1 Ohm benchmarks on my amp...

[Solid Snake]

I ran a triangle wave through my amp (with no load) so that the output measured 6.89vac with the DMM. I attached a one ohm load and the output measured 6.59vac. Also the triangle on the scope seemed to be logarithmically compressed, instead of a straight ramp, it was a slight curve. Is the damping factor of this amp good enough or should I shoot higher?

[markp]

If your amp has any protection circuitry it could be doing this as it clamps back on the current limit. It is a dangerous test for most amps to do at one ohm.

[Solid Snake]

There's no protection circuitry. Running at one ohm wasn't a huge problem except for the compressed triangle wave.

[sss]

imho u should measure the damping factor , above 100 is ok i guess , it also depends on the speakers u will be using with your amp .

btw if u got any dc blocking cap at the output u should revove it when u r testing with triangle wave

[Solid Snake]

I have no DC blocking cap. Damping factor changes with load impedance, in this case, it's about 23 @ 1 ohm. It should be a lot better at higher impedances right?

[Solid Snake]

So if the damping factor is 23 at 1 ohm then it should be 92 at 4 ohms and 184 at 8 ohms. The output impedance calculated from these figures is 0.043 ohms. Is this good enough?

[sreten]

The distortion you observe is not related to damping factor.

Ouput current related distortion depends a lot of the type
oF BJT output devices, the value of emiiter resistors fitted,
and for MOSFETs the gate voltage needed for high currents.

sreten.

[Eva]

I think that distortion is actually very related to changes in damping factor with output current [and its polarity], frequency and Vce of output devices

With BJTs, damping factor decreases with increasing output current, decreases with decreasing Vce of output devices and decreases with increasing frequency

So what you've seen is nothing but the fact that output impedance is markedly non-linear in your amplifier [this should cause measurable intermodulation and maybe even measurable subharmonics to appear]

I think that a closed loop amplifier having constant output impedance has constant open/closed-loop gains and therefore is unable to produce any waveform distortion [assuming a constant load impedance]. Dynamic changes in open/closed loop gain are allways required to be able to produce any distortion

[Solid Snake]

I think that when pushing lots of current, the transistor's hfe curve shows through on the output a bit. I'm trying to fix the situation without adding anymore active devices. The driver/output of my stage is entirely class AB. I'm thinking of modifying it so that the beginning of the stage is A and the end is class B, and use global feedback. If I do this, will biasing the B stage be unnecessary since the class A stage can cut down on crossover distortion?

[Eva]

The maximum linearity obtainable in an amplifier is determined by the linearity of the devices used. Topology may reduce more or less this maximum linearity but will never enhance it

Could you post your schematic?