Author Topic: Non-conductive non-corrosive liquids  (Read 6212 times)

0 Members and 1 Guest are viewing this topic.

Offline jmoschetti45Topic starter

  • Contributor
  • Posts: 16
  • Country: us
Non-conductive non-corrosive liquids
« on: September 25, 2017, 05:59:24 pm »
Due to temperature, mineral oil is out of the question for this one. Project will be exposed to a max of -55F.

So I'm looking for another liquid coolant that a raw pcb can be submerged in. No radiators or pumps. Due to size, the coolant will have to work via convection inside the finned box. Less than ideal, but that's the only option I have.

I've heard pure alcohol works, but I have a feeling that'd eat up the board over time.

Flammability isn't much of an issue here, neither is toxicity, metal box outside away from other things. If the box didn't have to be sealed and occasionally submerged, I'd use a fan, but that's out of the question.

Suggestions?
 

Online wraper

  • Supporter
  • ****
  • Posts: 16802
  • Country: lv
Re: Non-conductive non-corrosive liquids
« Reply #1 on: September 25, 2017, 06:11:57 pm »
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16272
  • Country: za
Re: Non-conductive non-corrosive liquids
« Reply #2 on: September 25, 2017, 06:20:49 pm »
If using Galden or any other perflourocarbon you cannot use electrolytic capacitors with a rubber bung, as this will swell with time. Same for wire, the only wire stable long term is PTFE insulated, or some Kynar wires, and you will have to avoid any ABS or other susceptible plastics, as they will eventually degrade with time or leach out into the fluid.

What is wrong with mineral oils, you can get them with wax temperatures below -55C, though in that case they tend to be more like JP1 rocket fuel with a similar viscosity to water. Same caveats as to compatibility as well, but at least the JP1 will be cheaper as well, though will wreak havoc if you have any controlled impedances on the board or fast signals ( will increase loss and capacitance), or just use regular Jet A1, which will at least not be solid at -55C, though it will probably be a softish wax till it melts, which can be an advantage as it will sink heat nicely in the phase change. You will need to have a dewatering agent in there though at the bottom, as it will have some water in it, though boiling it and getting the water out will help, but you will want a nice big pack of water absorbing molecular sieve in there as well at the low point of the tank.
 

Online wraper

  • Supporter
  • ****
  • Posts: 16802
  • Country: lv
Re: Non-conductive non-corrosive liquids
« Reply #3 on: September 25, 2017, 06:33:30 pm »
If using Galden or any other perflourocarbon you cannot use electrolytic capacitors with a rubber bung, as this will swell with time. Same for wire, the only wire stable long term is PTFE insulated, or some Kynar wires, and you will have to avoid any ABS or other susceptible plastics, as they will eventually degrade with time or leach out into the fluid.
Where this information is taken from? As much as I see for second link I provided:
Not react with metals, plastics, rubber, and most of materials
 

Offline Marco

  • Super Contributor
  • ***
  • Posts: 6694
  • Country: nl
Re: Non-conductive non-corrosive liquids
« Reply #4 on: September 25, 2017, 06:58:19 pm »
If you put a heatsink on whatever is creating the heat might the melting oil not sufficiently cool it any way until the temperature gets high enough to start convection?
 

Offline nanofrog

  • Super Contributor
  • ***
  • Posts: 5446
  • Country: us
Re: Non-conductive non-corrosive liquids
« Reply #5 on: September 25, 2017, 07:13:43 pm »
An Overview of Liquid Coolants for Electronics Cooling might be of interest.  ;)

From what you've provided, I'd go with silicone oil (i.e. Syltherm XLT). It's relatively inexpensive, and will remain a liquid at -55F (down to -100C actually). Look for CAS 63148-62-9 to get it in smaller quantities (Syltherm only comes in drums IIRC). It comes in various viscosities, but you'll want it on the thin side; say 100cSt (centistokes).
 

Offline jmoschetti45Topic starter

  • Contributor
  • Posts: 16
  • Country: us
Re: Non-conductive non-corrosive liquids
« Reply #6 on: September 25, 2017, 10:33:24 pm »
Leaning towards Jet-A...especially since I have some. I'll look at silicone oils though.

Thanks!
 

Offline chuckb

  • Frequent Contributor
  • **
  • Posts: 339
  • Country: us
Re: Non-conductive non-corrosive liquids
« Reply #7 on: September 25, 2017, 11:12:03 pm »
Jet A has a freezing point of -40C, Jet A-1 has been the main Jet fuel since the early 70's, if I remember correctly. Jet A-1 has a -47C freezing point but at -40C my experience is it flows like jelly. Also Jet A-1 has a conductive additive so the flowing liquid does not build a static charge and cause electrical arcs. These arcs were the cause of a few aircraft fueling accidents in the 60's.
Jet B (impossible to find anymore) has a lower freezing point (-60C). We needed to do an aircraft test once with Jet B so we had to talk to a refinery to produce a few 55 gallon drums of Jet B for us.

I would go with the silicone oil discribed below.

Good luck!



 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21609
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Non-conductive non-corrosive liquids
« Reply #8 on: September 25, 2017, 11:23:12 pm »
Light mineral oil is basically petroleum ether (ligroin), although that's a bit too light, closer to gasoline, and would probably have bigger problems with compatibility.  There ought to be something inbetween ("odorless kerosene"?), but I don't know offhand.

PFC and silicone would seem to be the usual suspects.  Also, if there are low temperature varieties of dielectric oil.

A solid assembly would be a lot less messy.  Thermal pads can be used to clamp boards between heatsinks.  Potting is excellent, if it never needs to be opened for repair.  It fills gaps and voids, especially useful for high voltage and vacuum applications.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Offline jmoschetti45Topic starter

  • Contributor
  • Posts: 16
  • Country: us
Re: Non-conductive non-corrosive liquids
« Reply #9 on: September 26, 2017, 04:33:51 pm »
I'm actually debating redesigning the whole thing at this point. The silicone or Jet-A would work, but its liquid and just gets messy. Shoving a 60A regulated supply into a tiny box is just hard
 

Offline duak

  • Super Contributor
  • ***
  • Posts: 1038
  • Country: ca
Re: Non-conductive non-corrosive liquids
« Reply #10 on: September 26, 2017, 07:22:44 pm »
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3233
  • Country: gb
Re: Non-conductive non-corrosive liquids
« Reply #11 on: September 29, 2017, 02:19:46 pm »
Due to temperature, mineral oil is out of the question for this one. Project will be exposed to a max of -55F.

Do you mean a minimum of -55F, or is it really going to be exposed to temperatures lower than that?
 

Offline tooki

  • Super Contributor
  • ***
  • Posts: 11342
  • Country: ch
Re: Non-conductive non-corrosive liquids
« Reply #12 on: September 29, 2017, 03:57:21 pm »
Due to temperature, mineral oil is out of the question for this one. Project will be exposed to a max of -55F.

So I'm looking for another liquid coolant that a raw pcb can be submerged in. No radiators or pumps. Due to size, the coolant will have to work via convection inside the finned box. Less than ideal, but that's the only option I have.

I've heard pure alcohol works, but I have a feeling that'd eat up the board over time.

Flammability isn't much of an issue here, neither is toxicity, metal box outside away from other things. If the box didn't have to be sealed and occasionally submerged, I'd use a fan, but that's out of the question.

Suggestions?
Maybe this is a dumb question, but why wouldn’t mineral oil work? I assume the objection is due to viscosity, but if this board produces so much heat that it needs liquid cooling, wouldn’t it thaw out the oils fast enough to be useful? I somehow can’t imagine that you could have a boiling hot core while the outside is still frozen.
 

Offline gibbled

  • Regular Contributor
  • *
  • Posts: 102
  • Country: ca
  • VE7 call sign
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf