need a double check for a schematic

[add4]

So i'm in love with my polytone amp. it has a fat, dark tone that is great for old to modern jazz. sometimes i'm out jamming and i miss that on a regular amp so i thought i could try to fit the preamp into a pedal like all these dirt pedals emulating tube preamps.

worst case scenario, i get a clean boost pedal with a 3 band bandaxall tuned in a way that i like

So i looked at the schematic: and i made some changes, i would like to validate it with you more experienced guys.

my schematic:
https://drive.google.com/file/d/0B55lw2bfAg5cTjR3SEpBV1B0R3M/view?usp=sharing

original schematic (second one, on the bottom)

https://drive.google.com/file/d/0B55lw2bfAg5cQWU4aXR4UURsd0E/view?usp=sharing

other schematic, also official from polytone, but i guess from a bit later changes: distortion and reverb driver circuit are there (i'm not interested in them), mid pot, some changes in around the 'dark' switch

https://drive.google.com/file/d/0B55lw2bfAg5cdVIzS2I3eHZIUUU/view?usp=sharing

i basically want to use this version of the schematic, but only take the gain stage, the tone stack, and the last op amp stage i'm not sure.. it says summing stage. i guess if i don't have drive and reverb i'm not summing anything so it would just be useful as a kind of output buffer. but i guess i could remove it if i wanted and keep the volume pot as output, like in a normal pedal, right?

So my changes, that i would like to validate are:

- in the original schematic the opamps get +15v and - 15v and the bias is in the middle, to ground. I would like to first try to bias the opamps between 0 and 9v as usual to avoid using a charge pump if possible, and if i don't have enough headroom, between 9 and -9 volts.
So if i'm giving 0 to 9v to the opamps i have to bias them to 4.5V, using a voltage divider on the power section, but i would like to validate the bias with you guys. for the last 2 opamps they are inverting stages and it should just be a question of giving 4.5v to the non inverting input, that so i guess it's ok.

- For the first gain stage, it's an non-inverting stage, so i should bias the non-inverting pin to 4.5v using a resistor (here i kept the 680k of the original schematic) which lets enough current draw for the circuit needs, right?. My doubt is about the feedback loop. i used the more complicated one from the second schematic.. because i think it matches my amp, but it looks strange to me with a resistor to ground then the RC in series to the switch .. i just don't fully understand what's happening i guess.
changes i made


do you guys think it could work like this or do you see an obvious error?
i'll try to breadboard this tonight, but with the babies i'm lucky if i get one hour of concentration so i'd like to validate this before if possible.

thanks

[PRR]

Where did your R8 47K come from??

When you go from 30V total supply to 9V total supply, you will clip over 3 times sooner. For a guitarist, this is a total change of "feel". You should turn-down gain in the first stage. But that stage's gain is intimately mixed-up with the Brite/Dark switch network.

Your bias is incorrect. IC1A -In has a DC path to 0V, which will throw its Output to the positive rail. A similar thing at IC1C. Both need added caps. (IC1A "could be fixed" by returning NFB to VB, but that invites sneak-though from later stages.)

While you can fudge-up a plan which is "similar" to the true original, it may not be good-enuff until you get up to the design voltages.

I also wonder if 4558 is more similar to RC4739 than the wicked-clean '07X bifets.

> it says summing stage.

Yeah, it says that a lot. They lie. Only the II IV plan has summing (U4B and U6B).

[add4]

thanks so much for your answer

Where did your R8 47K come from??

on the third schematic i posted, there is the equivalent to R8 (R5 47k). that part of the schematic is simpler on the other version, i guess it's due to the fact that one schematic only has bass and treble and the other add mids.

When you go from 30V total supply to 9V total supply, you will clip over 3 times sooner. For a guitarist, this is a total change of "feel". You should turn-down gain in the first stage. But that stage's gain is intimately mixed-up with the Brite/Dark switch network.

good point. so i should calculate the frequency response and find the same response, for lower gain, right? i'll make a proposition.

Your bias is incorrect. IC1A -In has a DC path to 0V, which will throw its Output to the positive rail. A similar thing at IC1C. Both need added caps. (IC1A "could be fixed" by returning NFB to VB, but that invites sneak-though from later stages.)

that's exactly why i ended up asking! that DC path in IC1A NFB was bugging me and i wanted to have an opinion on this.
Also i did not spot the DV path on IC1C, but i guess a simple 1uF coupling cap right after the volume pot would be enough to fix that?

While you can fudge-up a plan which is "similar" to the true original, it may not be good-enuff until you get up to the design voltages.

maybe it would be easier to use a charge pump to get -9 and keep the bias to ground? that would also give 18v of headroom, which, in my experience, is already enough to have a very good headroom feeling in pedals. do you have an opinion? the headroom thing is definitely one of the features of the original amp that i'd like to capture.

I also wonder if 4558 is more similar to RC4739 than the wicked-clean '07X bifets.

well the original opamp is sometimes replaced by TL072 since they are impossible to find today. apparently a simple rerouting of the pins is enough to make it work correctly. so the wicked clean thing should not be too much of an issue i guess.

Thanks i'll try to improve based on your suggestions and come back here.