Uneven Cathode Bias Resistance

[HaroldBrooks]

I solved a slight Red Plating problem on one of my Gretsch (Valco) amps a While back (6162) by adding an additional 50 ohm 3 watt resistor to one of the 6973 tubes in the pair. I already adjusted the shared cathode bias resistor from the stock 250 ohm cement to a new 350 ohm Ohmite wire wound, but when I checked the temperature of both tubes with an infrared gauge, one of the tubes measured around 450F degrees while the other was cool at 300F. There was also a super faint red on the plate of the hotter tube after running for about 30 minutes, and adding the additional 50ohm resistor put both tubes approx. at around 360 degrees with no traces of red plating, and a Plate dissipation of roughly 12.5 watts on each tube.

Now for my question : Even though I checked the Plate dissipation using the output transformer shunt method and both tubes are very close in plate dissipation, will the additional 50 ohm resistance on the one tube change the function of that tube in any significant manner (audible or otherwise) ?

Let me say that the amp sound superb the way it is right now, but I was just wondering in a purely academic sense, how the dynamic balance will be between the two output tubes will be affected, given they both now run off of a shared cathode bias resistor, while one has an additional bias resistor on pin 7 (the 6973 cathode).

All comments welcome ! Thanks in advance !

https://music-electronics-forum.com/..."1554488349809

[Dave H]

A common cathode resistor exaggerates the bias current difference between the tubes.
I'd try using a separate 680 ohm cathode resistor and capacitor for each tube.
I tried it on one of my EL84 amps and the mismatch changed from 10mA to 2.5mA.

[eschertron]

caveat: I can't answer your question.

re: red plating. Have you checked the integrity of the coupling caps at the power tube grids? A leaky cap there could be a problem. Also, there' an odd grid leak arrangement with a split load on one side that feeds the NFB. Could that be causing an imbalance?

Please note that due to the seriousness of this issue, I've refrained from making another in my litany of "those Gretsch engineers must have been high when they designed this" jokes.

[HaroldBrooks]

All the coupling caps appear good, at least next to no hum in circuit even when dialed up to 10, and no measurable DC on the grid side of the coupling caps (ceramic caps), so they seem to be doing their job in that respect. I think the hot tube just have a larger draw than the cold one, but since I added the other resistor, they are now nearly perfectly matched at an idle, and no red plating even when the amp was played and idled for more than an hour. Glad you mentioned that freaky grid leak arrangement. I would like to understand that one day. I am overdue to get a scope, and maybe that would help to analyze such things.

I have three Gretsch amps, and this is the best of the lot (go figure) despite the initial red plating I now "Cured". But I would like to know if the balance between the pair is somehow changed by running two different resistances for each tube, and more over, a portion of the resistance (350 ohms) is shared commonly.

Hey, it's a slow news day, and I needed to ask about something...

Thanks for your help and advice eschertron.

Everyone is great at helping newbies like me on this board !!!

[g1]

Further to what eschertron mentioned about whether it could be a circuit issue (rather than tube). An important step that you did not mention would have been to swap the tubes around and see if the higher current stayed with the tube, or with the circuit.

[HaroldBrooks]

Forgot to mention that, the higher current did indeed stay with the tube, So I figured it's a tube related issue.

But my goal was to keep both of the original tubes from that amp if I could, and adding the other resistor to the hotter tube did the trick... For this combo of tubes that is. I am aware that with some tubes, the tube itself might have issues and have to be discarded because it doesn't play nice in a pair. The reason I figured I could 'Fix' the issue with this particular 6973 tube by adding some resistance, is it didn't look burnt like some of them get, and the original Plate dissipation that Valco set for those tubes was way higher than the RCA spec sheet from the beginning. Now I have both tubes pulling just a bit over 12 watts, and no red plating, but one requires a bit more resistance to bias that way.

Thanks for throwing that into the mix, I learned that trick only recently regarding swapping tubes to see if the current follows the tube.

It never hurts to remind me, as I have only been involved with fixing and modifying my own amps for about 2 years, and I am still a novice, but one who want's to learn from you guys, so keep up with the suggestions always, please !

Thanks g1.

[SoulFetish]

The added 50Ω resistor might be fine to stabilize this particular output tube, but could (and probably would) introduce some operating imbalance in any replacement tubes. You could change your circuit to a little to allow you to adjust the cathode resistance in either output tube. By connecting each cathode to the outside lugs of a power rheostat (say 100Ω, or whatever) and connect the wiper to the shared biasing resistor, you could adjust the cathode resistance for each tube for balance. But you would have to use a rheostat that can handle the power of the cathode current through it's resistance, and these are large and expensive. Diminishing returns in my opinion, and probably not a great solution.

I would recommend Dave H's suggestion of equal individual cathode resistors as a better modification for your amp. This is pretty effective at allowing imbalanced tubes to self bias to a more balanced DC operating point, regarding quiescent (idle) current.
Regardless, the simple fix might be that it's time to replace your output tubes. Without any way of monitoring the output waveform, and mains current, it would be difficult to put these tubes through any kind of stress test to see if one of them is weak, or faulty.

[Tony Bones]

My guess is that if the tubes are mismatched at idle, then they're almost certainly mismatched dynamically. What's the advantage of adjusting a pair of mismatched to have the same idle current if they're going to show their mismatchedness as soon as you start playing?

Bottom line: if they sound good, who cares if they don't idle at exactly the same current?

Having said that, tremolo that varies the bias on the output tubes shows that the bias does affect how the tube behaves dynamically. So, maybe inject a signal and adjust the bias while observing the signal level at the two plates?

[SoulFetish]

DC offset in a push-pull OT can saturate the core if the imbalance is great enough. So one advantage is better operating condition for the OT. It also allows the tubes to more evenly share the load, and can keep on from prematurely failing or aging.
Having said all that, guitar amps are traditionally pretty resilient. They seem to hold up under a lot of shit we probably “shouldn’t” do. So, end of the day, if a tube fails - replace it.

[HaroldBrooks]

Good Point Tony Bones ! The amp does sound very good right now, an I've played the heck out of the amp since I did that adjustment without any sign of red-plating or any sonic issues that I could detect, so my questions is moot in the real sense.

I asked the question so I can get a handle on "what would happen" if things were even more extremely out of balance. My singular guess is that eventually it would cause an issue... But I'm just not sure what the sound would be like or how it would change the amp's character.

I mucked around with big changes to the balance on one of my other Phase invertor tubes (a Supro Supreme circa 1949) and I actually like the results of the imbalance at the PI tube level. The amp sounds a bit 'wilder' with a bit more gain, and more compression of the signal, and without adding any significant amount of 'dirt' to the sound. It's now a bit closer to a more modern rock amp, but sounds like it's still fused with it's old Blues amp character, and that's what I hoped for. So in essence, that was a big win tone wise to intentionally increase the PI tube Plate voltage and bias on only one side.

SoulFetish mentioned using a rheostat, and that would be easy on the PI tube as per the low voltage across that tube's cathode resistance, so I will experiment to see at what point it sounds either better, or worse.

Yes, the one 6973 power tube probably has some issue just starting, as per it's propensity to red-plate slightly, but the small resistance seemed to cure it for the time being, and I am happy with that for now. If it fails in the future, I have several tubes ready to take the slot and I can quickly unsolder the extra 50 ohm resistor I have there now.

Thanks for everyone's expert input so far !!! I love this site, and I am learning a lot from your guys !!!

[Tony Bones]

Harold, thanks for reporting on your work. I realized just now that my comments might have seemed negative, as though I were trying to discourage you or anyone from spending their time this way. That's not the message I intended! Really, my comments were just open questions meant to stimulate thought and maybe conversation. Please don't stop experimenting or reporting your results here!

[HaroldBrooks]

No problems whatsoever ! Working on amps is good therapy for me, and I am happy to be able to do what I can at this point, and always into learning new things and challenging what I 'think' I already know.

I've made a couple of mistakes along the way, and that's how you learn, so please don't hold anything back, I can take it !

I am still fascinated with how it all works, and encouraged by the results of the repairs and mods I have done over the last two years. I could barely solder a good joint just a short while back, but now I've done cap jobs, replaced resistors, frayed wires, fixed bad solder joints, shorts, and bad sockets, and added filters and gain to different stages. I would have never believed it back then, but realisticly, I still have tons to learn about amps as that's child's play to a real tech with experience.

I used to play in an electronics dump when I was 4-6 years old (No joke !) and probably smashed a bunch of great tubes back then (Yikes !), and it took many decades until I finally got enough free time to start fixing my own guitar amps, but here I am, and It's GREAT !

Once again, Thank for your help !

[Gingertube]

The extra 50 Ohms in the "strong" tube cathode will degenerate (reduce) its gm to be closer to the "weak" tube and hence give better signal balance in the push pull - PROVIDING that the extra 50 Ohms is not bypassed with a cap.

By how much? Typical impedance looking into the cathode is 1/gm which for a 6973 (gm = 4.8mA/V) will be 208 Ohms,. So extra 50 Ohms (unbypassed) will give 258 Ohms effective and that means new effective gm is 1/258 = 3.9 mA/V (about 80%).

If the extra 50 Ohms is bypassed than you are only fiddling with DC balance not AC (signal) balance.

Cheers,
Ian

[HaroldBrooks]

You have a window to my workbench ! I did bypass the 50ohm resistor a couple of days later, as it didn't sit well with me that it was all alone as an unbypassed leg of the pair. In any event, the sound is still brilliant, and I wouldn't change a thing tone wise (famous last words !). Thanks.