Heatsink as a "wire" for + DC bus / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Heatsink as a "wire" for + DC bus

1 2 next

Move Thread

LAN_403

Tonic

Wed Jun 25 2008, 06:54PM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

Now I'm trying to design half-bridge layout with the smallest inductance as possible. Transistors have separated heatsinks, so they don't have insulating pads. I'm wondering if there are any disadvantages in connecting filter capacitor directly to one heatsink that has upper MOSFET (where current from + is drawn) - as the metal backside of transistor is a drain.

Beside the fact that touching this heatsink is a big no-no

HV Enthusiast

Wed Jun 25 2008, 08:16PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

the problem is you have a thermal interface between transistor and heatsink. even when using a conductive (electrically) grafoil pad, you still don't have a good high current electrical connection.

also, it may not provide a low inductance connection depending on geometry of heatsink etc...

my advice - don't do it

Tonic

Fri Jun 27 2008, 08:04AM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

Right, I forgot about thermal interface. But let's guess few options:

1). There aren't any thermal interfaces (no pads, greases etc.) and cooling is sufficient - still a no-no for current flow?

2). If no, then what about grinding heatsink's surface with fine sand-papier (800 to 1000)?

3). If still no-no, then what about using heatsink as "support" wire? I mean, there's a one screw (which mounts PCB with capacitors on heatsinks) I would like to use for connecting filter capacitor's terminal with transistor's drain - it would be simplest way without fiddling with insulation. There you have pic to get idea what I'm babbling - also let me tell, if this geometry won't give problems with inductance:

Another question, a bit off topic, will this insulation sustain 325VDC? It's for insulating TO-220's tabs from heatsinks.

Insulationtube

Or perhaps I should make a better insulation?

HV Enthusiast

Fri Jun 27 2008, 10:52AM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

You won't get sufficient cooling without a thermal interface. The surfaces of both device and heatsink are not perfectly flat or smooth. This leads to a poor thermal impedance.

Its probably okay for low power devices or non-critical, but for your switching devices where cooling is essential, you need a very good thermal interface.

Tiberius

Fri Jun 27 2008, 02:02PM

Registered Member #1484 Joined: Wed May 14 2008, 03:24PM
Location: Cary, NC, USA
Posts: 27

Mercury is one of the best thermal interface materials I am aware of, and it is as I'm sure we're all aware also rather electrically conductive.

You should only use it on copper heatsinks, though, it will dissolve aluminum.

Myke

Fri Jun 27 2008, 03:21PM

Registered Member #540 Joined: Mon Feb 19 2007, 07:49PM
Location: MIT
Posts: 969

...but Mercury is very toxic...
I would not suggest you use Mercury as a thermal conductor because it's so toxic.

Tonic

Fri Jun 27 2008, 05:28PM

Registered Member #528 Joined: Fri Feb 16 2007, 10:32PM
Location: Warsaw, Poland
Posts: 166

Mercury is a pretty exotic solution ;) In any case, I'm going to grind heatsinks with sandpaper and use thermal grease for PC CPUs. You can see in photo that I'm using another heatsink to mount transistors and improve cooling. Also, there will be a small fan blowing air through those three heatsinks.

Still, two questions remained without answers. Will this screw insulation sustain 325VDC? Can I use heatsink as "support" wire, like as picture from above shows?

Please, this half-bridge is going to work in high power, 1kW to 1.5kW, and perhaps in DRSSTC. But for start I'm going to try with classic SSTC, unless I buy more powerful MOSFETs.

Sulaiman

Fri Jun 27 2008, 08:15PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

I frequently use the heatsink as a 'wire'
there are many advantages, especially if you use the expensive silver-loaded heatsink paste.

Hockey-puck thyristors and diodes often use the heatsinks for electrical connections for hundreds of amps.

HV Enthusiast

Fri Jun 27 2008, 08:41PM

Registered Member #15 Joined: Thu Feb 02 2006, 01:11PM
Location:
Posts: 3068

Tonic wrote ...

Mercury is a pretty exotic solution ;) In any case, I'm going to grind heatsinks with sandpaper and use thermal grease for PC CPUs. You can see in photo that I'm using another heatsink to mount transistors and improve cooling. Also, there will be a small fan blowing air through those three heatsinks.

Still, two questions remained without answers. Will this screw insulation sustain 325VDC? Can I use heatsink as "support" wire, like as picture from above shows?

Please, this half-bridge is going to work in high power, 1kW to 1.5kW, and perhaps in DRSSTC. But for start I'm going to try with classic SSTC, unless I buy more powerful MOSFETs.

thermal grease is not conductive.

Tiberius

Fri Jun 27 2008, 08:54PM

Registered Member #1484 Joined: Wed May 14 2008, 03:24PM
Location: Cary, NC, USA
Posts: 27

Not to derail, but I don't see what the big fuss about the toxicity of mercury is. It's commonly found in neon signs, fluorescent lights, dental amalgam, etc. Dermal contact and ingestion have both been found fairly innocuous. In 1993, the United States Public Health Service reported that "amalgam fillings release small amounts of mercury vapor," but in such a small amount that it "has not been shown to cause any ... adverse health effects". If there is controversy over adverse health effects from the presence of mercury in a dental filling, I am pretty sure wetting a couple heat sinks with some mercury is not going to measurably impact your quality of life. In fact, I'm absolutely positive, as I've used mercury as a thermal interface material in several different projects over the years and have yet to feel any negative effects ;)

To quote Wikipedia:

"[Liquid metallic mercury] is poorly absorbed by ingestion and skin contact. It is hazardous due to its potential to release mercury vapour. Animal data indicate that less than 0.01% of ingested mercury is absorbed through the intact gastrointestinal tract; though it may not be true for individuals suffering from ileus. Cases of systemic toxicity from accidental swallowing are rare, and attempted suicide via intravenous injection does not appear to result in systemic toxicity. Though not studied quantitatively, the physical properties of liquid elemental mercury limit its absorption through intact skin and in light of its very low absorption rate from the gastrointestinal tract, skin absorption would not be high. Some mercury vapour is absorbed dermally but uptake by this route is only approximately 1% of that by inhalation."

It also offers thermal conductivity of 8.3W/K*m which is around 300%-1000% greater than the thermal conductivity of silicone oil based greases.

1 2 next

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