Change bias resistor or add voltage dropping resistor - or both

cottontails1959

Tele-Meister
Joined
Jan 23, 2019
Posts
151
Age
62
Location
Pittsburgh, PA
I modded a Fender MM (cathode bias) bass amp (6V6 version) by swapping out the stock PT and OT for a Hammond P-T290AX (275-0-275 taps) and P-1760E respectively, in order to drive the tubes a little harder. My intent was to run 315VDC on the 6V6 plates, and take whatever comes after that. So with my 123VAC wall voltage, I'm getting 289-0-289VAC, which threw me off at first, but if the V-bump up is about 4.5x, then 3*4.5 about 14 volts, so 275+14=289. Unfortunately for me, that's giving 361VDC out of the SS rectifier ... B+1 = 361 ... B+2 = 349 ... B+3 = 308 ... 6V6 plates = 353 ... 6V6 screens = 348 ... cathode = 20.7 (actual K-resistor = 216 ohms) ... according to robrob's calculator sets my output at 108% dissipation. If I was at my intended plate voltage, dissipation would be 95%.

So I can deal with the bias at the existing voltages by going to 270 ohm cathode resistor (easy), but still leaves more voltage on the 6V6 plates and screens. I could run a ~15K voltage dropping resistor from the rectifier to B+1 (not as tidy, but do-able). Or do a bit of both. The dropping resistor would be 15K*100mA=1.5 watts, so a 3 or 5 watt resistor?

I see Fender schematics with a lot higher B+1 and plate voltages on 6V6s, so I'm hoping I good with the first solution.

Any thoughts on adding screen resistors and grid stoppers to the 6V6 tubes?

Cheers,

Peter
 

King Fan

Poster Extraordinaire
Ad Free Member
Joined
Jan 1, 2013
Posts
8,184
Location
Salt Lake City
I may be missing something; lots of details there. But in simplest terms are you saying you have a cathode bias amp running at 108% dissipation? Usually we'd say that's fine — cathode bias is 'self-biasing' within reasonable limits.
 

Phrygian77

Poster Extraordinaire
Gold Supporter
Joined
Apr 30, 2016
Posts
5,480
Location
Crawfordville, FL
If what you trying to get is slightly more output power and headroom, then you might want to cool the bias a bit. If you're simply concerned about being over 100%, then I wouldn't change it. A lot of 5E3s run like that.
 

cottontails1959

Tele-Meister
Joined
Jan 23, 2019
Posts
151
Age
62
Location
Pittsburgh, PA
@King Fan - no, you're not missing anything ... just trying to confirm what (I think) I already know. What wigged me out at first was the high transformer voltage - I confirmed the 325-0-325 co-winding was actually running 355-0-355.

Thanks both!
 

JamesAM

Tele-Holic
Joined
May 13, 2020
Posts
738
Age
36
Location
Virginia, USA
Generally, (!) if the 6v6es aren’t red plating, you’re ok- even over 100% dissipation. Ask me and my home brew 5e3 how I know.

If you do want to cool it a bit, buy a 300 ohm and a 330 ohm 5w bias resistor and test them both out in place of your 270r. Your plate voltage actually may not go up that much, or at least enough to mess with performance or safety. Mine only went up a few volts, but it took my b+ down like 20 volts to a sane number. Still running the same jj 6v6es I’ve had in there since 2011 at about 95%. Sounds a peach.

You could also run a string of zener diodes off the high tension center tap as per Robrob to cool the b+. You’re trading heat for voltage at that point, and your amp is probably gonna be pretty hot anyway. I believe the dropping resistor is the least recommended method due to safety.

Edited to add: I also used a Hammond pt in mine and it was like 12 Volts hot. Also, what tubes are you running? If you’re red plating, get some JJ 6v6s. They’re way overbuilt and I think can take something crazy like 18 watts dissipation based on some unscientific study I think someone on another forum did years ago. Tung sols immediately red plated for me even at 94% dissipation- ymmv.
 

cottontails1959

Tele-Meister
Joined
Jan 23, 2019
Posts
151
Age
62
Location
Pittsburgh, PA
@JamesAM - I'm using Reflektor 15-15 (6P6S) tubes - I'll dim the lights to make sure they're not red plating. I might order in some of the higher value bias resistors ... I should just order 10 of everything ...
 

JamesAM

Tele-Holic
Joined
May 13, 2020
Posts
738
Age
36
Location
Virginia, USA
@JamesAM - I'm using Reflektor 15-15 (6P6S) tubes - I'll dim the lights to make sure they're not red plating. I might order in some of the higher value bias resistors ... I should just order 10 of everything ...
It looks like some old forum posts here indicate that those 6p6s may not be able to handle higher plate voltages, but some folks apparently have seen upwards of 410v with no problems. A 300r might get you where you want to be. 5w or higher would be the safe bet imo.
 

cottontails1959

Tele-Meister
Joined
Jan 23, 2019
Posts
151
Age
62
Location
Pittsburgh, PA
It looks like some old forum posts here indicate that those 6p6s may not be able to handle higher plate voltages, but some folks apparently have seen upwards of 410v with no problems. A 300r might get you where you want to be. 5w or higher would be the safe bet imo.

Here's a pic of the chassis - I redesigned the eyelet board to accommodate the first two filter caps, moved the SS rectifier to the left edge, and added a bleeder / heater elevation (33vDC) to the B+1 cap. And a pic of the tubes ... the camera makes them appear brighter than they actually are.
 

Attachments

  • 20220509_220006.jpg
    20220509_220006.jpg
    180.7 KB · Views: 37
  • 20220509_220245.jpg
    20220509_220245.jpg
    27 KB · Views: 37

rdjones

Tele-Holic
Joined
Nov 6, 2010
Posts
568
Location
Music City, TN
Make note of the voltage rating on your power supply filter caps.
The stock values leave very little margin above the B+ voltage at idle.
Anything done to increase the high voltage will exceed the cap's rating if they are still at the factory specs.

edit - you posted your images while I was composing my post.
I see now that you have made more changes to the circuit than I first realized, including the caps.
 
Last edited:

cottontails1959

Tele-Meister
Joined
Jan 23, 2019
Posts
151
Age
62
Location
Pittsburgh, PA
Make note of the voltage rating on your power supply filter caps.
The stock values leave very little margin above the B+ voltage at idle.
Anything done to increase the high voltage will exceed the cap's rating if they are still at the factory specs.

edit - you posted your images while I was composing my post.
I see now that you have made more changes to the circuit than I first realized, including the caps.
Yes, I made made some "modernizations" the circuit. I believe the stock filter caps are 350V. I'm using two MOD 500V caps for B+1 (30uf vs. 20uf stock) and B+2 (20uf stock). B+3 is a 450V radial cap (22uf vs. 20uf stock). I changed C1 to 0.022uf from 0.01uf. The tone stack is Blencowe's tone tilt control.
 

chas.wahl

Tele-Meister
Joined
Dec 8, 2012
Posts
424
Location
NYC
Forgive me if this is an ignorant question, but: isn't the voltage rating of a power transformer "at rated current load"? Are you measuring the 289 (vs 275) and 355 (vs 325) loaded or unloaded? Also, I notice that you're using solid state rectification rather than a tube rectifier. In a 5E3 or 5F2-A, the 5Y3 rectifier (one of the more lossy ones) will drop about 60 V (maybe less if not vintage), resulting in about 320 V or so on their power tube plates. Though all this is stating what's pretty obvious, it leads me to wonder why, to begin with, you thought that using 275-0-275 voltage supply would result in 315 V on the plates.
 

andrewRneumann

Tele-Afflicted
Joined
Mar 22, 2020
Posts
1,809
Location
Cincinnati, OH, USA
Seems like we left a lot of questions in the OP unanswered.

So I can deal with the bias at the existing voltages by going to 270 ohm cathode resistor (easy), but still leaves more voltage on the 6V6 plates and screens.

I would still try it. Vak of 330V (anode voltage minus cathode voltage) is totally fine. Agree with @JamesAM to buy 300Ω and 330Ω. My money's on 300Ω.

Any thoughts on adding screen resistors and grid stoppers to the 6V6 tubes?

Adding some screen grid resistance is another way to drop the dissipation. It can affect the dynamics of the amp, so you have to be choosy in your values. I think in your situation, you would have to lower the screen voltage at idle so much (use a lot of resistance) that the dynamics of the amp would be severely hampered. A small amount of control grid and screen grid stopping resistance is usually advised for reasons of stability... but that is unrelated to your dissipation question.
 

JamesAM

Tele-Holic
Joined
May 13, 2020
Posts
738
Age
36
Location
Virginia, USA
Here's a pic of the chassis - I redesigned the eyelet board to accommodate the first two filter caps, moved the SS rectifier to the left edge, and added a bleeder / heater elevation (33vDC) to the B+1 cap. And a pic of the tubes ... the camera makes them appear brighter than they actually are.
Yeah, those are cooking. You’ll either need to get them cooler with:

a higher bias resistor, or

new tubes that can handle the voltage, or

a new PT that gives you the right b+, or

the zener diode trick.

You may be able to stick some jj 6v6s or other more hardy NOS tubes in there as is now and call it done at 350v to the plate if the bias current is reasonable. My guess it is and it's those old soviet tubes that can't deal with it.

My first port of call would be bias resistors and then tubes.
 

cottontails1959

Tele-Meister
Joined
Jan 23, 2019
Posts
151
Age
62
Location
Pittsburgh, PA
Forgive me if this is an ignorant question, but: isn't the voltage rating of a power transformer "at rated current load"? Are you measuring the 289 (vs 275) and 355 (vs 325) loaded or unloaded? Also, I notice that you're using solid state rectification rather than a tube rectifier. In a 5E3 or 5F2-A, the 5Y3 rectifier (one of the more lossy ones) will drop about 60 V (maybe less if not vintage), resulting in about 320 V or so on their power tube plates. Though all this is stating what's pretty obvious, it leads me to wonder why, to begin with, you thought that using 275-0-275 voltage supply would result in 315 V on the plates.

I'm new to this, so no ignorant questions ... my measurements are unloaded. This MM amp transformer mod is suggested by @muchxs in a post in the Music Master Mafia amp owners thread to make better use of the 6V6 (vs. 6AQ5 in the original MM builds). The stock PT is 245-0-245, and is supposed to give 293V B+1 and 288V on the plates. I have verified these numbers are very close measured on an unloaded amp. 315V on the plates with the higher output transformer may have been optimistic, but I was not expecting 353V. Try and learn ...
 

cottontails1959

Tele-Meister
Joined
Jan 23, 2019
Posts
151
Age
62
Location
Pittsburgh, PA
Bias resistors on order ... will start at 300 ohm 5W.

@andrewRneumann - although this drifts away from the MM circuit, if an arbitrarily large filter cap, let's say 100uf/500V, was added after the SS rectifier (making it the B+1, and the existing B+1 cap into B+2), could a dropping resistor then between it and the next filter cap without mucking things up? Or will there still be too much sag?
 

JamesAM

Tele-Holic
Joined
May 13, 2020
Posts
738
Age
36
Location
Virginia, USA
my measurements are unloaded.

just to furhter clarify what @chas.wahl was getting at: if your measurements are unloaded (ie., with no tubes installed), are you trying to estimate the potential bias voltage before you plug tubes in? Be sure your tubes are installed and pulling a current load to obtain true voltage measurements. with these old circuits/tubes and today's wall voltages, estimation won't get you anywhere. Your numbers will be noticeably higher than under load, which could account for your discrepancies.

If your measurements in post 1 are with tubes installed and drawing current, your numbers are what they are and you'll need to do one of the corrective actions we've described above to cool the bias.
 

cottontails1959

Tele-Meister
Joined
Jan 23, 2019
Posts
151
Age
62
Location
Pittsburgh, PA
just to furhter clarify what @chas.wahl was getting at: if your measurements are unloaded (ie., with no tubes installed), are you trying to estimate the potential bias voltage before you plug tubes in? Be sure your tubes are installed and pulling a current load to obtain true voltage measurements. with these old circuits/tubes and today's wall voltages, estimation won't get you anywhere. Your numbers will be noticeably higher than under load, which could account for your discrepancies.

If your measurements in post 1 are with tubes installed and drawing current, your numbers are what they are and you'll need to do one of the corrective actions we've described above to cool the bias.
oh, got it, yes, all measurements with tubes in ... and not through light bulb limiter :)
 

King Fan

Poster Extraordinaire
Ad Free Member
Joined
Jan 1, 2013
Posts
8,184
Location
Salt Lake City
Side remark. I think those *are* redplating, almost for sure, but I had a weird moment with the 'night mode' on my new iPhone camera -- which automatically switches to time-lapse when taking pictures in near darkness, and sets the time to multiple seconds to get sufficient light. With a 3 or 4 second exposure it chose, the plates in my 5E3 (which don't glow at all to the naked eye) took on a faint rosy glow. Bias was around 97%. Raises the question whether the eyeball or the camera is more 'correct'. In the meantime, I'm going to a conventional camera, wide aperture, exposure under a second.

For your amp, @cottontails1959 , I wonder if a pair of JJ 6V6s wouldn't put out the fire.
 




Top