Max allowable DC offset for power amplifier

[anatech]

Hi xrk971,
It's not related to this issue at all. If you set your bias current properly, it will be 50 mA, which is on the high side. As Steve mentioned, you probably set the DC offset instead of measuring across the resistor as should be done. There should be no load on the amplifier when making these adjustments.

Hi Charlie,
You are absolutely correct. Good catch. That just shows you that posting with a low caffeine level can lead to all kinds of errors!

-Chris

 

[xrk971]

sregor,
Thanks for the tip on how to adjust this - I will give it a try. Is there a certain typical value of where to set the pots initially? It's good to know that too low of a bias can cause the harshness in the sound. Maybe it's not a bad amp but just me not setting the bias correctly.

Edit:

Anatech: thanks for the input - I am new to discrete component amps so this is all a good learning experience.

 

[xrk971]

Thanks for the tips on how to adjust this amp! I measured the voltage across the resistor rather than referenced to ground as the vendor's (vague Chinese-English) instructions said to do. And sure enough, VR2 controls the bias very smoothly but the voltage is negative when using the test point they specified for the test probe (just means the sense of the current is opposite). Then I adjusted the DC offset to close to 0 mV.

Anyhow, that worked! The harshness is gone and the amp runs slightly warm now as expected for a class AB amp - before it was dead cold (probably indication of too low a bias). The amp actually sounds very nice - deep, easy, powerful bass, smooth mids, and clear highs with no fatigue.

Thanks again guys.

This amp is quite a deal, now that it works. $28/pair shipping included.

 

[anatech]

Hi xrk971,
You're welcome.

That schematic has errors. Don't take it as gospel! Your DC offset control doesn't work like that. Each end goes to an emitter, the wiper goes to the "current source". That's just from a quick glance.

-Chris

 

[xrk971]

There is an ideal bias current where HD is lowest. I hooked up the amp to my speaker and set a sine generator to sit where HD was worst and used RTA to observe HD changes as I slowly adjusted bias. At 6mV the HD peaks disappeared - go too low or too high and they rise up.

 

[anatech]

Hi xrk971,
Yes, this is normal actually. The problem is that bias is highly variable, so the bias is set higher as that is the lessor of the two evils. You need to avoid falling below the optimal bias setting. I like amplifiers that have a slight negative temperature coefficient for bias vs temperature.

-Chris