Bias precision?
September 16th, 2016, 10:39 am
Hi, I'm wondering how precisely I need to set the bias on my amp. I have a dynaco st-70 clone (http://dcaudiodiy.com/phpBB3/viewtopic.php?f=4&t=703). I'm aiming for 20mA cathode current. Right now I'm reading 20mA at three of the test points and 23mA on the fourth. Is this acceptable? How close do I need to get? <1mA? <5mA? <10mA?
I'm asking because unfortunately I made it kindof hard to adjust the bias between the tubes on one channel. I basically have to flip the amp upside down. I can easily adjust the bias between the channels. Hope that makes sense. I'm thinking about ways to make it easier.
I'm asking because unfortunately I made it kindof hard to adjust the bias between the tubes on one channel. I basically have to flip the amp upside down. I can easily adjust the bias between the channels. Hope that makes sense. I'm thinking about ways to make it easier.
Re: Bias precision?
September 16th, 2016, 12:03 pm
I did a ST-70 re-build in my evil past (see scanned photo) and found that the sweet spot for EL-34s was 38-ma cathode current for each tube. Any less and the sound gets edgy -- any more the sound gets dark plus you end up pushing the tubes harder than needed and shortening their expected useful life.
On the subject of balance -- the devil is in the details -- I have always used matched pairs in push-pull amps. The closer the match the better.
On the subject of balance -- the devil is in the details -- I have always used matched pairs in push-pull amps. The closer the match the better.
- Attachments
-
Re: Bias precision?
September 16th, 2016, 3:10 pm
That is one good-looking amp! My 'st-70' isn't a real st-70; it uses 7591s as output tubes which have lower bias current.
Still wondering about how matched is matched...
Still wondering about how matched is matched...
Re: Bias precision?
September 16th, 2016, 3:53 pm
I don't mean to be pedantic, but doing the math gives me a 15% mismatch. (I think better in percentages....)
That sounds like too much.
I'm thinking 5% is more reasonable, which is 1mA in your case.
That sounds like too much.
I'm thinking 5% is more reasonable, which is 1mA in your case.
Re: Bias precision?
September 16th, 2016, 4:00 pm
Compliments are always welcome. I could not duplicate that effort again. The full differential boards are no longer available and Maryland EPA ended chrome plating.
A company in the city did mil-spec plating and anodizing. When I picked up the steel parts all plated and polished I was blown away. They far exceeded my expectations.
I did that amp. over 35 years ago. It is still running and spending its retirement years in Spain with its original owner who became a very good friend.
A company in the city did mil-spec plating and anodizing. When I picked up the steel parts all plated and polished I was blown away. They far exceeded my expectations.
I did that amp. over 35 years ago. It is still running and spending its retirement years in Spain with its original owner who became a very good friend.
Re: Bias precision?
September 17th, 2016, 1:34 pm
FerdinandII wrote:I don't mean to be pedantic, but doing the math gives me a 15% mismatch. (I think better in percentages....) That sounds like too much. I'm thinking 5% is more reasonable, which is 1mA in your case.
That sounds reasonable. I'll go with that. Thanks!
Pelliott321 wrote:A great source for info
http://www.tubebooks.org/
Thanks!
Re: Bias precision?
September 21st, 2016, 9:38 am
I have found that as little as 1/4 mA can affect transformer saturation and distortion in the bass and sound in general. That is why I aim for the lowest bias current that is acceptable for crossover distortion because it is easier to control 1/4 of a mA out of 5 mA than it is to control 1/4 mA out of 30 mA. It is the absolute difference, not the percentage here. Also, a separate bias pot should be installed for each output tube(s) for the pull and the push. This is also one of the driving factors for me to have autobias, but it is tricky to do for class A-B or class B.
--David--
--David--
Re: Bias precision?
September 21st, 2016, 10:35 am
Based on what David just advised my guess is that the sweet spot I talk about is where the crossover distortion is minimized. When I go lower than the sweet spot the sound gets edgy not unlike a 1970s solid-state amplifier. Of course with single-ended there is no push-pull crossover to worry about.
Re: Bias precision?
September 21st, 2016, 10:53 am
dberning wrote:I have found that as little as 1/4 mA can affect transformer saturation and distortion in the bass and sound in general. That is why I aim for the lowest bias current that is acceptable for crossover distortion because it is easier to control 1/4 of a mA out of 5 mA than it is to control 1/4 mA out of 30 mA. It is the absolute difference, not the percentage here. Also, a separate bias pot should be installed for each output tube(s) for the pull and the push. This is also one of the driving factors for me to have autobias, but it is tricky to do for class A-B or class B.
--David--
Thanks, David. I have one bias pot for each channel. I have noticed that when I adjust the bias pot on one channel, it affects the bias of the other. I think this is because my negative bias supply is not regulated. If I install 4 pots, won't it be even more difficult to set my bias correctly? Should I regulate my bias supply to get around this?