In Reply to: Re: Question for Mark Gilmore posted by Marty on March 4, 2000 at 09:44:52:
Marty,I agree with your conclusion. I biamp with 4 homebrewed OTLs with 12 Svetlana 6AS7Gs each. In order to use as many tubes as possible, I sorted 60 tubes into 8 groups of 6 that had similar bias requirements. I set the total bias current conservatively at 750 mA (62.5 mA average per section), and my plate voltage typically runs between 140V to 145V, which resulted in bias voltages for the different groupings that ranged from -57V to -66V. I rejected any tube where one section draws more than 90 mA or less than 30mA, which resulted in a yield of over 90%.
Of course, Mark's approach is best, but I'm too cheap to reject as many tubes as he does. The drawback to my approach is that the voltage where the amps clip will vary according to individual bias voltages of each group. But unless you routinely drive your amps into clipping, there should be no sonic penalty.
Since, after almost 2 years of service, the only failure I've had was an open section in one tube (which I suspect was a defective tube, as I had discovered two others that failed the same way after only one day), it seems this approach offers good reliability, too. And this is with very "OTL unfriendly" speakers (Sound Lab A-3 with the "Ultimate" toroids, which drop to 2.5 ohms @ 1 kHz). Despite having measured current peaks of over 10A coming out of the amps (almost 1A per section!), I have found the bias to be rock solid, which indicates that the tube currents have remained stable over time.
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- I agree... - Kevin 12:29:02 03/04/00 (4)
- Yow! - Dan 17:19:16 03/04/00 (3)
- Re: Yow! - Kevin 13:25:20 03/05/00 (2)
- Excellent response! - Dan 15:27:02 03/05/00 (1)
- Re: Excellent response! - Kevin Covi 13:35:29 03/06/00 (0)