Silpads are risky business for high performance.
When you measure, you'll most likely find a spread of the temperature from mosfet to mosfet. That would be due to the mosfet-heat sink interfaces not offering identical performance.
Silpads are for serial production, where better solutions are too time consuming and/or too expensive.
Mica and goop, phase change goop and Kapton, or Alumina and goop, are all tried and tested at high performance.
All of those are much harder to mess up as well.
Magura
When you measure, you'll most likely find a spread of the temperature from mosfet to mosfet. That would be due to the mosfet-heat sink interfaces not offering identical performance.
Silpads are for serial production, where better solutions are too time consuming and/or too expensive.
Mica and goop, phase change goop and Kapton, or Alumina and goop, are all tried and tested at high performance.
All of those are much harder to mess up as well.
Magura
Silpads are risky business for high performance.
When you measure, you'll most likely find a spread of the temperature from mosfet to mosfet. That would be due to the mosfet-heat sink interfaces not offering identical performance.
Silpads are for serial production, where better solutions are too time consuming and/or too expensive.
Mica and goop, phase change goop and Kapton, or Alumina and goop, are all tried and tested at high performance.
All of those are much harder to mess up as well.
Howdy Magura,
The Kapton tape and grease are currently what I like and you can bias it till it smokes,and the Kapton doesn't curl or distort or of course burn up!
I had mine up to 2.5 amps per mossfet,that's all my transformer could do,bridge is heatsinked too I stopped at 65 C,I am in process to add a radiator into my water cooled system and upgrade the transformer ,I am using a spring clamp to hold them down and also the pressure is the same on the fets all the time,but to be honest I am also concedering to keep it between 1.0 amd 1.5 I like the sound at that level,just my 2 cents
NS
Attachments
NS, 2.5 Amps? Wow.
Well, supposedly the K-10 Silpads are Kapton. I know they are not cheap...I think they cost more than the Mosfets! As soon as I get an IR temp gun I'll measure them.
Should the N and P Channels be the same temp? It does seem like one side of the heatsink is hotter than the other.
Kapton® -Based Insulator
Sil-Pad K-10 is a high performance insulator. It combines special film with a filled silicone rubber. The result is a product with good cut-through properties and excellent thermal performance.
Well, supposedly the K-10 Silpads are Kapton. I know they are not cheap...I think they cost more than the Mosfets! As soon as I get an IR temp gun I'll measure them.
Should the N and P Channels be the same temp? It does seem like one side of the heatsink is hotter than the other.
Kapton® -Based Insulator
Sil-Pad K-10 is a high performance insulator. It combines special film with a filled silicone rubber. The result is a product with good cut-through properties and excellent thermal performance.
hi Hikari1,
Check this guy out look for water cooled amps http://www.inlowsound.com/ ,IIRC he said he was putting 10 amps through his,LOL,Check ebay for kapton tape use the 1mill thick stuff.
You can run as much bias as your heatsinks can stand and your transformer can stand,heat wise,I got fans on the trans and bridge,and of course water cooled heatsinks,
I read earlier in this thread 85 C ,
As far as temps go one side runs hotter and will run close bias wise if you have matched components other wise it will be off a little ,I keep a eye on the dc output ,I set it close to 0 and let the amp run till it's as hot as it's going to get then tweak the bais some more for a good temp and set the DC and play music,I like papa nelson tool for that "if you can keep your hand on it for 5 secs that's a good temp to leave them at ,Experiment that's what we all do here,It's fun too you learn by doing,You won't put but one cap in backwards,lol.BANG,poof,lol
Have fun!
NS
Check this guy out look for water cooled amps http://www.inlowsound.com/ ,IIRC he said he was putting 10 amps through his,LOL,Check ebay for kapton tape use the 1mill thick stuff.
You can run as much bias as your heatsinks can stand and your transformer can stand,heat wise,I got fans on the trans and bridge,and of course water cooled heatsinks,
I read earlier in this thread 85 C ,
As far as temps go one side runs hotter and will run close bias wise if you have matched components other wise it will be off a little ,I keep a eye on the dc output ,I set it close to 0 and let the amp run till it's as hot as it's going to get then tweak the bais some more for a good temp and set the DC and play music,I like papa nelson tool for that "if you can keep your hand on it for 5 secs that's a good temp to leave them at ,Experiment that's what we all do here,It's fun too you learn by doing,You won't put but one cap in backwards,lol.BANG,poof,lol
Have fun!
NS
Last edited:
I'm using these pads, and I measured the Mosfets using an instant-read food thermometer at 143.5F biased at .59v.
http://www.bergquistcompany.com/thermal_materials/sil_pad/pdfs/sil-pad-900s/PDS_SP_900S_12.08_E.pdf
http://www.bergquistcompany.com/thermal_materials/sil_pad/pdfs/sil-pad-900s/PDS_SP_900S_12.08_E.pdf
Silpads work best with a well applied, reasonably high, pressure. That is in comparison to the small, offset, screw mounting hole used on the typical TO-3P package. However this can easily be accomplished with a clamping type arrangement using 2 other screws/bolts and a bracket contacting the top of the xistor surface.
I'll check this weekend when I can get it good and hot and measure inside and out again. I went to 220R resistors since I last measured the temp, not sure if this affects it.
Most probably the underlined voltage in your post above is a typo and needs to be 3.6V per schematic. 114f = 46 Celsius for the sink. and 121f is 49 Celsius for the MOSFET.
. 
It's the standard case from the store.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
