How to implement Baxandall tone stack

[jwblant]

Hey ya'll.  I'm still pretty new to building pedals, but I'm designing a pedal based off an OCD/VooDoo OD pedal.  I'm taking what I like from both pedals and adding a 2 knob tone stack to it.  I'm wanting to go with the James/Bax stack route but I'm not sure how to actually implement it.

Below, I've added the schematic for the OCD and for a schematic for the BAX stack.  The problems I've run into is where to place the stack.  In the picture of the BAX, you can see it comes after a buffer but before a amplifier.  But if I already have the first half of the op-amp and the fet/diodes for the distortion as seen, do I still need a separate op-amp for a buffer or will the amp for the gain section suffice?  And should the stack go before or after the fet/diodes? Or even after the amplifying section of the OCD? 

Thanks in advance for the help!!

[Derringer]

Regarding where to place the tonestack in the OCD circuit is really a matter of taste.
You may prefer the way it sounds before the diodes, after the diodes or after the second gain stage.

The bax tonestack will attenuate the signal so the OCD may or may not have enough gain to give you the volume you want without extra alteration.

I beleive that the simplest place to try the BAX in the OCD would be after the 1uf Cap coming from pin 7 of the TL082.
Try it with just the tonestack and forget about the input buffer and gain-recovery stage in the Bax in a Box schem. Ditch the HP/LP and tone control of the OCD and just use the BAX instead. Have the output of the BAX feed into that 500K volume control.
See if that nets you enough volume and controls the frequencies how you like. If it's not loud enough, you can either raise the value of the 150K resistor going from pin 7 to 6 of the TL082 or reduce the value of the 39K resistor. You may or may not then need to adjust the value of the 220pf or 100n caps to compensate for the change in frequency response ... the new tonestack might do this for you though.

Trying to place the tonestack between the two TL082 stages would involve heavier modification because those two stages are DC coupled. Adding the tonestack between them would interrupt this coupling and would require you to reconstruct the 2nd TL082 stage so that it has its own bias point.


and please, anyone who sees where I misspoke, feel free to correct 

[jwblant]

In that case, I guess it's just a matter of building it and seeing what happens.  I'm pretty sure I'm going to end up with it after the last coupling capacitor.  But to be clear, I won't really need a unity gain buffer, correct?  So if I'm not happy with the output, I can just crank up the gain of the 2nd stage?

Also, what do the caps in the feedback loops do exactly? And what do you mean by saying that the two gain stages are "dc coupled."

Thanks for your help!

[Derringer]

In that case, I guess it's just a matter of building it and seeing what happens.  I'm pretty sure I'm going to end up with it after the last coupling capacitor.  But to be clear, I won't really need a unity gain buffer, correct?  So if I'm not happy with the output, I can just crank up the gain of the 2nd stage?

correct. And if there's too much bass coming through the output no matter where you put the bass control, reduce the value of the volume pot

Also, what do the caps in the feedback loops do exactly?

The 220 pf cap in the loop along with the 150K resistor form a low pass filter that attenuates the frequencies starting at about 4.8Khz on up at 6db per octave.
So if you bumped the resistor up to 220K and you wanted the same high frequency roll-off you'd have the replace the 220 pf cap with 150 pf.
To do these calculations I'm just using the boxed-up-calculator here:
http://www.muzique.com/schem/filter.htm
along with the conversion chart (pf to nf to uf) here:
http://www.justradios.com/uFnFpF.html

the 39K resistor along with the 100nF cap creates a high-pass filter that attenuates frequencies starting at about 40Hz on down at 6db per octave.
The two filters together create a band pass filter (the middle stuff gets through and the extreme lows and highs are rolled off)
so if you were to reduce the 39k to maybe 22K and did not change the 100nf cap, your low roll-off point would be at 72Hz so you'd lose a little boom on the bottom end of things

And what do you mean by saying that the two gain stages are "dc coupled."

There is DC coupled vs. AC coupled. If stages are AC coupled, that means that there is a capacitor between them that blocks the flow of direct current from one stage to the next. Capacitors allow AC (alternating current), the guitar signal, to pass.  When this is the case and each stage needs a bias voltage, then that voltage must be supplied to each stage individually.

Take a look at the OCD schem and look at where the 9V+ enters the circuit. It encounters a 10K resistor in series with another 10K resistor to ground. This is a voltage divider and is there to create the bias voltage of the first stage at 1/2 of 9V = 4.5V.
Look how there are no capacitors blocking the DC (direct current) from the output of the first stage at pin 1 to the input of the next stage at pin 5.
Now notice that there is no connection of that bias creating voltage divider to the second stage.
How is that second stage biased? It's biased by the 4.5V DC coming out of Pin 1 that the guitar signal is also riding on.

Thanks for your help!

glad to be able to help!

[jwblant]

Thanks! You've been a great help so far. I'm an electrical engineering student so we've covered all the stuff you're talking about in our circuits class but it's all been theoretical so I'm still working in recognizing the stuff in a practical circuit!

My next question: what's the advantage of putting the freq. filters in feedback loops? Wouldn't the tone stack take care of the filtering?

[Bill Mountain]

Thanks! You've been a great help so far. I'm an electrical engineering student so we've covered all the stuff you're talking about in our circuits class but it's all been theoretical so I'm still working in recognizing the stuff in a practical circuit!

My next question: what's the advantage of putting the freq. filters in feedback loops? Wouldn't the tone stack take care of the filtering?

Well.  Not every pedal has a tonestack so for some builders its just habit.  The old Ross distortion I have doesn't have a feedback cap but it does has a LPF in front of and behind the opamp to avoid nasty highs and oscillations.

My only concern with your design is that if you increase the gain in the second stage too much you may get some unwanted op amp clipping.  I would keep everything as is on the OCD (up until the 1uf cap after the 2nd gain stage), remove the input buffer of the Bax-in-a-box (not needed in this setup) and use the last opamp as a recovery stage like in the original Bax-in-a-box but set it to the max gain without clipping, then put the volume pot after that.

[jwblant]

Well I'm actually working on a custom design for a friend so I'm trying to figure out all this along the way. But he specifically asked for a high and a low eq knob and so that BAX seemed like it would be the best fit for what he wanted as opposed to the single knob that comes stock on the OCD. 

This schematic shows that the 2nd gain stage has about a gain of about 5 so I think that even after the attenuation from the bax, the opamp should have plenty of gain to make up for the loss.  If not, I'll check it out and see if I can boost it a little more.  If I do start clipping the opamp I'll look into adding a recovery stage. 

While I'm at it, I might as well talk to you guys about a few other mods I was wanting to make to the circuit.  I was going to play around with the values of the cutoff frequencies in the feedback loops.  The guy said he wanted a tone like the OCD except with a more treble sound to it so that it really cuts through the mix.  Also he wants a slightly compressed sound to it.  He doesn't want it as transparent as a Timmy, but nowhere near as compressed as a Tubescreamer.

So would playing with the input caps and cutoff frequencies affect the treble like I want it? And how do I get that "compressed" sound without overdoing it?

Thanks!

[ashcat_lt]

The Bax section could be referenced to the 4.5V point if you wanted to stick it between the two OCD stages without adding coupling caps.  Or you could put a big fat cap on that ground leg and accomplish about the same thing.

The Bax is somewhat sensitive to source impedance.  Very large values seem to cause a loss of highs.  A straight opamp output is way lower than necessary.  The 10K after the first OCD stage lost a tiny bit of treble in simulation, but not a huge deal.

It is extremely sensitive to the load impedance.  As that Z drops you start to lose bass fast!  You definitely want to control that, rather than leaving it at the mercy of whatever is plugged in after.  The second stage of the OCD is super high, so no problem there.

To get more treble out you can decrease the value of either or both of the 220pfs.  You might also mess with that 1nf before the clipper.

[jwblant]

It is extremely sensitive to the load impedance.  As that Z drops you start to lose bass fast!  You definitely want to control that, rather than leaving it at the mercy of whatever is plugged in after. 

So are you suggesting that I should use a buffer/recovery stage after the bax as well? How else would I control the load impedance? Also, are you saying that it will be fine to place it right after that 1uF cap after the 2nd gain stage?

Thanks!

[ashcat_lt]

Yes, you want to know for sure what the load will be, so you need a buffer, and preferably an opamp or JFET for high-Z.  That buffer can also provide makeup gain if you want it.

Yes, after the coupling cap is a fine place to put it.  Impedance there will be very low for guitar frequencies.

[jwblant]

Would you prefer a JFET or an op-amp buffer? I'm leaning toward the JFET due to the smaller size. I don't want to try to cram another IC onto this board haha

Will this buffer work well enough?

[ashcat_lt]

Me?  I'd prefer an opamp, but only because I have an irrational fear of transistors.  I'd probably go to a quad opamp in place of that dual, even if it meant leaving one stage unused.

[Derringer]

Would you prefer a JFET or an op-amp buffer? I'm leaning toward the JFET due to the smaller size. I don't want to try to cram another IC onto this board haha

Will this buffer work well enough?

that buffer will work just fine after the BAX if you tack the Bax onto the output of the OCD's second stage.

I kinda like what ashcat_lt said though about putting the BAX in-between the OCD's two opamp stages (perhaps between where the 1n34 branches off and the 10K before the 2nd gain stage starts) and instead of referencing the BAX to ground, reference it to the bias voltage (4.5v)
... pretty sure that would work

[Davelectro]

Be careful, passive Baxandall has a 20db signal loss with both controls on five.

Besides, it works better with type-B Log pots.

[Derringer]

Be careful, passive Baxandall has a 20db signal loss with both controls on five.

true, but the gain recovery stage can be altered to compensate for that

Besides, it works better with type-B Log pots.

ehhh ... whaddaya goanna do?

the baxandall explained on this page uses Log pots and looks fine
http://www.duncanamps.com/technical/tonestack.html

[BubbaFet]

Do a search here (diystompboxes) for James tonestacks. (James are passive, Baxandalls are active). There are some good links to the actual papers by James (1949) and Baxandall (1952). Great explanations of how they work. Definitely worth the reading and understanding. These tonestacks have been around longer than both of us, and are still IMHO the 'best'. 

[jwblant]

I'm being really tempted to place before the 2nd gain stage because that would really help me keep my schematic simple.  Plus, I can crank the gain as much as I want on the 2nd stage to make up for the drop in the stack.  This also allows me to keep the Z_o high ("infinity" for the opamp) 

I considering going with the quad for a minute, but that just takes away my main choices for opamps. (i.e. 4558, TL082, plus a couple of Burr Browns, etc...)  It also takes up more real estate and the PCB layout I have is already pushing it as far as size goes. 

And while we're on the topic..... which type of pot would you guys recommend? I know linear is the general tone choice for most, but have any of you guys actually fooled around with this type of stack to know?

[PRR]

> I know linear is the general tone choice

Passive James "requires" 10% Audio taper to work as expected.

20% taper is not fatal; the "5" position is significant bass/treble boost and true Flat position is more like "3" on the knob (making any Cut very twitchy).

Linear puts "Flat" at "1" on the knob, so Boost takes-up 90% of a turn and the entire Cut zone is cramped-up in the "1-0" crack.

IMHO the James is for tracks which are already mixed and balanced and tone-shaped, but need mild general tweaking (either way) for the listener's room and taste. Playback Systems. It's not as useful for music Creation. Yes it has been used in electric guitar amps, but not much, and perhaps mainly because it was SO common in Hi-Fi. (You could buy the complete network as a unit.)

[jwblant]

> I know linear is the general tone choice

Passive James "requires" 10% Audio taper to work as expected.

By "10%" are you referring to the tolerance of the pot?

[PRR]

No.