1100D cold finger sensor cooling with TEC and water cooling

[nightvision]

I bought a couple of TECs last year, still haven't got round to doing anything .. but seeing this great mod has inspired me to have a go. BTW came across these devices: http://www.ebuyer.com/search?q=corsair+hydro

Tony.

[Gina]

I bought a couple of TECs last year, still haven't got round to doing anything .. but seeing this great mod has inspired me to have a go. BTW came across these devices: http://www.ebuyer.com/search?q=corsair+hydro

Tony.

Yes, I've seen those advertised but thought the CPU cooler was rather heavy and went for a much lighter North Bridge cooler. Glad you've found my thread inspiring

[Gina]

The Peltier TECs have arrived this morning so I'll probably put them in today and try them out

[Gina]

Put the pair of Peltier TECs on the camera this afternoon. Had to find some longer screws as there are now two TECs instead of one. Maybe test it this evening.

[Gina]

Just come in from testing dual Peltier cooling system. Hmmm... disappointing - not as good as the single 50W Peltier. Could only get down to EXIF T of -5C and couldn't even hold that - it settled at -4C. Ran at 12v supply ie. 6v each and the temperature went down nicely, flattening out as it reached -5C. Turning up the volts to 17 produced no more cooling and at 5.4A was the limit of my bench PSU. At that point the EXIF T went up to -4C and stayed there when I turned the volts down to 12.

Results :-

1. 12v 50W single Peltier TEC at 10v gave EXIF T of -9C :- input power :- 10v @ 3A = 30W
   
2. 2 x 12v 136W Peltier TEC could only manage to maintain -4C :- input power :- 12v @ 5A = 60W
   

Conclusion :- Dual Peltier cooling was actually WORSE

[nightvision]

Interesting, I think the 2 Peltier approach would work if the finger Peltier contact area was extended so that each contacted to the finger separately (I gather they are currently stacked together on to one contact area). Another enhancement could be more finger mass to increase thermal inertia (outside the camera).

[Auntystatic]

Try them in parallel instead of series, they might work better, however those tec's have a maximum current draw of 9Amp @ 15.2volts, your bench supply just isn't up to the job of supplying 18Amp.

Of course doing the math on the acclaimed 136W / 12v = 11.33A maths makes my head hurt.

I ordered a couple on your recommendation, I'll hook them up to my 30Amp variable radio psu and take measurements when they arrive.

Wouldn't stacking them cause 1 just to act as a heatsink to the other, or am I missing something?

[Gina]

I used the stack method since there is a limit to the temperature difference a Peltier TEC can produce. I gather some cooled astro cameras use stacked Peltiers as well.

Now I need to decide what to do next. I guess I can try one 136W TEC. If that proves worse than the single 12v 50W, the logical conclusion would be that going for higher power is wrong and that a 5v 20W one may be better I have one of those too that I was going to use inside the camera before I decided to go external to retain the camera's display.

[Gina]

Try them in parallel instead of series, they might work better, however those tec's have a maximum current draw of 9Amp @ 15.2volts, your bench supply just isn't up to the job of supplying 18Amp.

I don't think parallel would be any better than series. If both get 6v at 4A it shouldn't make any difference whether they are in series or parallel if they have the same characteristics. Increasing to 17v for the two should give them 8.5v each - this produced no more cooling. Actually, on thinking about it I doubt they are properly sharing the voltage as the outer one has the heat from the inner one to get rid of as well as it's own heat and the sensor heat. With that thought in mind, maybe the outer one should be capable of greater cooling.

Of course doing the math on the acclaimed 136W / 12v = 11.33A maths makes my head hurt.

I ordered a couple on your recommendation, I'll hook them up to my 30Amp variable radio psu and take measurements when they arrive.

Yes, I know my bench PSU won't supply their maximum current. I'll be very interested in your results

Wouldn't stacking them cause 1 just to act as a heatsink to the other, or am I missing something?

Yes, in a way. It's true that all the heat generated by the cooler one needs to be taken away by the hotter one.

[Gina]

I've decided to re-run the test tomorrow and measure the volt drop across each TEC with my digital multimeter. The effective operating voltage is reduced by the Peltier voltage generated by the temperature difference when run from a constant voltage source. This is why the current decreases as the cold side cools down when running a single Peltier TEC from a constant voltage source. With TECs in series one may drop more temperature than the other resulting in unequal sharing of the supply voltage so in this respect you may be right Danielle.

As for PSU, I have several ex PC PSUs so I'll set one up so I can run TECs in parallel at either 5v or 12v, or even one at 12v and the other at 5v if I like. I'll order MOSFET and other bits to make a controller tomorrow too - I should have done that before really

I intend to crack this one!!

[yesyes]

I'm watching this with much interest as I have a few problems with my peltiers as well.

[nightvision]

Just been reading up on Cascaded Peltier cooling. It appears they should not be the same power, the one closest to the heat source should be the lower power.

However cascading puts greater emphasis on the heat-sink capability.

When cascading Peltiers the heat-sink contact area and efficiency eventually become the limiting factors at which point also using parallel physical arrangements and larger heat-sinks (for more contact area and dissipation) makes sense.

Tony.

[Gina]

Just been reading up on Cascaded Peltier cooling. It appears they should not be the same power, the one closest to the heat source should be the lower power.

That's very interesting - what I thought really

However cascading puts greater emphasis on the heat-sink capability.

Oh yes, quite true. My water cooling is capable of shedding a lot of heat. Even with 92W input power to the Peltiers, the water was still cool - I could only just feel the difference between the in and out tubes on the cooler.

When cascading Peltiers the heat-sink contact area and efficiency eventually become the limiting factors at which point also using parallel physical arrangements and larger heat-sinks (for more contact area and dissipation) makes sense.

Tony.

Thank you Tony - that's all very helpful

[Gina]

OK then... Test runs to try next :-

1. Single 12v 136W Peltier - 40mm square
   
2. Cascaded 12v 50W and 136W Peltiers - both 40mm square
   
3. (Possibly) Single 5v 20W Peltier - 25mm square
   

I'm guessing no.2 should give the best results. If this gives a good result I might not bother with no.3 though the results would be interesting. There is a 4th possibility of using a combination of 20 and 50W Peltiers though the difference in size complicates things.

On a side note, I'm thinking perhaps I should make myself a 20A bench PSU

[RogerTheDodger]

I'd go for the home-made PSU Gina, 20A linear psu is a doddle to make, and higher currents are achievable with the same design but using a bigger transformer and more OP transistors.

[Gina]

I'd go for the home-made PSU Gina, 20A linear psu is a doddle to make, and higher currents are achievable with the same design but using a bigger transformer and more OP transistors.

Yes, I've made a 10A battery charger - a 20A variable bench PSU is almost an extended version of that with manual control of voltage and current limiting rather than automatic.

[mindburner]

looking forward to more if your results with the peltiers. The one i ordered has not arrived yet from ebay. I noticed that RS components do a range of high efficiency peltier coolers, They are pricey but may be worth investigating.

[Auntystatic]

If you use your 5A psu the peltiers will only be running at half power, you need something heavier duty.

[Gina]

I have a PC PSU I plan to use. 30A @ 5v and 15A @12v. I have a couple of power MOSFETS and other bits on order to make PWM controls for the Peltiers.

[Gina]

looking forward to more if your results with the peltiers. The one i ordered has not arrived yet from ebay. I noticed that RS components do a range of high efficiency peltier coolers, They are pricey but may be worth investigating.

Ah yes, I wondered if there were better Peltiers available. I'll check that out. Thanks

[Gina]

Been on the RS web site but they quote their Peltiers as being for power generation rather than cooling

[Auntystatic]

Mine have just arrived 2x 12709, I put one of them on my radio psu 13.8v@ upto 30A, it said it was drawing 10 Amps at first but soon settled down to 8 Amps it was very cold to the touch after just 30 seconds.

I'm dead excited now as I have another thing to play with during this dank weather, my 2x 4n25's arrived too so I can control my modded 5D from APT, looks like I'm going for APT as it can control more than one camera, so I'll be buying it soon, I'm now half way to saving for a 190mn or c9.25 so I better stop spending and start selling more gear.

Hmmm I wonder if I cool my 5D it might give the atik 11000 a run for its money he he

[nightvision]

A great Peltier reference site, some really practical information that might help:

http://thermal.ferrotec.com/technology/thermal/thermoelectric-reference-guide/

Tony.

[Gina]

A great Peltier reference site, some really practical information that might help:

http://thermal.ferrotec.com/technology/thermal/thermoelectric-reference-guide/

Tony.

Thank you - I'll have a read

[Gina]

Mine have just arrived 2x 12709, I put one of them on my radio psu 13.8v@ upto 30A, it said it was drawing 10 Amps at first but soon settled down to 8 Amps it was very cold to the touch after just 30 seconds.

Yes, they seem to work quite well on their own

I'm dead excited now as I have another thing to play with during this dank weather, my 2x 4n25's arrived too so I can control my modded 5D from APT, looks like I'm going for APT as it can control more than one camera, so I'll be buying it soon, I'm now half way to saving for a 190mn or c9.25 so I better stop spending and start selling more gear.

We certainly need something to take our minds off this dreadful weather. I was able to dodge the rain and bring my camera indoors to take one of the TECs out leaving the one but not to get out again to test it. I need to connect up the water cooling and most of that is attached to the mount. If the rain eases off this evening I might be able to do a test run. Forecast is for heavy rain and strong winds though

Hmmm I wonder if I cool my 5D it might give the atik 11000 a run for its money he he

Yers!!