Device that circulates mineral oil and cools it?

[Plasmateur]

I purchased an induction heater off of amazon. I want to push this thing kinda while keeping it cool.

I figured I could submerge the whole board in mineral oil, and circulate it so it would act as coolant. Has anyone ever done something like this? If so, are there some hobbiest options for this type of device?

[Nerull]

Look up mineral oil PCs. You can generally use water cooling radiators and aquarium/pond pumps.

[tsman]

The mineral oil will wick up cables so expect it to get everywhere if you're not careful.

[Plasmateur]

Look up mineral oil PCs. You can generally use water cooling radiators and aquarium/pond pumps.

The mineral oil will wick up cables so expect it to get everywhere if you're not careful.

Hey, thanks for the help! I was a little worried about using a water cooling radiator with mineral oil. I'll check into the PCs. And thanks for the wicking advice.

I have a feeling tupperware might be the enclosure. 

[Berni]

Yep generally stuff for water cooling can be used with mineral oil.

Tho you will find out how big of a mess this stuff can be as it gets everywhere and is impossible to simply wipe off. The cooling properties of mineral oil are also not all that impressive. Its specific heat capacity is quite a bit smaller than water and its considerably more viscous compared to water. This means if something is simply submerged then the oil next to hot surfaces will heat up and mostly stick to them, it needs to be forcefully moved across the surface just like air.

In practice oil submersion is mostly done for high voltage isolation, tho with large HV transformers it does also aid in cooling because the internal structure is designed with large gaps for oil to get trough and made to pick up strong convective currents. This helps move heat from deep inside the core to the outside walls. The large scales make conductive heat flow painfuly slow while enhancing convective effects due large mass of fluid and effects of viscosity diminishing with distance.

[Cyberdragon]

Wonder how good one of these would work as a cooling tower.

http://historysdumpster.blogspot.com/2014/01/rain-lamps.html

Sorry, not sorry. Had to do it.

[Nerull]

You do want to be careful about heat dissipation if you're running continuously. Mineral oil can soak up a lot of heat so people often declare success after they get a system running with seemly stable temps for an hour or so, and then leave the system running with insufficient cooling. 8 hours later, the system is cooking itself. Mineral oil is a heat transport, not a replacement for proper cooling. You need just as much cooling capacity as you would with air or water cooling.

[james_s]

Generally you cool the work coil by running plain old water through it. The electronics can be cooled with forced air up to >10kW provided they are designed accordingly.

[tautech]

Google transformer oils.
They are lightweight and suited for just convection circulation and have good dielectric strength.

[MagicSmoker]

OP wants to spend several hundred dollars on immersion cooling of a $30-$80 induction heater?

Upgrade the heatsinks, MOSFETs, add a fan, buy the next most powerful heater; anything besides immersion cooling, and definitely not with mineral oil. Fercryinoutloud that stuff only works as a decent heat transfer medium when there is a massive temperature differential, like going from windings at 180C in a distribution transformer to a 30C ambient (and using the thermosiphon effect to pump the fluid).

Oh, and mineral oil does, indeed, wick up stranded wire cables like a mo-fo.

[james_s]

I agree there, mineral oil is a huge pain, I've used it for small HV transformers and it's almost impossible to keep it contained. It wicks up inside wires and leaks out every conceivable place, it makes a sticky mess all over everything, once it's been immersed you pretty much have to just throw it all away if anything fails. I don't see any advantage of using it over forced air in this application.

[OM222O]

literally any water cooling loop would work just fine. don't use plastic tubing as it will get brittle and breaks.rigid tubes like copper/glass/etc are better options.

[james_s]

Plastic tubing works fine, metal is conductive and glass is brittle, it doesn't seem very practical to be running fragile glass tubes full of water around electronics. Cars and trucks use rubber hoses for cooling water without big issues so long as it is maintained.

[amyk]

OP wants to spend several hundred dollars on immersion cooling of a $30-$80 induction heater?

If he really wanted to, he could use Galden or other fluorinated heat transfer fluids... nonflammable and less viscous than mineral oil, but costs even more.

[Plasmateur]

Hey thanks everyone. It appears forced air is probably going to be the best way to go then. But then I'll probably have to mod the induction heater by switching out the caps.

The heater is here: https://www.amazon.com/Yosoo-Voltage-Induction-Heating-12V-48V/dp/B01C70G7Y8/ref=sr_1_4?keywords=induction+heater&qid=1555120176&s=gateway&sr=8-4

Those caps get super hot, and I don't feel like forced air is going to do the trick since I'm going to have to run this over some length of time.

[CatalinaWOW]

You can always up your forced air game by chilling the air.  TE Cooler or conventional air conditioning.  Delta T always talks on heat transfer.  Lots of cost relative to your heater, but far less trouble than oil or OMG fluorocarbon fluids.

[james_s]

You're expecting too much from a cheap Chinese toy. If the caps are getting too hot with forced air then they are being overloaded and will not last long no matter how you cool them. Either mod that unit or build your own from scratch using high quality film capacitors and appropriately beefy mosfets or IGBTs.

You can't polish a turd.

[coppercone2]

yes, if you look at pictures of professional induction heaters, they have MASSIVE banks of foil capacitors. There are some pictures on this forum, I mean like a (Fairly) small PCB connected to capacitors mounted to thick anodized aluminum plates (for mechanical stability AND current handling). Like red bull cans.

[Berni]

Hey thanks everyone. It appears forced air is probably going to be the best way to go then. But then I'll probably have to mod the induction heater by switching out the caps.

The heater is here: https://www.amazon.com/Yosoo-Voltage-Induction-Heating-12V-48V/dp/B01C70G7Y8/ref=sr_1_4?keywords=induction+heater&qid=1555120176&s=gateway&sr=8-4

Those caps get super hot, and I don't feel like forced air is going to do the trick since I'm going to have to run this over some length of time.

You can always get a 10 000rpm high power server fan. Those things can almost take off verticaly with how much air they push (Tho they tend to be 4 times the weight of a normal fan so they dont). Im pretty sure that would force quite bit of air between those caps.

And the caps used there are actually the good kind of caps for induction heating. The massive currents required trough them would cause a lot of typical foil caps to catch fire. You ether need more of these smaller caps to reduce the ESR even more or get a proper capacitor designed for induction heating. such caps are flat in construction and have massive copper terminals on each end to bolt onto a bus bar. One of those proper caps new likely costs about 10 times what your entire induction heater kit cost hence why the real deal capacitors are not used. But you can probably find one on the used market.

[OM222O]

Plastic tubing works fine, metal is conductive and glass is brittle, it doesn't seem very practical to be running fragile glass tubes full of water around electronics. Cars and trucks use rubber hoses for cooling water without big issues so long as it is maintained.

plastic reacts chemically with mineral oil and breaks down. there are purpose built rigid tubes such as copper and glass which you cut down to size for this application. I'm not quite sure what it has to do with copper being conductive or glass being too brittle? this is the standard method used for years in the PC industry

[coppercone2]

Plastic tubing works fine, metal is conductive and glass is brittle, it doesn't seem very practical to be running fragile glass tubes full of water around electronics. Cars and trucks use rubber hoses for cooling water without big issues so long as it is maintained.

plastic reacts chemically with mineral oil and breaks down. there are purpose built rigid tubes such as copper and glass which you cut down to size for this application. I'm not quite sure what it has to do with copper being conductive or glass being too brittle? this is the standard method used for years in the PC industry

look at CPU voltages vs induction heater voltages. you have to be more careful and use higher quality insulators and the failure mode will be worse. It's going to be very high current 170-300VDC @ 1000A in the tank vs like 3-12VDC @ much less A-DC (the processor does not use high voltages). And the PC supply might have a e-fuse or current limit or something. That tank is just going to blow something up immediately if there is a low impedance formed some where.

I think you will want to up the ante compared to cooling a computer unless its a piss weak low voltage heater.

[OM222O]

that's true but AFAIK mineral oil is a good insulator so unless there is something extremely wrong with the installation where the tubes are touching the supply, it should be fine? in case you want extra safety, it's super easy to just ground the tubes as well?

[james_s]

I was assuming using water, not mineral oil. We've already established that immersion in mineral oil is pointless for this thing. Just pump water through the work coil.