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

[daz]

This may go long, but here goes. I have long pondered this question, and i think i even asked here once. But i have never really found an answer. It's not just about my amp, most all amps do this but some much worse than others. What i'm talking about is about the change in tone AND feel as the amp is turned up, and i am mainly talking here about the one style of amp i have always used....master volume with cascaded preamp that generates a fair amount of OD, lets say about the amount a blues junior or peavey classic 30 does with the gain all the way and master to taste. When i play mine at low volume at home, the tone is super squishy on the attack and frequency response is even all the way across the board. Harmonics swirl like theres no tomorrow and at times you'd almost swear theres some sort of modulation effect it's so rich. Turn it up to loud home level and it changes radically. The rich harmonics and sag are gone and the attack becomes hard and brittle and it sounds like 1k is ruling. Now that may be a bit exaggerated for the sake of example, but it's a very different amp.

But before you start telling me what you think the problem is with my amp, let me say i have doe everything you can imagine and then some, and any suggestions you have are likely to be met with "I've tried that" or "I've tried that many times". Cutting treble doesn't work, it's not that. I also have NFB on a pot and can go down to 33k, but just the same as i find if i use that or a cut control or treble in the tone stack, when i use any of those methods to tame it they work but the volume drops back to where it was before I turned it up. Then when i turn it up to match the volume i was at before i "fixed" the harshness with whatever method i used, it then is back to the same harsh state. In other words, no matter what method you choose that cures it, it also drops the volume bacl to where it was when it was good in the first place, and turning it up makes it harsh again.

I've come to realize It's not that my amp has flaws, but that it's in this design to some extent. It's also partially playing it in a small bright room which is a big part of it. Speakers don't matter, i've tried several. But what i find is that output tubes make a difference. they don't eliminate it to a huge degree, but some stay much smoother than others. I have used EL34 which is what it's default design was, 6L6, 6550, and 6V6, and it has dual bias with a range to cover them all. I also have 3 OT's, 2 PT's and and 2 chokes i have all tried at various times.

Anyways, i am certain this phenomenon is normal but varies a fair degree depending on certain factors and could be improved, and thats why i posted. I'd love to hear you guys discuss this because this forum has some of the best minds around when it comes to amp circuitry and I'm hoping that i can get some ideas from listening. I think the fact the biggest improvement i have found so far was with output tubes may be a hint as to what this phenomenon stems from. By the way, 6V6 helped more than any other tube. My fav tubes fared worse, tho at the low volumes i mostly play at they sound best. Discuss please...

[g1]

Have you looked at the Fletcher-Munson curves and considered how different the ear hears at higher spl's?
Also, for any particular speaker, I wouldn't necessarily expect it to have the same frequency response at low spl's as high spl's.

[daz]

Have you looked at the Fletcher-Munson curves and considered how different the ear hears at higher spl's?
Also, for any particular speaker, I wouldn't necessarily expect it to have the same frequency response at low spl's as high spl's.

Yes, but note i'm not talking bedroom volume vs stage volume. In fact, i can't tell you how much louder that harsher tone is than the lower volume "good" tone, but i can tell you i'm not talking a lot. Well under 1 click on the knob, say from 8:00 to 8:30. Heres the thing tho, speakers and ears/munson curves aside, i CAN get it to do this 50% less with 6V6 so that tells me theres something else going on in the power section that is responsible for at least part of this.

[cooldude666]

Whats the topology of the phase inverter? Is it possible that at higher volumes you are seeing some kind of phase shift/harmonic cancellation as the push pull stage works harder?

[Chuck H]

I do hear the effect daz is talking about, but only across a much broader volume shift. As in actually going from bedroom to bar levels with an even tonal alteration across that expanse. I've always chalked it up to speaker drive and ear fatigue because the shift I hear is quite gradual. In daz case, based on the info given, I'm thinking something very different. IMHO daz can be trusted with an objective consideration. Even on his own amp. Many things he's tackled here have seemed like obsessive particulars. Then when he finally solves for these problems the solutions and the changes they made always correlate with the original issue. Daz is one of the few ultra picky tone guys who's observations I believe without ever thinking it's his iMOJOnation.

Daz, is your master volume post PI??? I'll bet it is!

[daz]

Daz is one of the few ultra picky tone guys who's observations I believe without ever thinking it's his iMOJOnation.

Daz, is your master volume post PI??? I'll bet it is!

Thanks Chuck, I do know a lot more about tone than i do about circuit theory. And i know you have come to understand me and where i'm coming from. But no, it's a pre PI master. I have tried a post PI a number of times, both crossline and the dual pot divider type. But they always cause blocking or cutoff or some other sort of issue where past 1/2 way it becomes unusable. I know that immediately triggers the thought that theres too much gain at that point. But i have tried cutting gain at that point many ways with no change, plus the gain is not at all what you'd call high gain. If you are familiar with a blues jr, thats about what it's like. Maybe a tad more. The 6V6 JJ's are a pretty good fix, but i know things would be much better if i could figure out what is really going on. The circuit is a jcm 2204 from the cathode follower on back except for the adjustable NFB. I think this is a bit worse than normal but the same sort of thing all amps like this suffer from as you insinuated. I think it comes down to a fine balance of things. Maybe something that will eventually become clear to me as often happens. I have spent years tweaking this amp and this is the last imperfection i have not been able to figure out. Otherwise it sounds fantastic. If i could just get this change to happen in a much more linear fashion instead of so abruptly i'd have no reason to screw with it anymore.

[teemuk]

Turn it up to loud home level and it changes radically. The rich harmonics and sag are gone

Inquiring minds need to know:

How can there be less sag with more current draw? How can there be less harmonics with more overdrive?

I have tried a post PI a number of times, both crossline and the dual pot divider type. But they always cause blocking or cutoff or some other sort of issue where past 1/2 way it becomes unusable.

Inquiring minds need to know:

How can you drive power tubes to blocking distortion or cutoff state when you are actually reducing their grid drive amplitude and increasing the grid resistance?

I'm sorry, but to me some things in your posts just don't make any sense at all.


Here's a suggestion, instead of using your ears and recollection of clichés to evaluate what's going on in your amp and its power amp section use an oscilloscope instead.

[pdf64]

I've noticed this effect too, especially when I used to play out; we'd have to keep the volume down for the first set but if I got it all set up right it could sound good, the notes could get some bloom and sustain. It might not need particularly high gain settings to achieve it.
But as the volume (ie master) snuck up as the night went on, the sustain would seem to get lost and I would need to turn the gain up and maybe hit some sort of boost to get any sustain but that could lead to pickup squeal etc and a loss of any subtlety and dynamics.
I theorised that it could be something to do with class AB, with the signal level altering the load on the power tubes as they move into cut off but never adequately got to the bottom of it. It's really difficult to replicate the rig and the perception thing on a test bench.
It occurred with all my amp types, 6L6 / EL34 / EL84 / Fender / Marshall / Vox.
These days I overdrive the power amp and alter the sound level on the Fluxtone speaker but with that the tone seems to gel sometime better than others; I've been trying to keep notes of my preception of this alongside environmental / atmospheric conditions but haven't found any correlation so far, though barometric pressure is next on my list!
Pete

[soundguruman]

The design of most guitar amps is very primitive.
They are not designed to maintain consistent tone at all settings of volume.

Hi Fi amps ARE designed to maintain consistent tone at all volume levels.

So, it is possible to design such an amp. Just much more expensive and complicated in design...

[daz]

Inquiring minds need to know:

How can there be less sag with more current draw? How can there be less harmonics with more overdrive?
I'm sorry, but to me some things in your posts just don't make any sense at all.

Well, if you think i'm nuts or lying or whatever, i guess your participation here is done because i'm sure not going to try and convince you. Apparently others here are experiencing the same, so you may wanna rethink your assumption that it's not possible.

[Chuck H]

Easy there. The cliché descriptions are useful because we all know what daz means when he uses them. I know what daz is talking about without the need to get semantic.

Regarding "sag", I think there's a phenomenon happening that implies that affect apart from actual power supply sag. And how could we know what daz wishes to convey if he doesn't describe what he hears?!?

And as to blocking distortion... Who's to say he's not talking about the PI grids blocking!?! A post PI master can do that for you.

At any rate, my intention is to listen to what the OP of any thread has to say and then, hopefully, be able to help. There may be forums that specialize in antagonism but I don't know about them.

[jazbo8]

Perhaps it has something to do with the output power vs. the distortion spectrum, i.e., you may experience nulls iin distortion when you turn the volume up/down since it is not constant as shown in the example below (ignore the distributed loading condition, pentodes do the same thing just differently). Does the "sweet spot" appear in any other setting besides at low volume?

[Chuck H]

Well I do have a second theory that applies to pre PI masters. I've heard this phenomenon myself in very clear detail since I own an amp very similar to yours (modified Marshall preamp, pre PI master, stock Marshall power amp). There is definitely a shift in tone at low MV settings, not just the higher volume shift I mentioned earlier. The lower volume shift is actually a circuit anomaly that plagues ALL pre PI Marshalls. I believe it has to do with the MV causing an impedance change that drastically affects the EQ. Compound this with the fact that the presence circuit, and in your case other NFB loop alterations, have no pre master attenuation in their affect. So the balance of the NFB controlled parameters against the pre PI signal also changes. There's no good way around the problem as I see it. An amp that didn't do this thing would be a very different amp in other ways. And I don't think you want THAT.

[teemuk]

Easy there. The cliché descriptions are useful because we all know what daz means when he uses them.

You can speak for yourself. I sure didn't understand what he meant with those descriptions since I regard them as highly contradictory.

e.g.

Regarding "sag", I think there's a phenomenon happening that implies that affect apart from actual power supply sag.

Which means we are NOT talking about sag and he should have not used that term because it confuses idiots like me. In fact, it probably confuses all the guys who are not idiots but have a clear understanding of what "sag" is.

Basically the issue is this: If I want to talk to you about motorcycles do I refer to them as "cars"? Thought so.

nd how could we know what daz wishes to convey if he doesn't describe what he hears?!?

There's a saying, "picture is worth thousand words". I often regard data like oscilloscope captures far more informative than vague descriptions of something someone's hearing.

And as to blocking distortion... Who's to say he's not talking about the PI grids blocking!?! A post PI master can do that for you.

Well, we don't know what he means because his descriptions are way too vague. He might as well refer to grid blocking in ANY stage... or he might simply be using a "sexy" term to describe tone that commonly pops up in forum discussions. Truth is, we don't know if he is actually talking about grid blocking at all.

Basically, I find it quite challenging to come up with a solid understanding of what takes place in Daz's amp when the only real data about it is "master volume with cascaded preamp that generates a fair amount of OD". Thanks, that helps a lot. A description like that could refer to just about any modern amp.

If we are going to discuss this seriously wouldn't it be more beneficial to provide cold facts, theory and measurements instead of subjective descriptions, which could be all over the map? I do stand by my words that sometimes it's more educational to plug the amplifier into an oscilloscope and examine what really happens in it instead of trying to guess what happens based on what you hear.

Most of the reasons behind tonal changes vs. different volume settings are easily explained by well known phenomenons discussed in dozens and dozens of books written about tube amplifier design. Heck, there are dozens of books about tube guitar amplifier circuits alone and about the non-linear stuff that seldomly gets discussed in books covering more "HiFi-ish" design. Oh, and even if these things are not discussed in books nothing beats the fact that they are damn sure measurable with proper instruments. The problem is, too much talk and guessing, not enough action and solid, real data.

For example,

I believe it has to do with the MV causing an impedance change that drastically affects the EQ.

Why "believe". You can easily calculate how much the impedance change actually effects the response and evaluate if the effect really is that audible. Is it a few decibels on audible range or not? If you are short on math (like I am) you can always have a software simulator do the math for you in few seconds. Or you can take a real amplifier and a few measurements to plot the response and the change in it. You can easily replicate the conditions of this impedance change and then use your ears to test if that thing was the thing you were hearing. People are keen in throwing around theories and assumptions but these things are easily measured or calculated and then you can actually present us something that is informative instead of "belief".

This is science, not religion.

[daz]

Well I do have a second theory that applies to pre PI masters. I've heard this phenomenon myself in very clear detail since I own an amp very similar to yours (modified Marshall preamp, pre PI master, stock Marshall power amp). There is definitely a shift in tone at low MV settings, not just the higher volume shift I mentioned earlier. The lower volume shift is actually a circuit anomaly that plagues ALL pre PI Marshalls. I believe it has to do with the MV causing an impedance change that drastically affects the EQ. Compound this with the fact that the presence circuit, and in your case other NFB loop alterations, have no pre master attenuation in their affect. So the balance of the NFB controlled parameters against the pre PI signal also changes. There's no good way around the problem as I see it. An amp that didn't do this thing would be a very different amp in other ways. And I don't think you want THAT.

Yeah, I've thought about that and i've often wondered if you could add a tube after where the master is and have a set voltage divider instead of the master set low where it sounds good, then use the new tube to amplify the signal back up. You'd still have a master in front of the PI, but i wonder whether it would work different than coming after a cathode follower and tonestack. Or maybe even a tube after the PI using each side to amplify each power tube's grid. Probably way off here in some way theoretically but as u know i have little theory to say the least.