QHY5L-II Mono Options for Passive / Active Cooling

[SnakeyJ]

yeah...the driver issues... -.-

and maybe a temperature readout of the chip!

somewhere here was a discussion regarding low gains on these chips...

...there seem to be a problem under 9% gain...but i haven´t tried it yet

I'm told (via the qhyccd forum) that the QHY5L-II ascom driver works fine with LX mode, though this does limit the choice of capture program. I think Sharpcap support ascom drivers, as do Nebulosity and MaximDL - Nebulosity seems to be worth the 80USD, but it does seem over kill for this camera! MaximDL looks an incredible product, but is way too much for my budget or needs at the moment.

I was not aware of reported issues below 9% gain with the QHY - with 10% being the lowest level I have used. There forums have been down for the last two days so haven't been able to check. However, I do recall a post from JamesF about reduced dynamic range at low range on his ASI120 - though from memory I think he has the colour version based on the Aptina AR0130CS sensor (may be wrong and he might have both).

I think the 'offset' feature/control allows the ADC response to be shifted to enhance low end sensitivity, though as with many of the QHY controls/features there doesn't seem to be any official documentation.

[Jannis]

Hmm, after reading this thread, i'm still a bit unsure. I'm having my QHY5L on it's way from Bernard now. As i live in Norway, and it's usually always below freezing here while i do imaging. How much would additional cooling actually help?

I mean, let's say it's -10c outside - how hot does the sensor actually get then?

I know it's a saying that colder is better, but after reading up on the sensor it's supposed to have an operating temp from -30C to +70C.

If i attach a peltier on the outside of the camera body, and leave the camera housing original, i'm guessing this should cool down the sensor an additional 10-15c, maybe a bit more? Would this help much on the noise if the sensor is allready below 0 due to the cold outdoor temp?

[SnakeyJ]

Hmm, after reading this thread, i'm still a bit unsure. I'm having my QHY5L on it's way from Bernard now. As i live in Norway, and it's usually always below freezing here while i do imaging. How much would additional cooling actually help?

I mean, let's say it's -10c outside - how hot does the sensor actually get then?

I know it's a saying that colder is better, but after reading up on the sensor it's supposed to have an operating temp from -30C to +70C.

If i attach a peltier on the outside of the camera body, and leave the camera housing original, i'm guessing this should cool down the sensor an additional 10-15c, maybe a bit more? Would this help much on the noise if the sensor is allready below 0 due to the cold outdoor temp?

Jannis,

Hopefully the drivers/sdk will be updated very soon to show the CMOS temperature as has already been done with the ASI120MM. These sensors definitely generate some heat and quite a lot of it at high frame rates. How much is being generated during LX imaging is a more difficult to quantify. I think the basic design of the camera is good and will dissipate heat out to the shell effectively, though this may be improved by replacing the thermal pad with a more efficient material. Working with an ambient temperature of -10C, I think the gains of additional cooling would be small and probably not worth the trouble. However at high frame rates some external passive fins might help to dissipate the heat more quickly.

In UK I'm currently imaging in Ambient temps of 16-18C overnight and in winter average overnight temps drop towards zero. I haven't taken any measurements of internal temps on the device, but I would guess these are typically around 5-10C above ambient even at low frame rates.

There's an interesting article on cooling the older QHY5 by THX8411 @ http://thx8411.over-blog.com/pages/Cooling_the_QHY5_imagerautoguider_step_1-3170312.html and a useful comparison of noise reduction by cooling produced in Tonedeafs own cooling experiments with the QHY5L-II earlier in this thread

My target in this is to cool and hold the sensor to a setpoint of -10C, which I expect to give around 95% less noise than an unmodified camera operating at temps of approx.25C with an ambient air temp of 18C. This will require substantial modification, invalidation of the warranty and in truth I would probably be financially better off buying an Attik Titan, Brightstar Mamutt or QHY6 - but I do like to play, get a better understanding and enjoy the challenges.

[Jannis]

I see. Thanks for the answer. I'm mostly going to use it for longer exposures for deep sky, and probably 1+sec for guiding.

Acording to the sensor specs:

Power

• Supply: I/O: 1.8V

Core: 1.8V

Analog: 2.8V

• Consumption: ~300mW

0,3W doesn't sound like a lot though. And complete camera is rated at 0,64W Shouldn't require much, if any additional cooling in the winter time. I'm guessing a small peltier would cool the sensor quite a bit even with an unmodded camera?

Or is the stock heat transfer pad under the sensor really just that poor?

[SnakeyJ]

Jannis,

I think the design is quite good, though the thermal pad is quite thick and the cold finger could be bigger cross sectional area. Thermal pads are typically rated 0.3mWK, which is quite poor in comparison with standard thermal pastes and modern TIM - I've replaced mine with the Phobyas self adhesive XT Thermal pad, with a claimed 7W/mK which if the blurb is correct is about as good as standard thermal paste. This is a straightforward and cheap mod to do, with the only risk being dirt contamination and the invalidated warranty. So after a few weeks bed in period, its probably worth re-running the tests I did on the unmodded camera to see if there was any benefit - though I think any benefits would be small, perhaps 2-3C best, as the die construction is probably the major limiting factor for heat transfer. It's also difficult to get absolute numbers without direct driver/api access to the on chip temp sensor, though hopefully this will be rectified quickly.

I've just managed to find a working link to the full Aptina MT9M034 data sheet, which I've been trying to get for a long time - http://ghgtechn.com/...T9M034_DS_B.pdf (just the register reference left of my collection list).

Not sure if any of this will make much difference for guiding purposes, as I doubt you will want to go beyond 1-3 sec subs and dark frame or hot pixel maps should take care of the worst offenders - but for longer exposures it would make a significant difference for me operating at higher ambient temperatures. The other major plus would be using set point to control temperature so that you can work with a fixed set of dark frames - this allows more time devoted to capturing useful light and flat frames!

[Jannis]

Aha, interesting to see that in THAT data sheet the rated sensor power is completly different:

Power consumption <620 mW (parallel data output), 45 fps, High Dynamic Range Mode.

Personally i'll be using it mainly for guiding. But i'll hopefully also use it to capture IR data, and maybe Ha, OIII and SII in the futore, so long exposure is for sure going to be used. I'll give it a go stock first to see how it performes with no extra cooling at -10c, and then at -30, for comparison. Too bad i have to wait probably 2-3 weeks for it to arrive in the post though. So darn slow post systems here in Norway. >_<

[SnakeyJ]

My main point of interest from the data sheet is is the -30C lower operating limit, not that I'll be pushing it that far but might open it up for a two stage peltier approach! Though important to remember that this is the the CMOS sesnor itself and does not cover the other electronic components. QHY don't publish a temperature range for this camera, but they do give 640mW Power consumption that is in line with the 620mW upper limit on the sensor.

The narrow band filters will be interesting from an SNR perspective and I'm not sure even with active cooling if you will be able to reduce noise levels sufficiently. OIII will be most favourable with a QE of 73%, QE for Ha @656nm will be approx. 53%, QE for the SII @672nm looks sub 40%:

Not saying its impossible and will be interested to see some results, but think a true astro CCD might be an easier route for narrowband.

[Jannis]

Yeah, i guess it's really pushing the QHY5L a bit futher then what it was designed for, but if it works, i'm willing to give it a try - as i really can't aford an expencive CCD now anyway.

And this thread really proved that it can perform significantly better if kept cold - so there are hope for the Norwegian winter with very often -15c or colder at night.

QE at SII is low indeed, but not sub 40. Actually it's at 48% - if the graph is accurate.

I'm actually more worried with the IR data where i'm going to use an IR-pass filter at 742nm, where the QE is about ~32%, and droppign fast.

But it's got to be a lot more then just the QE that matters anyway i guess. Concidering the, for example SXVF-H16 with the Kodak KAI4021M CCD have a QE of sub 30 at SII/672nm.

I have to admit i haven't really put myself too much into the sensor details too much though - about what means what. But I guess i better just get started on that, hehe.

Edit: but i guess this is a bit off topic though ^^;

[SnakeyJ]

QE at SII is low indeed, but not sub 40. Actually it's at 48% - if the graph is accurate.

I'm actually more worried with the IR data where i'm going to use an IR-pass filter at 742nm, where the QE is about ~32%, and droppign fast.

Sorry for my misread off the horizontal axis/bum steer - 48% certainly looks a lot more attractive!

I've only used my Astronomiks 742 on Saturn with the QHY and to be honest it wasn't very useful on this target - Once Jupiter season starts I'll have a good go and have some comparable data shot with the 742 and Pt Grey Firefly MV from last year to give a side by side comparison (I think the firefly was around 45% QE for this). There are some really nice e2v ruby sensors that offer good NIR sensitivity, though this is more of riklaunim's area of expertise - worth checking out his website at http://www.rkblog.rk.edu.pl/w/p/astrophotography-and-observations/ as it's ram packed with useful info and tutorials.

I'm in the same boat re the CCD - will get one in the not too distant, but loads of useful stuff to learn first.

[SnakeyJ]

Following up on earlier posts in this thread, re getting temperature readout from the QHY5L-IIs MT9M034 sensor and a quick post in Firecaptures Yahoo forum, I posted a thread in the QHY development forum last night and got an almost immediate and very encouraging response back. Seems the Chinese SDK is already released and an english version is imminent and LZR will add the link for temperature readout so it should be available in our favourite capture programs in the not to distant.

[SnakeyJ]

Taken a long time to get back to this thread/mod - but in my defence I have been a bit pre-occupied with clearing and building my observatory, now pressed in to service even if not fully completed.    My other project of the last few weeks has been getting my lathe up and running so that I can turn/mill a new case and heatsink for the QHY5L-IIM to improve the design for active cooling.

All now in place and have (thanks to a tip from Gina) sourced the last components, some T0-92 temperature sensors to control set point operation - 3 in total for hot-side, cool-side and ambient temp.    I've also booked out two weeks from work so have no excuses for not pushing this through.

Almost forgot - there's a new release of Firecapture (23 Beta 14) due on Friday that should extend the max LX Exposure from 60 to 600 secs and will also add the new QHY SDK DLLs to log on chip temperature.   Hopefully this will allow me to calibrate my setpoint cooling and confirm the effectiveness of the set point cooling.     I have also purchased Neb 3 and although the ascom driver camera control is rather limited (8/12bit, Gain and Exposure time), this offers another route for longer exposures and allows me to process at full bit depth (as opposed to Gimp 2.8.6).

[tonedeaf]

mmmhh observatory...maybe in three years   (there are other needs my wife wants me to do first )

hmm temperature readout from qhy? niiiiceeee

looking forward...

also on my side to much work...

...and when i would have time, the weather gods are in my way

[tonedeaf]

ok had made an additional mod to the QHY5II...

i hope that the weather will change soon (no observation time since weeks/months )

to prevent dew/freezing of the cmos sensor, i´ve made a heater:

(ok this is the first version, with 2W resistors, maybe i´ll change them to smaller 1W resistors -> cause the 2W are a bit too big!)

ok now to the mod 4 100ohm 2W resistors 2/2 parallel/serial -> 1,44W @ 12V

ran a test @ ~ -5-5,5°C (that´s ~the max now -> with 20°C room temperature -> -25°C down from ambient)

here the result....and yes no freezed chip, not even dew on the chip -> successfull mod!

ok the downside of the mod -> 620g camera! (incl controller)

[JamesF]

I've just managed to find a working link to the full Aptina MT9M034 data sheet, which I've been trying to get for a long time - http://ghgtechn.com/...T9M034_DS_B.pdf (just the register reference left of my collection list).

Resurrecting this thread... Did you ever manage to find a copy of the register reference, Jake?

James

[SnakeyJ]

I didn't find this James, though haven't searched recently. Though I do recall QHY's linux developer posting on their forums, so could be worth posting any specific questions there. I think TE has more or less given up on further feature support for the qhy5-ii, due to poor developer support and frustration!

typed on my mobile with Tapatalk

[JamesF]

I didn't find this James, though haven't searched recently. Though I do recall QHY's linux developer posting on their forums, so could be worth posting any specific questions there. I think TE has more or less given up on further feature support for the qhy5-ii, due to poor developer support and frustration!

I can understand that.  The more I look at the open source library the odder it seems and questions on the forums aren't really being answered in the necessary detail.  Such is life...

James

[Jannis]

This thread is probably a bit old now i guess. QHY already came with a new version of the sensor with active cooling and stuff... Anyway, i've been working a bit on my own cooling setup here.

I managed to find some heavily rusted scrap metal plate laying in the woods, and thought i could try to use it. So i cut off a suitable piece from it, and used the remaining of the metal to clean off the rust with electrolysis.

After a day in the bath i brushed off and cleaned the metal piece. Even though it was heavily rusted it looks like it will do the job, and the before and after speaks for it self!

I have very limited tools, but a vise grip, hammer, and a pipe socket of the correct size was all i needed in addition to an angle grinder to cut off the excess metal.

I tested it with an actual EP and drilled a hole in it which i'll later use for clamping the metal around the QHY5L for a good contact.

Although the end result looks a bit weird, i'm hoping it will work.

I found a heatsink laying around in my basement, and i found a nice Delta AFB0712HHB fan that will do the trick although it's slightly larger then my heatsink. It's capable of working from 4V to 13,8V, have a high airflow of up to 37,5 CFM, generates extremely little vibrations and is less then 50 gram.

For my peltier i'm using a 40x40mm TEC1 60W peltier i bought from ebay at a very low cost.
I mistook my thermal glue for paste the first time, and so when i found out the next day, that peltier was unusable as it now had a 1cm tall heatsink firmly glued onto it... I won't do that mistake again... Luckily i had orderd 2x peltiers.

I hooked up a temperature meter, and my power supply and gave it a test run. I tried it at several different currents and voltages, but it seems this one likes it best at around 6V pulling around 3A (18 Watt) even though i've having no problem keeping the heatsink rather cold at full power... Here i just started it up:

And soon enough it shows:

I left it running after taking the pic, and it's still running as i type this, but an hour after the pic it had only dropped down to 4,1c (same room temp).

Next step will be to solder the metal together where it's folder together to close teh gap and increase the thermal transfer, and then to use thermal glue instead of thermal paste before mounting the peltier permanent. I'll update on the progress and results later.