PedalPCB Darkglass B3K Clone question involving biasing and tone shaping

[bushidov]

Hi Guys,

I've noticed something that I am not sure of "why", on a schematic for the Darkglass B3K clone from PedalPCB. I've seen this schematic thrown around on several sites, but I grabbed PedalPCB's as it was easy to grab.
https://www.pedalpcb.com/docs/Obsidius.pdf

If you see below:


We have an overdrive style pedal being used as a pre-amp, which is also designed to target bass players, albeit, it does work well for really down-tuned guitars as well (baritone folks). It consists of TL072 op amps and a CD4049UBE inverter.

After it leaves the one inverter, it goes to a non-inverting op-amp "IC3.2" and then does what appears to be some RC filtering.

After that, is where my confusion begins. IC4.2 and IC4.1 appear to also be "non-inverting op-amps", but don't their non-inverting pin going to bias. They do have a small capacitor going to ground, but from a DC perspective, that means nothing to bias.

Now, I've build one of these clones, and made one as well, both with the same schematic in mind. When I have the pedal plugged into power, I get a really high pitch ringing/oscillating, even though I have the pedal disengaged. It's faint, and I really have to turn my amp high to hear it, but if I have any pedals in front of it, that sound does get amplified in that case. When I add a 1M resistor to bias on the non-inverting inputs of both of those op-amps, the ringing goes away, almost completely.

So it seems that's a problem and a fix. But here's the part that confuses me even more: I could go with, "well everyone just copied off the same schematic, and it was wrong, and that's why it is there". However, I have a Darkglass B7K. It's just a Darkglass B3K + a 4 band EQ section (and a couple of capacitor values changes that are slight. So, it has that same part (IC4.1 and IC4.2). And I tore the pedal down to its PCB and put it back together again successfully. It's laid out the same way in that spot. However, no ringing. So, what the heck?

So, the questions are:
1. What is that section doing, as far as tone shaping is concerned?
2. Are they non-inverting op-amps?
3. If so, why are they not tied to bias?
4. Why did tying them to bias with 470K resistors fix the mild, but high-pitch ringing?

Thanks again for all your help!

[PRR]

> IC4.2 and IC4.1 ....don't their non-inverting pin going to bias.

Each stage "follows" the one before.

[Rob Strand]

After that, is where my confusion begins. IC4.2 and IC4.1 appear to also be "non-inverting op-amps", but don't their non-inverting pin going to bias. They do have a small capacitor going to ground, but from a DC perspective, that means nothing to bias.

The opamps are buffers.  They are wired as Sallen and Key low-pass filters.  (Well more or less as the bridge-T network R21, R22, C16, C17 messes with the first filter's response a bit.  That's on top of the bridge-T response.)

Bridge-T part provides mid-dip.
Following 2x low-pass filters remove a lot of high frequency junk.

See the Sans-amp Bass-Driver.  There was also a simplified version "BD21 Lite" or something which used a similar filter to the Darkglass B3K.

https://s232.photobucket.com/user/alembicplyr/media/Schematics/bd21_lite_v10.png.html

You can see a lot of similarities in the two circuits.
(Notice also the 1M on the Bass/Treble controls, like the Darkglass 7K!)

[Crowella]

1. What is that section doing, as far as tone shaping is concerned?

Lets look at everything *after* the CMOS.

IC 3.2 or rather, specifically after it in this case is a Bridged-Tee Notch Filter. The way the schematic is drawn disguises it to look like it's part of the Op-amp when in reality, IC 3.2 is just a unity gain buffer and the resistors and caps make that filter. It knocks out somewhere around the 800Hz mark a bit.

IC 4.1 and 4.2 are both forms of Sallen-Key Low Pass Filters. If you want to do a bit more reading and get some nice maths involved in it.
https://sound-au.com/articles/active-filters.htm#s3

Here I plotted IC 4.1 and 4.2. You can also note that dip around the 800Hz point. (I should note this is using the schematic provided)


To summarise, the B3K tone starts with a mu-amp driving the signal heavily notch filtering out around~300Hz and about 1.5KHz before driving it, sending through a HPF (the grunt switch), then followed by the CMOS stage, then everything I listed above to clean up the noise a bit.

2. Are they non-inverting op-amps?

Yep, they are!

3. If so, why are they not tied to bias?

None of these stages are ever decoupled with a capacitor since each stage works fine while sitting at 4.5V. PRR shows a great picture showing the signal flow if that makes sense. There's about 4 parts not needed by doing it this way. In fact, even keeps that way for the blend

4. Why did tying them to bias with 470K resistors fix the mild, but high-pitch ringing?

Ahhh, I think this is the question you probably want answered most. I don't know. It should work fine with the 1M resistor covering all that.


Side note:  I'm sure the B7K doesn't use TL072 but uses MC33178 and MC33179 (ooh, a quad chip) so I wonder if that's something to throw into the pot.

EDIT: Rob beat me to it.