Too much gain?

[R.G.]

I've been thinking about this for a while.

Everyone who messes with electronics knows about resistive dividers. Put in a voltage through two series resistors and the voltage at the middle is in the ratio of the two resistors if unloaded/etc/etc.. Not many people are aware of capacitive and inductive dividers. Reactances can divide voltages just like resistors do. A 10uF and a 1uF cap will divide AC voltages across them in the ratio of 11 to 1, just like resistors would.

What makes this important is that every conductor is capacitive to every other inductor in the universe. The only variable is how big the capacitance is. If a conductor has a voltage on it, and has a capacitance of a few pF to another capacitor (or resistance, but stay with me for a minute), then the 10pF lets through a voltage using the other impedance as the lower leg of a divider. The division ratio is brutal - maybe a division of hundreds or thousands to one, but some voltage gets through.

If the circuit has enough gain around that sneaky capacitance, you can get oscillation. But it's also enough that there just be enough sheer volts of signal on the top side of the divider to get an unfortunate signal fed back to somewhere it shouldn't be. You're describing a lower gain, but bigger signal situation.

I think it might be worth your while to set the thing up, get it doing the ugly sound, then move wires around with some chopsticks. If you get any change in operation at all from the wire moving, then at least some of your problem is capacitive feedback from wiring capacitance, and the solution will be either a change in wire routing or some shielded wire runs.

This may not be the actual problem, but it might be.

[daz]

Interesting. I'll try that. I suppose i should just leave a crossline master in it so I can try a number of things. I was also thinking if i put a 12AU7 in it, that may be telling in some way you guys would understand more than i would.

I've been thinking about this for a while.

Everyone who messes with electronics knows about resistive dividers. Put in a voltage through two series resistors and the voltage at the middle is in the ratio of the two resistors if unloaded/etc/etc.. Not many people are aware of capacitive and inductive dividers. Reactances can divide voltages just like resistors do. A 10uF and a 1uF cap will divide AC voltages across them in the ratio of 11 to 1, just like resistors would.

What makes this important is that every conductor is capacitive to every other inductor in the universe. The only variable is how big the capacitance is. If a conductor has a voltage on it, and has a capacitance of a few pF to another capacitor (or resistance, but stay with me for a minute), then the 10pF lets through a voltage using the other impedance as the lower leg of a divider. The division ratio is brutal - maybe a division of hundreds or thousands to one, but some voltage gets through.

If the circuit has enough gain around that sneaky capacitance, you can get oscillation. But it's also enough that there just be enough sheer volts of signal on the top side of the divider to get an unfortunate signal fed back to somewhere it shouldn't be. You're describing a lower gain, but bigger signal situation.

I think it might be worth your while to set the thing up, get it doing the ugly sound, then move wires around with some chopsticks. If you get any change in operation at all from the wire moving, then at least some of your problem is capacitive feedback from wiring capacitance, and the solution will be either a change in wire routing or some shielded wire runs.

This may not be the actual problem, but it might be.

[daz]

I think i figured this thing out ! I just tried something i have tried many times before. But this time it worked differently for some reason. It's a trick you all know...adding a 100k from the pre PI master's wiper to the coupler on the PI grid. As u know it's said to make the tone sound the same thru the master's full rotation. Of course we all know thats not going to happen, but i think they simply mean a much smoother transition and LESS difference. I tried it again tonite and i never noticed this before, but the tone became VERY thin and bright al ow volume. That tells me that the amp is VERY bright NORMALLY and that it's always sounded not too bright at all because i play it at low volume and there is no treble bypass cap on the master so......

It never really hit me, but my tone was muddy all this time because the master set low like on a guitar with no treble bleed cap is uber muddy as 2 or 3 ! So like conner said, and it's not like i never thought of this, but i designed/tweaked this circuit at that volume so i tended to make it very VERY bright, which with the treble robbing master sounded normal ! So with that resistor on the master it was simply making it sound at low volume like it really is. So now i'm thinking the answer is this.....leave the resistor there and re-tweak the rest of the amp to compensate. this should not only make the master far more linear in tone, but that nasty harshness i get as i turn up is no doubt a huge over abundance of treble as the master gradually brings back in all that treble it cuts when low. This may be the design flaw i mentioned above ! But the question is, why don't marshalls w/o the resistor not have this problem? In any case, tomorrow nite i know what i'm doing ! It's no wonder a 100pf across the plates of V1 is as high as i could go without muddying up the sound ! With the resistor off the master i bet i will have to up it to 250pf or more ! I'm excited about the possibilities.....maybe i should play sick and stay home tomorrow.


EDIT: well, i took the day off and first thing in the morning put the resistor back on the master's wiper and began experimenting. This has been a revelation to me. Everything has changed, and all my questions as to why certain things in my amp work when the values are so different in other amps was answered. So many things are answered i don't know where to begin. First, i now know this is the reason why i always feel the amp sounds worlds better with no or very little NFB. The amp was so blanketed by the fact the master was muddying the tone that now after rectifying that, i can't use the amp WITHOUT full NFB ! It made the amp muddy and lifeless before, now theres not even enough NFB to kill the piercing high end ! I knew this was going to be the case and that i would now have to change many things to cool the high end and round it off. A couple .0047uf couplers i put back to .022. Upped the value of that 100pf cap on the V1B plate to 250pf. I had removed the grid stoppers on V1B and V2A because the amp was too dark.....now they're back and work great ! One test i use to see how well my tone is doing is to play the B string about the 10th fret with my typical amount of gain and see how it sustains. If it sustains and starts to feed back into a higher octave i know things are getting good. It's now doing that. In fact, i could stop right now and have a very good marshall sound. But i built this thing to find a killer marshall sound, and i can now see all the possibilities there are to try. The way the amp was before i was at a brick wall which forced me to learn a lot to try and advance it. All those things i leaned will now i believe help me take this amp to a higher level than it's ever been. Just in the 30 minutes i tweaked it this morning the tone has come quite a ways.

Whats odd however is that when i first built this design it was like it is now at the beginning. bright, needed plenty of NFB, fighting the brightness, etc. Just like now. Yet it started out with the same pre pi master with no resistor on the wiper ! So i have no idea how it got that way. But whatever it was i think now the possibilities are endless. Many of the tweaks i experimented endlessly with now need to be done again and many will no doubt improve it with values or designs that before went the opposite direction. The goal now is to get the same sort of round top end and other good attributes it had before when at it's best, while retaining the best of what this change has brought. I have my work cut out, but i'm having a blast with this !