PCB heat shield between cap and heat sink?

[coppercone2]

I have a few prototypes I made on perf board, I think they are quite good and reliable, the only problem is that one of the caps is close to a heat sink TO220. I measured the thermals and the heat sink has a fair bit of margin, and its a high temp cap, but I was wondering if maybe I could stick a sheet of some kind of material between the cap and the heat sink (the cap does nothing much).

Like some kind of semi stiff fiberglass stuff I can epoxy in there.

Is this a thing? Just to give it a little easier condition. There is like 5mm between the cap and the TO220 that gets to 75C. I don't plan on using this much, I just made one of the boards a little wrong, the other have about 1cm of space.

[Gyro]

Most of the thermal coupling will probably happen through the lead connections rather than the air.

[coppercone2]

well they are both tall structures and its a static enclosure. I could test it with a thermal cam though, just want to know the material. sounds like a sheet of whatever I want should be at most like $10.

interesting experiment I think, even if the benefit is low

lets call it thermal myth busting

[Gyro]

Somehow I misread 'heatsink' as 'regulator'. Yes, if they are both tall, vertical, and close then there will be coupling through the air as well as the board.

As a thought experiment, I suspect that anything of low thermal mass is going to be heated to some mid point temperature between what is radiated by the heatsink at one side and what is absorbed by the capacitor on the other, possibly with some overall increase due to any restriction in airflow. If the sheet was thermally conductive and attached to something cool then I'm sure it would make a difference.

Time for the thermal camera.

[wraper]

How hot that heatsink gets? Does it even matter? Also you risk interrupting the air flow thus increasing the temperature. Heatsink does not necessarily mean hot, free standing parts often get much hotter.

[T3sl4co1l]

Either soft fiberglass (felt?), or maybe more likely flame retardant polyester or other poly felt material, I suppose would be better. Or aluminum foil tape, reflective to keep IR away.

But way more likely is sheer overthinking.

Could very well be that the tiny amount of heat the cap soaks up, while it raises its temperature marginally, it also has more area to dissipate, lowering temperatures overall.  What about that?

Introducing thermal insulation anywhere is probably a mistake made elsewhere in the design, not being able to sink it directly instead.  (Aside from anything that's intended to get hot, of course.)

[coppercone2]

ok yeah a aluminum foil sounds like a good experiment, I don't want to rework this board

[coppercone2]

You are right, the traces conduct the most heat, the capacitor is at a balmy 55C, so I don't think it will effect it (its a 115C polymer cap).

[mariush]

Not sure why you'd care then ...   the lifetime would be big ... something like 

Lifetime = Original lifetime x 10^z  where z = (Tm-Ta) / 20    and Tm = rating of capacitor and Ta = ambient temperature

So a 2000h@105c would be 2000x10^(105-55)/20 = 2000 x 10^2.5 = 630k hours

[coppercone2]

Not sure why you'd care then ...   the lifetime would be big ... something like 

Lifetime = Original lifetime x 10^z  where z = (Tm-Ta) / 20    and Tm = rating of capacitor and Ta = ambient temperature

So a 2000h@105c would be 2000x10^(105-55)/20 = 2000 x 10^2.5 = 630k hours

thats ok for a prototype I guess, but I need to get it to 6k years before the engineering team goes into stasis