Lets discuss amp tone and how it changes w/volume and why

[daz]

Thanks for the update Daz.

I for one, am in awe at how you stuck at it.

What a cool Christmas present.

An amp that you can like & enjoy.

I just like the chase too much to ever give up. And having a lot of free time and being a fanatic about great tone and feel, i'm very persistent. I keep telling myself if i ever get it to the point where i feel i'm done i'd start building a new amp, a fender style one and spend whatever time it takes to get that one where i want it. Maybe now's the time....maybe not. This is one of many times i felt I've finally got it where i originally intended. So you never know with me ! Just ask Chuck But i really did find a few things that gave me what i've been hunting for with this amp since day one. It's amazing what simple things they were, things i should have figured out thousands of tweaks ago. But isn't that how it always is? Sometimes clarity in respect to certain things takes years. But it's been a fun journey, one that has no real end. Well, it does but at that point it won't matter anymore...

[Dave H]

...The whole key here seems to be losing the presence and maybe NFB altogether and tweaking the preamp circuit to compliment the PI/PA like that.

If you lose the negative feedback the frequency response of the power amp changes so it really needs a different tone control set up. When the amp has feedback and it’s driving a speaker load (not a resistor) the frequency response will be flatter than the amp without feedback. The amp without feedback drives the speaker harder than the feedback amp at high and low frequencies where the speaker impedance is higher so it is ‘scooped’ already to a certain extent and having the usual scooped tone stack and bright caps can overdo it making it sound harsh. All my EL84 no feedback amp needs is a smaller cap feeding the volume control to control the bass and my 1M log, 2n2 ‘backwards’ cut control.

Have a look at the plots below. The top one is a simulation of an EL34 amp with feedback driving a Vintage 30 speaker. You can see it’s nearly flat to within 1dB. It’s not perfectly flat because a guitar amp doesn’t have all that much feedback. The bottom plot is the same amp without feedback.

[daz]

All my EL84 no feedback amp needs is a smaller cap feeding the volume control to control the bass and my 1M log, 2n2 ‘backwards’ cut control.

That seems to be the key for mine. Not that exact tweak, but as i also mentioned, the trick to using no NFB on this amp at least seems to be to tweak other areas to work well with no NFB. And thats what i did. For example, if the tone gets too bright even for the tone controls with no NFB, by just removing one of the treble peaking caps in the preamp it no longer sounds overly bright with no NFB as it used to. As for low end, my amp has always been tweaked to reduce flab and i think thats why i don't seem to have that issue with NFB off. The tone stack works fine with the stock marshall values, and even when i turn it up they seem to be able to handle the changes in loud volume. But if i ever find they don't i have nfb and cut on no load pots and can activate either with a twist. I never gig anymore with rare exception, but if and when i do those will be available to me if needed. With the preamp changes i made i'm not hearing a scooped sound at all like i have in the past when i disabled NFB. But it IS more so than with NFB. But i prefer it, as NFB makes it sound nasal. And i can easily change the mids with the tone controls. When i turn all 3 down equal amounts, certain mid frequencies aren't cut as much as others the result is the same degree of mids as with NFB but much more open sounding.

Anyways, at the moment i can't find any reason to change it because i'm getting what i want and have the ability to take it most anywhere within it's inherent voice.

[pdf64]

'The amp without feedback drives the speaker harder than the feedback amp at high and low frequencies where the speaker impedance is higher so it is ‘scooped’ already'
My thought is that the degree of boost at high and low frequencies is dependent on the amp's output impedance?
On a class AB amp, won't that be affected by the signal level? Due to one of the push-pull pair being cut off for much of the waveform at high signal levels, whereas at low levels, both will be conducting. This will affect the OT primary to secondary impedance ratio.
And then there's the output tube plate impedance - is that the same for small and large signals?
Pete

[Dave H]

“My thought is that the degree of boost at high and low frequencies is dependent on the amp's output impedance?”

I’m sure you are right. That frequency response simulation I posted for an amp without feedback is almost exactly the same shape as the speaker’s impedance curve which tells me that the amp's output impedance is high enough to approximate constant current drive.

“On a class AB amp, won't that be affected by the signal level? Due to one of the push-pull pair being cut off for much of the waveform at high signal levels, whereas at low levels, both will be conducting. This will affect the OT primary to secondary impedance ratio.”

I’m no expert Pete. Does the amps output impedance change by a factor of two when one of its push pull pair cuts off? If so it should still be current driving the speaker so the speaker’s impedance curve will determine the high and low frequency boost.

“And then there's the output tube plate impedance - is that the same for small and large signals?”

Sorry, I don’t know the answer to that.

[jazbo8]

There are several moving parts in the above question - output impedance, frequency response, NFB, and Class of operation! Taken separately, the answers would be pretty straight-forward but mixing them all together scrambles my brain Perhaps some of the gurus here could provide a tutorial on how to approach this...

[Chuck H]

I got tired of fussing and compromising with master volumes and started making attenuators to go with my amps That's this guys answer to the problem. Even power scaling is too fussy with it's variable bias and grid drive requirements. Not to mention that tubes sound different at different voltages. An attenuator is as late in the chain as you can get, alters the least about amp function and therefor is the most transparent.

I'm not being snide or anything, and it's fun to consider and try to solve for the less obvious problems, but it seems to me that with all the variables and the fact that the variables will be variable for different amps and operating conditions I just don't see a good way to solve for all the snarky problems associated with master volumes. They're only good for one thing IMHO and that's feeding the signal from a self sufficient preamp into a clean power amp. Too often players expect too much from master volume controls and it always results in disappointment.