Heatsink thermal compound with silicon pads?

[diy didi]

OK. I know that i read somewhere that adding thermal compound paste to silicon pads aren't necessary.
However everytime I repair an amp from a major brand, the devices are mounted on silicon pads with white thermal compound grease.
Why is this? I'm I mistaken? Should one always use thermal compound Grease??

[norazmi]

hi didi,

silicon pad are always necessary for tansistor with sink on back and not for full plastic to-220f, to-3pf, and others. its to protect transistor for being short circuit with others and shorts circuit to ground. and adding thermal grease white compound is a advantage for fast heat dissipation from the transistor into heatsink.

regards.

[rob g]

It depends on the circuit.
Most commercial designs will use a conductive case material therefore will need a properly devised earthing scheme, just in case something goes wrong in the electrical circuit.
The silicon pads are doing nothing to prevent electrical conduction if the screws used touch both output device and the heatsink. Nylon screws can be used or top hat washer that prevents contact between the two.
Silicon pads do not transfer hear as well as thermal grease.

[butch_vlad]

The man is asking about thermal grease....not insulation
Some cheap sil-pads do require a bit of thermal grease, the proper ones are soft enough to make a proper contact without the need for grease.
Remember, best heat transfer is by direct contact, the more things you put in between the less heat you transfer in a given amount of time.
"Thermal Compounds

One of the most common mistakes made by hobby electronics enthusiasts (and quite a few professionals too), is to assume that if a little thermal compound is good, a lot must be better. Absolutely not so! The amount of thermal compound should be exactly that amount which ensures that an air-free join is made between the mating surfaces. If too much is applied it will cause an increase in thermal resistance, since it is not really that good at conducting heat. Generally speaking, any electrical insulator is also a thermal insulator, so the thinner the final composite insulation - including thermal "grease" - the better. " ESP

Have a look here: Sil-Pad Products ~ Thermal Materials, Thermal Solutions ~ The Bergquist Company


DO NOT CONFUSE SILICON PADS WITH MICA PADS.
Mica pads are rubbish!!

Hope that helps
Vlad

[AndrewT]

I cannot agree that mica is rubbish.

1mil = 0.001" = 1thou ~ 0.025mm, thickness mica insulators are very good.
There are few other insulators that can beat a 1thou mica for Rth c-s.

[butch_vlad]

You are right, but I tried to find decent mica insulators in UK and failed.
As you said the secret of mica insulators is to make it as thin as possible but at the end of the day mica is not a good thermal conductor.
The usual thickness is 0.5mm, maybe slightly thinner (cheap chinese imports sold at %500 profit), what I normaly do is split it into 3 thinner foils.
Compared with high performance insulators, thickness for thickness mica is rubbish but in all honesty will do a decent job for most amplifiers.

[Elvee]

Originally Posted by butch_vlad

The usual thickness is 0.5mm, maybe slightly thinner (cheap chinese imports sold at %500 profit),

Typical thickness for standard 1KV mica insulators is in the range 0.05mm to 0.1mm:
- heatsink interface material.

Their thermal performance, when used with compound is similar to standard Silpads, Chotherms, etc.
Very much better performances can be achieved with ceramic powder coated synthetic film insulators.

[tvrgeek]

We did a big study on this subject in the FA lab when I was in industry. Berquest, who made the glass fiber reinforced sil-pads had a great reference. Every pad has a different thermal conductivity. This is just engineering. Quick summary of the choices:

#1 Bare anodized aluminum heat sink is the best, but manufacturing repeatablity makes it aerospace only.

#2 Graphite is next, as it beds in and gets better with use. Unfortunately it is conductive so the heat sink is at case potential.

#3 Next in line in performance is mica-grease. It does real well when first installed if done correctly. It is almost never correct. The second problem is with thermal cycling, the oxide will pump its way out from under the device and over time, the performance goes down. Messy and hard to repeat performance in manufacturing. Another side effect is the migrating silicone grease can pick up a layer of dust and insulate the surface. There is no reason to use mica-grease today.

#4 Next is the sil-pad technology. Not as high a performance, so you need to do proper derating. But, easy to reproduce (there or not) and like graphite, they settle in giving better performance over time. Most can be reused, where graphite should not be.

#5 Kapton is sometimes used ( gold plastic). Performance fair, pretty tough, but not what I would choose. Low power and more just for electrical insulation.

#6 Exotics. We had places where the sil-pad just could not hack it and used a Be oxide wafer that was about 1.5mm thick. This relied on the extreme flat surfaces of aero-space, but was less prone to electrical failure if it was not perfect. Mounting was with exotic bevel washer stacks. I have seen anodized aluminum wafers used like this too with very soft aluminum to bed in. These are not for consumer products and no way for DIY.

Speaking of washer stacks. It is a very bad idea to use the same hardware that mounts your device as the current carrying path. I HATE TO-3 packages as much as I hate sockets. We found a pigtail of a lug soldered to the board held in contact to the case with a spring assembly the only mostly reliable way to mount TO-3. Before we redid all or our assemblies, I would estimate 25% or our power amplifier ( big motor servos) were due to hardware stack issues. They went away. ESD damage was another 25%, and the rest other design faults leading to overstress. Bad parts almost don't exist.

While I am ranting, the star washer is one of the most evil inventions ever. They don't work, generate little metal flakes, and break. There is no place inside electronics for them.

There, 10 years of failure analysis lab experience.

[sofaspud]

I thought Kapton's big benefit was it's thermal stability? I haven't ever used it as a heatsink insulator, but I have considered it. Here it's described as low power, electrical insulator. I'd like to read a little more about why it isn't often used commercially in this app, and how that might affect DIY activity. I can see that it wouldn't "settle in" as you say, but might it's thermal conductivity neutralize some of the cons?

[SY]

Thermal conductivity of Kapton isn't that fabulous. Heat resistance is excellent, as is dielectric strength. The downsides are cost and really poor tear resistance.