Hacking a mid-boost mod into a MOSFET based OD

[mordechai]

I'm about to build up the pedalpcb "Mosfet Drive" project...it's a Butler Tube Driver that uses two BS170s instead of the tube stage.   An ingenious idea and I was thrilled to buy a PCB to build this up.  Here's the schematic (second page of the document):

https://www.pedalpcb.com/docs/MOSFETDriver.pdf

I'm trying to figure out where I could add a mid-boost mod to this thing, so that with the turn of a knob I can get a sort of TS-808 type of midrange bump.  I'm also not clear on how best to do this.  I was looking at the 470K/470pf filter stage as a place where futzing with the cap and using a pot instead of a fixed resistor might be a place to do this, but I wonder if this is too early in the signal chain for this to work effectively, and if maybe there's a better way. 

[garcho]

but I wonder if this is too early in the signal chain for this to work effectively, and if maybe there's a better way.

if you want mid boost, basically, you put all the frequency through a band pass filter, and then add gain; theoretically, the bass and treble go back to where they were, and the band that gets passed by the filter is now "boosted". if you don't do that before a gain stage all you can do is subtract from the signal.

futzing with the cap and using a pot instead of a fixed resistor might be a place to do this

that's a great place to futz. it might not do exactly what you want but it might get you close enough. and it might be the cleanest and easiest way to alter the sound without redesigning the tone stack at the end, or adding an active stage or whatever, not a big deal but means you have to make a daughter board or redo the whole thing yourself.

another way to get something like mid boost is to play with your neck and/or mid pickups and then when you want the mid boost sound go to humbucker. i see a lot of people noodle with a million pedals on stage and then never once switch pickups. just sayin

[mordechai]

Actually, the prospect of adding an active mid boost via daughterboard isn't necessarily unappealing.  If I'm experimenting, I might as well go in whole-hog.  But I've never made an active mid boost circuit, so this is uncharted territory for me.  Any suggestions on examples I could look at to get a sense of where to give it a try?

But I will surely try messing with the 470k/470p filter section.  To get a higher frequency out of that, I would have to decrease the 470p, correct?  Down to, say, 220p?  I'm just thinking about how that combination works in the clipping stages of a Muff, where if the cap values is higher, the tone gets darker....so I'm guessing that lowering the the cap would add greater upper-mid presence?  Or, maybe just keeping the 470p and increase the value of the 470K to 1M?  (I'm clearly still grappling with how the parallel cap/resistor network actually functions.)

[antonis]

470k/470p form a high pass filter with 720Hz cut-off frequency..

For a MID boost, you actually need another filter (low-pass) to form a band pass filter (mid band) of your taste..

An easy way to do that is to place a switchable cap of 120pF in parallel with C4 (120pF) so as, when activated, total capacitance should equal the double of 120pF, effectivelly lowering low-pass cut-off frequency to about 1.32kHz..

Another way is to place a series resistance with a shunt cap between C5 & Q1 Gate..

[Mark Hammer]

The MosFet stages are a sort of melange of the ZVex SHO and the AMZ MosFet booster.  The tonestack is essentially a slightly-modded Fender tonestack, with the addition of C10, and R14 to shape the taper of the Treble control.  YOu may want to plug the values into the Duncan tonestack calculator and see if the response you want can be achieved with a component-value change or two.

The TS-808 mid-hump is produced via the clipping stage itself.  In the case of the present circuit, you'd be either shaping the response before anything hits the MosFet stages, or after them.  I suppose there are circuit changes one could insert to those stages, but before or after strikes me as simpler.

One approach to use might be an adaptation of what the Ibanez Super-Tube used, as a variable "Bite" control (Maxon re-legended this as "Mid-Boost").  Since it would be a small daughterboard, you have the latitude to insert it either before or after the MosFet stages.

[mordechai]

Hi Mark and thanks for posting that schematic.  The topography of that part of the circuit reminds me a bit of the fixed mid notch in the BD-2 circuit.  So, if I understand correctly, the mid boost portion could be hacked into this larger circuit as such:



(I think if you click the image it will get bigger...)


If I have this right...would the additional components (the parallel 1.5K coming off of pin 7 and the 20K pot) affect gain or risk introducing oscillation?

[Mark Hammer]

Hi,

I don't know enough about the math involved, but it may require an additional op-amp stage, and may not be a simple supplement to the existing gain stage.  But you should probably try and find some YT videos of the Supertube that vary the "Bite" setting.  It may be something you like, but may be insufficient.  Or, it may be what you want, but needs some different component values.  Try before you buy.

[antonis]

As Mark supposed, it should need an extra op-amp..

In your diagram, you've placed BITE circuitry in positive feedback loop with 2 issues:
Mainly, you've broken non-inverting input DC bias path and also you've made a prone to oscillation configuration (in case of  bi-polar supply working op-amp..)

[mordechai]

What about this option?  It's copped from the Green Rhino "curve" control, which I've placed here in a negative feedback loop: 





I've heard this work in the Green Rhino very nicely, though I'm not certain how this network of components somehow does NOT effect the gain/drive of the circuit.  Is that because the gain is determined by the 1.5K off of pin 5 independently? 

In this network, I can see that the caps are determining the corner frequency in different ways due to their different values in series with the 3.3K and the resistance from the 5K pot...but I'm not clear on how the resistance variation on one side (with the 22K resistor) creates an effective parallel to the other side without the additional resistance changing the gain structure.  With the pot centered, I am assuming this would be a flat mid response, but why would the 22K be required to have the curve change be consistent when the wiper is rotated to that position?

[Frank_NH]

I humbly offer a bipolar mid boost/cut circuit for your approval.  The image is a LTSpice simulation, but I actually built this for a bipolar version of the Klon I had recently on my breadboard, and stuck it at the end.  It worked pretty well - by turning up the mids, you can change a Klon into a Tube Screamer.  But then, who wants to do that?? 

You can adapt it work with a single sided supply with suitable biasing.

[mordechai]

That looks very interesting.  So, if I understand you correctly...this circuit could be inserted after the mosfet stages, but before the volume pot?

Also, I am not familiar with the LT1056...is it a single op-amp?  Would any single op amp work in this configuration (like a TL071)?

Actually...I have a few quad op amps leftover from another project.  Could I configure this somehow to use the quad op amp for the TL072 stages, the Mid boost stage, and then somehow stabilize the last op amp stage to not be used?   

[Frank_NH]

That looks very interesting.  So, if I understand you correctly...this circuit could be inserted after the mosfet stages, but before the volume pot?

Also, I am not familiar with the LT1056...is it a single op-amp?  Would any single op amp work in this configuration (like a TL071)?

Actually...I have a few quad op amps leftover from another project.  Could I configure this somehow to use the quad op amp for the TL072 stages, the Mid boost stage, and then somehow stabilize the last op amp stage to not be used?

Yes, you can stick this after your gain stage(s) and then tack on a volume control at the end.

Don't pay any attention to the op amp model - it's just something generic I use in LTSpice.  A TL071 will work just fine!  Again, if you are using a single sided supply voltage for your circuit (and thus for the op amp), you will need to bias the op amp (@ the + input) to 1/2 supply voltage.

Note: my circuit topology is essentially the same as the three band EQ discussed in this old thread (I'm just using mid-boost/cut):
https://www.diystompboxes.com/smfforum/index.php?topic=24359.40

[mordechai]

Frank, thanks for pointing me to that old thread.  Since the larger circuit is just using a single 9V source, would it be more advisable, then, to get rid of the tone section of the original circuit, and just have the second mosfet stage go into this, which seems to incorporate elements of your mid circuit and also tackle the highs and lows:



It looks like a single quad amp could take care of all of this, and I have both a few OPA4134 and TLE2074 chips in my stash that could work great if this could just be used instead of the original lo/hi tone section on the original circuit...

[garcho]

It looks like a single quad amp could take care of all of this

I've found that unless I'm doing multi-layer routing, quad op amps aren't really more convenient, a lot of times they're more of a headache, sometimes two dual op amps are easier.

Here's a simple single op amp Baxandall roughly around 1kHz, cut/boost, just like the 3 band eq but on its own

[antonis]

IMHO, C4 value is marginally low for low Vol pot settings, Gary..

[mordechai]

It looks like a single quad amp could take care of all of this

I've found that unless I'm doing multi-layer routing, quad op amps aren't really more convenient, a lot of times they're more of a headache, sometimes two dual op amps are easier.

Here's a simple single op amp Baxandall roughly around 1kHz, cut/boost, just like the 3 band eq but on its own

Garcho...what is setting the 1K corner frequency?  Is it the network of components coming off of pin 3? 

[bateriaselectronicas2000]

Thank you very much, I was looking for something similar and with this I'll be fine.

[antonis]

what is setting the 1K corner frequency?  Is it the network of components coming off of pin 3?

Not at all...
(that network is setting op-amp bias voltage to halfway the supply voltage.. - common bias method for single supply powered circuits..)

Corner frequency is set by R1,R3,C1, C2 & Pot items..

[garcho]

Garcho...what is setting the 1K corner frequency?  Is it the network of components coming off of pin 3?

Pin 3 (non inverting input) is connected to a voltage divider (with a bypass cap) that halves the DC voltage of the pedal. Standard guitar pedals are at roughly 9VDC, so in order to give the AC audio signal maximum headroom, the op amp is biased (via pin 3) at half voltage, 4.5VDC. That way the upswing and downswing of the audio voltage each have 4.5V to occupy before being squared off by the limits of the power supply (basically 9VDC in this case) and sounding distorted (turned into square-ish wave signal). There is a ton of information about this here, all over the internet and also in this lovely reference from Texas Instruments, check out section 4 "single supply op amp design techniques":

https://web.mit.edu/6.101/www/reference/op_amps_everyone.pdf

I'm not sure how to exactly calculate the values for a Baxandall, neither it seems are most people without degrees and much experience because it's complicated. According to the gurus SPICE simulations are worth more than doing the math.
I just look at other proven designs from Rod Elliott, Doug Self, commercially produced gear, etc. This is based off of a Doug Self tone control, adapted for 9V and the most common values, which I'm sure slaughtered his careful considerations. He uses 1K2 instead of 1K resistors, a 15N cap instead of 10 and a bipolar op amp presumably at ±15V. Try a 22N cap instead of the 10N if the mid sits too high for you.
Filters, basically, are about finessing the capacitive reactance of capacitors; the change in resistance that happens at different frequencies (a consequence of AC and how electric fields store energy). The ratio of cap to resistor can affect the Q, as well as cascading stages, and resistors affect impedance, gain, etc, so they matter, too. The controls (or control in this case) interact with each other, making the math more complicated than passive RC filters or MFB filters.
In this case, the potentiometer is part of the feedback gain structure. When it presents 10KΩ of DC resistance to the feedback loop, the mid band is boosted because low frequencies are at unity gain from C2 (and the two 47KΩ resistors), and high frequencies are bypassing the pot via C1, at unity gain. The band between C1 and C2 is what becomes amplified more than unity. When the pot presents 0KΩ to the feedback loop, low and high are still at unity, but the middle band range is now attenuated, below unity (it's a inverting amplifier).

IMHO, C4 value is marginally low for low Vol pot settings, Gary..

That schematic is thrown together really quickly, probably could be improved all over the place. But if the pot is 100KΩ to ground, corner is 16Hz, no? 50% is 34Hz, 25% is 66Hz. It's only -3dB, yeah? Takes a while to hear it. And why would the volume be at low settings, realistically? I'm probably not following or missing something obvious.

[mordechai]

In your schematic...would I not need to do something with the other pins on the TL072 (that is, pins 5, 6, and 7)?  If they're not in sue, can they be ignored or do I have to otherwise stabilize them somehow?  I'm thinking of the old Fuzz Central "liquid drive", where the second op-amp in the IC isn't used in the circuit but is treated this way.