Biasing a submerged diode / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Biasing a submerged diode

1 2 next

Move Thread

LAN_403

Meatball

Mon Jul 22 2013, 06:44PM

Registered Member #2401 Joined: Mon Sept 28 2009, 04:25PM
Location:
Posts: 74

Ok. Lets run a thought experiment here before I ruin a perfectly good constant current supply.

Lets say I'd like a compact method of cooling 60W from an ordinary TO-247 power diode, as I run 50A of DC current through it.

Can I cool this device effectively without having to worry about a CC supply trying to force current across the submerged diode leads instead of through the diode junction? I don't want to deal with hydrogen and oxygen being produced. It would be neat to keep the thing in a little sealed container of water + some small air gap.

OR - will the leads underwater allow a current to bypass the diode? Is there a water additive that can prevent this?

What about other liquids that can soak up some heat in place of water?

Any other clever passive liquid cooling methods you guys know of?

Ash Small

Mon Jul 22 2013, 06:58PM

Registered Member #3414 Joined: Sun Nov 14 2010, 05:05PM
Location: UK
Posts: 4245

Most people here would try cheap cooking oil. A container large enough to allow convection currents to circulate should do the job (ie you won't need a lot)

Edit: it may be worth using a small aluminium heatsink as well. Place the fins vertically to allow for convection currents.

You can also drive off any moisture in the oil by heating in a saucepan.

Ben Solon

Mon Jul 22 2013, 08:49PM

Registered Member #3900 Joined: Thu May 19 2011, 08:28PM
Location:
Posts: 600

I've experimented with submerged cooling of switching devices as well. As a general rule, don't have the water as any part of the circuit. Standing water, even the cold stuff doesnâ€™t do much as convection currents are slower than in air. Unless done with careful thought the temperature will only rise near to the heatsink or tab.

I highly recommend water flow. Even slow flows are extremely efficient. Something as simple as a small length of copper pipe soldered to a copper plate where your device bolts to is able to dissipate that kind of power with no problem. Even so, just take simple thermodynamics into account when designing it and you'll have no problems.

Mattski

Tue Jul 23 2013, 05:39AM

Registered Member #1792 Joined: Fri Oct 31 2008, 08:12PM
Location: University of California
Posts: 527

Meatball, It's unavoidable that you'd electrolyse some of the water. When the diode is on and forward conducting the voltage is relatively low so most of the current will flow through the diode, but even with ~0.7V you're likely to see some electrolysis and H2 and O2 gas production. But when the diode is off and holding a high voltage it will leak more current through the water and electrolyze faster. Not good.

Deionized or distilled water would be best as it reduces the conductivity of water. But better still as Ash and ben say are oil cooling or traditional water cooling with a water cooled cold plate.

Patrick

Tue Jul 23 2013, 09:09AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

use oil.

food or petroleum, but please use oil. it saves so much trouble. Even DI water is dam near worthless.
Then a radiator, like the tiny transmission type, i have a thread somewhere here on the forum.

Daedronus

Tue Jul 23 2013, 10:16AM

Registered Member #2329 Joined: Tue Sept 01 2009, 08:25AM
Location:
Posts: 370

With 60A I would use something like this:

Houdini0118

Tue Jul 23 2013, 11:23AM

Registered Member #8558 Joined: Thu Dec 06 2012, 11:38PM
Location: usa
Posts: 57

DI water is very non conductive. Its used to cool the biggest of generators (on the order of 100's of MW) but they use a rather expensive system to keep it completely de-ionized and ideally it stays completely inside the windings but its still has to be a good insulator.

Shrad

Tue Jul 23 2013, 01:43PM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

coaxial water capacitors also use deionized water as a dielectric with one of the best strengths and the ability to self-heal in case of breakdown

there are coaxial cap based TE lasers which use this as the spark gap achieving the highest power for this kind of laser configuration so far

Patrick

Tue Jul 23 2013, 05:11PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

you might look at Helmholtz electrolyte type behavior, but oil is just easier and safer....

Meatball

Tue Jul 23 2013, 05:41PM

Registered Member #2401 Joined: Mon Sept 28 2009, 04:25PM
Location:
Posts: 74

So while this diode sits on this tiny heatsink (cooled by two centrifugal blowers), I could build a small aluminum encapsulation around it (since the height is not allowed to pass the heatsink fit height), weld some corner joints, fill it with oil, and then weld a cap on it.

This would be to absorb what the tiny heatsink and fans cannot keep up with. The oil would take some time to increase in temperature while absorbing heat, giving the HS and fans a chance to catch up until the diode stop conducting again.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob