Help understanding Peltier Modules.

[Enceladus Dan]

Hello, 

a project I’ve had on my mind a while now is a cooler for my DSLR. After testing my car battery power box for 6hrs on Friday 17th June running a mini pc, 14” monitor, dslr power, guide scope, tracking motors and 3 dew heaters (not attached just running for a test) with good results as I still had 70% battery, however the images were not so good. The Ambient temp was still over 20c at midnight and APT was reporting sensor temp of 34c after 300s exposures. So now I think I need to get moving with the cooler. 

my plan is to adopt the cooler box idea from Gary Honis, but slightly deeper so my 600D screen can be open slightly. Allowing cold air to reach the warmest part of the camera. I do like the simplicity of the cold plate butted to the back of the camera like with  Martin Pyott, but think all round cooling with its own little ambient temp would be better. I have drawn up an idea to add a small plate, from the cold heat sink to the back of the camera. In an attempt to see if like a cold finger takes the sensor heat out to the Peltier, if this plate will take some heat from the back of the camera away to the cold circulating air or even to the cold heat sink. See if I can create the best of both. 

Not wanting to just copy parts from others I’ve seen, I wanted to look into and understand what I was doing. Im thinking of buying separate components rather than a kit, as most posts I have read indicate bigger heat sinks and fans are required. I have read up and watched videos from overclockers about Peltier and have managed to wrap my head around most of the technical details on how it all works, but I have a few questions for those who specialise in this area. 

After reading many data sheets, I finally realised adding more volts, amps or going to a bigger module was only using more power and not increasing the delta temp as the hot side was hotter. So I settled on a tec1-12704 at 11v 2A giving a -40c drop. Assuming the hot side is (Peltier power 36w + Qc (camera heat unknown) then multiply by heat sink C/W 0.5 (actual is unknown) and add ambient 20c) I’m looking at a guess of 40c on my hottest night. 

1. Peltier power - for the hot side, is it just the rated power 36w or the power used ie. 11 x 2 = 22w?

2. power supply - My limited knowledge from a past career in vehicle electrics 25 years ago, is telling me if I apply 11v with the module being 3 ohms will draw 3.6A. I found a buck/boost module that has constant current, but I don’t know if it somehow limits current or just alters the volts to keep Amps steady. 
 

3. Heat sink - thermal resistance? Overclockers said assume 0.5 if it’s unknown, I was looking at an arctic alpine 12 co 100w rated. Smaller and lighter than a tower cooler but bigger than the kit sink and fan. The website doesn’t give a C/W rating for this but a graph showing a 65w cpu running at 70c with this cooler, one of their larger tower cooler is rated at C/W0.8. Am I just over thinking this, as I can’t get my head round how at 0.8 I won’t even reach 10c, but on YouTube a setup with a 60mm heat sink and fan or a cooler from a 5v mini fridge has formed ice at the cold side. 

Any advice or experience welcome, especially on the Peltier power and power supply. For the heat sink, my gut is saying just do it, As the cpu cooler must be better. 
 

Thanks

Danny

[skybadger]

There's a good guide to peltier calculations on the rswww.com website.

 

[Enceladus Dan]

Thanks I will have a look. 

[tomato]

Hi Danny, this thread might be of interest:

 

Edited June 23, 2022 by tomato
Correct link added

[Enceladus Dan]

Thanks tomato I will search it up. I found an other video, Explains a bit more and dt is driven by the amps used. So I believe the volts will be of less concern if I get amps right. So the buck/boost should work, and I can play with settings t fine tune it to get the air in the box to 0 or just under. I’m only looking to keep sensor in single figures if I can. 
 

maybe it’s time to play and see what happens. 
 

the rs online has some clearer data sheets, the ones I found online were in let’s say broken Chinese English. 

[Enceladus Dan]

Interesting to see the comments in that topic on fan push and pull. I was debating pulling heat away in my drawings, not for the airflow but on the fact the hot air leaving the heat sink is directed back towards the hot side of the Peltier. By pulling its taken away from it and possibly even picking up some cold air radiating round from the cold side and drawing that through the heat sink. 
 

I like a good tinker, and will have fun with this. More so than my diy dew heaters I think. 
 

thanks danny. 

[david_taurus83]

I made a water cooled box for my 6D as a side project over lockdown. Pretty sure I was inspired by @Adam J much tidier project than mine.

https://stargazerslounge.com/topic/354732-yet-another-dslr-coolbox-thread/

[VectorQuantity]

8 hours ago, Enceladus Dan said:

Hello, 

a project I’ve had on my mind a while now is a cooler for my DSLR. After testing my car battery power box for 6hrs on Friday 17th June running a mini pc, 14” monitor, dslr power, guide scope, tracking motors and 3 dew heaters (not attached just running for a test) with good results as I still had 70% battery, however the images were not so good. The Ambient temp was still over 20c at midnight and APT was reporting sensor temp of 34c after 300s exposures. So now I think I need to get moving with the cooler. 

my plan is to adopt the cooler box idea from Gary Honis, but slightly deeper so my 600D screen can be open slightly. Allowing cold air to reach the warmest part of the camera. I do like the simplicity of the cold plate butted to the back of the camera like with  Martin Pyott, but think all round cooling with its own little ambient temp would be better. I have drawn up an idea to add a small plate, from the cold heat sink to the back of the camera. In an attempt to see if like a cold finger takes the sensor heat out to the Peltier, if this plate will take some heat from the back of the camera away to the cold circulating air or even to the cold heat sink. See if I can create the best of both. 

Not wanting to just copy parts from others I’ve seen, I wanted to look into and understand what I was doing. Im thinking of buying separate components rather than a kit, as most posts I have read indicate bigger heat sinks and fans are required. I have read up and watched videos from overclockers about Peltier and have managed to wrap my head around most of the technical details on how it all works, but I have a few questions for those who specialise in this area. 

After reading many data sheets, I finally realised adding more volts, amps or going to a bigger module was only using more power and not increasing the delta temp as the hot side was hotter. So I settled on a tec1-12704 at 11v 2A giving a -40c drop. Assuming the hot side is (Peltier power 36w + Qc (camera heat unknown) then multiply by heat sink C/W 0.5 (actual is unknown) and add ambient 20c) I’m looking at a guess of 40c on my hottest night. 

1. Peltier power - for the hot side, is it just the rated power 36w or the power used ie. 11 x 2 = 22w?

2. power supply - My limited knowledge from a past career in vehicle electrics 25 years ago, is telling me if I apply 11v with the module being 3 ohms will draw 3.6A. I found a buck/boost module that has constant current, but I don’t know if it somehow limits current or just alters the volts to keep Amps steady. 
 

3. Heat sink - thermal resistance? Overclockers said assume 0.5 if it’s unknown, I was looking at an arctic alpine 12 co 100w rated. Smaller and lighter than a tower cooler but bigger than the kit sink and fan. The website doesn’t give a C/W rating for this but a graph showing a 65w cpu running at 70c with this cooler, one of their larger tower cooler is rated at C/W0.8. Am I just over thinking this, as I can’t get my head round how at 0.8 I won’t even reach 10c, but on YouTube a setup with a 60mm heat sink and fan or a cooler from a 5v mini fridge has formed ice at the cold side. 

Any advice or experience welcome, especially on the Peltier power and power supply. For the heat sink, my gut is saying just do it, As the cpu cooler must be better. 
 

Thanks

Danny

Hi Danny,

1. Apologies it's a bit ambiguous are you able to re-phrase the question?

2. Is it a buck convertor (step down) or a boost convertor (step-up)? It could theoretically be both but that's unlikely. The majority of output protected convertors will current limit at which point it appears as a constant current. You shouldn't be running at the current limit point, this exists to prevent destruction of the PSU and the attached load.

3. The peltier isn't a cooling device, it's an inefficient heat pump, in addition to having to get rid of the pumped heat you also need to get rid of the applied power that provides the heat pump action, if you don't then the temperature delta falls eventually to zero then goes positive and you have a heater. The most important thing is getting rid of the heat which is hard to do with a non-blown heatsink. Your example of the cpu running at 70 degress unfortunately doesn't tell you much as the cpu is additionally connected to a significant PCB heatsink not just the added heatsink. The degC/W rating assumes perfect heat transfer into the heatsink, you'll also need to factor in the degC/W of the heatsink transfer material probably silicone pad or thermal oxide both of which have poor thermal transfer. The metal oxide is superior but likely to add 0.4 degC/W, the silicone thermally filled pads maybe 4 to 8 degC/W. As far as the fan goes you're interested in the total volume of air it moves rated in either cubic metres per minute or per hour.

Your fridge example is a thermal mass that's not being heated so easier to cool, and it's probably been on for many hours to get down to temperature, your camera heats itself so a comparison isn't that valid.

I'd start by having a peltier that doesn't have to be driven hard then you don't have added heat to get rid of, thermal oxide to the heatsink then a heatsink shape that ducts onto the fan and then a high flow/volume fan.

The other option is a non peltier and purely fan blown one, that may keep you close to ambient without all the hassle.

 

 

[Adam J]

49 minutes ago, david_taurus83 said:

I made a water cooled box for my 6D as a side project over lockdown. Pretty sure I was inspired by @Adam J much tidier project than mine.

https://stargazerslounge.com/topic/354732-yet-another-dslr-coolbox-thread/

Thanks, here is my thread. 

https://stargazerslounge.com/topic/289309-adam-js-dslr-water-cooling-thread/

[Enceladus Dan]

10 hours ago, VectorQuantity said:

Hi Danny,

1. Apologies it's a bit ambiguous are you able to re-phrase the question?

2. Is it a buck convertor (step down) or a boost convertor (step-up)? It could theoretically be both but that's unlikely. The majority of output protected convertors will current limit at which point it appears as a constant current. You shouldn't be running at the current limit point, this exists to prevent destruction of the PSU and the attached load.

Thank you for the reply, 

1. Was for the equation to figure out the hot side temp. Where ((Pin + qc)xC/W)+Ambient. the Pin figure I was unsure if I use the Peltiers rated power 36w or power applied. I went back to overclockers and it would appear to be power applied, so In my case 22w (which is good to reduce heat I need to extract). 

2. The adjustable buck/boost I have found claims to be both while offering constant current. Max Amps is 4A and I intend to only use 2 amps. I hoped this would give me the 11v 2A supply that the Peltier data sheet states. I guess I could just get it and test on a 12v 21w brake light bulb which should draw 1.75 amps, then try and limit to 1A and see if voltage remains at 12v. 
 

3. is again going to be a buy and see, I can’t find any data showing the efficiency in c/w, so will be some trial and hopefully not error. 
 

im trying to work it all out before I buy any of the parts, but with not knowing the C/W or the Qc values, it’s going to be a best guess. But at least I will have fun experimenting. 
 

thanks

Danny

[VectorQuantity]

1 hour ago, Enceladus Dan said:

Thank you for the reply, 

1. Was for the equation to figure out the hot side temp. Where ((Pin + qc)xC/W)+Ambient. the Pin figure I was unsure if I use the Peltiers rated power 36w or power applied. I went back to overclockers and it would appear to be power applied, so In my case 22w (which is good to reduce heat I need to extract). 

2. The adjustable buck/boost I have found claims to be both while offering constant current. Max Amps is 4A and I intend to only use 2 amps. I hoped this would give me the 11v 2A supply that the Peltier data sheet states. I guess I could just get it and test on a 12v 21w brake light bulb which should draw 1.75 amps, then try and limit to 1A and see if voltage remains at 12v. 
 

3. is again going to be a buy and see, I can’t find any data showing the efficiency in c/w, so will be some trial and hopefully not error. 
 

im trying to work it all out before I buy any of the parts, but with not knowing the C/W or the Qc values, it’s going to be a best guess. But at least I will have fun experimenting. 
 

thanks

Danny

re: 1. Yes you are correct, it's the applied peltier power.

re: 2. You would expect the o/p voltage to stay at its set point unless it's in current limit, i.e. not to change until you reach a current draw of 4A. Do you have a part number/link to the psu?

Do you know how much power your camera consumes?

Edited June 24, 2022 by VectorQuantity

[Enceladus Dan]

32 minutes ago, VectorQuantity said:

re: 2. You would expect the o/p voltage to stay at its set point unless it's in current limit, i.e. not to change until you reach a current draw of 4A. Do you have a part number/link to the psu and the peltier?

The buck boost module is this one. 

DollaTek ZK-4KX CNC DC Buck Boost Converter CC CV 0.5-30V 4A Power Module Adjustable Regulated https://amzn.eu/d/dMVOCbE

And the chosen peltier is a tec1-12704. 

On a side note I have disregarded the sensor temp for my bad image on friday gone. It was bugging me and darks were bright red which made me think temp. But seen another post and this is normal when opened in a programe (my case gimp) I popped sd card in my 600d and sure enough darks are actually dark and zoomed in sees the noise. My darks were at 32c as was my last 2 x 300 subs which were the best 2 as background was grey. I went to earlier sub at 32c before sensor peeked at 34c and the background was red. 

It turns out we had some saharan sand signals in the sky when I was imaging. Explains why my car looks like I've been in the Dakar rally. 

Thanks

Danny

[Enceladus Dan]

1 hour ago, VectorQuantity said:

Do you know how much power your camera consumes?

Not sure but I have been using a dummy battery on a 12v3a ac adaptor till now. Just having made a 12v power box so I can go off grid to a bortle 4 site. I could put my multimeter inline and measure current used while doing some test long exposures. Then Average them out. 
 

I did find a post with someone who did lots of tests with a 700d and 30s Timelapse with screen off is 1w of usage. 

[VectorQuantity]

26 minutes ago, Enceladus Dan said:

The buck boost module is this one. 

DollaTek ZK-4KX CNC DC Buck Boost Converter CC CV 0.5-30V 4A Power Module Adjustable Regulated https://amzn.eu/d/dMVOCbE

And the chosen peltier is a tec1-12704. 

On a side note I have disregarded the sensor temp for my bad image on friday gone. It was bugging me and darks were bright red which made me think temp. But seen another post and this is normal when opened in a programe (my case gimp) I popped sd card in my 600d and sure enough darks are actually dark and zoomed in sees the noise. My darks were at 32c as was my last 2 x 300 subs which were the best 2 as background was grey. I went to earlier sub at 32c before sensor peeked at 34c and the background was red. 

It turns out we had some saharan sand signals in the sky when I was imaging. Explains why my car looks like I've been in the Dakar rally. 

Thanks

Danny

Hi Danny,

PSU. Appears to be OK and it's a good price so not the end of the world if it turns out not to do the job.

You should indeed be able to use it to adjust the peltier heat pump to an optimal value.

here's a link to a more detaied description. https://www.aliexpress.com/item/4000197872662.html

 

[Enceladus Dan]

2 minutes ago, VectorQuantity said:

Hi Danny,

PSU. Appears to be OK and it's a good price so not the end of the world if it turns out not to do the job.

You should indeed be able to use it to adjust the peltier heat pump to an optimal value.

here's a link to a more detaied description. https://www.aliexpress.com/item/4000197872662.html

 

Thanks, good to hear. I was hoping that’s the case to fine tweak power input to suit the need, rather than just run it on more power and waste energy maybe and have to use temp control switch as read its best not to keep switch on and off. PWM was ok but can use low cycle so thought I would avoid both. 
 

thanks for the help. Hopefully have a picture and some successful temps to show soon. 
danny 

 

[VectorQuantity]

3 minutes ago, Enceladus Dan said:

Not sure but I have been using a dummy battery on a 12v3a ac adaptor till now. Just having made a 12v power box so I can go off grid to a bortle 4 site. I could put my multimeter inline and measure current used while doing some test long exposures. Then Average them out. 
 

I did find a post with someone who did lots of tests with a 700d and 30s Timelapse with screen off is 1w of usage. 

If it's only 1W then that's great, your peltier should have no problems.

[Enceladus Dan]

7 hours ago, VectorQuantity said:

If it's only 1W then that's great, your peltier should have no problems.

I have just done 4x 300s darks indoors at ambient 18c. For ease I measured amps at the 12v input to the dummy battery lead which has a 12-24v to 8.4v convertor. The dc jack after the converter is so small and goes in the side of the battery. 

My battery was at 12.4v as not charged since Friday. The 4 subs drew 0.169, 0.168, 0.167 and 0.167A seems the hotter the less power which makes sense. So average power is 2.08w but this includes power used to step down. Under no load (cam off) it was using 0.01A /0.02w but I guess the more load more wastage. Guessing efficiency is no better than 80% camera draws 1.6w. If it's lower I guess it will equal out as with electronics cooler will use more. 

Still 2w or lower still good. I used 4w as a worst case in first calculations. 

[Enceladus Dan]

Quick update. 

I have the parts and have constructed the cooler box. I ran a quick test, although I have not assembled the project box with buck/boost module or the temp controller. I just soldered some wires so I can test it first incase I need changes. 

1st it did cool, but not as I hoped. 26c ambient and 18c at the heatsink inside the box. The arctic alpine 12 co fan is not very good. It's so slow and doesn't seem to be moving much air. The outside heatsink was hot to the touch. So next step is a better fan and some frankinstien attachment. I have an old tower cooler in the loft with a 92mm fan (same as the arctic, but should be faster) 

2nd. The buck/boost module can be set with both voltage and current limits. But it limited the volts to keep current (as I thought but was hoping some kind of magic was happening inside) but I did see the volts increase as the peltier was working ( guessing the resistance change with temp change) 

So now it's a fan exchange, then a different heatsink if I need to. 

Regards

Danny. 

[Enceladus Dan]

The prototype mk2 is finished. With a 100mm x 70mm aluminium heatsink and a 120mm computer fan . I made a bracket that pulls the heatsink down onto the holes for the Arctic Alpine holes (had to elongate a bit) and also holds to fan onto the heatsink. 

first test with 2A at ambient 22.3c gave me a drop of about 14c to 8.5c

then at 2.5A it dropped fast but from 9.6c it then heated, so at 2.5A the heatsink cant disapate the heat fast enough   

Finally at 2.25A at ambient 21.5c it dropped about 16c to 5.7c. 

all tests were indoors at night, with no camera inside just the probe taking inside air temp. The box is not yet complete as i need to seal around the nosepiece (i just plugged it for the test) and i need to make sure the lid seals good. 

all in all im much happier with a 16c drop in ambient temp. It Might improve with  a better design at the heatsink than my bodge up from using old holes. With the added weight of a bigger heatsink and fan i decided to put the project box with controlls on a long lead and not on the back of the box. I might even take out the temp control for a simple temp display, as i was able to tweak the temp drop with current limit. 

thanks

Danny

Edited August 19, 2022 by Enceladus Dan

[Enceladus Dan]

Test Results are in, the box isn't complete and I have further issues. I have made a handset on a 1m cable as that was length of temp sensor cable (will extend when finally finished). now my buck/boost module overheats, so I need a small fan on the side of the handset. with the box empty at 22c I can get the inside air to 4.9c. I still haven't made a seal for the lid yet, this is because I realised I never left space for the data cable between the camera and the wall of the box. now I'm planning to make a new box. while I had this one almost finished, I thought I would test it while taking darks. 

when setting up the ambient temp was 15c and the box reached a low of 2.4c, by the time I started taking darks the temp had risen to 20c and after my tests it was 22c. I set Magic lantern to take a 300s dark frame every 10 mins. I tested if having the LCD screen open helps by having it open as my intended plan, then quickly open the box and close it, then check the data after further cooling. results as below. 

peltier was running at a current limit of 2.2A variable voltage between 7.5-8.5v

Time from cooling started (Sub Started)                           Box Temp at start and end of sub                    Camera Temp from EXIF Data

10mins                                                                               13c / 11.5c                                                         22c

20mins                                                                               10c / 9.4c                                                           19c

30mins                                                                               8.8c / 8.5c                                                          17c

40mins                                                                               8.1/ 7.9c                                                             15c

50mins                                                                               7.6c / 7.4c                                                           14c

60mins                                                                               7.4c / 7.3c                                                           13c

70mins                                                                               7.3c / 7.1c                                                           12c

Screen Closed

80mins                                                                               8.6c / 8c                                                              12c

90mins                                                                               7.8c / 7.8c                                                           13c

100mins                                                                             7.7c / 7.6c                                                           13c

110mins                                                                             7.6c / 7.5c                                                           13c

120mins                                                                             7.5c / 7.6c                                                           13c

130mins                                                                             7.6c / 7.6c                                                           13c

140 mins                                                                            7.6c / 7.7c                                                           13c

Dark Frame comparisons, first are the frames which lead me to want to make this box. on the 7th of July during a heat wave I tried to image the veil, the ambient was 20c at midnight and by 3am when I started my darks it was still 17c. this is my first and last dark frame from that night. the Dark frames when viewed on the DSLR or my PC photo app don't look as red, this is how they open in GIMP. not sure if it is how the data actually looks and other viewer's are applying something to reduce noise. (no idea) . this is a crop of the bottom left corner of each frame (worst corner) 

these are the frames from the cooling test, edited in exactly the same way as above in GIMP, no changes just the same crop of the bottom left corner. this is the 10min, 70min and 140min frames listed above. 

while not completely free of noise they are remarkably better than non cooled. if the night is cooler than 20c it can only be better than this. now its time to order more Aluminium and start the Mk2. 

Regards

Danny. 

 

Edited August 26, 2022 by Enceladus Dan