Brent, I'm sorry if this seems a bit pedantic but with peltiers, the process of cooling (as I understand it) is not to absorb the heat from inside the enclosure and direct it out the back. The idea, as I understand it, is that the bottom side of the peltier device gets cold and the trick is to push that cold into the enclosure via a heat sink/fan on that side. A very different process. And since the peltier produces more heat energy than cold as part of that process (because of the power/enegry used), there is more heat to be blown away by the external fan. So the internal fan, heatsink etc is always smaller in proportion to the out(hot)side one.
So my idea is to generate as much cold air as I can from the peltier and the heat sink in contact with its cold side and push that via the internal heat sink and fan onto the back of the camera.
That's how I understood the workings of TEC units Peter.
Just revisiting post from Rowland below, there seems to be a case for the TEC to be attached directly to a larger conductive surface rather than to have a smaller fan pushing the cooled air around, provided that conductive surface is conforms to the body contours of the camera to get even cooling.
Therefore, I wonder if we may get a better cooler box by building a copper or aluminium "armour" (rather than a box) around the camera body and attach the TEC unit directly to the large surface area of the armour. This would provide more efficient cooling than a fan pushing the cold air around inside a box. The armour can then be encased in a box filled with foam like the link provided by John K below with an external fan extracting the hot air from the reverse side of the TEC unit.
What do you think?
Bo
That's how I understood the workings of TEC units Peter.
Just revisiting post from Rowland below, there seems to be a case for the TEC to be attached directly to a larger conductive surface rather than to have a smaller fan pushing the cooled air around, provided that conductive surface is conforms to the body contours of the camera to get even cooling.
Therefore, I wonder if we may get a better cooler box by building a copper or aluminium "armour" (rather than a box) around the camera body and attach the TEC unit directly to the large surface area of the armour. This would provide more efficient cooling than a fan pushing the cold air around inside a box. The armour can then be encased in a box filled with foam like the link provided by John K below with an external fan extracting the hot air from the reverse side of the TEC unit.
What do you think?
Bo
I'll throw in my 2 cents worth. I agree that conductance via metal foil wrapping is likely the most efficient means for heat/cold transfer to where you need it without worrying about cooling the air and then moving it within a chamber. It may also reduce the risk of icing at the peltier because of the large surface area created by the foil.
Surrounding the foil in foam insulation would allow you to make the package smaller or indeed use more insulation for the same size package.
Attaching the cold side of the sensor to the foil armour effectively would be really important.
I would have doubts about foil being part of any robust long term solution, can someone point to where it has been used successfully?
Re the cooled base idea, if you were to attach the camera to a broad cold finger platform that is cooled by the TEC that should cool the camera body pretty effectively. The copper/alumnium cold finger platform that I have from a laptop can be bent to serve as a base for the camera - screwed to it via the tripod mount. A box of thin foam can then surround for insulation. It may not be as effectve as an internal cold finger on the sensor but you would not have to mod the camera in any way.
Just to clarify, I was not referring to covering the DSLR body in foil (although it may prevent aliens taking control of my camera ), I am suggesting to cover the camera body with a copper/alu sheet that's been cut and formed around the camera body.
I have sourced a piece of copper (0.3mm thick) and I am in the process of making a template using thick paper before transferring it to copper.
Will upload photos later.
Bo
Brent, I'm sorry if this seems a bit pedantic but with peltiers, the process of cooling (as I understand it) is not to absorb the heat from inside the enclosure and direct it out the back. The idea, as I understand it, is that the bottom side of the peltier device gets cold and the trick is to push that cold into the enclosure via a heat sink/fan on that side. A very different process. And since the peltier produces more heat energy than cold as part of that process (because of the power/enegry used), there is more heat to be blown away by the external fan. So the internal fan, heatsink etc is always smaller in proportion to the out(hot)side one.
So my idea is to generate as much cold air as I can from the peltier and the heat sink in contact with its cold side and push that via the internal heat sink and fan onto the back of the camera.
I hope I don't have this concept wrong.
Peter
Peter,
Which ever way you look at it its all about shifting energy from one place to another. Heat is energy, BTU's. My internal heat sink is chilled by the TEC and internal air is blown across it to chill the air and thus the camera. The BTU's from the internal air being circulated is transferred by the TEC to the outside Radiator element ( inside or outside, they are both heat sinks just working in opposite phases.)
In Physics you don't say you push cold, what you are doing is removing energy thus lowering it's internal temperature.
My system works as you describe or any other way for that matter. It's just not very efficient, that is all. A cold finger or a metal to metal contact via the base plus some better insulation and a bigger heat sink inside to make the Peltier more efficient is the answer. Trying to cool down a well protected sensor inside a plastic bodied camera is never going to be an easy job. A bigger heat sink inside will allow more heat transference by the TEC and therefore a greater colder air volume to circulate. The outside radiator on that version is way oversize for the tiny inside fan and unit. Worked much better on the AS! 120. It was just that I'd already modified it to mount together for a previous version and it was easier to use.
I'll keep looking for a better enclosure after Xmas when the big electrical companies are back in action. Roll on Mk III version ...
Just to clarify, I was not referring to covering the DSLR body in foil (although it may prevent aliens taking control of my camera ), I am suggesting to cover the camera body with a copper/alu sheet that's been cut and formed around the camera body.
I have sourced a piece of copper (0.3mm thick) and I am in the process of making a template using thick paper before transferring it to copper.
Will upload photos later.
Bo
I'll give this idea a go next, I have plenty of copper downstairs. May even wrap the copper up past the rear screen. I wonder if Alum angle will be usable, I have some 50mm x 4 mm thick alum extrusion.
Mk III A coming up
Still chasing the original concept, and onto design #2: no peltiers arrived yet but what the heck!
Here is the tub liner model - in the first shot, it's shown assembled and in situ - but minus outside insulation.
In the remainder, you'll see the inside lining, the lid lining and the outside lining as well. All of it has come from one of those car-fridge things.
The hole in the back is sized to take the TEC I have on order from the states. The chinese one is a bit bigger but had a 60W rated peltier. If I have insufficient cooling with the 15 Watt one I'll swap them around.
Just to clarify, I was not referring to covering the DSLR body in foil (although it may prevent aliens taking control of my camera ), I am suggesting to cover the camera body with a copper/alu sheet that's been cut and formed around the camera body.
I have sourced a piece of copper (0.3mm thick) and I am in the process of making a template using thick paper before transferring it to copper.
Will upload photos later.
Bo
I think you just want to cool the camera in the box by circulating cooled around it or by conducting energy by brute force, per my diagram - though uneven cooling might be a problem. Adding more metal inside will diminish the TEC performance overall. TEC performance is basically a matter of moving as much heat as possible from the hot side irrespective of how cooling is performed on the inside. Ambient /Fan exhaust deltaT <10C is good.
EDIT: If the hotside heat sink/fan combo is too small/inefficient, best TEC performance occurs at a lower voltage/current than for a more efficient combination. In this case adding more energy further reduces the cooling performance of the TEC. This needs to be resolved before considering how to obtain thermal performance on the inside. Heatsink, fan and good clean flat junctions will lower thermal resistance, by comparison, and improve the chance of better internal cooling
Thanks Rowland,
So just to summarise, there is no advantage in having a large conductive surface to the cold side of the TEC, as it would only decrease the efficiency of the module.
Therefore, a copper strip size of a camera base plate (similar to your drawing below) would be sufficient and more effective than a large copper sheet covering the lower half of camera body (which was what I planned).
I will also need a lower current than the one rated on the TEC and a slightly larger fan over the hot surface to increase the efficiency.
Thanks,
Bo
Thanks Rowland,
So just to summarise, there is no advantage in having a large conductive surface to the cold side of the TEC, as it would only decrease the efficiency of the module.
Therefore, a copper strip size of a camera base plate (similar to your drawing below) would be sufficient and more effective than a large copper sheet covering the lower half of camera body (which was what I planned).
I will also need a lower current than the one rated on the TEC and a slightly larger fan over the hot surface to increase the efficiency.
Thanks,
Bo
Hi Bo
I guess whatever additional metal you add inside the box must be cooled. A plate to the bottom of the camera is merely a way of covering the area as evenly as possible, but the size of the TEC will determine this as well, because all the surface of the TEC should be in contact with whatever medium you use.
The hot side heatsink/fan combo is more complex as it must dissipate the energy produced by the TEC and the energy required to drive the TEC. Theoretically, if your TEC is rated at 60w the heatsink should be capable of dissipating 120w (rule of thumb, ignoring inefficiences). This will get you the magic ambient/fan exhaust <10C deltaT - best efficiency.
If the hs/fan is only capable of 80w (for example) running the TEC at 60w will add energy that cannot be dissipated - heating the system. In this case, running the TEC below rated produces best efficiency. Instead of 12v you might run the TEC at 9 - 10v - in any case, the result is less cooling. End of day, hs/fan/TEC ratings should be complimentary. If not play, around with the TEC voltage to get the <10C ambient/fan exhaust deltaT, if this produces adequate cooling.
It gets more complex because cooling energy requirements are not linear, particularly over the last 5C deltaT. I have spent considerable time plotting various cooling combo's to arrive at a sensible compromise in performance, weight and size. My system is cooling the sensor, not the whole camera - although conduction is unavoidable and the camera gets cold.
Make sure your camera electronics are kept free of condensation using the cold plate design. In some ways this is more complex than building a dry air freezer box. The cold plate gets very very cold and you dont want condensation in the bottom PCB near the camera base plate - or anywhere else for that matter.
There is an optimal ratio of internal heat sink to external radiator. My heat sink inside is too small to be effective it can only cool the air that passes over it and in this case that is not enough to be effective and efficient. Therefore no matter what voltage I apply and how cold the chilled TEC surface gets the sink cannot collect enough BTUs to really make the big external radiator system work.
For air chilling the internal HS needs to be pretty big actually, more surface contact with the airflow, more BTU's sucked up to be moved away.
I've just built an external base chiller using 50 x 50 Alum Angle and insulated it with Styrofoam. The Angle goes up the back of the 450D with a wee gap and is a sealed cavity and the base is also insulated plus I've used silica heat grease to even get some transfer through the plastic bottom surface. I even extended the foam up the left side and sealed any gaps to minimise any air exposed chilled surfaces.
It works but it is slow and it only got 4* cooler than ambient after nearly 60 mins of 12volts. There was dew dripping off the rear exposed alum plate and the plate was very cold. It is just that the minimal contact are only allows minimal heat transfer. The big radiator on the back was barely warmed, not enough coming through to make it work.
A hermetically sealed, well insulated enclosure with a big internal heat sink and fan as per Gary Honiss's build seems the best option.
Does anyone have any idea roughly how much heat (in Watts) a DSLR will generate over a long session. Assumptions:
Using an external power supply not battery
Saving to PC not internally;
Monitor turned off.
The context of the question is that I know the cooling capacity of my peltier assembly. With the assembly I have on order from the States, if I require a DeltaT of say -15C and with an external/ambient of 25C, I would have about 9 watts of cooling capacity available. So if a DSLR produces say 5 watts (a guess entirely), I have about 4 watts of cooling capacity.
There is an optimal ratio of internal heat sink to external radiator. My heat sink inside is too small to be effective it can only cool the air that passes over it and in this case that is not enough to be effective and efficient. Therefore no matter what voltage I apply and how cold the chilled TEC surface gets the sink cannot collect enough BTUs to really make the big external radiator system work.
For air chilling the internal HS needs to be pretty big actually, more surface contact with the airflow, more BTU's sucked up to be moved away.
I've just built an external base chiller using 50 x 50 Alum Angle and insulated it with Styrofoam. The Angle goes up the back of the 450D with a wee gap and is a sealed cavity and the base is also insulated plus I've used silica heat grease to even get some transfer through the plastic bottom surface. I even extended the foam up the left side and sealed any gaps to minimise any air exposed chilled surfaces.
It works but it is slow and it only got 4* cooler than ambient after nearly 60 mins of 12volts. There was dew dripping off the rear exposed alum plate and the plate was very cold. It is just that the minimal contact are only allows minimal heat transfer. The big radiator on the back was barely warmed, not enough coming through to make it work.
A hermetically sealed, well insulated enclosure with a big internal heat sink and fan as per Gary Honiss's build seems the best option.
Brent.
I built something similar and it was readonably effective, but the camera and cold side assembly were insulated with a neoprene glove lined with foil. I used silica gel as a drying agent. A small 40mm fan circulated air through the camera - albeit slowly. I made a foot to cover the base of the camera and supercooled it. Hence my earlier suggestion. But the foot does need to be hard against the camera body. On second thoughts, leaving a gap was hypothetical wishful thinking. The camera must be isolated. Any exposed surface is just harder to cool, but a little is useful as a nucleus for condensation. Your prototype is too exposed.
Does anyone have any idea roughly how much heat (in Watts) a DSLR will generate over a long session. Assumptions:
Using an external power supply not battery
Saving to PC not internally;
Monitor turned off.
The context of the question is that I know the cooling capacity of my peltier assembly. With the assembly I have on order from the States, if I require a DeltaT of say -15C and with an external/ambient of 25C, I would have about 9 watts of cooling capacity available. So if a DSLR produces say 5 watts (a guess entirely), I have about 4 watts of cooling capacity.
Peter
Not sure Peter - the camera operation is probably negligable. But heating of the sensor on newer cameras is in the order of 30C and more. Some get outrageously warm. Dealing with the sensor only and ignoring warm up, what power is required to produce a steady ambient of +35C. Keeping in mind that older Canon DSLR sensors run only marginally warmer than ambient ~10C. Depends on camera model.
Can you approximate your wattage based on the following data - though cold finger?
Typical TEC running flat out at less than manufacturer ideal and listed power rating of 60watts, say ~50watts - temp differential from ambient -28C, with a sensor wanting to run 10C warmer than ambient. Therefore, a differential of -38C with reference to the sensor. You will need to do the maths.
This doesn't account for conductive losses, where the camera iself is cooled over time to around 5C. Over 20 minutes, say.
Not sure Peter - the camera operation is probably negligable. But heating of the sensor on newer cameras is in the order of 30C and more. Some get outrageously warm. Dealing with the sensor only and ignoring warm up, what power is required to produce a steady ambient of +35C. Keeping in mind that older Canon DSLR sensors run only marginally warmer than ambient ~10C. Depends on camera model.
Can you approximate your wattage based on the following data - though cold finger?
Typical TEC running flat out at less than manufacturer ideal and listed power rating of 60watts, say ~50watts - temp differential from ambient -28C, with a sensor wanting to run 10C warmer than ambient. Therefore, a differential of -38C with reference to the sensor. You will need to do the maths.
This doesn't account for conductive losses, where the camera iself is cooled over time to around 5C. Over 20 minutes, say.
The variables seem to be multiplying. Add to the mix that CMOS sensors generate a lot less heat than other types. I don't know of any way to monitor the actual sensor temp - at least not in a cooled box design. Any ideas on that?
Peter
Rowland your comment regarding older dslrs running cooler - what is that based on? Is there a list or reference somewhere? I'd like to know how hot my 450D gets.
Rowland your comment regarding older dslrs running cooler - what is that based on? Is there a list or reference somewhere? I'd like to know how hot my 450D gets.
Me too: but I have tried this experiment this morning:
Using a 350D (the CMOS from Canon), I used a block of polystyrene and made a plug to sit over the T-Ring and inserted a temperature probe through that and into the camera - as close to the mirror as I could without fouling it. That's the best I could do to get a reading of the internal temperature of the camera - as isolated as I could make it from the outside air.
I then ran a series of 50 x 30 second subs back-to-back (I only had a USB cable with me and not the DSUSB that is in the obs) and finally a series of open BULB exposures of 300 secs by manually holding down the shutter. Ambient was 21.75 give or take. The temperature inside the camera did not exceed 22.4 in either series of tests.
I concede I may not have isolated the probe from ambient as well as possible but even allowing for that, the internal temperature did not rise more than about 0.75C above it. So I'm inclined to think that internally generated heat will not be much more than 1C. That is encouraging for my cooling box plans.
I found a web page several years ago with sensor temp plots for long exposures, for the 1000D and 450D (I recall the 40D also). The 1000D is ever so friendly, the 450D is very similar, if slightly warmer. Peter, the 350D should be similar to the 1000D or 450D +10 - 15 at the outside, without cooling. I'm not sure about 1C though. You may be right.
I guess this can be tested with a cold finger mod by taking a series of 3 min darks with cooling off to measure the dark noise. Then set the cooling to maintain ambient and take another series and measure the noise. Dark noise increases 2x / 6C.
There are number of threads on sites such as cloudy nights about heating of various sensors. Some, such as the 5DMKII run +35C warmer, others are reported to be worse but I forget exactly. It seems to depend largely on the model. I found the 5DMKII, when I had one to be a beast.
than a fan pushing the cold air around inside a box. The armour can then be encased in a box filled with foam like the link provided by John K below with an external fan extracting the hot air from the reverse side of the TEC unit.
), I am suggesting to cover the camera body with a copper/alu sheet that's been cut and formed around the camera body.
