Everyone doing peltiers, so here´s my next version.
Had a 4x40w peltier earlier that actually worked in my Swedish cold climate.
This one is 4x80w
But, what if we water cool the hot side too?
Well thats what i´m trying to find out, if i can grab some more cooling efficiency out of it
Setup will have three loops, one to cool the hot side with a 240mm radiator, and the other loop will go from the peltier down to the lasers coolant tank. And the third from the coolant tank trough the laser. Totally 3 pumps.
Do not want to hook it up directly to the tube just yet
Will do a couple of tests
One block with air cooled heatsinks on the hot side
Three blocks with water cooled hot side
And then with and without running the machine.
Just have to make a clamp to hold everything together, tried with cable ties but they wasn´t enough. Shoot a idea if you have one on the clamping situation
With radiator cooling the hot side won’t get lower than ambient though will it? My problem is ambient is now in the 90F.
Not clear in the picture but the middle plates are intended to be threaded side rails the top plates bolts screw into. Pulls from top to middle and bottom to middle.
@donkjr Did some quick calculations, a water block can carry the heat away from the peltier quicker, than a heatsink with a fan can do. Both cooled with ambient air though and it will never get below ambient correct, but it´s how quick i can get there and how close to ambient i can go
It depends on the heat sink too of course, and it´s ability to pick up the heat. Same reason why water cooling is popular at computers over air cooling, it´s more efficient. Both still using ambient air.
That got me thinking if i watercool it, will i gain anything worth doing it this way
If i can get the temp on the hot side down closer to ambient, i would gain on the cold side too.
Just a crazy idea yet, and i love ideas, there´s where facts are born
Plate with bolts tightening it would work, great pic, got my ideas flowing again
@Steve_Clark@HP_Persson looks like it would be better to stack modules horizontally across a heat exchanger vs vertically in a stack.
Use longer exhangers vs smaller stacked?
From what I’ve read stacking is only if you want to increase you cold to heat ratio a lot and way below what we need. Efficiency wise, apparently if one wants to only reduce the ambient temperature say between 10 or 30 degrees it’s better to run them parallel and in series.
Oh…before I forget…are you using thermoconductive grease between your layers?
The tests begins now.
Went with this clamp-style from @donkjr idea earlier.
Any suggestion on data to keep track of?
Right now i´m measuring coolant temp, hot side temp, ambient temp.
And i will time cool down and heat up times when machine is running aswell.
Got some more TEC´s, so after this i´ll try some stacking/cascading too
Had to rebuild a old computer PSU to keep up with the power
@donkjr Smart, i got a arduino somwhere around with a flow meter on. I´ll hook it up!
Nice to see if increased flow really give any better cooling too, or at what flow it´s not helping.
First test was back-asswards, had the peltiers flipped the wrong way
But, here is the first test results.
4x80w peltiers, connected to a 750w PSU at 12V.
Pulled about 34A continuously
40x80mm water blocks with four TEC 12708.
How i tested it
27.5c ambient, identical pumps in each tank.
All water was 27C when the test started.
Temperature of the hot and cold side measured in the tank.
Test 1
Hot side on 2 water blocks - 10L tank, 240mm radiator
Cold side on 1 water block - 10L tank
After 30min - cold side temp: 16.2c
Test 2
Hot side on 1 water block - 10L tank, 240mm radiator
Cold side on 2 water blocks - 10L tank
After 30min - cold side temp: 12.7c
Bonus test
2x60w peltiers, one 40x40mm water block. Hot side air cooled
After 30min - cold side temp: 22.4c
