The "huge cathode cap method" on 18watters... does it work?

[Silvio55]

I will try a 2.2uf cap Chuck, but do you think itīs still fully bypassed with a 10uf? I can note a big difference now, and Iīm not the guy who can hear small and subtle changes in the sound, for expample the amp sounds the same to me whatever tubes I put in it (I have tried 12AX7's sovteks, Mullard RI and JJ's, I donīt say it is the same tone but I canīt hear any significant difference) but this change is not subtle to me, it has far less buzz now and a tighter bottom end.
I was watching TW Express videos on youtube and itīs like you said, it has that swirl but itīs not buzzy, I have heard it on some records too, itīs a cool effect when itīs not buzzy!

[Chuck H]

itīs a cool effect when itīs not buzzy!

Ahhh. But if you listen real close, it's always a little buzzy.

Two reasons for the 10uf sounding significantly different. One is that your no longer boosting LF harmonics that the amp can produce, but never make it out of your speaker. This LF, though not heard, does change the way the amp feels and the way it reproduces higher frequencies (because there is now more current available for them). The other reason is that there is a small amount of attenuation at the 82Hz (low E) point. Just under 2dB. And that's not much. It's worth trying a smaller cap to see if you get more improvement without sacrificing too much low end in the final EQ. Too much bass in the pre clipped stages just adds flab and buzz, not more bass. But you don't want to cut too much or you can lose bass in the final EQ balance and the amp can start to thin out. No one wants thin tone. and no one wants flabby, buzzy tone. You must create balance grasshopper.

[Silvio55]

Hey, look at this, this is the first and only guy I know who actually likes the crossover distortion buzz/fizz, listen to that clip, itīs not just "swirl", itīs pure buzz, I just canīt take that amount of buzziness, I really canīt!
EL84 Fizz? Forgot how much I loved that tone… Ŧ Guitar Gear

[Chuck H]

Ha... Ewww. I couldn't even listen to it all the way through. It offended my sensibilities.

[Wombaticus]

Ha... Ewww. I couldn't even listen to it all the way through. It offended my sensibilities.

I'm with Chuck on this, that's pretty horrible. If one my amps sounded like that, I'd have the top popped off and be tweaking within minutes...

[Chuck H]

Silvio's amp wasn't that buzzy in his clip. And he say's it's less buzzy now. Has to be way better than that last clip of a "boutique" amp.

[Silvio55]

Yes, itīs much better now, but didnīt had time to try a 2.2 cap, maybe tomorrow...
My amp did sound like that when I had a 150R bias resistor, When I was building the amp I was following the instructions in the liteIIb schematic that recommends to change the original 125R to 150R for a "more realistic bias for newer tubes". Many people, specially at 18watt.com suggest to go even higher in value, like 180R, (to avoid exceeding the 12W of dissipation at idle) I canīt imagine the level of buzz in those poor amps! Maybe some people donīt crack their 18watters...
Iīll try also putting a switch to the PR mod diodes, I donīt think theyīre making a big difference here, even with the PR mod, the 150R bias resistor made the amp sound almost like that clip.

[Chuck H]

I have 355Vp and am using a 130R cathode resistor. A pair of el84's lasts about 300hrs to 400hrs. Not very good. But certainly worth it. el84's are cheap and bad tone sucks.

[Silvio55]

Tried the 2.2 cap for a few minutes today, but I think I liked the 10uf more, I donīt know, maybe Iīll need to play it a bit more and see.
I put a switch in the PR diodes and they indeed make a difference! Itīs real nasty without the diodes, a little bit of swirl with them on. The lower value zaners I have are 11V, my cathode voltage is 10,2V, but I wonder what happens if I try lower voltage zeners, like 10V or lower. If I put a lower voltage value (or the same) than the cathode voltage does the diode clipping become audible?

[Chuck H]

It certainly should. Paul Ruby used under voltage zeners without issue though. I think the swirl is brilliant if it's not buzzy. I have almost none in my design and I just can't bring myself to tolerate the amount of buzz that comes with it. As to the preamp bypass cap... There are values between 2.2uf and 10uf. In fact, that's where all the range is. So it may be worth experimenting with 3.3uf, 4.7uf and 6.8uf. The standard values in between. But I guess the 2.2uf didn't reduce the buzz if you didn't like it better. Maybe just getting out of the subsonic was the biggest improvement.

[Silvio55]

Well, I think Iīm gonna stay with the 12V zeners for now. I donīt have other caps to try, maybe a 4.7 rescued from an old pc speaker I think, got to check. What I want to try is the zener across the cathode, I have to buy some 5W zeners and experiment. What would be the recommended voltage for that zener? With the 2.2 cap I recheked the voltage rise across the cathode resistor and it was from 10.2 up to 16V (it was up to 15V before, when I had the 47uf cap, and the other coupling caps, donīt know if it has something to do with that). A 13 or 14V zener maybe?

[Chuck H]

Wait.?. Are you using that 10uf bypass cap on the power tube cathode? I've been talking about the preamp tube cathode.

[Silvio55]

No Chuck! The 10uf and the 2.2 cap now are the V1 bypass cap! I have a 47uf cap on the EL84's cathode resistor. I mentioned the voltage swing because I meassured it yesterday and was 1V higher than I had meassured before, I think it has nothing to do with the cap changes, but thatīs the only thing I have modified in the amp.

[jimboyogi]

With 2 * EL84s (shared cathode string) I have been using a 7.5V, 5W zener to ground, and using a resistor between the zener and the cathode.
The resistor value is adjusted by measurement to get your required plate dissipation. Eg. I have ~ 330V plate voltage, and a 47R resistor gives me ~ 80% of max. dissipation. If your B+ is lower then your could try a lower value resistor, higher value for higher B+.

I keep the resistor unbypassed, this has 2 effects here.
1 - this adds a little bit of compression to the signal peaks as they approach max. output, which is an effect that i like the sound of,
2 - the unbypassed resistor dramatically reduces crossover distortion.

With the above values, the bias supply voltage is ~ 70% "fixed" by the zener, and ~ 30% cathode bias. You may prefer a different proportion of fixed/cathode bias, so it's worth getting a range of zeners of between 5V and 10V, and a range of resistors from say 10R to 150R. Then play around and see what gives you your favourite combination.

[Silvio55]

With 2 * EL84s (shared cathode string) I have been using a 7.5V, 5W zener to ground, and using a resistor between the zener and the cathode.
The resistor value is adjusted by measurement to get your required plate dissipation. Eg. I have ~ 330V plate voltage, and a 47R resistor gives me ~ 80% of max. dissipation. If your B+ is lower then your could try a lower value resistor, higher value for higher B+.

I keep the resistor unbypassed, this has 2 effects here.
1 - this adds a little bit of compression to the signal peaks as they approach max. output, which is an effect that i like the sound of,
2 - the unbypassed resistor dramatically reduces crossover distortion.

With the above values, the bias supply voltage is ~ 70% "fixed" by the zener, and ~ 30% cathode bias. You may prefer a different proportion of fixed/cathode bias, so it's worth getting a range of zeners of between 5V and 10V, and a range of resistors from say 10R to 150R. Then play around and see what gives you your favourite combination.

So, do you use a zener in series with the cathode resitor? I've never heard of that! very interesting, one thing I donīt understand is how do you meassure the dissipation without the cathode voltage, just with the plate voltage and resitor value, or is the cathode voltage the zener voltage?
If you keep the resistor unbypassed, how does that affect the overall volume (or PA distortion) of the amp? I think itīs reduced in some way, cause you're reducing the PA gain, and in the 18 watt most of the distortion comes from the PA.