For what it's worth, and that's not much, there is possibly another technique in bias/balancing that may be worthy of consideration.
Personally, I've always biased for as close to zero imbalance as possible. Mr. Berning has lectured on the need to keep that imbalance as close to zero as practical. He tells us that measured distortion increases rapidly with as little as 2 mA imbalance, as the core saturates. I take him at his word! He's spent an awful lot of time creating Class AB capable auto bias circuits just to remedy that ill.
But, Tubelab George had a very interesting post. I HAVE NO FIRSTHAND KNOWLEDGE, I just read it and am passing it along. Bias for as close to zero imbalance as possible. Then, using a dummy load and signal generator set to ~1kHz, listen to the transformer "sing". Adjust bias in one tube to minimize the singing. He claims that the result is improved sound. Kind of an AC balance trick I suppose.
Please don't flame me, if this is terrible, I'm just passing along something that sounded interesting and easy to try.
Stuart
|