Capacitor selection for Power stage ripple filtering (TCE Spark Booster)

[matopotato]

... and it doesn't fix the problem with the output being tied to ground when the input is tied to Vref.

How would that problem manifest itself?
Still, madbean and pedalpcb let it be like that. I suspect it is traced off the original. Which would surprise me if the impact is big?
I am not questioning your concern or conclusions, I just can't make it add up.

[Matthew Sanford]

Looking at the layout from guitarpedalbuilders, the power supply resistor is 1R, maybe that is why 2 x 220u instead of a 100u to match a higher resistance? It does seem the 100ns are there next to where the power goes to each op amp, but kind of silly so far from the power pins...so maybe that is just bad layout that works anyway as peeps have mentioned?

Still, what would be the advantage of a low power supply resistance compared to the usual ways? Does it ensure this pedal on a power chain gets a bit more so has more volume to it or something?

[ElectricDruid]

... and it doesn't fix the problem with the output being tied to ground when the input is tied to Vref.

How would that problem manifest itself?

Distortion, I would guess, or much more limited headroom, at least. The input and the output are fighting. Perhaps one wins easily and there's no problem. In that case, it would have to be the input that wins and the op-amp simply regards the 10K on the output as "extra load". That's quite possible, since a 10K load isn't huge for an op-amp.
But the question remains - what the hell is it doing there?!? I'd remove it and see if anything changes. My bet: "No".

Still, madbean and pedalpcb let it be like that. I suspect it is traced off the original. Which would surprise me if the impact is big?
I am not questioning your concern or conclusions, I just can't make it add up.

Neither can I. Why would you have the output of an op-amp dragged to some voltage that the input is trying to force it away from? That's the bit I don't get. Maybe there's some reason why it doesn't matter/isn't the case in this particular situation, but I'm not seeing it.

Hey Bean! Shout out! What do you know about it?

Thanks,
Tom

[matopotato]

Looking at the layout from guitarpedalbuilders, the power supply resistor is 1R, maybe that is why 2 x 220u instead of a 100u to match a higher resistance? It does seem the 100ns are there next to where the power goes to each op amp, but kind of silly so far from the power pins...so maybe that is just bad layout that works anyway as peeps have mentioned?

Still, what would be the advantage of a low power supply resistance compared to the usual ways? Does it ensure this pedal on a power chain gets a bit more so has more volume to it or something?

Ok, I just thought the 1R in that layout was used as a sort of fuse protection thing since that build had no protective diode on the 9V

[Rob Strand]

What *is* that input stage, anyway? It's not something I recognize.

It's just a Sallen and Key filter but with a few things that obscure it.

Normally when you have a single supply LPF you put a bias resistor at input resistor, in this case before the 4k3.  You can see this method on Boss pedals like the CE2.  However in the case of the TC pedal they have added bootstrapping to removed the effect of the bias resistor loading at the input.  That doesn't change the fact it's a Sallen and Key Filter.  (You can even have resistors at the input of opamp with no boot-strapping but you get a LF gain loss.)

The bootstrapping doesn't appear to affect the filter, in the sense that it doesn't load the filter but also requires the same part values as when the bias resistors are moved to the input.    (There is a difference between the two in that bootstrapping provides a high input impedance but the common biasing scheme sets the input impedance with the bias resistor.  However that can be made equivalent as well by selecting the grounded input resistor value (1M in this case).)

The bootstrapping and the funky part value makes me think the person who designed this doesn't have much filter design experience.  I could be wrong there may be a motive behind it that's not obvious.  I haven't put the filter through it's paces to see if I can spot anything.

The other thing that obscures the look of the filter is the 10k resistor to ground.  That biases the opamp into class A and doesn't affect the filter.  That's a drawing issue, it would be clearer moved to the opamp output.

[ElectricDruid]

Thanks Rob!

[R.G.]

I am a bit surprised that it seems that TCE use them there then. Both schematics seem to concur.

Yeah, it's that "mostly get away with it" thing. In a copying world, sub-optimal and non-fatal mistakes get copied. Non-fatal mistakes get copied, and varied upon. It's kind of Darwinian.

I noticed in both schematics that R3 is 10k2. I am trying to understand what difference and impact that 10k2 will have vs a 10k resistor?
If it would have any impact on RC filtering, then 2% would probably not be enough for me to tell. Or does it have other purpose(s)?

The oddly specific value 10.2K probably mean's it's a 1% resistor, and that in turn probably means metal film. Metal film will have lower noise in general than a carbon film type, although not a whole lot less for modern resistors. Maybe for lower noise?
A possible secondary point might be detection of copying. An oddly specific value appearing unnecessarily is a good indicator of that. A friend of mine used to work at RCA back when RCA was a real company and made broadcast equipment. It was the practice of the engineers to "sign" their work with the values of doesn't-matter-just-need-a-resistor location. One would use 10K, one would use 8.2K, one would use 91.K, etc.

[antonis]

One would use 10K, one would use 8.2K, one would use 91.K, etc.

Due an order of magnitude difference of the later resistor, I presume you probably mean 9.1K (9k1)..

[Rob Strand]

I did a comparison of the filter with 4k3 + 10k2 and 4k7 + 10k.   There is very little difference, see plots.

As mentioned before there is a very significant difference when different impedances are presented to the input of the tc filter.   I looked at two scenarios: the first is a single coil guitar plugged into the pedal and the second is the pedal is fed by the buffered output of another pedal.   In the buffered case the response is essentially the response of the tc filter itself, which has a cut-off around 13kHz.

When the pickup is directly connected to the pedal, the filter has about -10dB more attenuation above 4kHz.

The guitar volume pot was maxed out.    We aren't looking at the effect of the backed-off volume control impedance.  That will be more severe.  We are just looking at the interaction of the filter with the pickup + guitar electronics + cable.

In the plots I've removed the response of the pickup + guitar electronics + cable.  What is shown is the response of the filter with and without the interactions with the pickup.