This came from a a very large commercial UPS.
Had bunches of TO-247 and TO-220 devices
attached to a large heat sink. I doesn’t stretch
and has fibers (see photo) embedded in it. If it
is Keratherm would it be better than the Alumina
Ceramic insulators that I have on order now?
Sorry if this is in the wrong forum but I am in
process of building an Aleph J. I’m a little lost
If it doesn’t have a glowing filament inside it.
Best Regards, Mark
Had bunches of TO-247 and TO-220 devices
attached to a large heat sink. I doesn’t stretch
and has fibers (see photo) embedded in it. If it
is Keratherm would it be better than the Alumina
Ceramic insulators that I have on order now?
Sorry if this is in the wrong forum but I am in
process of building an Aleph J. I’m a little lost
If it doesn’t have a glowing filament inside it.
Best Regards, Mark
Attachments
I would not re-use it. Better to go with something new. What's wrong with Mica and thermal compound?
I didn't think mica was as efficient and thermal compound can get messy.
Trying to get the best heat transfer. If mica is best then I'll go that route.
Thanks!
Trying to get the best heat transfer. If mica is best then I'll go that route.
Thanks!
Silpads and the like are probably mostly used becase they make assembly more efficient (= faster). Those considerations drive a lot in mass-production but less so for DIY'ers.
Ceramic insulators + thermal paste are the most durable and thermally efficient solution.
I remember Elektor recommending them in the Titan 2000 to prevent oscillation.
Hugo
Well, call me old fashioned. But i think it looks plain, cheap, ugly and dirty with dirt collecting toothpaste on and around my active devices and heatsinks.
Yea. So. Keratherm all the way, when its possible, just for the pure Class and Beauty of it all. 🎺🙂🎸
Yea. So. Keratherm all the way, when its possible, just for the pure Class and Beauty of it all. 🎺🙂🎸
I replaced shiitty keratherm in most of my classA amps (about 20 and counting) with mica and white thermal paste. Looks worse, but runs much cooler. Who cares how it looks inside, as long as longevity is much better.
Old fashioned? Mica and goop is old school. Keratherm and the like seem too prone to failure for me, despite any thermal advantages they may have.
The sinks run cooler, or the devices?
I agree Kheraterms are fragile. But they aren’t messy, an advantage for beginners but also if swapping amp boards often.
But, I take care in mounting, and always use a torque wrench, keep the devices stable and ensure they are torqued to spec before I do any soldering. And when used once, I usually discard them. Expensive, yes.
They're installed in a more controlled environment and with better process control than I can provide in my basement. If I was manufacturing electronics in any sort of volume, I would use them, too. I think they are better suited to it. For my personal hobby use, I'll stick with mica and compound. I've managed to puncture a few Keratherm like insulators without too much difficulty.
The Keratherm 86/82 pads are the best (read: lowest thermal resistance) silicone thermal pads on the market. At a few bucks apiece they're not exactly cheap, but they're worth it in my view.
Mica + goop can beat the Keratherm pads on thermal resistance if you get the goop smeared on correctly. You want just enough goop to fill the voids and that's it. If you have too much goop you'll end up with a higher thermal resistance.
If you use spring clips to mount the devices, mica + phase change material can be a nice option. The phase change material is basically thermal compound in solid form that you apply to the devices and the mica washer. It turns liquid above a certain temperature and distributes itself into the voids as an ultra-thin film. But it only works if the devices are mounted with spring clamps or if you torque the mounting screws once the thermal compound is liquid.
The phase change compound is pretty expensive. I have a stick of it that's about the size of a large pencil eraser. That set me back around $30 as I recall.
Tom
Mica + goop can beat the Keratherm pads on thermal resistance if you get the goop smeared on correctly. You want just enough goop to fill the voids and that's it. If you have too much goop you'll end up with a higher thermal resistance.
If you use spring clips to mount the devices, mica + phase change material can be a nice option. The phase change material is basically thermal compound in solid form that you apply to the devices and the mica washer. It turns liquid above a certain temperature and distributes itself into the voids as an ultra-thin film. But it only works if the devices are mounted with spring clamps or if you torque the mounting screws once the thermal compound is liquid.
The phase change compound is pretty expensive. I have a stick of it that's about the size of a large pencil eraser. That set me back around $30 as I recall.
Tom
The best on avail is AlN. Despite having conductivity is on par with aluminum, cheap Ali ones have quarter of that. This still beats alumina.
Look @ NP designs with lots and lots of devices in parallel - it's much wiser than few with the greatest and latest.
FUI Alumina is 25 Wm/K, AlN - 170(50??) Wm/K, 86/82 - 6.5 Wm/K, mica - 0.71 which is on par with most SilPads.
That's said you can get 2 mil mica. Keraterm is 10 mil still being 2 times better. You can compare the rest.
It depends on isolation requirement as well. Keraterm is no good for HV (like SMPS primary switches) but the other mentioned are.
Be aware of power level and heatsink. In case of higher power (some IXYS 247 packaged have 300W specified) it will do on good forced air
heatsink with AlN but will not with Keraterm not mentioning SilPad. If you switch to water cooling then it becomes more forgiving, pad wise.
Ceramic is fragile and requires really flat surface. I would second phase change material - no mess being quite forgiving during application.
Did you take into account of the grease layers on both sides of the AlN or mica or .... ?
How thin do you think those layers can be ?
Do you have a flatness specification on you device, ceramic wafer, and heatsink ?
Are they sub-micrometer flat ?
What is the average size of the filler particles inside your grease ?
Patrick
How thin do you think those layers can be ?
Do you have a flatness specification on you device, ceramic wafer, and heatsink ?
Are they sub-micrometer flat ?
What is the average size of the filler particles inside your grease ?
Patrick