high b+ and bias

[wagdog]

Greetings all,

In my never ending quest to not leave well enough alone, I've put my twin project amp back to fixed bias (i.e. adjustable bias voltage) from cathode bias (which sounded great, but sort of lacked that glassy shimmer I love from a twin).

This project amp started out as a super twin reverb which had a very high b+. I gutted it and built an ab763 style twin inside the chasis. The b+ however was still over 500v, so I dropped it a bit with some zeners (4x10v 5w ea) placed after the first filter cap (so as to limit the ripple current) and before the node that goes to the center tap of the OT and through the choke.

With the bias set at about 40ma through ea of the 4x6L6's (EH's), I'm seeing about 455v on the plates - 455 x .040, ~18w dissipation per tube. This sounds good on paper, however it doesn't translate to a great tone (sounds almost "strained"?) and there is a definite low freq "hiss" going on w/the amp at idle.

Bringing the bias down to 22ma, I'm seeing about 465v on the plates, 465 x .022, ~10w dissipation per tube. This will supposedly sound "cold", but it actually sounds very nice! There is no low freq hiss (this disappears at around 26/27ma) either. Very glassy.

The 6L6EH is rated at 500v on the plates and 450v on the screens. The screens in my amp are seeing about 462v - probably ok. I could drop the b+ another ~10v, but don't see the point at the moment.

Here's the question: when does the ballpark 70% idle dissipation ballpark figure start dropping and what is the ratio that it drops? If I had kept the b+ at 500v, and biased at 70%, I suspect the tubes would red plate in short order. The tone I'm getting at 10w dissipation (465v/22ma) is really nice, so I'm going to stick with it, but I would like to know more of the details as to why it sounds good.

Thank you!

Wag

[MWJB]

It sounds good because you have biased it by ear, to your taste. This is the way that you should do it...then check the % dissipation to ensure you don't burn anything up.

The 70% rule of thumb is merely a starting point that you aim for when lacking any further details/info. If you bias to 70% dissipation at any reasonable voltage that the plates will stand, you shouldn't redplate the tubes in normal operation because as the voltage goes up, the current comes down (70% for a 25W tube is 17.5W, or 43mA at 400vdc & 35mA at 500vdc)...trick is knowing (not guessing/assuming) what your max plate dissipation is - 6L6GC types vary from low 20's to 30W per tube depending on brand. Susceptibility to redplating will also depend on how hard you drive the amp & as you have found out...you often do not need to hit 70% to achieve a desirable tone.

[imaradiostar]

This hiss is the anger of the tubes...if you let the anger out sometimes the tubes will go on strike during a gig.

jamie

[wagdog]

It sounds good because you have biased it by ear, to your taste. This is the way that you should do it...then check the % dissipation to ensure you don't burn anything up.

Yeah, that's the truth right there 'eh. I guess I was just alarmed at HOW much better it sounds using what is supposedly "cold", like around 35% of max dissipation (the 6L6EH is rated at 30w at idle and I'm running 'em at 10w). But the proof is in the tone and the tone is good.

This hiss is the anger of the tubes...if you let the anger out sometimes the tubes will go on strike during a gig.

Yeah, they let you know at the least inopportune time that they ain't happy.

[Enzo]

So the lesson is, don;t listen to people on the internet...

Do what sounds good.

[pdf64]

The b+ however was still over 500v, so I dropped it a bit with some zeners (4x10v 5w ea) placed after the first filter cap (so as to limit the ripple current) and before the node that goes to the center tap of the OT and through the choke

I'm not sure whether that arrangement would put significant impedance in series with the plate supply.
However, I doubt it would have any benefit in terms of ripple and it would be more 'normal' to put them either after the rectifier/before the reservoir cap or between the B+ winding CT and ground.
So, in the spirit of experimentation, I'd try that arrangement also, even though you're happy with how you have it at the moment.
Also, the cathode biased arrangement you had before is difficult to make compatible with such high B+; to get idle plate dissipation reasonable, large cathode resistors usually result in significant crossover distortion and a constricted tone at high signal levels.
Zener biasing can help to reduce that though.
So you may not have given cathode biasing a fair chance - good sparkle is achievable.

[imaradiostar]

Is it fair to say you want the bias, screen and plate voltages all to sag relative to each other? I was playing with some tubes on the scope last night (6v6's in push-pull) and it seemed to hold together best when the three values tracked linearly from rest to full overdrive.

jamie

[wagdog]

Thanks for the reply...

I'm not sure whether that arrangement would put significant impedance in series with the plate supply.

I'm going on what I read in Merlin Blencowe's book Designing Power Supplies for Tube Amplifiers, pg 213; "Dissipation can be reduced by placing the zener after the reservoir capacitor so that it only passes DC load current." He doesn't mention the zener adding impedance... so I'm curious if it is something I should investigate. Of course he's talking about a single zener, and I've got several strung together.

Also, the cathode biased arrangement you had before is difficult to make compatible with such high B+; to get idle plate dissipation reasonable, large cathode resistors usually result in significant crossover distortion and a constricted tone at high signal levels.
Zener biasing can help to reduce that though.
So you may not have given cathode biasing a fair chance - good sparkle is achievable.

FWIW, I did have the B+ reduced when it was it cathode biased, albeit configured a bit differently. I did it by lifting the ground end of the rectifier and running that through the zeners to ground. It did sound good, not quite as glassy though. The "squish" it had was nice.

I'm really having fun trying different things w/this amp and do appreciate your input and insight.

[wagdog]

So the lesson is, don;t listen to people on the internet...

Do what sounds good.

Well yes... but I do like to know why things sound the way they do.

[pdf64]

OK, go with what Merlin says; take no notice to some guy on the internet. I wasn't sure and hadn't seen it done like that before. Thinking about it, if the zeners are conducting (always the case) then their impedance will be low. It's just that the 'hiss' got me suspecting that something wasn't adequately decoupled.
Re cathode biasing, I'd look to get the B+ below 400V with 6L6. Otherwise large cathode resistors are needed and the bias/cathode voltage can double when overdriven, leading to an unpleasant variant of crossover distortion and a constricted tone.

[wagdog]

Re cathode biasing, I'd look to get the B+ below 400V with 6L6. Otherwise large cathode resistors are needed and the bias/cathode voltage can double when overdriven, leading to an unpleasant variant of crossover distortion and a constricted tone.

OK, will do. I'm sure I'll put it back to cathode bias at some point. I did have a switch to bounce back and forth between the two at one point, but liked the cathode bias so much I took it out for some reason... and now I'm back experimenting with fixed bias. Going from memory (I'm not at my workbench), w/cathode bias I believe I had something like 380v between the plate and cathode, and 38v on the cathode resistor... could be wrong though, going from memory.