In the article Nelson specifies a heatsink per output device of .6 Deg C / Watt. Then goes on to recommend a heatsink that's 6" x 8" with 2" inch fins per output device which seems larger than what he specifies.
Each channel would be 6"x 16" which is about 150x405 mm or 60,750 square mm.
Going by his recommendation the 4U 400mm case is about the same.
If Cviller is using a 3U case without problem then I guess Nelson leaves a good safety margin when he choses heatsinks.
I like peppe's suggestion of a little taller but shorter case. It may actually work well with the compactness of the board.
Each channel would be 6"x 16" which is about 150x405 mm or 60,750 square mm.
Going by his recommendation the 4U 400mm case is about the same.
If Cviller is using a 3U case without problem then I guess Nelson leaves a good safety margin when he choses heatsinks.
I like peppe's suggestion of a little taller but shorter case. It may actually work well with the compactness of the board.
heatsinks
So my thought is, and I cannot remember much of this entire thread, other than max 10A, if the unit is cranked, or someone sets the bias higher, I think bigger would be better.
Is it unreasonable to set nominal bias @ 2A instead of 1.3A? If so, I would imagine the unit would get hot if listening to Metallica through 82db Magnepans.(no I dont own them anymore).
Mike
labjr said:
So my thought is, and I cannot remember much of this entire thread, other than max 10A, if the unit is cranked, or someone sets the bias higher, I think bigger would be better.
Is it unreasonable to set nominal bias @ 2A instead of 1.3A? If so, I would imagine the unit would get hot if listening to Metallica through 82db Magnepans.(no I dont own them anymore).
Mike
During idle, all current will flow through the 2 output devices, 2 Amps times 48Vpp is 96 Watts dissipation.
At the peak of class A operation, your 2 Amps times 2 will flow only through 1 power mosfet.
With 20V out, Vds is only 4 volts, 4 volts times 4 amps is only 16 watt dissipation.
Biased at 2A, the die of the MOSFETs will be at 120C if you keep the heatsinks at 50C, chance of popping an output device will be 8 times higher.
At the peak of class A operation, your 2 Amps times 2 will flow only through 1 power mosfet.
With 20V out, Vds is only 4 volts, 4 volts times 4 amps is only 16 watt dissipation.
Biased at 2A, the die of the MOSFETs will be at 120C if you keep the heatsinks at 50C, chance of popping an output device will be 8 times higher.
Class A amps are different from Class AB ones. The latter become hotter as you play louder (and you increase the bias), while a Class A amp draw a constant current that can be turned into music or heat. Thus, if you play music, part of the energy goes to the speakers not generating heat.
We can then say: the louder you play, the cooler the amp stays
We can then say: the louder you play, the cooler the amp stays
