GPCAM Peltier mod advice

[cuivenion]

Hi, I have the GPCAM Colour and I'm wanting to add a peltier to reduce noise when doing long exposures. There is a 5mm x 5mm space behind the sensor on the camera board and I thought that this peltier would make a good fit:

https://www.alibaba.com/product-detail/TEC1-00701-Thermoelectric-Cooler-5-5_60250316700.html

The sensor (Aptina AR0130CS) connects to the shell of the camera via this space so that the metal shell acts as a heatsink, so I'm hoping that if I put the peltier between the case contact and the sensor, that the shell will work as a heatsink for the peltier and cool the camera.

Here's what I'm not sure of:

1. Will the camera casing be effective enough as a heatsink for the peltier to work properly?

2. The pelter will fit on the space on the camera board which 5mm x 5mm square. The camera case contact is round and slightly smaller than 5mm x 5mm, would this effect cooling?

3. I have no idea how to power the peltier. Is there a commercial power supply that would do the job? Maybe some kind of usb powered device as the peltier is only 0.7v 1amp?

The camera draws 270mw max so I'm assuming the peltier will be more than powrful enough if cooled correctly.

[Gina]

You need to cool the whole of the hot side of a Peltier TEC otherwise the uncooled part will overheat and burn out.  A square piece of copper or aluminium will suffice between Peltier and case.  Don't forget to use a good thermal conducting compound too.  Power could be a problem with such a low voltage rating.  USB will only supply up to 500mA.  Maybe someone else can suggest power supply.

[ChrisLX200]

I think if I were going to attempt this (and I'm not ;-) ) I would keep the Peltier external, and have a copper 'cold finger' passing through the side of the case which would be in contact with the underside of the CCD sensor. So you would have an L-shape piece of copper - as large as will fit under the sensor - and the Peltier attached to the outside part (which would be larger) and a small fan on top. You would have a better choice of components that way, a larger Peltier and maybe a 12v one.

ChrisH

[cuivenion]

Thanks for the replies. There's pros and cons to both methods and a 12v peltier would be much easier to power. All the heatsink and fan assembly would add a lot of weight though, plus finding a way to attach it to the camera. Getting the cold finger inside the camera would be problematic due to its compact design.

I'd probably be better off finding a way to transfer the cold side of the peltier to the camera case which would in turn cool the sensor. I imagine I'd have to insulate the electronics at this point though.

I still think the smaller peltier would be the more elegant solution, so if anyone has any ideas on how to power it, and wether it would provide strong enough cooling that would be much appreciated.

[Icosahedron]

You can power it using pulse-width modulated 12V. The important thing is not exceeding the maximum current of 1A. Start off with a very low duty cycle while measuring the current and increase the duty cycle until the desired current is reached.

I've no idea if the cooling will be sufficient.

 

[Dragon_Astro]

Doesn't Altair already sell a peltier cooling unit that attaches to the GPCAMS? Its USB powered too.
 

[cuivenion]

2 minutes ago, Dragon_Astro said:

Don't Altair already sell a peltier cooling unit that attaches to the GPCAMS? Its USB powered too.
 

Nah, It's not a peltier. Its a heatsink and fan. I was thinking of getting it and seeing what difference it makes though.

[Dragon_Astro]

Sample images I've seen suggests it makes quite a bit

[cuivenion]

Just now, Dragon_Astro said:

Sample images I've seen suggests it makes quite a bit

Cool, do you have a link. I've not seen any comparison images.

[cuivenion]

1 hour ago, Icosahedron said:

You can power it using pulse-width modulated 12V. The important thing is not exceeding the maximum current of 1A. Start off with a very low duty cycle while measuring the current and increase the duty cycle until the desired current is reached.

I've no idea if the cooling will be sufficient.

 

Just been reading a bit on this and according to a couple of sites PWM is not a good way of powering a peltier:

http://www.meerstetter.ch/compendium/pwm-vs-direct-current

Would there be a DC way of powering it?

[pete_l]

7 hours ago, cuivenion said:

Just been reading a bit on this and according to a couple of sites PWM is not a good way of powering a peltier:

http://www.meerstetter.ch/compendium/pwm-vs-direct-current

Would there be a DC way of powering it?

Yes, you don't want to be switching large currents at high slew rates close to a sensitive device like a CCD. There would be a lot of transient electrical noise generated. The Peltier you are looking at (0.7 Volts @ 1 Amp) wouldn't be easy to power with a traditional linear power supply as they control the voltage to their load. However, you could investigate using a constant current source instead. This can be made simply with a 7805 linear regulator and a couple of extra components.

It would not be very power-efficient, but I doubt if that is your main concern. Just be careful to filter the HF noise from the regulator chip and have it on a heatsink.

[ChrisLX200]

You may already know this but... Unless you are going to open up the camera case and install a fan I really wouldn't bother doing this... Peltiers don't actually 'cool' anything, they just create a temperature gradient and move heat from one side of the device to the other. This is at the expense of needing put a fair bit of extra power (heat) into the device which also needs to be removed. In a sealed case the net effect is an _increase_ in temperature and not a decrease, so you have to provide some effective method of removing that exess internal heat. Note that every camera that uses TEC cooling has a fan and big vent. If you are relying on the casing to transfer and dissipate that extra heat well, I'm not sure it would work? It isn't designed to do that it's not an efficient method. That was why I suggested having the peltier external with a cold finger going in through the side of the case - the net (internal) effect will be extra cooling with the generated heat already being outside the casing. Just my 2p's worth...

ChrisH

[Gina]

I totally agree with Chris.  Peltier TECs are extremely inefficient (particularly the cheaper ones).  I have a QHY5II-C in my all sky camera and this has similar cooling from chip to case and I have found that efficient conduction of heat from the outside of the case to the cool night enviornment helps with noise quite a lot.  I reamed out and slitted an aluminium pipe to fit the camera, added thermal grease and with a couple of metres of pipe in the air and a bit in the ground, this provides pretty good "green" coooling.  Thus the pipe has two functions - support and cooling.

[Gina]

I did think of possibly incorporating a Peltier TEC between case and mast pipe but decided it was too difficult.  I also looked into the possibility of a cold finger but that wasn't practical either.  I have used Peltier TECs in DSLRs with cold finger with some success but tests showed that cooling much below freezing point didn't improve the noise performance significantly.  A small more efficient and expensive TEC proved worthwhile.

[Thalestris24]

Hi all

I recently bought a minicam5s though not actually tried it yet. It's interesting that the style of case is more similar to the zwo and someone on here has previously posted a cooling mod for the asi120mm. There is a slight temptation to attempt to reverse engineer it.... But I don't want to risk it right now. Maybe when I've had it a year or so . Still, the minicam5s spec and picture give an idea of how it might be done

Louise

[cuivenion]

Hi, I see what you're all saying. This is the type of thing I'm going for:

http://thx8411.over-blog.com/pages/Cooling_the_QHY5_imagerautoguider_step_2-3239743.html

The thing about the gpcam is that the cold finger to the sensor is already built in and attached to the case, so thats one thing I don't have to bother with. Unfortunatley this guy is a lot cleverer than me and can build his own circuit to power the peltier.

[Thalestris24]

8 minutes ago, cuivenion said:

Hi, I see what you're all saying. This is the type of thing I'm going for:

http://thx8411.over-blog.com/pages/Cooling_the_QHY5_imagerautoguider_step_2-3239743.html

The thing about the gpcam is that the cold finger to the sensor is already built in and attached to the case, so thats one thing I don't have to bother with. Unfortunatley this guy is a lot cleverer than me and can build his own circuit to power the peltier.

I think you'd need to think in terms of decoupling the sensor from the case. The way qhy have done it gives a good indication. I think if it was simple to do with the cylindrical case then others would have attempted it. The cylindrical case as heatsink limits what cooling can be done.

Louise

[cuivenion]

I think what I'll do is test the passive usb cooler first and see what sort of results I get. I still think this is a viable option though if I can get the peltier to work.

[Thalestris24]

18 minutes ago, cuivenion said:

I think what I'll do is test the passive usb cooler first and see what sort of results I get. I still think this is a viable option though if I can get the peltier to work.

Hi

If you think about it, if you just try and tec cool the case you'll be working to try and cool whatever is connected to it and against the ambient temperature. I think that's probably at least one reason why they only offer the fan-cooled option for the cylindrical package. There is also the condensation to consider and how it might affect the electronics. Using as is at low temperatures i.e. in the winter is ok - not so good in the summer! Have a look at the asi120mm cooling mod - on here somewhere.

Louise

[nightvision]

Not so much advice but may be of interest.

Just completed some basic testing of various 40x40mm-TEC/heatsink/fan combos fitted to the back of an ASI120mm.  I would prefer to fit a 10mm TEC to the back of the sensor or use a cold finger but can't figure out how to get the PCB out of the casing (USB and ST4 sockets appear to lock the PCB in place).  

Info from TEC manufacturers indicates slow changing DC control rather than PWM control lowers junction stresses and extends the life of the Peltier, with that in mind I used one of these brilliant little devices: http://www.ebay.co.uk/itm/LCD-Constant-Voltage-Current-Step-down-Programmable-Power-Supply-Module-CS-R3U9-/222046508144?hash=item33b300d470

Preliminary findings:

Cheap 40x40mm 12V 120705/12706 TECs can work well (with small heatsinks and fans) when driven initially at no more than about 13W (voltage/current depends on TEC, typically 6.5V/2A) and can (slowly via the alloy case) drop the sensor temp by 20C (monitored via Sharpcap).  Running the TECs at 12v (typically at least 40W!) just generated more heat (and less heat transfer from the camera) and so required much larger heavier heatsinks and fans than are practical.  Once the casing got close to the dew point the power was dropped to around 6-8W to prevent moisture build up.  The test set up uses a 50x50x30mm heatsink with a 50mm 12V fan running on around 8V to keep things quiet.  The fan provided much better dissipation when blowing down through the heatsink fins.  

No thermal paste was used for the tests, no camera insulation was tested as I wanted to see how much moisture built up, I expect better performance (and lower power requirements) from the final implementation.  The camera case was much colder than the sensor so I don't expect dew problems directly on the warmer sensor (no misting up occurred during testing), however plan to add some desiccant inside the camera. 

[Thalestris24]

There's also the option of a two-stage tec  Some useful info here

Louise

[cuivenion]

Thanks both very helpful and just what I was looking for.