F5 power amplifier

1/4si is ~1.6 cm2

Advisable torque for a TO-220 is 0.5 Nm (both for a regular one, as for the F-version)
For convenience sake, assume an M3 screw has an effective diameter of 3 mm

T = C * D * F
0.5 = 0.34 * 0.003 * F
F = 490 N
Makes the mounting pressure 306 N/cm2 (>5 times 60 N/cm2)

If the screw is lubricated for the torquing, C is ~0.20 (the C factor also depends on the material), in which case pressure would go up by 70%
(screw bonus is sucker in domus)

For TO-220s, it makes sense to use clamps instead of the mounting hole because of the hole location and top of the package.
 
Last edited:
Keratherm 86/525 good enough

Short answer is that 0.5 mm sheet of 86/525 equals 86/82 if compressed to ~0.21 mm thickness.
(5.5 : 6.5 = 0.21 : 0.25)
If you dare to compress it to 0.1 mm, thermal resistance should be twice as good as 86/82.

Costs a fifth more though.
Why bother for 1C difference, read EUVL's article for Linear Audio.

(the older 86/525 datasheet only went to 1 mm sheet thickness, not 0.5 mm)
 
Short answer is that 0.5 mm sheet of 86/525 equals 86/82 if compressed to ~0.21 mm thickness.
(5.5 : 6.5 = 0.21 : 0.25)
If you dare to compress it to 0.1 mm, thermal resistance should be twice as good as 86/82.

Costs a fifth more though.
Why bother for 1C difference, read EUVL's article for Linear Audio.

(the older 86/525 datasheet only went to 1 mm sheet thickness, not 0.5 mm)

That is reasoning without any non-linearity factor in the ratio between thickness/thermal conductivity applied by the torque. Sure the more torque, the less thickness and better the thermal conductivity. But as it gets thinner, it won't necessarily increase conductivity by a linear factor. There is a limit in compression of these thermal sheets, where thermal conduction can't be raised by any factor.
Still a very good product. ZhouFang used to sell some in his thread.
Best,

nAr