A Mosfet boost with tonestack -- component value suggestions?

[mordechai]

I am about to put together a Mosfet (BS170) based booster much like the ZVex "Super Duper"....so two SHO type stages, one cascaded into the other.  With all that output, I thought it might be a good idea to include a tonestack afterwards, but I want to avoid the BMP style tonestack that was used in the Box of Rock. 

I've always liked the way the Baxandal/James tonestack works in the Colorsound Overdriver/Powerbooster.  Would the component values in that setup work well after two stages of BS170s, or would you suggest altering the values?  I'd like the end result to come relatively close to the Colorsound devices, but with a touch of that nice "sparkle" that come with Mosfet boosts.

[GibsonGM]

Post a schematic of your idea, mordechai...it is a good idea, by the way.

What I see...you need to know (approx) the output impedance of the last gain stage, so you can tailor the tone stack for this.  It may not be critical, though!    If I were you (and I have done this) I would just breadboard a tone control and tack it in, see what you get. 

Sim. software goes a long way for this.  Duncan's TSC (free) lets you change the impedances etc, so you can see what will happen if you put tone stack "A" into a circuit with impedance "Z".  Often, the results are not earth-shaking....don't let it hold you back, is what I'm saying. 

You could sim the 'ideal' set of conditions, the graph shows best power transfer - and the tone stack might stink with those values, lol.   Experimentation goes a long way...

[mordechai]

I'm pretty moronic with some of those programs (I can noodle with some features on the Duncan TSC), but as per your suggestion, here's a schematic of what I'm thinking...

[nocentelli]

I'm sure that will work fine: The only question is whether you want to have the second SHO acting as a gain recovery stage (like you have it here) or whether you want the full gain of the first SHO going direct into the second SHO, and have the tonestack after that (with maybe a buffer to separate the stack from interacting with the volume pot). This would mean the bass and treble controls wouldn't affect the amount of dirt produced, but would just affect the high and low frequency content. As you have it, I suspect the bass control will dramatically alter how filthy your signal gets at any given gain setting.

[mordechai]

Well, when I first thought of doing this, I pictured the second SHO really just doing gain recovery as pictured.  But the more I've been thinking about it, I do like the idea of the second stage being able to produce some dirt when cranked...sort of like the Colorsound Power Boost or Overdriver.  In which case, I wonder if it would make sense to have SHO 1 run directly into SHO 2, THEN have the tone stack, and then finally have a third boost stage (maybe another SHO or an LPB type stage) follow for gain recovery-- which might then help separate the tone stack operation from the volume pot settings...

The goal is to provide a nice sparkly sort of boost to a silicon fuzz face I put together (sort of the way Gilmour used to use the Colorsound device in his early, pre BMP live rig).  I've had some good results just following the fuzz with a regular SHO but this might give some nice flexibility.

[GibsonGM]

I would try it both ways, if I was you, moredchai      

I have tried it both ways, and there are pros and cons to each.  The 'ultimate' in my book would be SHO1, 2, tone stack, SHO3 for recovery.  This would be more of a distortion pedal than a boost, of course

[mordechai]

Well...I was always a little more partial to the Colorsound Overdriver than the Power Boost, so maybe a compromise, grit-wise, between the two can be obtained.  If I put a 5K trimmer on the third SHO stage and noodle with it and the trimmer on the second SHO stage, I think I can probably come close to the gain range of the Overdriver while still getting that nice Mosfet "shine" for most of the gain pot's sweep.  I think maybe putting another small pF cap before the volume pot might be helpful in toning down whatever high end might be a bit too spikey, too.

[GibsonGM]

Sure, those are all good ideas!  I'd use a trimmer on all of them (well, on the breadboard, then use fixed resistors except for whichever stage seems to be the best to control with a pot).   pF caps here & there - always a good idea!

[Gus]

Why do you show 2.2meg resistors at the 2nd stage?

[mordechai]

Hi Gus...I found that 2.2M gave me a little less noise but the effect and tone was pretty consistent with the first stage.  In my revised schematic with three stages, the first two have 10M, and the gain recovery stage after the tone stack has 2.2M.  I've only done this up on a breadboard/vero combination, so once it's boxed I'll use sockets and experiment with values to see if it really makes a difference noise-wise.

Question -- since I've only done this on the breadboard and haven't used on/off switches...does the SHO need a pulldown resistor to deal with switch popping, or does that 10M from the gate to ground take care of that?  I guess if I put a 1M pulldown resistor on the input, it wouldn't offset the value of the 10M-to-ground too much in relation to the other 10M coming off the gate...

[antonis]

Question -- since I've only done this on the breadboard and haven't used on/off switches...does the SHO need a pulldown resistor to deal with switch popping, or does that 10M from the gate to ground take care of that? 

Unfortunately, No..

You have to ground BOTH sides of input capacitor (otherwise the "floating" side should leak..)

I guess if I put a 1M pulldown resistor on the input, it wouldn't offset the value of the 10M-to-ground too much in relation to the other 10M coming off the gate...

Not at all..

The 10M resistor is for DC bias and there is a capacitor between it and the bleeding (anti-pop) resistor..
(but you'll have to take in mind that the input impedance will be dominated to about 909k..)

[mordechai]

Thanks so much for the insights.  Very helpful.  When I manage to put this together and box it up I will report back with how it sounds.  The goal, I think I said, was to have the functionality and oomph of the Colorsound device but the sparkle of the SHO type devices, so here's hoping it all works out...

[miech]

I would use so called silend biasing for all stages, and feed it to the gates using a single resistor. For the first stage that would be 2M2, for the latter stages it should be roughly 10 times the source impedance. That should cut down a lot of noise.

Silent biasing is done by making a voltage divider from two relatively low value resistors between the supply rails, and shunting the junction to ground with a capacitor. Any noise from this voltage divider is now gone. Biasing the transistors is done with a resistor between the voltage dividers junction and the gates or bases of the solid state componentes.

The proper way of limiting noise is to have all boost as much as possible from the first stage. If the signal gets attenuated from there on, so will the noise.
If you revert this, any noise generated by previous stages will be amplified in the last stage. Because attenuation always means a poorer signal to noise ratio, the results can be very significant.

You might consider using a build out resistor on the last stage between the coupling cap and the transistor. That should limit chance of oscillations due to less than perfect real world situations.