we are all nuts. Whatever works for you folks!

Just keep those images coming

Bert

[strongmanmike]

Nuts? who's nuts? I resemble that comment :P

What's wrong with loading your car up to the hilt and traveling 4 hours to a dark sky, spending 2 hrs setting up a whole lot of scopes, mounts, cables and cameras and then staying out in the dark pointing at the same spot of sky for three nights in a row with little sleep during the day, then spending 2 hrs packing up and 4 hrs driving home and then unpacking the car....only to have to do it all over again next clear dark night.....:screwy:?

http://www.pbase.com/strongmanmike2002/wiruna_trips

I'm not nuts...am I? :sadeyes:

[Peter Ward]

:lol:

Ok That's a good reason.

I can think of another CCD that has:

Full Frame Image Buffer
Multiple A/D Channels
Image and Guide While Downloading
Internal and External Self-Guiding (with Remote Head)
Simultaneous Dual CCD Guiding
Differential Guiding
Continuous Guiding During Autograb
Adaptive Optics Control with Internal or External Guider
-50 degrees C Delta Cooling with Air Only
Water Cooling Ready
Even-illumination Mechanical Shutter
Gas Purge Valve on CCD Chamber
User Rechargeable Desiccant Plug
Variable Speed Fan Control
Focus Mechanism for Tracking CCD
Opto-isolated Relays for Telescope Control

Doh! *I* Better hit the gong before its too late :)

[strongmanmike]

Wha?..all that and STILL only 50degC cooling? Ah sigh...we still gotta get out the buckets and pumps and tubes and fumble in the dark :rofl::rofl::rofl::rofl::rofl::rofl :


Com'on Pete loosen up man :)

Certainly sounds like a beaut camera :thumbsup:

Mike

[Peter Ward]

Hey I'm cool....

But for the newbies out there, delta T's are rarely constant (ambient temp and air flow can have a big influence) and often *reduce* once the thermal cooling/inertia of the camera head is lost after being heated by the peltier stack. (thermal runaway can happen at that point...not good)

I've personally found bugger all difference in well calibrated images taken at -15 or -25 degrees. The key is the quality of the calibration frames rather than the CCD temperature set point.

[strongmanmike]

You may be right Peter but luckily I have been able to image at a chip temperature of -35C on each and every night I have imaged since receiving my ProLine in mid 2007 right throuh summer, autumn, winter and spring and in fact could have gone colder on "every" ocassion and no water pumps and tubes in sight, so any variations in the dalta cooling of the FLI ProLine haven't been great enough to stand in my way in sunny Newcastle.

Mike

[Bassnut]

Ditto........:whistle:

[Peter Ward]

For those with further interest in this thread, Steve Cannistra (http://www.starrywonders.com/coolinghidden.html) has done some excellent work in this area. You can also download an excel spreadsheet from his site that helps you play around with some real world values

His take?

CCD cooling beyond -15 to -20 C produces minimal gains in the functional dynamic range for most imaging sites, because the sky flux becomes the limiting factor in this situation.

Roughly translated.....a dark site is way more valuable than a super cold chip! (Why do I have a backyard observatory again?? Doh! )

[Gama]

Mike, turning the power off at -35 is fine, as it will take time to warm up. But going from 26.5 deg to -35 in a minute or 2, isnt a good idea. Kodak may be basing there response on a warm up temp range, and not out right "Flat out" hot to cold thermal transfer.
Still, gota love them cameras..

Theo

[strongmanmike]

Bah humbug! You SBIG clowns are just jealous, ya'll say anything, like the neo Christians in the US have whole museums to varify Genisis :lol:

[strongmanmike]

Hmm? wonder how I'd know if something went wrong?..a big black jaggered line on the downloaded frame? I've been setting it at -35C for a year and a half now summer, winter, autumn and spring and it gets there in just a few minutes everytime...?

[AlexN]

Ahh now Mike.. Just because SBIG users get APODs... :P no need to fire up that old chestnut.. :) In all seriousness.. nobody can say that cooling that far is not an extreme achievement, although what does it mean in the real world? (That was not rhetorical.. I have no idea.) I know no camera I own will cool to -35c in a 26c ambient temp.. not even close... so yes. The FLI cooling wins against my arsenal of cooling setups. :)

[strongmanmike]

Ah Alex I think you have my measure, I'm a sarcastic fun loving bustardo fisherman and Peter SBIG always bites, can't help himself....like brown trout in a trout farm (or do they use nets there?). I was just posting about the cooling cause I thought it was ..?.."cool?" but the brown trout just had to jump outa that water again like always :rofl:

Mike
Super chip cooling rules :thumbsup:...dark skies are a myth

[strongmanmike]

DOH :doh:those pesky fussy moderators, moved us again :rofl:

[Peter Ward]

..Aye carrumba. :doh:

Clandestine promotions for FLI aside, this could have been an interesting discussion.

As for my spin? Dark sites do far more for CCD S/N ratios than almost anything else. Really accurate tracking/guiding comes next. Everything else? Well I suppose that's why we have these discussions....... ;)

[AlexN]

I've taken shots with a DSLR from dark skies that easily rival my CCD shots from light polluted city skies... In that situation, cooling is out of the equation, dark skies did all the magic there.

[gregbradley]

Ah but Peter there are more advantages to super cooling than merely noise.

As you know the STL11 KAI11002 chip along with others in the KAI family suffer from vertical lines caused by hot pixels. Now theoretically these should dark subtract out or a bias frame should remove it etc but often they do not. They then become a real pain to remove if you can during processing.

Richard Crisp - the arguer, has demonstrated clearly in a paper that high cooling gets rid of these lines. They don't fade out until you hit -30C and beyond.

Merely going down to -10C will not and they will be a pesk when you process your images.

I know from long experience the STL11 ideally is operated at -35C. It is quite noisy at -10 and even -20C isn't ideal. It starts to shine at -35C.

Yes dark subtraction works but it now becomes critical that it is done excellently with lots of darks sigma reject combined and with accurate matching temps and exposure times.

Cold chips have practically no noise in the first place, dark subtraction almost becomes only if you want to and certainly less critical and exact matching of darks is less important and use of adaptive darks is feasible with little or no difference to exact matching.

In my opinion excellent cooling is important and gets rid of a lot of problems that complex processing requires otherwise having used different cameras with different cooling abilities.

Of course a dark site is more important again but if you had 2 cameras and all other things were equal then the one with the superior cooling should be your choice.

Greg.

[gregbradley]

Apogee engineers take this line and not so much the sensor but the soldered connections etc - they don't want thermal shock.

But really I have seen no evidence in my camera or others that there is a real problem here. I think its a case of engineers being conservative and protecting against an imaginary problem.

Slow cooling and slow warmup is for the birds for those who use equipment that takes more than 2 seconds to work before they get impatient! - --- uhm thats me.!!

I love my Apogee camera but the slow cooling does annoy me and causes some delays at times. The new firmware has sped it up but still you have to plan if you want to take dusk flats. A warm camera gives useless flats.


Greg.

[Peter Ward]

Sorry Greg, (not surprisingly) I don't agree at all. All other things are rarely equal. Steve's analysis (http://www.starrywonders.com/coolinghidden.html) is quite thorough...what aspect of the physics would you care to take issue with or debate?

The few S/N percentage points you gain by extra cooling are easily swamped by having the option to use a device...well, like an AO...which can easily add 20% more flux under a stellar Gaussian curve (translation: brighter & tighter stars)...as does self guiding.... as the pick off chip is looking at the same aplanatic field as the imager. *All* other guiding methods are less terse, and to say otherwise is...as I like to say: "bollocks"

Dismissing these benefits I think is a bit like knocking back Cindy Crawford
because she has a mole.... :lol:

[gregbradley]

You missed my point. I was referring to ugly lines warmer temps on chips can leave behind in the KAI series.

Its no good having optional extras (which are often problematic to use and expensive although the benefits are there for long focal length scopes) if you have an ugly line going through your image due to inadequate cooling. Nice tight stars are wonderful but will be spoilt by the line which is hard/next to impossible to Photoshop out without leaving a trace.

Are you telling me you've never seen these ugly KAI lines in images?

Did you know they nearly disappear with stronger cooling?

Also I never mentioned anything about self guiding. But now you bring it up I think it is overplayed as a marketing point for SBIG.

I get far better autoguiding results using a separate guide scope and an ST402ME than I ever did with self gudiing using the tiny and extremely noisy ST237 chip. It picks up the edge of the FOV of the scope where most optics are weakest, is very noisy and has a very small FOV. Also blue filter often requires longer exposure times which for my setup meant less optimum where 1 second guide exposures have almost uniformly given the best result.
Also with Ha, S11 or O111 self guiding is virtually useless as often unless the object is super bright it requires something like 30 second guide exposures to register a guide star.

Also the way the camera ceases to autoguide when it is downloading an image and in the case of the STL that is something like 26 seconds means tracking errors are building up requiring time for the autoguider to catch up. I see this problem is addressed with the latest STX cameras which do not interrupt the guider when downloading. If you don't program in a delay between exposures you will get eggy stars from the initial errors having built up.

So if you are imaging LRG and no blue or Ha O111 or S11 there is no problem! As well as using a scope with lovely pinpoint stars right to the very outermost edge of the FOV where the pickoff prism collects the guide star. Having said that self guiding is still handy, especially with long exposures, but a guide scope and guide camera are really too easy to use to argue against them.

So self guiding is a bit ho hum. QSI with their offaxis guider built in before the filter would be way superior. If you get flexure you'd be way better off to use an off axis guider than self guiding.

Lets face it, if you weren't selling SBIG cameras and protecting your business we wouldn't be having this discussion which is about protecting SBIG's camera sales. The vested interest factor is too strong.

Greg.

[AlexN]

Awesome analogy...

[Peter Ward]

1) On class 2 (or lower) devices you can certainly see these...but that's why I use a class one chip: it dosen't have hot columns.

2) The hot columns I have seen fade with delta T....but...and here's the rub.... if you calibrate the data accurately...hot/dark pixels and columns disappear.....unless you have defect on the chip.

[gregbradley]

I have seen it one a class 1 chip.

Theoretically the line will go and often does but mysteriously I found it also often did not. Its also not a column defect as you know - there is a paper on it on the SBIG site - its caused by a hot pixel and the readout process of KAI family chips.

They can also appear later on as a CCD deteriorates over time.

Greg.

[Peter Ward]

Not saying you haven't...but if it didn't dark subtract out...it was no longer a class one and good reason for a warranty claim.[/quote]



Without exception ( I've tested hundreds of cameras) I have traced this to less than perfect dark frames....sure, lowering the temperature made the effect less obvious, but nailing the dark frames extinguished the problem.

[gregbradley]

No, it was brought to the attention of SBIG and they said it was normal and not a column defect and merely a hot pixel which does not cause it to not be a Class 1. You'll see these lines also on the 16803 chips as well although these chips are quite noisefree at -10C.

From memory a class 1 has no column defects and hot pixel groups are less than a certain number.



Without exception ( I've tested hundreds of cameras) I have traced this to less than perfect dark frames....sure, lowering the temperature made the effect less obvious, but nailing the dark frames extinguished the problem.[/quote]


Yes you would be right there. I wondered what that difference was sometimes. Perhaps the bias frame reduction on the flats caused it.
It sometimes seemed a bit random. My darks were quite good. Oh well a bit of a mystery, sometimes it went and others it did not. Perhaps using a bias puts it back in there as the dark subtracts it and then it is gone and then if you use a bias in your flats you put it back in the flat and it comes through again? Perhaps better not to use a bias frame in this instance when making your flats.

Greg.

[Terry B]

When I bought my camera (second hand) it has a secondary cooler and water cooler installed. The secondary cooler didn't work so I removed it and purchased a new peltier device to replace it. It needs a 30mm square peltier and the closest I could find is a called a TES1-12703 from these guys (http://www.thermoelectricsupplier.com/single.php).
I bought 2 for US$19 plus postage. (so now I have a spare).
I run it with 13.8V from a marine battery.
I think it is a bit more hefty than the one that SBIG uses hence the good temp drop.
My water is just a 20L drum that sits in the observatory and is at room temp. I don't change the water much as I'm on tank water. It seems to stay OK.

[gregbradley]

Some people use water chillers. I know Fred simply put chunks of ice in a bucket of water.

Greg.

[Peter Ward]

This really needed a separate response. ;)

http://www.atscope.com.au/BRO/gallery20.html

AO guided from the urban lights of Sydney.

Vested interest? I disagree as I simply could not have got that resolution (AO guided at 15Hz) with any other system....and invite IIS members to have a look at all other M104 images on the web and decide whether this technology works.

[Peter Ward]

Dark and dark-subtracted flat and frames automatically have bias subtraction ....to repeat the process buggers up the data.

[gregbradley]

Indeed a marvellous image.

Don't get me wrong I am not against SBIG and I loved my STL camera.

My point really was that the various manufacturers have their relative strengths and weaknesses. One of FLIs strengths is the best cooling of anyones and it is a valuable feature. SBIG has a large range of accessories.

Its a good time to buy a CCD camera as its never been this competitive before. So many good manufacturers with excellent cameras.

Greg.

[renormalised]

If Cindy Crawford was an autoguider, every guy on the planet would take up astronomy!!!!:P:D:D

[Geoff45]

Well look here (http://www.andor.com/learn/applications/?docid=253) for REAL cooling. Down to -100 deg C. I note they don't mention prices anywhere. If you have to ask...etc

[Terry B]

I always liked the stats of the temp used for the receivers on the ATCA dishes. They are cooled to about 18deg K :scared: and get oxygen frost in the dewer in front of the receiver.

[Omaroo]

You mean one of these Terry :)

This LS-band (~2GHz) receivers' cold finger in this unit is immersed in a nice cool 14-18 degK bath. It's all so critical that they hire a bloke just to maintain it and several others. He's our mate John - the cryogenics engineer at ATNF Narrabri. He's the one on the left. The other is Jenny, my wife. Confusing them earns me a slap across the chops. :lol:

[renormalised]

Now we know why you're wearing sunnies in your avatar....to hide the black eyes!!!!!:eyepop::P:D

[strongmanmike]

Well?....hard to argue with all that Peter, isn't it? ;)... and this is why so many imagers with selfguiding cameras are forced to purchase a guide camera and guide scope anyway.

Mike

[strongmanmike]

Well actually Peter, that image as presented is terribly low res and looks quite pixelated...? :whistle:

Mike

[Terry B]

Thats the one- although the one I saw was the 21cm version. It was just bigger.

[AlexN]

Self guiding, off axis and standard guiding through a separate scope all have their equal merits.. I think if anything, the best option is how QSI did it with the built in OAG in front of the filter wheel... However, an OAG before the filter wheel, but after AO is a great option too... I personally think that self guiding has its merits.. Though, Im an SCT user.. If anyone else commenting in here uses a large SCT that does not have a mirror lock for imaging, they will know that guiding through a separate scope is not the easiest thing to do.. no matter how good your guiding is, if the primary mirror moves during a 10 minute exposure, the exposure is ruined.. I had good results using an OAG in front of my QHY8... and I've had a good test of self guiding. I found the results from self guiding to be better. although that's plainly just because I was not using any filters. I've just ordered a set of custom scientific LRGB/SII/HA/OIII filters from the USA. (Nb @ 5nm) will see how that goes... In the meantime, if the rain here ever stops, I'll give my 13nm Ha filter a go and see what length exposure is required to get a guide star...