Less treble Fizz, more upper-mid Kerrang! in 2203-based hot rodded preamp?

[Burnsy]

Hi everybody,

I've built a marshall 2203-inspired amp inside the body of an old Traynor Voicemaster PA head. I've goosed it a bit for some more gain, and I think it sounds pretty good, but I would still like to calm down the treble fizz a bit and bring out a bit more of that upper-midrange kerrang. I pursued this a while back but I kind of just went in a circle. I've attached a schematic of where the circuit is now. Any weird values and configurations are probably due to me wandering amidst the wreckage of my past experiments. The tonestack and PI are pretty much Marshall stock. Any ideas greatly appreciated! BTW, the B+ on the preamp tubes is 230V.

[Chuck H]

Those buzzies aren't caused by too much, or the wrong high end, though you often see builders using late preamp stage bleeder circuits to remove them. They're caused by too much low end, grid blocking and excessive cutoff only preamp clipping. You have some circuit design in there that I know from experience to be buzzy in nature. Here are some things I would try. You should try them one at a time and decide for yourself before moving to the next.

1) Change C4 to 2.2n (2200pf) and change C11 to 22uf

2) Remove C10

3) Change R3 to 1.5k and R4 to 100k

4) Change R8 to 1.5k and R12 to 47k

You will notice a decrease in high end by removing C10. But these types of preamps only need one HP filter IMHO. If you like in general what these mods do you can then experiment with removing C2, C3 and C8 (again, one at a time) to restore brightness. These caps are top end bleeders that darken the sound of the amp.

The reason for doing changes one at a time is to give you a feel what each one does so you can pick what works for you.

Cheers

Chuck

[pdf64]

To remove some fizz, I'd move C9 so that it rolled off R15 rather than R8. Taking it off R8 would allow more high treble to overdrive V2. The combined effect might be all you need.
I'd move all the low pass caps (C2, 3, 8) to a proper ground, rather than rely on the high frequency performance of the B+ bypass caps.
daz put up a schematic a while back, that had a built in way of rolling off some bottom end at high gain settings - just add a resistor (100k-say 470k) between the gain1 pot R6 wiper and ground. That would limit blocking distortion, which can be buzzy at onset (though it would be more effective if R5 was also reduced or eliminated).

[Burnsy]

Thanks guys! I took Chuck H's suggestions 1, 2, and 4 and I moved the 2nd stage lowpass cap to the third stage like pdf64 suggested, and it's a big improvement. I'd say the fizz is whooped. Still got a little work to do to get it dialed in but it's much closer.

[Chuck H]

Forgot to mention it in my last post...

If your feeling ambitious you could lower the preamp volts a little. That will soften the attack and sound a little less harsh overall, which sometimes helps since a harsh amp really accentuates the buzzies.

I see you didn't do mod 3... I don't know if you tried it but IME high plate loads in early stages can sound fuzzy. Not fizzy or bad really just less defined. It might be really good with your higher plate volts though. It's all moot if you tried it and it didn't help or you didn't like the change it made.

Chuck

[SysCrusher]

I would go even further and lower your last two cathode bypass caps to 1uf.


Try raising C5 to 1000pf for more mid crunch. Those marshall 470k/470pf treble peakers can be spiky. Good call on removing C10. I always remove that one.

[HTH]

I would go even further and lower your last two cathode bypass caps to 1uf.


Try raising C5 to 1000pf for more mid crunch. Those marshall 470k/470pf treble peakers can be spiky. Good call on removing C10. I always remove that one.

yep, C5 to 1000pF was my immediate thought too - could try up to 2200pF here.

[Burnsy]

Hey, since HTH woke this thread up, I figured I ought to give a long-overdue update! I messed around with this circuit a zillion different ways, and, especially with the help of everybody here, learned tons about blocking distortion avoidance and tone shaping in this sort of preamp. I ultimately went back around almost full-circle to something closer to the stock 2203 type circuit. My big lesson- if you're really going for later-80's fully-saturated type distortion, you need either 1) another gain stage, such as one of those Soldano/Recto 39k RK cold-clipper things OR 2) smack it up front with an OD pedal.
That being said, I find there's a real satisfying amount of crunch and chug to the amp right now. I thought I would share my "final" changes, which still contain a number of oddball paralleled-up values, but not worth the trouble to dig out and re-do with fewer components. The changes in reference to my original schem are

R4 is now 117.5K (actually FOUR x 470K in parallel. )
C3 is gone
C4 is 2.2 nF
R5 is 330K
C5 is 680 pF
R8 is 6.8K
C7 is 0.32 uF (parallel 0.22 and 0.1 uF poly film caps)
C8 is 680 pF
All that crap between the second and third stage is replaced with a 220K to ground, and a 220K grid stopper. So I guess you could say I removed R10, C10, R11 (the Gain2 pot), kept R12, and upped R13 to 220K.
C11 is gone (V2a is now unbypassed)
R16 (CF cathode resistor wayyyy at the right edge) is nudged down to 68K.
My tone stack is "normal" except the slope resistor is up to 47K, and the treble cap is up to 680 pF.

The B+ supply voltage for the first two stages is 305V. V1a's anode is at 223V and V1b's anode is at 261V.

This was built in an old Traynor YVM-1 50W PA. These have four preamp sockets. I used the extra socket for an AB764 blackface clean channel. I put in a low-voltage supply and a couple DPDT relays to do the switching. Switching is done with a standard one-button SPST foot switch. This is my first (so far only) channel-switching amp. I call it the BlackMelt, because it has the Blackface channel and the Melt-Your-face channel.

I also went further and built an accompanying TS808-type overdrive pedal but with a switch matrix composed of four momentary DPST heavy duty footswitch buttons and a pair of 556 dual timers set up in bistable mode controlling a couple relays. The pedal both turns the OD off and on, but also connects right to the amp's channel switch and switches clean and dirty channels. The four buttons (and color-coded LEDs) correspond to the four possible states
1- Clean channel, OD off (blue LED)
2- Clean channel, OD on (green LED)
3- dirty channel, OD off (yellow LED),
4- dirty channel, OD on (red LED).

Sure, one can just use a basic overdrive pedal and the one-button channel footswitch, but then you can't go from full clean to full shred and back with one foot-tap. My pedal exists really just to solve that problem.

Again, big thanks to everybody for the help. This forum is GREAT!

[okgb]

Can you post a new schematic ? [ easier for me to visualize that ] tia
What program do most people use to create schematics ?
Also thanks for the people evaluating the circuits , I love learning from the
descriptions while looking at the circuit , Is there a chuck-like analysis of a 2203
or other typical marshall circuit ?