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

[daz]

Ok, tried it in series and no difference. But from what u r saying it sounds like you are curious about this WHEN the outputs are clipping. But this is much too low a volume for that to happen. When i played out regularly i used marshalls and i could always tell then the output started to clup, especially because i use a EV12L which eliminates the speaker as distortion source. And it was always much louder than this. This is lous for home use, but onstage it might be ok in a very small bar gig at most.

I was suggesting that the resistor be added in series with the speaker. When an output stage has global negative feedback, it lowers the output impedance. (Don't confuse output impedance with intended load impedance.) When the output clips, the feedback stops working and the output impedance gets fairly high in a tube amp. This impedance discontinunity has some effect on the tone, I'm not sure if it's good or bad. Many solid state amps made today have what is called mixed mode feedback. This raises the otherwise very low output impedance for small signals to several ohms, then when the amp clips the output impedance is low, opposite of a tube amp, but it's still an impedance discontinunity. Most people agree that this makes the solid state amp sound better, or more tube amp like.

I must say that I always think an amp sounds better when the output stage is clipping. I theorize that the speaker(s) create a better tone when driven with a discontinuous impedance. But it could be that the amp just feels better because the power supply is sagging. Most solid state amps miss the mark here because any compression is caused by distortion and not sag related gain reduction like in a tube amp.

Daz, I don't remember you saying much about the amp in this thread. I'll go back and re-read your posts. If you can provide links to other threads about the same amp, I would appreciate it.

[daz]

Oh, and the amp...from the cathode follower to the speaker it's a 2204 except the variable NFB pot. OT is a mag comp 3.6k marshall 800 replacement. Preamp up to the CF is similar to a 2204 but a regular gain stage in place of the cold bias stage. Some other changes too, but for all intents and purposes the only change worth nothing for this issue is the regular 100k/1.5k gain stage in place of the cold bias stage. Power tubes are biased to about 35ma each (dual bias) with el34 and plates have 440vdc on them. 8 ohm EV12L speaker. Main cap is 100/100uf and all preamp nodes are 20uf separated by 10k.

[nickb]

For clarity's sake, when you experience this effect, are you changing the volume control or the master volume?

[daz]

For clarity's sake, when you experience this effect, are you changing the volume control or the master volume?

The master. The gain control only give more OD as you turn it up. My issue is amp volume.

[Chuck H]

I'm going to explain what's happening in layman's terms. Someone tell teemuk that I wont be offering the supporting diagnostic specifics or math ...

With the master down low there are two things happening. One is that you are adding series resistance after the tone stack with a following load. This has the affect of shifting the tone stacks midpoint lower and reducing bass. The end result is an upper mid bump and a decrease in LF and the overall mids flatten out. The other thing that is happening is the input of the PI sees a relatively low load with a relatively high resistance preceding. This effectively creates a higher impedance signal going into a lower impedance input. This has the effect of rolling off HF since impedance increases with frequency here. Again, you get a dominant mid frequency. So the phenomenon your hearing can be attributed to a difference in EQ because of impedance anomalies caused by the MV circuit. This could be perceived as a difference in gain because your guitar and ears are more sensitive in this range. To get a similar tone at higher MV setting you would need a fairly aggressive mid boost. You could put a push pull pot on your mid or treble control and use it to add something like 220pf more capacitance parallel to the existing treble cap. Then, when you want to keep the low MV setting tone at higher volumes you can pull the pot switch to implement the increased treble cap value, turn down the bass pot a couple of digits, max the mid control and perhaps roll the treble down a digit or two. This should allow you more control over the EQ differences at different MV settings.

[nickb]

The master. The gain control only give more OD as you turn it up. My issue is amp volume.

Here is what I would expect to happen to the frequency response as you turn the master from 10% to 80% gain and with the tone controls mid point. Note that's not the same as "1" and "8" on the dial due to the log master vol pot. The low trace is scaled to overlay the high vol and low vol plots. I can run the simulation again if you give me your actual tone settings.



There is a drop of about 1dB around 3.2KHz when volume is higher. To a trained ear that would be significant, especially if you are overdriving the preceding stage and so are rich in high harmonics. I also think that ties into your description of "swirling".

Also I bear in mind the debugger's mantra "shut up and measure". This is largely speculation. You can use Audacity as a useful signal generator and scope / spectrum analyzer and your PC audio card to measure what is going on if you can get some cables and make an attenuator. I can give you more details if you are interested.

[daz]

I perceive it as a loss of LOW mids, (400-800hz) but a lot of hi mids, like i said sounds like 1k. The mids that create that smooth creamy sound i always think as just below that, around 700-800Hz. I get all these beliefs from playing with EQ's and hearing what those bands add. The biggest issue however more than what frequencies are changed is the loss of any sag and that brittle hard attack which is nite and day different and totally loses that whole feel thing that makes it so much more playable at low volume.

Also, note that i don't think the difference between the low volume goodness and the harshness i'm talking about when turning the master up is anywhere near that difference in gain, IE:20% to 80%. I'm talking turning the knob literally about 1/16" further. Maybe 8:00 to 8:15 to go from smooth and nice to that harsh state.

[Chuck H]

Yo nickb. Nice that you've provided a CAD generated plot there. But this is precisely what some of the other posters are talking about when they say that the measurements and math don't account for everything. To be fair, measurements and math CAN count for everything but only if everything involved is accounted for. My experience is more dramatic than what your plot shows. But I will likely be discredited for not providing a contrary visual. I'd be interested in your actual simulated circuit, your source impedance, input impedance for the circuit it's driving and the what constitutes "10%" power.

[Chuck H]

Daz, this post is one reason some posters would discredit you and claim that there's no useful information to work with. It's impossible to specify responsible frequencies without an actual reading from a piece of bench test gear. At lower master volume settings your EQ isn't acting the same or affecting the same frequencies as it does at higher settings. If you want to test for a drop in low mids do this... Set the amp up to the loudest volume where it sounds right. Then turn the MV up one digit, turn the bass down two digits, turn the treble down two digits and crank the mid pot. Does this help or even hint at going the right direction? Also...

We can't know from what setting you say your MV pot is just how high up it is. Meaning, pot taper has a lot to do with it as does the preceding circuit. So, how loud are we talking here? And how many watts is your amp? Is it a 50W or a 100W, I can't remember. And how loud are you when the good/bad jump occurs? As loud as a shout? As loud as a coffee shop gig? Quiet enough that you can still talk loudly over it?

[daz]

By the way Chuck, i have a 120k resistor in series with the master's wiper before the coupling cap. I did this to help with this issue from advice given at this or another forum long ago. If i remember correctly i did it because I believed the smooth darker low volume sound was due to the low master setting and the resistor removed that "benefit" to some degree. So i left it because i figured if it's closer to what it sounds like as i raise the volume, I can EQ the amp darker at low volume and expect less change as i turn up. It worked, but obviously not nearly well enough. But it was even worse before. I think, not sure but i think when i tried a larger R it got even brighter at low vol. Maybe i should try raising it way up, say 470k, then if it gets far brighter, redesign some preamp areas to be less bright. Seems odd tho because i would have thought that adding series R there would DARKEN it, not brighten it. Even what you said seems to suggest that.

[daz]

Daz, this post is one reason some posters would discredit you and claim that there's no useful information to work with. It's impossible to specify responsible frequencies without an actual reading from a piece of bench test gear. At lower master volume settings your EQ isn't acting the same or affecting the same frequencies as it does at higher settings.

We can't know from what setting you say your MV pot is just how high up it is. Meaning, pot taper has a lot to do with it as does the preceding circuit. So, how loud are we talking here? And how many watts is your amp? Is it a 50W or a 100W, I can't remember. And how loud are you when the good/bad jump occurs? As loud as a shout? As loud as a coffee shop gig? Quiet enough that you can still talk loudly over it?

I've done that, IE: adjust EQ to try and compensate. It does help, but the tone still suffers a lot especially in the lack of sag and complexity. And it's just very very different overall which is the problem. Like a extremely different amp, and i DO know some of that is due to room acoustics and EQ at higher volume etc etc. But it's just too much for such a small volume jump. And that brings me to your other question....as i said in a post a notch or two above, i'm talking about going from what i would cal high TV volume and turning the master maybe 1/16" at most, probably from 8:00 to 8:15 or 8:30 max. Volume wise i'd say to the ear what probably seems -30-50% louder. But no where near the amp's loudest, not remotely close. Thats the best i can explain it.

If you want to test for a drop in low mids do this... Set the amp up to the loudest volume where it sounds right. Then turn the MV up one digit, turn the bass down two digits, turn the treble down two digits and crank the mid pot. Does this help or even hint at going the right direction?

No, any increase of the mid knob at all when at the higher volume will result in more of the same nastiness. The mid pot contributes high mids, not at all anything in the lower mid register. Sounds sorta like the treble control but shifter 2-3k lower.

[eschertron]

I'm talking turning the knob literally about 1/16" further. Maybe 8:00 to 8:15 to go from smooth and nice to that harsh state.

Can you provide some readings for sound pressure levels at these two points? Even something like a phone app, or an input VU meter on a recorder that would give good relative levels, and a ball-park absolute figure.

I was thinking Fletcher and Munson could explain "brittle and harsh" at higher volumes, but probably not with the incremental change you describe. While you have your recorder out, commit to tape some of the smooth and creamy that you describe. Then play it back on your hifi. At the lower volume, where you have the smooth sound, is the playback as you remember? Now turn the playback volume up. Does it remain smooth and creamy, with the sag that you remember, at higher volumes? If it does, it will pretty well eliminate psychoacoustic phenomena as a cause for your audible effect.

[eschertron]

The biggest issue however more than what frequencies are changed is the loss of any sag and that brittle hard attack.

FWIW my impression of the sudden transition into a "brittle hard attack" sound like clipping and a loss of NFB somewhere in the amp, *just* on the transient attack of the note. If you play softer (or turn the guitar volume down) to compensate for the small increase in MV, what happens to the sound?

[Chuck H]

I haven't found enough control with the stock EQ to compensate for this effect either. But I like the tone of my Marshall in both "modes" and just consider it a difference, not a detriment either way. But I have another theory about what may be happening...

An AB1 circuit will operate without cutoff at very low drive signals. I would expect a shift in tone as the power tubes move from class A to class AB1 (for the nitpickers, yes I know this isn't an entirely accurate description). Perhaps the difference in tone you hear is caused by, or at least compounded by this.

Which brings me to another observation. With all these things happening at the same time, MV impedance shifts, any ear or speaker effects, power tube class changing, etc. it seems impossible to get on top of this issue without diagnostic equipment. Even then, would it be possible to solve for it in a satisfactory way for all volumes? I mean, the MV does what it does and the power tubes do what they do in this regard and there's not much to be changed about it. I'm not saying to stop trying. Never give up the fight! But this might allow you to relax about it a little so you can see/hear the issue without your shoulders tightening up

If it's convenient, Maybe try decreasing the negative voltage on your power tube grids a little (hotter bias). If the class shift is part of the problem this should get you into a little more volume before the shift occurs.

[daz]

If it's convenient, Maybe try decreasing the negative voltage on your power tube grids a little (hotter bias). If the class shift is part of the problem this should get you into a little more volume before the shift occurs.

You mean just up the tube's bias point? I have tried it fairly high so i don't think it's that, but i can raise it to close to max and see what that does. I bias them at 37mv/ma with the 1 ohm cathode R method and dual bias, but not accounting for screen current so i suppose thats what....32 ma?

FWIW my impression of the sudden transition into a "brittle hard attack" sound like clipping and a loss of NFB somewhere in the amp, *just* on the transient attack of the note. If you play softer (or turn the guitar volume down) to compensate for the small increase in MV, what happens to the sound?

see, thats kinda what i was thinking too and i'll tell you why. I have a 250k variable NFB pot in series with a 33k off the 8 ohm tap. I like it with min NFB, but as i turn up and the tone gets harsh i adjust the NFB pot to smooth it out. It works...BUT, after i turn it down to the point the smooth rich tone is back, the volume is now low again where it was to begin with. If i then turn the master up to get to that higher volume again, the tone changes again, so i turn the NFB pot up even more to smooth it out again, and same thing. No matter ho many times i turn the master up and then adjust nfb to compensate, it always drops the volume back down before the tone gets good again. So by the time the pot is at max NFB and the tone is good, the master is much higher but the volume is the same as the beginning where i started with great tone at low volume.

So if i turned it up to the harsh point and could compensate with adjusting for more nfb, that would be a 1005 satisfactory solution if the volume stayed at the higher level when i did that. So it DOES lead me to feel theres something going on with NFB. But the NFB circuit is just like a 2204 aside from my adjustability tweak.