Recommendations for TO-3 thermal pads

jindra

jindra

Well-Known Member
I'm replacing outputs on my GFA-565 and would like recommendations on the proper thermal pads to replace the old ones with. I was thinking of using this:

http://www.digikey.ca/product-search/en?x=0&y=0&lang=en&site=ca&KeyWords=HF115AC-0.0055-AC-05

Maybe the mica pads would be best? The Sil pads have a shorter life-span, it looks as though the originals were mica as well? The one I linked looks like a fiberglass/silicon combo not sure if that would need to be replaced much sooner than mica.

What do you guys think?
 
Last edited:
Register to hide this ad
Done and done. Thanks for the confirmation, was looking at the exact same part after reading a bit.
 
Sil-Pads might be OK for production line work or when one does not like the messy-ness, but I thing the mica with the grease works best when done properly.

Mark T.
 
I'd love to see some long-term comparisons on mica with properly applied quality compound versus good quality silicon pads, such as Bergquist K-10.

Small initial differences in thermal conductivity, price, etc. seem less important than prolonged ability to keep filling those little gaps and keep conducting heat well.

There are some threads here and at DIY, but I haven't found a long-term comparison.

I use mica and grease because I've always done it that way. But I'd be hard pressed to come forward with some data.
 
Last edited:
Just make sure you match the hole spacing up. THere are 2 commonly available TO3 mica sheets available, one is a bit bigger than the other. I forget which measurement it is, but I ran into that when I rebuilt my friend's Phase Linear. One was the right size, the other wouldn't have lined up with the screws or something.


but yeah, I do the mica and grease thing too. Once in a while, in a particularly moody circuit, you'll find that they can get stupid if you change the insulator material. The insulator forms a capacitor between the collector and the heat sink. Sometimes that value actually plays into whether or not a unit will oscillate. Most things aren't quite so touchy though.
 
I recently did a study thinking that the silpads would be a cleaner way to go vs. mica and thermal grease. Turns out that the silpads and similar just are not quite there yet in terms of thermal conductivity. Best performance is still mica and grease. Silpads may be ok if you are not dissipating a lot of power and can live with higher case to sink resistance.
 
Problem with mica is the delineation. You have to look at it and not see any abberations in it for performance. When they get bent, the layers split and this delineation is very bad for thermal conductivity. This is why I do not remount power transistors just for the hell of it. If you see what almost looks like a bubble in it after it is cleaned, it is separating.

The gray rubber type are not as good as mica, as long as it is thin mica. Used to be you could order different voltage ratings, the lower voltage ones were thinner. Of course this was on the manufacturing level, we got what we got. I have seen thinner ones break down in high voltage applications. Kinda rare but does happen. But then the thicker ones have a higher thermal resistance. Six of one, half dozen of...

It is a tradeoff.

One of the main things though, is there could be air pockets in the thermal grease. You do not spread it, you put a dolliup at the center of the force to be applied. On a TO-3 this is almost betweent the B and E pins, but to the side where your dollup is more to the center of the case. Then you tighten the screws evenly and make sure it squeezes out evenly. This assures there are no air pockets unless they were in it in the tube. (store the tube with the outlet up , like caulk)
 
If there were one answer that was superior in every way, that's all anybody would use. Unfortunately there isn't. The old gray Sil-Pads are just OK, but should never be used where you need the absolute best conductivity. Or even second best. They probably gave Sil-Pads a permanent bad name, but if you spend the bucks for their better materials, they're pretty good. Probably good enough for most audio work. Probably not as good as thin mica and grease. Thin is the key. Mica isn't that great in terms of bulk conductivity, nor is thermal grease, but they lend themselves to very thin layers.

I have to disagree about spreading grease. Most people way over-torque transistor mounting hardware. If you adhere to the data sheet torque values, a blob of grease probably won't spread to as thin a layer as you want, especially on a TO-220, where the force goes way down away from the screw. I do recommend spreading the grease to a very thin layer. Most people use way too much of the stuff. In this case, more isn't better.

The best thing I've found, though you have to use a lot of care with flatness and burrs, are the heavy hard anodized aluminum insulators, combined with Aavid Ultrastick, a wax type compound that actually liquefies when heated, giving the thinnest possible layer. It also maintains conductivity over time better than grease, at least if you believe the manufacturers charts.
 
Pads, be they Sillycone or Mica under devices is typically to isolate the things electrically. Heat transfer is good IF you can get that as well. Not a given tho.
Don't be forgetting that one of Mica's uses is in High Heat isolation /insulation . ie; as in view port windows in furnaces/burners of all types

Kapton is a Good (better?) electrical isolator and heat transfer media than Mica,
If only because it's a consistent product and can be produced as very thin sheets or tape rolls
Try some?
 
https://www.onsemi.com/pub/Collateral/AN1040-D.PDF

The conclusions to be drawn from all this data is that
some types of silicon rubber pads, mounted dry, will
outperform the commonly used mica with grease. Cost
may be a determining factor in making a selection.
Click to expand...

The biggest problem with mica is that most are way too damn thick. Consistency is a big issue. I've gotten pretty good at thinning them out with a hobby knife, but if cost isn't an issue and the surfaces are clean and flat, the Bergquist K-10 pads will kick the traditional mica & grease in the nads.

People tend to lump all thermal pads into the same category. Bad idea if you're looking for reliable information. The cheap crap sold at Parts Express doesn't begin to compare with serious attempts such as those marketed by companies like Bergquist.
 
Last edited:
What swayed me from the Bergquist pad is i'm not sure about lifespan. I'm not exactly up for experimenting with these amps, or having to replace the pads in 5 or 10 years. Would these need to be split?

Any idea what the lifespan should be for the Bergquist pads?
 
ConradH said:
.....
I have to disagree about spreading grease. Most people way over-torque transistor mounting hardware. If you adhere to the data sheet torque values, a blob of grease probably won't spread to as thin a layer as you want, especially on a TO-220, where the force goes way down away from the screw. I do recommend spreading the grease to a very thin layer. Most people use way too much of the stuff. In this case, more isn't better.
.....
Click to expand...

I also spread the grease, usually on both sides of the insulator, thin but still visible over the entire surface. Unfortunately, the best tool for this is your finger, and the stuff does not wash off with anything. You just have to rub a lot with a clean dry rag and hope you get all of it.
 
jindra said:
What swayed me from the Bergquist pad is i'm not sure about lifespan. I'm not exactly up for experimenting with these amps, or having to replace the pads in 5 or 10 years. Would these need to be split?

Any idea what the lifespan should be for the Bergquist pads?
Click to expand...

I use these: http://www.digikey.com/product-search/en?x=0&y=0&lang=en&site=us&keywords=4662K-ND

The thicknesses are all over the map, regardless of the exact mica part you go with. I generally order like 100 of 'em and hand pick the thinnest for use.

Not sure why you think you'd have to replace the Bergquist pads somewhere down the line.
 
I can't imagine a good Bergquist pad breaking down or changing over time. OTOH, white thermal grease, probably depending on who made it, seems to have a definite lifetime before the silicone oil creeps out and the joint goes dry. The clear stuff used in vintage receivers seems unchanged after several decades.
 
FWIW, I've been using Wakefield 120 for years and never seen it dry up. Maybe after 35 years. :scratch2:
 
I've used Thermalloy Thermalcote for years with good results as well. Comes in a number of formulations, some updated. I believe AAVID now owns Thermalloy.
 
You must log in or register to reply here.
Back
Top Bottom