2 Tubes fixed bias/2 tubes cathode bias

[lowell]

Hi all,
Working on Twin Reverb... 4 6L6s. I'm trying to figure out a way to have 2 tubes fixed bias and 2 tubes cathode bias. I think I'll need to split each of the phase inverter signals into 2 signals in order to separate the bias voltage from the cathode biased signal. Does anyone know how I could do this w/o tainting the signal?

Thanks!

[cbarrow7625]

The easiest way to do that would be to run two seperate coupling caps off of each half of the phase inverter. One cap on each side would go to your 220K grid resistor to ground (cathode biased side) and the other cap would go to a different set of 220K grid resistors that have a junction at the existing bias circuit.

The 2nd coupling cap would block the DC bias from getting to the power tube grids on the cathode biased side.

This does require some comprimises & requires you to make some decisions & tweaks based on listening to the amp. By splitting each phase inverter output 2 ways, you are going to be halving the load each half of the phase inverter sees. Instead of each side of the phase inverter running into a single 220K resistor, it is now running into a pair of 220K resistors that are essentially in parallel (110k ohms). Have you ever changed the power tube grid resistors in a Fender amp to 110K? Does it make much difference on the sound? Well, what we do know is that you will change the LF roll-off frequency somewhat but not drastically. Loading down the output of the phase inverter with half the value of resistor usually does sound different though (to me it ends upsounding more constricted with less voltage swing & headroom).

You could compensate for that by increasing all of the 220K resistors to 470k to get the load on the phase inverter & the LF roll off back to where you started. But how will the power tubes like having 470K grid leak resistors? I don't know. They may like it just fine or it may cause premature power tube failure when running at high volume levels. Those resistors are usually kept fairly low in value...Anybody have a bead on that one? I would guess that it would probably be fine. However, will it affect the tone? Maybe a comprimise of 330K power tube grid resistors would be the best solution.

So, splitting the signal for a mixed fixed / cathode biased scheme is not too tough. The tough thing is making sure you know what trade-offs you are making and how (and where) to adjust for them if you are trying to keep or acheive a particular tone.

Sounds like fun. I've wanted to do that mod myself but never have. I have an amp that i can switch to fixed on one side & cathode on the other & I like that a lot. Let us know how it comes out. I expect it will be pretty nice.

Chris

[lowell]

Chris,
Awesome thanks for the info. Yeah I forgot that simply adding 2 more caps would block the DC bias voltage... nice. I will try w/ the 330k's, sounds like the best compromise at this point. I want to have some warm sponginess/sustain of the cathode bias but also still would like to keep some of the headroom.

Thanks!

[stokes]

Heres a simpler way to accomplish a similar effect.Leave the bias set-up the way it is.Put a 100ohm resistor on each pair of cathodes to ground instead of grounding each cathode.I used this on a 2x6L6GC parallel SE output.You will have to tweak the bias voltage and try different value cathode resistors.I ended up with a 120 ohm cathode resistor.Sounds pretty good,and a lot easier than adding coupling caps etc.

[lowell]

Yeah that's definitely simpler... I guess now I'm curious if mixing all four tubes biasing versus 2 cathode bias and 2 fixed... I wonder if both create the same dynamic and sonic results... anyone??

Thanks

[Old Tele man]

...you could DOUBLE the size of the separate 1500-ohm grid resistors (if used) since they'll be effectively in parallel (per side), if you want to "keep" same tonal effects as existing circuit.

...conversely, you could HALF the size of the two BIAS isolation resistors, since now they'd been working into half as many tubes.

[stokes]

Yeah that's definitely simpler... I guess now I'm curious if mixing all four tubes biasing versus 2 cathode bias and 2 fixed... I wonder if both create the same dynamic and sonic results... anyone??

Thanks

Dont quite get what you are getting at here.With my suggestion you are using all 4 tubes in a sort of hybrid fixed/cathode bias thing.Fender did it on some amps during the BF to SF transition.Dont remember which amp I saw it on,think it may have been an early SF Bassman head I did some work on some time ago.

[lowell]

What I'm going for is not 4 tubes w/ mixed bias. I'm going for 2 tubes cathode bias and 2 tubes fixed. Not sure if this will be sonically similar to 4 w/ mixed. Make sense? I see what you're saying about mixing all 4. One reason being that I swapped out 2 of the 6L6 w/ 6V6 and would like the 6V6 to be fixed bias and the 6L6 to be cathode.

[stokes]

Okay,I see where you are going now.I cant comment on how it will work sonically.But I think if you do it as you describe you may want to have one 6V6 and one 6L6 on one side of the OT and one and one on the other side.If you were to put a pair of 6V6's as a pair on one side of the OT and then a pair of 6L6's as a pair on the other side you will have a big mismatch between the two sides of the OT.If you put a cathode biased 6L6 and a fixed bias 6V6 on one side of the OT,and the same on the other side you will have a more balanced output.See what I mean?

[lowell]

Yes! My plan exactly... you're very right that it'd be unbalanced the other way. This silverface twin is sounding killer already!! I converted to cathode bias, tubes are at around 36ma per tube, doubled the tone stack slope resistor to 200k (reduced bass a bit), and have converted 2 6L6s to simulated triode mode. I will be chaining both channels together so vibrato can be used in channel one also, this will also put both channels in phase w/ eachother making it easy to daisy the channels to use all gain stages and tone stacks. I will possible be substituting 12ay7s for the 1st 2 tubes... we'll see. I'm thinking about lowering the preamp voltages for a more tweed dynamic response.

Lowell

[DirtyGeorge]

The grid reference resistor can be much higher with cathode bias. So if you stay with 220K on the fixed bias tubes and use 470K on the cathode bias tubes you will not increase the load on the PI by as much. You will want to go to smaller coupling caps for the cathode biased pair to keep the same time constant.

DG