Can anyone explain why this is?

[daz]

You may have read that thread where i mentioned using a 3.3uf cap in place of the typical .1uf as the NFB cap. I went on to say i had done this to my 2x6V6 converted 18 watt and that it gave it a much thicker and very nice tone. I also said i intended to do this to my 2xEL34 amp. And i did, but it is not creating the same beautifully thick tone. It seems a tiny bit thicker, but not in a good way like with the 6v6 amp. In the 6V6 amp it's thick and fluid while in the EL34 amp it seems to make the highs that remain less smooth and fluid and it all just gets stiffer.

So what i'm asking is do you know why this would be? Would it simply be a case of the way a 6V6 reacts to NFB vs a EL34? Both amps are cathode biased, both have the same values in the NFB loop, both have the same preamp, and both have NFB taken from the 16 ohm tap. The differences that i figure may matter are the PI for the 6V6 has a 1.2k cathode and 22k tail resistors vs the EL34 amp's 470R/10k, and the 6V6 amp uses a single cathode resistor and cap while the EL34 amp has seperate resistors and caps for each tube. Same .1uf on the unused side of each
amp's PI tho. (well, unused is not for the NFB loop) Any ideas why they react so different to this mod?

EDIT: for those who read this before i fixed the 123,567 typos, sorry. geez...gotta stop relying on spell checkers....damn things don't check context do they !

[loudthud]

I think I found the thread you mentioned but in that thread you used a .0033uF cap, not 3.3uF. You finally posted a schematic in that thread but I'm now seeing the "picture has been moved or deleted" graphic. It's difficult at best to comment on your circuit without knowing exactly what is in that circuit. A schematic would go a long way towards removing any confusion or uncertainty

Most power amplifiers use Voltage feedback. The voltage gain and frequency response of the circuit after feedback is applied are highly dependent on the gain and frequency response of the circuit with no feedback. Changing from 6V6 to EL34 and/or changing the impedance of the output transformer changes the voltage gain of the circuit and thus the operation of said circuit with your feedback loop.

[daz]

I didn't post a schematic, so you must have seen the wrong thread. in any case, it's a 3.3uf, not .0033. This is the thread i meant...

http://music-electronics-forum.com/s...ad.php?t=12378

As for the gain in the loop, i'm running the NFB from the highest gain tap, the 16ohm. At least thats what gives the most effect. But still nothing like the 6V6 amp. tho it's not as much HOW MUCH each amp's NFB affects the tone, but in what way.

[loudthud]

Ok, in that thread it sounds like you are just replacing the typically .1uF cap with a 3.3uF cap. The difference between the 4 ohm tap and the 16 ohm tap is a 2 to 1 change in voltage. But the Primary Impedance also comes into play in the overall voltage gain of the PA. Try reducing the feedback resistor (normaly 100K) by half and doubling the cap. The amp may start oscillating at an ultrasonic frequency. If you don't have a scope, look for a big AC voltage across the speaker with a DVM.

[daz]

Well, thing is, if i go lower on the FB resistor the amp gets real soft and loses a lot of cut. Sorta like adding the proverbial blanket over the speaker instead of removing is as is often mentioned when describing a punchier more in your face sound. I know this because i used to have a 100k pot in series with a 22k resistor in place of the 100k thats there now. this way i could vary the FB. i removed it and put the 100k resistor there when i realized that the tome got too mushy when going much lower than 100k.

[loudthud]

Ya, but did you try it with 6.6uF? Reduce the 100K as much as possible then try to find the value cap that gives you what you want.

[daz]

No, but my point is that when i reduce that resistor the amp gets much less in your face. the brightness is subdued and it loses cut and definition. And then doing this cap mod would make it much too dark. If i reduce that resistor it makes me want/need more highs, so if anything i'd need to go back to the .1uf.

[Chuck H]

What Mr. thud is saying is that you can fine tune the NFB loop the same way you can fine tune a tonestack, RC decoupling circuit, etc. You could leave the 100k in place and try other resistor capacitor combinations added to the NFB loop as bypass circuits. The same way the presence control works. Theres no limit as to what kind of circuit combinations you could try there. How about:

Bypass the 100k with a 4.7uf in series with a 22k resistor bridged by a 1uf/22k resistor and leave the presence control with the stock .1uf for some top end fine tuning...See?

Thats just an example. The amp is your oyster.

Stick with the 100k fixed feedback on the very bottom end since your amp loses definition with a larger value. But you could bypass it with other circuits that effect the mids and highs without effecting the very bottom. You can experiment with caps to about 10uf without getting into the mud. But let your ears be your guide (use the force Daz).

FWIW I have even experimented with inductor circuits in NFB loops with good results. With inductors added to the mix the possibilities are endless.

Chuck

P.S. Yes, the reason the mod has a different effect in each amp is because the tubes being effected by the mod (the power tubes) are different. They have very different character when overdriven and even clean. Thats why the PI arrangements are different for those amps. You are gaining a wealth of knowledge toward future mods by coming to these realizations. Knowledge of how different tubes react to different circumstances will eventually pay off when you get to that point where you are "choosing" power tube types for their sound as well as their output wattage in new builds.

[Enzo]

And to restate CHuck's closing statement, the NFB resistor, with or without a cap, is not a thing in and of itself, it is part of a whole circuit. So unless the PI is identical, neither will be the results of NFB component changes.


Or: I changed the tire pressure in my Buick, but it still doesn;t ride the same as my Chevy.

[daz]

I have already come to realize the potential here in playing with all the values and even thought of adding parts. But i guess you could say i don't have a clear picture in my head of the NFB generally like i do with say the preamp. I just need to play around with it and use common principles such as i use elsewhere in the amp, but i need to think about it more i suppose. I was just surprised at how different the 2 amps reacted to a 3.3uf in the NFB. If i could get that in the other amp i think it would be incredible. It's already the better sounding amp by a wide margin.

[daz]

I had a thought.....when i went to cathode bias on the EL34 amp i went with seperate cathode resistors and caps because thats the way the matchless chieftain did it. I tried a single R/C at one time but felt the seperates were better. However, that was long ago before i even nailed the preamp section and it was a different amp at that point. So considering that the tone was very different with a signle R/C cathode for both tubes, and considering thts what i have on the 6V6 amp which benefitted so much more, maybe i should try the single R/C cathode again on the EL34 amp. I think i shall. But what value resistor would i then use to get the same idle current? Would i use the same value as i now use on one side or 1/2 of that? I have 450R on each tube right now, so would i 1/2 that to 225R and use that value as the single resistor ?

[Chuck H]

For a two tube amp where each tube has it's own RC cathode bias, to change to a single shared RC but keep the all other things equal you would half the resistor value (but expect double the current) and double the capacitor value. So a simple way to see if you like it would be to add a jumper between the power tube cathode pins. If you do like it you could then change to a single RC if you want to bother with it for aesthetic reasons.

Chuck

[daz]

Well, i wanted to do it to both see if what i remembered still holds up, (remembered separate resistors sounded better) and see if that bigger NFB cap works closer to how it does in my 6v6 amp. What i remembered did hold up....it sounds/feels better with separate resistors. And the bigger cap wasn't any better with one shared resistor. By the way on thing i notice was that with a shared resistor, lifting the cathode cap out of circuit made no noticable difference. Yet is you recall from past posts i had a switch that lifts the caps with my seperate resistor/cap circuit and the power goes way down and the tone gets richer/squishier. Kinda odd. I'm sure theres a reason that over my head tho.

in any case, the amp has sounded fantastic since i nailed the preamp months ago. i have not since been disappointed at all. best marshall style amp i've ever owned, honestly. But i just have this feeling in my gut that the power section could be even better. When i nailed the preamp after months of experimenting, it was like a bolt of lightning. It just went from really nice to wow, and i had no idea there was even room for that much improvment. And I just think there is something in the PA waiting to be discovered too, and if i find it this amp will be beyond amazing. I think i can too, and i doubt i will ever stop tweaking this thing till i do. But i'm probably close to a year into this at this point ! So i hope i have enough years left to find what i'm looking for between them 2 big ol' tubes.