| Pages:
1
2 |
Twospoons
International Hazard
Posts: 1282
Registered: 26-7-2004
Location: Middle Earth
Member Is Offline
Mood: A trace of hope...
|
|
And epoxy has a higher expansion coefficient than either aluminium or ceramic, by a factor of two.
Graphite powder would probably make a good thermal interface material in this case - its lubricating, highly conductive,, and already dry so it wont
thicken over time or with low temps.
And as with all thermal compounds, thinner is better. In some cases using a thicker layer is worse than using none at all.
Helicopter: "helico" -> spiral, "pter" -> with wings
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
Unless you have sealed modules (they are sold but significantly more expensive), or seal the modules yourself with a electronics grade silicone, I
would highly advise not to use anything electrically conductive as your thermal compound.
What happens is, as the heat ex changer cools and heats up, the peltier plate start rotating/walking, then you get exposed thermal interface
materials.
Your best bet is the silpad, which is expensive, but has a higher performance then grease, typically, because it won't dry out.
You also want to join the module strings together with flexible wire so there is some slack so they can rotate and move around a bit over the years,
otherwise they will get tension on the solder joints and either open circuit or short out the power supply to the heat sink. For this reason sealed
modules are better, because typically the seal will hold on to the wire, so even if the pad is ripped off, the wire will stay in the sealed area so it
can't take out your power supply with a short circuit (though you should have short circuit protection).
The other reason for sealed modules is because of condensation, the thing gets hot and cold and its difficult to prevent moisture condensation from
occurring, which can short things out.
The sealing process typically requires temperature controlled curing ovens.
What I recommend doing is either making (with silicone) or getting presealed modules, then sealing up the entire heat exchanger with a secondary
moisture seal.
The best way to do this is to get a gasket that fits between your heat exchanger and the heat sink, with a small cut away for wires, and to seal with
quality adhesive here.
You will want to use robust wires. Be sure to calculate how much wire you have inside of your heat exchanger, know its gauge, and calculate the
thermal dissipation that occurs in the wire. If you wire the peltier strings up for high current , the thicker wire you use the less thermal losses
there will be in the wire, this way your peltiers are not fighting the wire dissipation.
I would actually recommend high gauge high strand count silicone wire to connect the peltier modules together, as it is rather flexible, higher gauge
wires with other types of insulation may pull considerable tension on the pads of your peltiers, which can rip off rather easily.
If the sealing compound has difficulty adhering to your wire, simply place a sticky glue heat shrink (3M makes it) over the wire at the point where it
will be glued, the surface of the heat shrink material will be more adherant to the glue then certain wire insulation types.
|
|
|
CouchHatter
Hazard to Others
Posts: 146
Registered: 28-10-2017
Location: yes
Member Is Offline
|
|
As far as projects go, there are a few that haven't been mentioned that I would find useful. Not all strictly electronic. I found a bunch of these on
somebody's sig, can't remember...
UV lamp for TLC visualization
http://www.sciencemadness.org/talk/viewthread.php?tid=29250#...
Gas scrubber by BromicAcid
http://www.sciencemadness.org/talk/viewthread.php?tid=70003
Rotary evaporator
http://www.sciencemadness.org/talk/viewthread.php?tid=23139
Syringe pump
http://www.sciencemadness.org/talk/viewthread.php?tid=22713#...
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
Another idea similar to the syringe pump is a powder auger, though these are both motor controller projects..
|
|
|
DavidJR
National Hazard
Posts: 908
Registered: 1-1-2018
Location: Scotland
Member Is Offline
Mood: Tired
|
|
Quote: Originally posted by coppercone  |
The other reason for sealed modules is because of condensation, the thing gets hot and cold and its difficult to prevent moisture condensation from
occurring, which can short things out.
|
I have also heard of TECs being damaged by condensation freezing inside and cracking the ceramic plates.
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
This can most derinatly happen frozen water has immense force
Btw if you like peltiers a hack i figured out i think to save you money on custom components.
If you are handy with a dremel diamond saw and a soldering iron you can probobly cut yourself out a peltier zigaraut to make something like a tiny
multistage peltier without having to buy anything special.
You will have higher thermal resistance then a commercial one because the ceramic will be thicker since the joints will be twice as thick. Might be
good for prototyping though since you can make whatever you want. Keep in mind the peltiers are provobly lapped you need to maintain even tension on
the semiconductor doped bits
This is more useful for electronics... But you can make a multi stage peltier heat exchanger like described before to do something like chill alcohol
or maybe something else so long you got a gear pump to move dense fluid.. Ithink you can get like -20c but you will need to move alot of heat. I dont
think it will be practical to do more then like 100w without a liquid heat exchanger and a big radiator to dissipate heat from a water coolant loop.
But i suspect that if you use a decent heatsink, like the ones meant for cpu, and make a small diamond cut multistage peltier, and get your self
quality insulation, you can make a decent freezing point tester.
Things i can think about it:
See if glyoxylic acid you made is ready to solidify
Determine freezing point of p-xylene to see isometric purity..
[Edited on 16-5-2018 by coppercone]
[Edited on 16-5-2018 by coppercone]
|
|
|
stamasd
Hazard to Others
Posts: 133
Registered: 24-5-2018
Location: in the crosshairs
Member Is Offline
Mood: moody
|
|
Let me chime in with a couple of projects I've had on the backburner for a while now. One day though...
1. peristaltic pump with MCU control - useful for those reactions with variable addition rates over time
2. induction oven
|
|
|
Melgar
Anti-Spam Agent
Posts: 2004
Registered: 23-2-2010
Location: Connecticut
Member Is Offline
Mood: Estrified
|
|
| Quote: Originally posted by coppercone | another biggie for those peltiers is to provide proper torque to mate them with the heat sink for the thermal compound to work well.
|
I actually used an obsolete CPU fan for the one I made. It would easily scale up to multiple CPU fans, and the one I used could fit two 40x40mm ones
under it. If you need more cooling though, much higher than 60W will generate so much heat you won't be able to get any actual cooling.
Anyway, I used the spring mechanism from the clamp that would normally hold the fan onto a CPU. Then I used my Dremel to notch the underside of a
screw head, such that I could put a screw into either side of the fan, fit the two sides of the spring into the notch, then push down the lever to
apply pressure. The stuff between them was just regular heatsink silicone grease:
http://www.sciencemadness.org/talk/viewthread.php?tid=80597
It works well enough for 14/20 distillations of solvents that are liquid at room temperature. When you're running it, the coolant will eventually
reach approximately the boiling point of your solvent anyway, then the TECs will drop it below that somewhat. If running it in a loop with nothing
actually being cooled, condensation will form on the tubing.
The first step in the process of learning something is admitting that you don't know it already.
I'm givin' the spam shields max power at full warp, but they just dinna have the power! We're gonna have to evacuate to new forum software!
|
|
|
coppercone
Hazard to Others
Posts: 133
Registered: 5-5-2018
Member Is Offline
|
|
Yes but just technically if you look at the data sheet for your thermal grease, if its a decent thermal grease, not some fly by night chinese one, you
will have a specification that shows you thermal transfer vs surface pressure.
If you put bolts around the thing, you can ensure good surface pressure by using the correct torque on the bolts and calculate it. Tighten it like a
car rim.
The thing is though, sometimes its rather high, and if you have a large heatsink, you can actually cause deflection/bending in the heat sink, so you
might need to add special thick steel plates or such to prevent bowing.. if it bows you can crack the modules and thermal dissipation is uneven. Easy
to do if you are bonding water heat exchangers together, not sure how you would do it for air cooled heat sinks (I would recommend just making the
heat sink modules small enough that you can rely on bolts around the 4 corners providing adequate pressure without needing to reinforce things like
copper/aluminum.
I am actually working on an induction heater but the control system is a pain in the ass, the big problem is basically, even if you want to use a
external thermocouple or pyrometer to control it with really slow ramp rates (to eliminate difficult programming and tuning), you still need to make
some kind of fucking phased locked loop to detune it.. basically what can happen is if you put a metal of different conductivity into one, the
induction coil begins to look like it has shorted turns (i.e. aluminum), so then you are running the tank circuit at a higher frequency then the
resonance, so it can cause your transistors to explode violently (i have seen this on a induction stove when I put an aluminum pot on it by accident,
with a steel thing in it, pressure cooker while trying to make a pastrami)
so then you need electrical feedback from the isolated (i made a transformer for mine) side, to measure phase, so I need to figure out how to do that
reliably, I think a silicon laboratories isolator will do the trick but it basically put an annoying electrical feedback requirement on it. I did not
want to make a damn feedback circuit for something water cooled running at hundreds of amps, I was hoping I could just aim a pyrometer/IR sensor at
the pot and link em together to make a good high temperature solution with some analog electronics.
[Edited on 26-5-2018 by coppercone]
|
|
|
Texium
|
Thread Moved
27-11-2023 at 12:18 |
| Pages:
1
2 |
![[*]](images/xpblue/default_icon.gif){kind=icon}
