-50C 3 stage pielter cooler
..., Sun Jul 23 2006, 05:42PMWell record high's last weekend, it hit a wooping 115.5F (46.3C) at my house, so I decided to make some cold...
I started with a 4 stage stack with a big huge pielter I have that is about 2" on a side, but it was putting out too much power into the heat sink I was using (about 4" by 12" with 3.8" fins on a .5" spacing...part of a telecom laser with a 4" fan blowing through it in the middle) and wasn't giving me any better temps, so I used only 3.
The first stage takes it down from 70f to about 30f, the second from 30F to somewhere arrounf -10Fish, and the last one from -10F to -64F (my max record). For those temps I put some fooam arround each layer to prevent ice forming and robbing all of my cold... Running about 1W into the top pielter (.5' by .25") 4w into the middle one (.5" .5') and 25w into the bottom one (1" by 1") for a grand total for 20w of power into the stack (which made my heat sink about 75F, so in theory I could have got considerably cooler if I put the bottom one in ice water...
But it seemd sorta boring to post a pic of a block of foam... So I decided to tke off the foam and freze something... Sice I had stepped on a ucc27321 the night before (pin side up
) it seemed like a good canadate...peter
Re: -50C 3 stage pielter cooler
Alex, Sun Jul 23 2006, 06:39PM
Neat, nice pics. Now try freezing a positively charged HV electrode and spraying water mist from a grounded container.
Alex, Sun Jul 23 2006, 06:39PM
Neat, nice pics. Now try freezing a positively charged HV electrode and spraying water mist from a grounded container.
Re: -50C 3 stage pielter cooler
Electroholic, Sun Jul 23 2006, 09:17PM
Cool idea, but somehow I don't think HV insulation and heat transfer is a good match.
I used to have a 4cm by 4cm 65W pielter, it was rated for 50C diff between the two sides.
so I put the hot side to a cpu heatsink and i reached -35C on the other side.
now if you put the heatsink into ice water, you should be able to get -50C.
Electroholic, Sun Jul 23 2006, 09:17PM
Alex wrote ...
Neat, nice pics. Now try freezing a positively charged HV electrode and spraying water mist from a grounded container.
Neat, nice pics. Now try freezing a positively charged HV electrode and spraying water mist from a grounded container.
Cool idea, but somehow I don't think HV insulation and heat transfer is a good match.
I used to have a 4cm by 4cm 65W pielter, it was rated for 50C diff between the two sides.
so I put the hot side to a cpu heatsink and i reached -35C on the other side.
now if you put the heatsink into ice water, you should be able to get -50C.
Re: -50C 3 stage pielter cooler
Avalanche, Sun Jul 23 2006, 09:23PM
How would this perform in reverse? ie by applying heat to generate electricity.
Avalanche, Sun Jul 23 2006, 09:23PM
How would this perform in reverse? ie by applying heat to generate electricity.
Re: -50C 3 stage pielter cooler
..., Sat Jul 29 2006, 10:33PM
Well I am glad to see that at least one of my projects got some attention (this is the first of my projects thread to get any comments about the project)... I took the first pic with my cannon powershot a85 that I got for free, on fixed focus mode at the highest optical zoom and focused as close as it could go (20cm) and then holding the camera in my hand I took a series of about 10 shots and took the one that was in focus
The other one was using the autofocus.
I did try adding some hv to the mix (why not?) by taking a 30kv hvac supply (out of a big plasma globe) that gives about 1/3" corona and pointing it at the forming ice crystals... Even from very far away (like 2") the crystals on the side of the tec closer to the supply would melt down... Didn't really get to far though since I didn't want to spend too long at it. It really just showed how delicate the crystals are; the ion wind from the supply (which I couldn't notice) managed to melt them
To make energy from this stack all you would need to do is heat up the top pielter. You would probably get better results with just a single stage (or at least stages that are all the same size)... If I turn off the middle stage with the other two going I got about 1v across it (I didn't check the current)... So don't expect to get a whole lot out of it
..., Sat Jul 29 2006, 10:33PM
Well I am glad to see that at least one of my projects got some attention (this is the first of my projects thread to get any comments about the project)... I took the first pic with my cannon powershot a85 that I got for free, on fixed focus mode at the highest optical zoom and focused as close as it could go (20cm) and then holding the camera in my hand I took a series of about 10 shots and took the one that was in focus
The other one was using the autofocus.I did try adding some hv to the mix (why not?) by taking a 30kv hvac supply (out of a big plasma globe) that gives about 1/3" corona and pointing it at the forming ice crystals... Even from very far away (like 2") the crystals on the side of the tec closer to the supply would melt down... Didn't really get to far though since I didn't want to spend too long at it. It really just showed how delicate the crystals are; the ion wind from the supply (which I couldn't notice) managed to melt them
To make energy from this stack all you would need to do is heat up the top pielter. You would probably get better results with just a single stage (or at least stages that are all the same size)... If I turn off the middle stage with the other two going I got about 1v across it (I didn't check the current)... So don't expect to get a whole lot out of it
Re: -50C 3 stage pielter cooler
Marko, Wed Aug 02 2006, 11:09PM
bump.
My olympus crap is famed for it's poor autofocus, and totally bad ability to focus onto close objects (I could never take a pic like your first
)
You made me into digging out my old big peltier and powering it up a bit.
It is on heatsink immersed under cold water, and made some nice snow tough temp wasn't much under zero.
I also froze some old SMD IC, led got a bit frosty but didn't freeze.
Later I cooled diodes wich showed increase in their voltage drop, and IRFP450mosfet wich had it's ON resistance fall to some 270 miliohms at few degrees over zero.
Maybe one could make efficient cryo-fullbridge that way and overrate the mosfets slightly while keeping them at some -30 degrees..
Marko, Wed Aug 02 2006, 11:09PM
bump.
My olympus crap is famed for it's poor autofocus, and totally bad ability to focus onto close objects (I could never take a pic like your first
)
You made me into digging out my old big peltier and powering it up a bit.
It is on heatsink immersed under cold water, and made some nice snow tough temp wasn't much under zero.
I also froze some old SMD IC, led got a bit frosty but didn't freeze.
Later I cooled diodes wich showed increase in their voltage drop, and IRFP450mosfet wich had it's ON resistance fall to some 270 miliohms at few degrees over zero.
Maybe one could make efficient cryo-fullbridge that way and overrate the mosfets slightly while keeping them at some -30 degrees..
Re: -50C 3 stage pielter cooler
..., Thu Aug 03 2006, 01:11AM
While it might be possible to keep a bridge at -30degres, I think the amount of power put into the pielters would kill any efficiency gains... I tried putting the whole stack (about 25w of energy) on a cpu sized cooler but even putting in 50w to the bottom pielter on that big ole heat sink I couldn't get the big pielter to freeze up
I suppose with a huge p4 heat sink with plenty of fans you might be able to get it arround freezing, but to get it below that you will need water cooling (which I happen to have all of the parts for, but haven't had time to try it out).
..., Thu Aug 03 2006, 01:11AM
While it might be possible to keep a bridge at -30degres, I think the amount of power put into the pielters would kill any efficiency gains... I tried putting the whole stack (about 25w of energy) on a cpu sized cooler but even putting in 50w to the bottom pielter on that big ole heat sink I couldn't get the big pielter to freeze up
I suppose with a huge p4 heat sink with plenty of fans you might be able to get it arround freezing, but to get it below that you will need water cooling (which I happen to have all of the parts for, but haven't had time to try it out).Re: -50C 3 stage pielter cooler
GimpyJoe, Thu Aug 03 2006, 01:42AM
How about putting the heatsink in LN2? *Cue TDU*
A quick thing to try would be using dry ice on the heatsink. You can buy it easily at a lot of grocery stores.
GimpyJoe, Thu Aug 03 2006, 01:42AM
How about putting the heatsink in LN2? *Cue TDU*
A quick thing to try would be using dry ice on the heatsink. You can buy it easily at a lot of grocery stores.
Re: -50C 3 stage pielter cooler
..., Thu Aug 03 2006, 03:09AM
I have acess to LN2 these days
(working at a semiconductor fab does have its advantages...)
I might try and see how well pielters work at really cold temps the next time we bring some home...
..., Thu Aug 03 2006, 03:09AM
I have acess to LN2 these days
(working at a semiconductor fab does have its advantages...)I might try and see how well pielters work at really cold temps the next time we bring some home...
Re: -50C 3 stage pielter cooler
dan, Thu Aug 03 2006, 04:29AM
Just remember the colder the hot side of the peltier gets the less efficient they are at moving heat. Also heat load is also a determining factor of the temperature difference. The max deltaT only happens under ideal conditions and under no load. Stacking peltiers is generally bad practice since each one down the line needs to pump more power, thus lowering their deltaT. Ideally you would just want one big ass peltier and jam as much juice through it as possible. But with TEC don't expect to get get past -40 or so with any decent efficiencies. I think once you get down to LN2 temps they stop working completely and actually at that point only generate heat.
LN2 is pretty cheap (but the storage dewar isn't.) and -196C is about as cold as you'll probably ever need. Unless your dealing with conventional superconductors then you'll need something way colder.
dan, Thu Aug 03 2006, 04:29AM
Just remember the colder the hot side of the peltier gets the less efficient they are at moving heat. Also heat load is also a determining factor of the temperature difference. The max deltaT only happens under ideal conditions and under no load. Stacking peltiers is generally bad practice since each one down the line needs to pump more power, thus lowering their deltaT. Ideally you would just want one big ass peltier and jam as much juice through it as possible. But with TEC don't expect to get get past -40 or so with any decent efficiencies. I think once you get down to LN2 temps they stop working completely and actually at that point only generate heat.
LN2 is pretty cheap (but the storage dewar isn't.) and -196C is about as cold as you'll probably ever need. Unless your dealing with conventional superconductors then you'll need something way colder.
Re: -50C 3 stage pielter cooler
Marko, Thu Aug 03 2006, 01:13PM
m, isn't peltier just going t be more efficient whille hot side is cooled more?
Marko, Thu Aug 03 2006, 01:13PM
Just remember the colder the hot side of the peltier gets the less efficient they are at moving heat.
m, isn't peltier just going t be more efficient whille hot side is cooled more?
Re: -50C 3 stage pielter cooler
Wilson, Thu Aug 03 2006, 01:46PM
Its more efficient when the hot side is cold, to a certain degree. After a certain point. peltiers just stop moving any significant amount of heat
On a side note, i can only get my dual peltier stack down to minus 10 degrees
Wilson, Thu Aug 03 2006, 01:46PM
Its more efficient when the hot side is cold, to a certain degree. After a certain point. peltiers just stop moving any significant amount of heat
On a side note, i can only get my dual peltier stack down to minus 10 degrees
Re: -50C 3 stage pielter cooler
Marko, Thu Aug 03 2006, 03:14PM
My single peltier on the pic was on heatsink in ice water (cooled in refrigerator) and I managed to get maybe few degrees under 0 (I didn't measure it precisely).
If I had some smaller peltiers to stack I could get to even lower tempšeratures..
Marko, Thu Aug 03 2006, 03:14PM
My single peltier on the pic was on heatsink in ice water (cooled in refrigerator) and I managed to get maybe few degrees under 0 (I didn't measure it precisely).
If I had some smaller peltiers to stack I could get to even lower tempšeratures..
Re: -50C 3 stage pielter cooler
..., Thu Aug 03 2006, 06:23PM
There is nothing really wrong with stacking pielters, you just have to make sure that the bottom is much larger than the top one.
For applications like this, it is actually much more efficieny to use a stack. Lets see you get a 130F diferential using 20w of power from a single peilter
..., Thu Aug 03 2006, 06:23PM
There is nothing really wrong with stacking pielters, you just have to make sure that the bottom is much larger than the top one.
For applications like this, it is actually much more efficieny to use a stack. Lets see you get a 130F diferential using 20w of power from a single peilter
Re: -50C 3 stage pielter cooler
dan, Thu Aug 03 2006, 07:11PM
It would seem to me that each layer of your peltier stack doesn't have a heat spreader to couple the smaller pelteirs to the larger ones. If that's the case then it's about the same as stacking 3 equal sized pelteirs. I can see a little silver thing over the bottom pelt but I can't tell if that’s thermal compound or a very thin heat spreader built into the pelteir. If it is a heat spreader plate it's too thin to do very much anyway.
With no/small load conditions it's ok to stack peltiers to a point; after which you start getting diminishing returns. But you usually want to actually cool an active heat source. This is when stacking pelteirs is bad. I have gotten close to a ~75C difference with one big 320W peltier being water cooled on the hot side to ~25C. The cold side was cold enough to keep a few drops of liquid propane from boiling at 1atm. ~-50C(no load). I had to feed it 30A at 16V though (480 watts!!).
Peltiers are fairly cool devices ( pun not intended) They have no moving parts and can provide many years of constant operation without problems or leaking refrigerant. They are also small for the temperatures they can reach.. I'd like to see someone fit a cascaded phase change system in their pocket.
(hmm seems to be something wrong with Google temperature conversion.. with -64F I get -550C!?!)
Edit: I see what I did wrong.. To get the correct conversion with Google you need to enter '(-64)F to C" not just '-64F to C'
dan, Thu Aug 03 2006, 07:11PM
It would seem to me that each layer of your peltier stack doesn't have a heat spreader to couple the smaller pelteirs to the larger ones. If that's the case then it's about the same as stacking 3 equal sized pelteirs. I can see a little silver thing over the bottom pelt but I can't tell if that’s thermal compound or a very thin heat spreader built into the pelteir. If it is a heat spreader plate it's too thin to do very much anyway.
With no/small load conditions it's ok to stack peltiers to a point; after which you start getting diminishing returns. But you usually want to actually cool an active heat source. This is when stacking pelteirs is bad. I have gotten close to a ~75C difference with one big 320W peltier being water cooled on the hot side to ~25C. The cold side was cold enough to keep a few drops of liquid propane from boiling at 1atm. ~-50C(no load). I had to feed it 30A at 16V though (480 watts!!).
Peltiers are fairly cool devices (
pun not intended) They have no moving parts and can provide many years of constant operation without problems or leaking refrigerant. They are also small for the temperatures they can reach.. I'd like to see someone fit a cascaded phase change system in their pocket.(hmm seems to be something wrong with Google temperature conversion.. with -64F I get -550C!?!)
Edit: I see what I did wrong.. To get the correct conversion with Google you need to enter '(-64)F to C" not just '-64F to C'
Re: -50C 3 stage pielter cooler
..., Fri Aug 04 2006, 07:05PM
I was just relying on the ceramic plate on the tec's as a heat spreader as I didn't have any sutable pieces laying arround. I probabyl could get considerably cooler with them, but the ceramic seems to have a suprisingly high thermal conductivity (I tried to solder on a wire for one that broke and I had to heat the whole thing up to soldering themp to melt the solder)
..., Fri Aug 04 2006, 07:05PM
I was just relying on the ceramic plate on the tec's as a heat spreader as I didn't have any sutable pieces laying arround. I probabyl could get considerably cooler with them, but the ceramic seems to have a suprisingly high thermal conductivity (I tried to solder on a wire for one that broke and I had to heat the whole thing up to soldering themp to melt the solder)
Print this page