Negative feedback help (in an amp)

[midwayfair]

I thought I had a handle on this, but I'm having some serious trouble with something.

I went back and made a modification to my "Woodchipper". Current schematic is here (it's big, so I'll just post the link):
https://dl.dropboxusercontent.com/u/9878279/Jon%20Patton%27s%20layouts/Woodchipper/Woodchipper.png

I recently changed R7 (cathode resistor on V2) to 4.7K from1.8K, to make the mid pot work better.

The mids pot action is greatly improved -- there's a much bigger boost than before -- but the behavior of the negative feedback switch completely changed. I used to get a noticeable bass and gain boost when the 680nF cap was switched in on the negative feedback (switch position 3). The gain would jump up and it was a different sound at basically all settings of the mid pot compared with the 33K alone (switch position 1) or the negative feedback lifted entirely (middle).

Now, though, the switch positions sound identical.

So I tried a bunch of stuff:

-I tried making the capacitor smaller so that it forms the same high-pass filter cutoff with R7 as it did before I changed the value of R7, but it still didn't sound any different between the two settings.

-I tried connecting the capacitor so that it bypasses the 33K, and that was very bad ... the speaker gets farty, the sound is extremely dark, very little output, and at lower "mids" settings (i.e., when the cathode was not bypassed) it actually got low frequency oscillation. I added a little bit of resistance and ended up with something very dark and kind of squishy, so I'm not sure if it's just lacking highs or what. It certainly wasn't louder and gainier like it had been in the past when the cap was switched in. It's as if this setting was now creating MORE negative feedback, because I was getting less signal overall, even though the capacitor should limit the negative feedback and result in more signal! Arg.

So what might I be missing in trying to make this work? I know for a fact that there are amps that use a capacitor as part of the negative feedback (like the Deep switch on the SLO, and some hi-fi amps), so it really doesn't seem to me that my method of cutting low frequencies from the negative feedback to create a bass boost is fundamentally flawed.

Just for clarity, my goal here was to create a setting with a small mid scoop around 200 Hz by boosting the bass below that and using the mids pot to boost above it. The only other way I could think of doing this involves more complicated switching that changes out a tone stack or something.

[Groovenut]

I'll take a stab at it...

You've decreased the amount of neg feedback getting to the cathode at R7 by reducing the resistor value, now more neg feedback is being shunted. Reducing the 33K feedback resistor should restore this to a point. As you reduce R13 you may need to increase C12 to maintain the freq knee. You could also look into putting a resistor in parallel with C12 to enhance the bass boost. Most amps with fixed resonance circuits tend to use 220K//4n7 here.

Just some ideas...

[midwayfair]

I'll take a stab at it...

You've decreased the amount of neg feedback getting to the cathode at R7 by reducing the resistor value, now more neg feedback is being shunted. Reducing the 33K feedback resistor should restore this to a point. As you reduce R13 you may need to increase C12 to maintain the freq knee. You could also look into putting a resistor in parallel with C12 to enhance the bass boost. Most amps with fixed resonance circuits tend to use 220K//4n7 here.

Just some ideas...

Thanks for the sample values. I'll give them a shot. If worst comes to worst, I drew up a version that switches in a modified BMP stack with a mid scoop centered at ~250hz.

I'm not sure I understand one thing, though. Increasing R7 decreases the negative feedback? I would have thought it would increase the negative feedback, since the frequency cutoff from the cathode bypass cap uses that for their "R". A bigger resistor there = lower frequencies and a bigger difference in dB between bypassed and not bypassed (by like 15dB in this case, according to the ampworks schematic).

[jatalahd]

The resistor R7 is parallel to the mid-filter (C5 - MID - C6). Increasing R7 will cause the mid-filter to stand out more in the total impedance R7 || mid-filter. So as you have noticed,  the mids will be more pronounced, but the downside is that then the negative feedback is controlled more by the mid-filter than R7.

So, if there was only R7, then increasing it would increase also the negative feedback, but I think in this case increasing R7 will let C5-MID-C6 combination control the feedback and interact with the low boost from the feedback path.

As Groovenut said, the easiest fix would be to tweak R13 and C12 to feed more signal back to compensate the mixing with the mid-filter. But the situation in any case will not be that same as having R7 as 1k8.

Best solution would be to completely isolate the "bass-boost" feedback from the "mid-boost".

[midwayfair]

Okay, I've tried a few things with this knowledge, and tinkered with Groovenut's values, and I just can't make it work similarly to before. So I ended up pulling out the switch and replacing it with a 3PDT to switch to change the tone control's behavior and switch out the negative feedback. One position has a variant of the BMP and the other is the same cut currently in there. It works well now. I got a nice deep notch near 250Hz, while the treble is more like a shelf. It never gets unusably dark, and can sound very thin with the treble maxed and the mids all the way down. The mids fatten it up nicely in either mode, and I can get something closer to top-boost type sounds with both knobs all the way up. It was a real challenge getting the extra components securely in place, but it seems to be worth it.

Thanks as always for the help, folks.

EDIT: I replaced the schematic file if anyone's wondering why none of the above posts make any sense now.