a75 bias question


Account Disabled
2002-01-07 7:20 pm
Newton Falls, Ohio
I am planning on building an A75 to drive my MTM's, which have a 4 ohm impedance (two Vifa p17's paralleled). My question is, what happens if you run the output stage on the 37 volt rails as usual, but with a 4 ohm load? Does the bias have to be adjusted up? With 37 volt rails, it would seem to me that the idle current should sit at about 4.4 amps to bias the amp to run 150 watts into 4 ohms. My heat sinks can only dissipate a power level equal to about 3.5 amps bias. What happens when I bias it to this level? Does it just leave class A earlier? My guess is that it will make it up to about 135 watts into 4 ohms. Is this correct? Any clarification on this would help me greatly! TIA.

Nelson Pass

The one and only
Paid Member
2001-03-29 12:38 am
Whatever you set the bias at, the amp will do
twice that peak in output current when it falls
into AB operation. 3.5 amp bias (per channel)
will allow a 7 amp peak, which is 49*4, or about
200 watts peak into 4, 100 watts rms.

This is plenty, and the amp will continue to operate
very well at higher output levels in AB, probably
having less distortion than the speaker or your
ears at these levels.


Account Disabled
2002-01-07 7:20 pm
Newton Falls, Ohio
Thanks Nelson!

I really appreciate the help. I'm trying to understand exactly how this works. A few more questions... How do you know when the amp will clip? Is it when the output voltage excedes the supply rails? Will it try to deliver more and more power into lower impedances? If I were to increase the voltage rails and the signal delivered to the output stage keeping the same bias, would it then be considered a class AB amp of higher power? Since I am using a 4 ohm load without biasing for 4 ohms, would the amp now be considered class AB, biased toward the A side? Thanks again for the help!
Clipping occurs when you ask the circuit to swing more voltage than it can. Exactly when, voltage-wise, this will happen depends on the topology of the circuit, but in general you can think in terms of topping out somewhere in the vicinity of the rail voltage.
Note that this is peak, not RMS. My meter reads RMS, for instance, so if I read the voltage at clipping, I'll see a lower number than the rail. The best way to determine clipping is with an oscilloscope.
If you test an amp with an 8 ohm load, then change to a 4 ohm load, the amp will swing the same voltage, but twice the current. (We're assuming that the power supply has sufficient muscle.) If you then hook up a 2 ohm load, it will still swing the same voltage, but with four times the current that it delivered at 8 ohms; twice as much as it delivered at 4 ohms.
If you take a circuit biased at class A and lower the bias current, it's AB. The relative "AB-ness" so to speak, of the bias, depends on how much you lower the bias. There are sub categories of AB: AB1 and AB2, depending on the relative percentage of time that the output devices are conducting.
The bias current determines how much current is available to deliver to the load. If it's insufficient, the amp will 'current limit' which means that the amp won't be able to deliver all the current that you're asking it to. All real world amps (with the arguable exception of the old Mark Levinson ML-2, which would drive a dead short with infinite current...) will reach a point where they current limit. It's only a question of how low an impedance you need to hook to them before they run out of current. If you're drawing enough current that the rail voltage collapses, clipping will occur earlier in the process, because you can't exceed rail voltage.
Sit down with Nelson's A-75 writeup and a calculator and run through a few test cases. It's all pretty much Ohm's Law stuff. Just bear in mind that if you're calculating RMS voltage, then you're dealing with RMS current. Ditto for average or peak values.