Class A/B Biasing

[bowdown]

Hi i was just wandering if i am running +/-95V rails and 4 output pairs(4 npn + 4 pnp) per channel and the idle current draw through the rails is aprox 110mA per channel, how can i figure out how much heat it is dissipating? and does it mean that up to the dissipation level the amp is running class A mode??? If i increase the bias to 400mA+ will i get more power in class A mode??
Please help me understand biasing issues and how it works.

Cheers
Bowdown

[djk]

"Hi i was just wandering if i am running +/-95V rails and 4 output pairs(4 npn + 4 pnp) per channel and the idle current draw through the rails is aprox 110mA per channel, how can i figure out how much heat it is dissipating?"

190V X 0.11A = 20.9W

" and does it mean that up to the dissipation level the amp is running class A mode??? "

NO

"If i increase the bias to 400mA+ will i get more power in class A mode??"

Yes

"Please help me understand biasing issues and how it works."

In a push-pull amp take the idle current times two, then times 0.707, square it, and then take that times the load impedance. That will be the point the amp goes from class A to AB.

0.11A X 2 = 0.22A X 0.707 = 0.15554 X 0.15554 = 0.0241926916 X 4 ohms = 0.0967707664W

Your amp will dissipate 20.9W to drive 0.096W in class A into 4 ohms.

400mA is 1.28W at 4 ohms for 76W of heat.

[AKSA]

Hi Bowdown,

You wrote:

i am running +/-95V rails and 4 output pairs(4 npn + 4 pnp) per channel and the idle current draw through the rails is aprox 110mA per channel, how can i figure out how much heat it is dissipating? and does it mean that up to the dissipation level the amp is running class A mode??? If i increase the bias to 400mA+ will i get more power in class A mode??

95V rails are very high, and will permit a full on output around 175Vpp. Into 8R, this represents 478 watts (175exp2/64) and I would probably opt for five pairs, or even six, to avoid stressing each device and risking destruction. This is a very large amplifier, much, much more than you need in all but sound reinforcement. Into 4R, the possible power is close to double this again, up round 850 watts, if the power supply can bear up.

Dissipation on each device is simply the product of current (110mA/4 or 27.5mA, rather low in fact) multiplied by the voltage across it. At idle this would be the full rail voltage, since the output, at the emitter, is at 0V. Thus dissipation is 2.61 watts per device, truly a walk in the park.

If you increased bias to 400mA per channel, then each pair would run 100mA, and thus dissipation in each device becomes 95 x 0.1 = 9.5 watts. With 8 devices, you need to get rid of 76 watts at idle, and this is no mean feat, as even a 0.3C/watt heatsink will rise in temperature approximately 23C above ambient. Under around 60% of full output, where dissipation is greatest, you will be dissipating roughly four times more per channel, resulting in intolerable temperatures.

Class A would require at least one amp of total bias, resulting in a dissipation of 190 watts at idle. This would require forced air cooling in most situations, with liquid cooling worth considering.

Cheers,

Hugh

[bowdown]

Thanx guys you have been a real help , I really appreciate ur help and i think im getting the biasing thing alot better. again thank you very much for both ur help.

Regards
Bowdown