Optimum class AB bias current (again)

[PMA]

Just read Douglas Self book. Useless thread.

[GerryCo]

Absolutely Daniel, I would fully agree too.
Thanks for all your feedback. ........ (Can't find where they've hidden the emoticons)

[nigelwright7557]

Originally Posted by nigelwright7557

I design my own amps so there is no standard setting for them.
I apply a sine wave and monitor the output.
I turn the bias up slowly until crossover distortion just goes.
The amps sound good so I can only assume what I have done is OK.

I have found since that setting bias this way at lower voltage output doesn't always work at higher voltage outputs. However it probably doesn't matter because you cant hear cross over distortion at higher power levels.

[GerryCo]

Originally Posted by nigelwright7557

I have found since that setting bias this way at lower voltage output doesn't always work at higher voltage outputs. However it probably doesn't matter because you cant hear cross over distortion at higher power levels.

I have heard others say exactly that too, and I wonder if it is due to too low a standing current in the VAS -


Another point is trying to set bias without a 'scope; setting to a calculated voltage or current level or, by ear, will not be accurate - and I think this is another reason for over-biasing. Too much bias is certainly better than too little.


My last 'big' power amplifier, many years ago, was a Yamaha M50, which has an 'Auto Class A' function - up to 20 Watts. When I first tried it (it could be switched into A via the panel), I felt I couldn't hear any difference, and the amp became quite hot, which is not really helpful here on a hot South African summer's evening .... and I wondered why they had added a switch that didn't seem to do anything useful, so left it playing in class A while listening more intently, then began to realise that in fact, it really did sound better - smoother and 'cleaner'.
The fact is, that whilst it is not immediately obvious - to me anyway - class A does sound better, even if it is only a low power level of class A function - gain-stepping distortion notwithstanding. And it makes bias setting so easy too, in fact, it becomes more practical to the source the bias current as a separate entity, rather than from the VAS, and it need not be thermally monitored - just current limited.

[Svitjod]

I'm probably reiterating what many others have said before me.

My philosophy; set the bias as high you can without overheating the amplifier.

Mr Self has indeed shown that the distortion will be lowest at appr. 20-30mA per transistor pair. But will it? Under what circumstances?

For example, if you bias the amp to 300mA means that it will operate in class A up to appr. 1W and in a normal living room that's a pretty high level.
Even if the "twitch" that occurs during overlap is lowest at 20mA, the masking effect will make it "drown" in the music signal if that twitch happens att a higher bias.

Did Mr Self ever ponder this fact?

[GerryCo]

Hi there Svitjod!

Thanks for your input. It seems that the majority preference is certainly for higher bias levels - and I agree with that too, mostly for two reasons: 1) The fact that loudspeaker impedance dips can and sometimes do make a mockery of classic class B bias levels; & 2) Because I cannot knowingly set the bias optimally without the use of proper equipment, such as an oscilloscope, and setting to a pre-calculated voltage or current level might not always achieve the optimal bias condition. This malady is made worse in Bi/Tri-amping applications.

On the other hand, very high levels of bias make for a most uncomfortably high temperature house guest here in sunny (and very hot and humid) South Africa ! But also, I feel that it is quite unconscienceable to wantonly/irresponsibly generate practically unproductive energy.

That said, I would want only as much over-bias as produces a true, clean class A crossover condition as can be achieved without producing significant heat.

[Svitjod]

Yes, a commercial amplifier must balance the calculated subjective benefits with the actual price level. Many audiophiles simply are happy with a low bias.
I experimented with an old Rotel - a very straight forward construction and I found that a lower bias took away some of the "sharpness" and introduced a forgiving fuzzyness.

Nowadays I usually construct amplifiers for my self and I'm only in to maximize the sound quality. My current class AB amp has around 250mA of bias. This also helps warming up my apartment a bit.

My subjective impressions about this subject is that a higher bias gives a feeling of solidity to the presentation. Other parameters such as imaging and the overall tonal impression tends to be just about the same.
With solidity I mean that the instruments sounds more like tonal "bodies" so to speak.

The subjectivist in me has spoken

[GerryCo]

Hi again Svitjod !!

You are absolutely right, fully agreed.

Personally, my mission is to achieve least distortion - to get the best sound that I am able to. That's quite a big ask, because my abilities in audio are somewhat limited !!

So I ask a lot of questions and pose a few ideas to get feedback.

Thanks again for responding,