pwr tube cathode question

[daz]

i recently posted about this elsewhere but the thread became so confusing i had to try here with a clean slate after learning what i have so far. I cathode biased my EL34 amp and it redplated a bit. Not bad, but a bit. I was using two 270 ohm resistors, one per tube. I did this because two amps i have copied or studied to some degree, the matchless chieftain and the hotcat both use a 270R per tube. I finally realized that with one per tube it was the same as using one for both tubes at half the value. in other words, a 270 on each cathode was like using a single 135 tied to both cathodes. So i figured no wonder it redplated. However, before i realized that i stayed with a seperate R per tube but increased them. i ended up at about 450R on each tube. the high gain tone was unreal. i mean, i know i've said i had great tone before, but this was insane. I had 2 players try it. one who is a world class player and they both flipped.

Now hopefully that all sunk in because it may get more confusing. At that point i was told i was likely getting very low output. So i tried a separate R tied to both tubes cathodes. I used a 270R. The amp got a LOT louder and cleaner and that unreal hig gain tone went right out the window. It still sounds good.....about like a good JCM800, but that amazing gain tone was lost big time. This is where i get really confused and need an answer. If i was correct in that using 2 resistors, one per tube would be the same as using a single R of 1/2 the value of either of the two R's, then changing from the two 450R's i had with that killer tone to a single 270 tied to both tubes should have given me the same specs. But not only was it way louder, but that unbelievably spongy and rich tone was gone.

Can anyone explain this to me in a clear way? Is there something about using seperate R's on each tube that makes for a very different much richer distortion, and even if thats true, why would using a single R make it so much louder if i halved the value? Is that clear? I'm trying to figure out how to get the output up like it was with the single 270 but retain the complexity i have with a 450R on each tube.

[Robert M. Martinelli]

Hi Daz,
you're understanding is correct from a DC standpoint, which involves only resistors and DC currents and voltage drops, but I suspect the change in the tone you' re talking about has something to do with the bias resistor(s) interacting with the tubes' inter-electrodes capacitance and the circuit stray capacitance as well - the bias resistor could form an RC circuit which could be the reason for the tone change IMHO. Even if DC current flows at all times, looking at it from an AC standpoint, this RC circuit is present only when the relevant tube is conducting, and, with different values ( one resistor per cathode or only one resistor ) the cutoff freq changes accordingly, and this could be the reason for the tone and gain change. You could have some change in the harmonic content or some high frequency oscillation....

I could be wrong, but it would be interesting to take a look with an o-scope in both cases to confirm or negate this.

Hope this helps

Best regards

Bob

[tubeswell]

Well I don't know how simple this is going to be, but I'll give it a shot.

If you have a single common cathode resistor for two triodes/tetrodes/pentodes, it only needs to be about half the value of a cathode resistor used on a single tube of the same type for the same bias level, all other things being equal. This is because it is passing twice as much current through the resistor (all other things being equal) to get the same bias voltage as you would with 1 tube drawing half as much current, ergo;

E = (2 x I) x (R/2) or (1 x I) x (1 x R)

2O V = (80mA) x (250R) or (40mA) x (500R)

E is cathode voltage

I is tube current

R is cathode resistance.

Not sure why your practical experiment yielded different results - Did you have the cathode resistors in both versions bypassed (or unbypassed, as the case may be)?

[Robert M. Martinelli]

Hi tubeswell,

( How are you doing ? )

I suspect he has them un-bypassed, given the tone and gain change he's witnessed

Regards

Bob

[Steve Conner]

Hi all,

If we compare, say, independent 500 ohm cathode resistors to a shared 250 ohm resistor, we get exactly the same DC conditions: the tubes bias up to the same idle current.

However, if the cathode resistors are unbypassed, the AC conditions are different. The independent resistors give negative feedback that lowers the gain, power output and damping factor: and all the interactions with the NFB loop that implies.

A shared resistor doesn't give negative feedback, because the AC components of the two tubes' cathode currents cancel each other. One tube is turning on as the other turns off, so the total current through the resistor stays the same.

In Class-A at least. If you drive it into Class-B on signal peaks, then you get a similar effect to independent resistors, but with twice the NFB.

[tubeswell]

Hi Bob

The sun has shining alot here lately despite the doom and gloom in the headlines. (Who needs news when you've got the weather?) My day job and family stuff has kept me away from the workbench for the last month, so I'm using the time to accumulate more bits and pieces. You?

[Robert M. Martinelli]

Hi Bob

The sun has shining alot here lately despite the doom and gloom in the headlines. (Who needs news when you've got the weather?)You?

@tubeswell :
Amen! As a matter of fact I am keeping ( or trying to keep ) myself away from all the craziness that surrounds us, and trying to make a better use of my time rather than wasting it listening/watching some bad news I cannot change....And, yes, the weather helps somehow, as here we're having a very mild November....

@Steve : Thanks, Steve, I tried to express the same concept you explained so clearly ( the differences between DC and AC conditions ) but as a matter of fact your explanation is way better than mine ( I guess I still have a lot to work on my English )

Best regards

Bob

[daz]

Thanks all. And Steve, thanks for that. I knew i had to be right that for example a shared 250 should be the same as far as DC bias as two 500's, one for each tube. And YES, i am using them unbypassed. The reason being is that is sounds/feels better, but there is a catch....it sounds/feels better when i have seperate resistors, but with a single resistor it seems to sound as good or even better with a cap. And adding caps to seperate resistors seems to cause the tone to revert to how a single resistor sounds, that of course being much less complex and rich. That seems to speak to what was said about bypassing them.

So then here is the question...i still don't understand why using 2 seperate resistors sounds so much better, (actually it FEELS much better even more than sounds) But can anyone suggest how i can get that sound w/o the power loss? heres what i'm thinking. At 450R per side it's like using a single 225 as far as DC goes. The amp was slightly redplating on one tube with two 270's which would be like using a single 135R as Dc goes. So i'm thinking that if i go above 270R with two resistors till it it stops redplating and am able to get it to stop and go a bit further for safety, if at that point i'm not close to the 450 per side i have now i should have my tone still and no redplate. If anyone has any futher thoughts about that please let me know. In any case, thank you all very much for your help. i wish you guys could play this thing. (clips won't do because a huge part of it is feel) i really got lucky and found the magic tone i never in a million years thought i'd find. But only with 2 seperate resistors ! weird !

[daz]

Ok, bear with me, as you all know i'm no tech but here goes. It just hit me that tho it won't red plate with a bit larger than 270R's on each tube. So i added 68k R's to each giving each tube a 338R. And sure enough no redplating. But the volume is still rather low. secondly, the bias calculator i use at weber is confusing because i'm not sure whether i'm getting the correct results due to the fact i'm using 2 resistors. It DOES indeed ask how many tubes is shared by the resistor, and i can enter "1". But i'm not sure whether i'm measuring things right. i enter the voltage across the cathode R, then the actual resistance of the one resistor, then the plate to cathode voltage. I get 30 watts dissipation and 76MA. So i assume each tube by itself is dissipating 30 watts which is too much. but it doesn't redplate. So you can overshoot the tube's specs w/o redplating, right? And the tube will die before long, correct? The reason i had settled on 450R on each side before is that it's what the calculator said would get me 25 watts.

I'm kinda lost here as to what to do, so any help would be appriciated. I suppose i can just leae it as it was with 450Rx2, but i really would like the output to be loud and clean like it should be. Some elsewhere suggested i may be only getting 10 watts output and that the dissipation spec is not related to actual output. they may be right.

[Steve Conner]

Well, how about a compromise. Instead of taking the individual cathode resistors to ground, just hook them together, and put another resistor from this node point to ground. Then you can mix the sonic flavors of individual and shared resistors.

For even more tweaking which could verge on OCD, you could try putting bypass caps on some of the resistors and not others, or even using cathode bypass caps that have resistors in series.

However, I suspect the good high gain tone is caused by the same factors that ruin the clean one. The shared resistor, or individual resistors bypassed, give the most transconductance (aka "gm") and output power. By adding the resistors, you hurt the power a little, but the gm takes a big hit.

This means that, while the tubes may still give very nearly the same power, they need more drive from the PI to do it. So ultimately, maybe what's making your high-gain tone rock and your clean suck is PI distortion.

Footswitchable cathode bypass caps, anyone?

[daz]

But Steve, the thing is neither scenario sucks ! the amp sounds great even with a single R. What you must understand and believe when i tell you is that it REALLY DOES sound phenominal with the two resistors as opoosed to one. I had 3 JCM 800's, one that was better than the others by a good margin. With a single r my build sounds about like that amp. With two it literally sounds a lot better. one of the guys i mentioned who played this amp owns a carr and a suhr badger and he was floored by it. I'm no tech by a longshot, but i have been playing since about 1970, about 2-1/2 decades of that in bars, and i do have an ear for tone. So it's not a case of OCD, i'm talking a major tone/feel difference.

Not sure how your idea with seperate and common R's would sound, but if i will give it a try and let you know. thanks.

[daz]

damnn, no luck on the "Y" setup. tone went about in between the single and dual setups i suppose, but output wasn't any better that i could tell. I tried 2x225R with a common 180R, then a 120R.

Ok, we're making progress here, in fact i think i'm ok with what i have now. I tried bypass caps already as i have said, but i tried it again with the 2 resistor setup but this time lowering the value to 25UF. result=tone is closer to what i have w/o them but the output is now way up. I can live with this, but because the no cap tone is so unbelievable i cannot go w/o that. So in goes a switch. This will do fine like that. Thanks for the help guys. By the way, suggestions as to cap voltage minimum?

[pdf64]

Unbypassed cathodes increase the output impedance, and therefore speaker resonance is undamped, though global neg feedback would act to counter that. But maybe that's why you're liking the tone?
You shouldn't lose output power, rather it's lost gain, so just have to put a bigger signal in from the phase splitter to the power tubes.
How about just bypassing the independant cathode resistors a bit, ie put a resistor in series with the cathode cap. Say equal to the cathode resistor in the first instance, and fine tune from there. Don't need high power resistors for these, I think, as there's no dc dissipation. Peter.

[daz]

Interesting concept. I will certainly try it. Thanks. (wish i'd seen you post b4 i drilled the hole for the switch tho !

[tubeswell]

And (its been said millions of times) if you use pots for the series resistors for the cathode bypass caps, you can fine-tune the AC freq boost/limit. And if you want more highs then add another (small value) bypass cap from cathode to ground (for multi freq boost). The series resistor/pots need to be between the bypass cap and ground to avoid scratchiness when you dial the pot