This is far too thick and far too expensive.
£0.98 per 30mm x 25mm
whereas the 1thou Kapton above @ ~£13 is enough for 1100 pads @ ~£0.012 each.
Thanks, I'm still looking for power transistor insulators here in South Africa.
RS Components is reasonable here, but Mouser costs to much here.
My local component shop, doesn't stock this size at all.
After that esp link I see that he is selling 1m strip of 25mm wide 1thou Kapton @ usd$7.50
That is an OK price for 33 off @ 30mm long.
Alternatively team up with Ian Finch and start a new business for us Members.
That is an OK price for 33 off @ 30mm long.
Alternatively team up with Ian Finch and start a new business for us Members.
The seller, CaplinQ, actually dispatches from Holland for destinations outside North America . That should be an even better deal for UK/EU guys. The tapes even go down to 12uM and less thickness, at which point the grease becomes a more significant factor.
This is suitable? It's Very thin Yes an adhesive type. A wipe with thinners removes it, although the adhesive ensures tailored placement.
Radio Control Planes, Helicopters, Cars, Boats, FPV and Quadcopters - Hobbyking
Usually 2.99$ shipping, worldwide. ~3 weeks though.
Several Fleabay vendors as well sell this or it's cousin, some with free shipping.
Really easy to source.
Radio Control Planes, Helicopters, Cars, Boats, FPV and Quadcopters - Hobbyking
Usually 2.99$ shipping, worldwide. ~3 weeks though.
Several Fleabay vendors as well sell this or it's cousin, some with free shipping.
Really easy to source.
If the seller quoted thickness, sure. It's the same basic DuPont film product as all the Kapton films discussed here but 1thou, mil. or 25uM is mighty thin. You need a micrometer, paint thickness gauge, or dial gauge to be accurate unless it is quoted or marked on the product, since most widely available tapes are actually more like 2-6 thou, as stated.
Unfortunately, this can be enough to knock it out of contention with mica and silicone rubber pads for critical applications, particularly where a designer cramps too many devices close together to save costs.
AndrewT might be glad to know thinners will remove the glue layer without damage, though.
Put thermal performance aside, what is wrong with thermal pads being thick? I think thickness may be something actually needed.
I am using L-MOSFETs TO264 ECW29N20-Z and ECW29P20-Z which have the back metal of the package electrically connected to the source of the L-MOSFETs. With thin thermal pads the capacitance between the L-MOSFET source and the heatsink exceeds 1nF capacitance, because of the large parallel plates between the back metal and the heatsink. Now the heatsink is connected to the chassis which can be quite some distance from the amplifier star ground point (say, 250mm or 260nH) with very little resistance. In other words, this is a 1nF capacitor in series with an inductor 260nH with extremely small resistance (i.e. almost no damping) from the L-MOSFETs source to the ground. With those values the resonant frequency will be at around 10MHz. We know that series (vs paralleled) LC resonant circuit creates impedance peak, rather than impedance null (in paralleled LC). This is something we don't want.
Now there is another current path from the L-MOSFET source to ground - via the 0.1R source resistor in series with the Zobel of 4-8R in series with a 0.1uF capacitor. Would this damp the LC resonance via the thermal pads - heatsink - chassis? Probably not. The characteristic impedance of 1nF and 260nH will be at around 16R. So we need a 16R resistor in series with a, e.g. a 10nF snubber to connect the heatsink to the star ground at short distance.
If a thicker thermal pad is used it will reduce the capacitance to push the resonant frequency much higher, making it less harmful to audio, perhaps.
Correct me if I am wrong.
Regards,
Bill
Ceramic washers may be expensive but if capacitance concerns you, that would be one way to decrease it whilst maintaining thermal conductance - at a price. Using thicker insulators like mica, silicone or plastics is self defeating - it just kills dissipation and output power.
aluminium oxide (ceramic) Thermal Washers are very good, if thin enough.
I have seen 2.5mm thick and even thicker.
These struggle to pass sufficient heat.
Very thin ceramic may just about match (and even better) the best of mica and the best of all "soft" insulators, except the Keratherm which are exceptionally good.
I am still surprised that the rest of the insulator market has not brought out a competitor to the Keratherm. How long have they been available?
I have seen 2.5mm thick and even thicker.
These struggle to pass sufficient heat.
Very thin ceramic may just about match (and even better) the best of mica and the best of all "soft" insulators, except the Keratherm which are exceptionally good.
I am still surprised that the rest of the insulator market has not brought out a competitor to the Keratherm. How long have they been available?
I did find some equally thick (2.5mm) ceramic insulators when I was breaking an American amp. I've got no confidence in ever using them in my own projects, my mind always says - the thinner the better.
Didn't NASA use ceramics for heat insulation ?
Didn't NASA use ceramics for heat insulation ?
If ceramic = aluminium oxide then I am wondering if the entire sheet is aluminium oxide or only the external surface, if it was the later, that means inside the sheet it would be conductor, and if that were the case, the capacitance would be worse.
I have just searched the digikey catelog and I can find aluminium oxide thermal pads but I can't find any for TO264. What else can be used for TO264? Many thanks.
p.s. actually, further check on digikey all of the aluminium oxide pads are non-stocked items and the minimum quantity is 1000.
I have just searched the digikey catelog and I can find aluminium oxide thermal pads but I can't find any for TO264. What else can be used for TO264? Many thanks.
p.s. actually, further check on digikey all of the aluminium oxide pads are non-stocked items and the minimum quantity is 1000.
Last edited:
the whole thermal washer is a ceramic.
Anodised aluminium is a metal and oxide doublet, or triplet is both faces are anodised.
Anodised aluminium is a metal and oxide doublet, or triplet is both faces are anodised.
no. two insulating layers would be equivalent to two series capacitors. That would equate to less capacitance if your circuit is susceptible to this.
Heres some ceramic insulators being sold on ebay. Any comments
Ceramic Insulator Plates 10 Pieces FOR TO 3P TO 247 Good Heat Conductiviy | eBay
Ceramic Insulator Plates 10 Pieces FOR TO 3P TO 247 Good Heat Conductiviy | eBay
I think that both the heatsink and the semi would need to be tested and surface ground as necessary before you tried a few thermal cycles with 1 mm ceramic. 1.5 mm samples I retrofitted a few years ago, shattered before I even got them torqued down. 2.5 mm seems fine with NJW21193/4 at full throttle, several nights per week. Case temperature averages only ~ 5 degrees C higher than the low quality mica used previously but I don't know exactly what the type is other than it isn't BeO, thankfully.
Used my Micrometer... 1.5 thous. Newspaper is typically 3 thous thick.
Damp turps rag wipes off the mild adhesive.. instantly.
Laquer thinners seem to do no film damage But unsure of it's appropriateness.
PS: isn't ceramic a V good insulator? as in re-entry heat shielding and Kilns and furnaces.
Sure, some ceramics like typical kaolin pottery and tiles are great insulators but synthetic types based on Alumina, Zirconia and some rare earth oxides are fine heat conductors whilst they remain electrical insulators - that's why we use 'em. Wikipedia compares the powdered ceramic forms used as filler in heatsink grease and incidentally, in silpads too, for some comparison of related materials.