Boss SD-1 C6 mod???

[holio cornolio]

First of all sorry if I'm posting this in the wrong place. - a few questions here, but here's the background:
I've built a couple of Boss style overdrives lately - actually both are iterations of the OD-1 although one has a number of component value swaps. I have to say, I really really like them. I used to have a SD-1 which I recalled also enjoying a lot. So I picked one up off ebay for not too much with the intention of taking the soldering iron to it... and I will soon, although I do really like it a lot as it is (it's a plain vanilla MIT version from about 2003 I think and sounds good, but it's very easily replaceable). So I'm researching recommended mods, which seem to consist mostly of capacitor upgrades which I get, one or 2 resistor value changes, some diode swaps and IC swaps. I won't be swapping diodes or ICs, because I like the sound of the clipping as is, and I'm sceptical about the real perceivable value of IC swaps (Full disclosure, I don't really know what I'm doing, I am picking up some knowledge and I now have a basic understanding of what the component swaps do for the most part, but circuit architecture is a closed book to me... but I do understand the concept of batch variability, and it strikes me that the subtle differences that people describe in the sound of different brands of 4558 chip could easily be explained by a small amount of batch variability)
One thing though that almost every mod thread suggests, is to remove C6. C6 is a 0.01uf cap which runs in parallel to a 10k resistor in the second stage (tone section?) of the IC. Mods suggest that removing this cap introduces 'clarity'. What does that even mean? I don't understand what purpose C6 is playing in the circuit, if it's totally removable. Is it just filtering off some (very) high frequencies? Anyway, a lot of chat for a simple question. what's the point of C6? And will I regret removing it when I start playing at gig volume?

[Steben]

Swapping the IC is indeed only to think of if you experimented changing all other components first. In other words: subtlety squared.
C6 forms a low pass filter with the 10k with a corner freq of 1.6kHz. That is indeed not quite bright.
Cutting it will make for a very bright tone compared to the original. Better would be experimenting with swapping it for a smaller value. 4,7nF for example or 2,2nF. Or make them switchable if you like.

What I did with my SD-1 is making clipping options.

[ElectricDruid]

+1 what Steben said. It's just a basic high frequency roll-off, like you thought.

From a noise point of view, it's better to limit the frequency range of the op-amp (even if that's just cutting out ultrasonics) so don't go lower than 1n or so (15KHz). Total removal is easier, but not best practice.

[holio cornolio]

Thanks both. Useful information. I had the feeling that simply removing c6 would be sub optimal. Have some other values that I can sub in and see if I like them instead.

[holio cornolio]

So I pulled c6 and replaced it with a 4700pf capacitor. All I had was a radial vishay polyester film cap which was a bit of a squeeze to fit in. I did give it a listen without c6 installed and I think it was a bit bright. I can imagine that into a cranked ampnit would start to sound a bit harsh, but at bedroom volume I can see that it might lift a little muffle from the sound. I put the 4700pf cap in there and it doesn't sound a lot different from stock, maybe just a touch brighter. I also replaced a few of the polyester caps with metal film caps of slightly higher quality and slightly higher values wh :)ich has made a subtle but audible difference imho. One swap that I made that seemed pretty much valueless was increasing c17 from 100uf to 220uf. It's a mod that I read about somewhere and I thought I'd give it a whirl because I couldn't see why it would make a tonal difference and frankly, I'm pretty sure now, having done the mod, that it doesn't.

[holio cornolio]

the vishay is worth a giggle. Doesn't quite fit.

[Rob Strand]

All I had was a radial vishay polyester film cap which was a bit of a squeeze to fit in. I did give it a listen without c6 installed and I think it was a bit bright. I can imagine that into a cranked ampnit would start to sound a bit harsh, but at bedroom volume I can see that it might lift a little muffle from the sound.

The existing 10n value isn't bad so it's best do these mods in small steps.  I actually suspect 8n2 will do it.  However, it's very useful try steps 8n2, 6n8, 5n6, 4n7.    If you can compare them quickly against each other you can make a better choice.

[ElectricDruid]

C17 is a power supply smoothing cap, so it might help reduce hum a bit if you've got a noisy power supply, but it won't affect the tone, no.

I think Rob is being very cautious in his recommendation of going from 10n to 8n2, 6n8, etc for C6. Even dropping straight to 4n7 is only changing the cutoff by an octave. For -6dB/oct filters like we're talking about, that's not a huge difference.

Personally, I'd drop the value in octaves (10n, 4n7, 2n2, 1n), and if I found I'd gone a bit too bright, I might go back up by a half-octave. Tastes differ, I guess. Maybe it's just a question of how much time and patience you have! A bit of experimentation is definitely a good idea though - we'd all agree about that.

[Rob Strand]

I think Rob is being very cautious in his recommendation of going from 10n to 8n2, 6n8, etc for C6. Even dropping straight to 4n7 is only changing the cutoff by an octave. For -6dB/oct filters like we're talking about, that's not a huge difference.

Only because the original value isn't bad and going to 4n7 starts to change the character of the pedal to something more "transparent".

[Rob Strand]

I think Rob is being very cautious in his recommendation of going from 10n to 8n2, 6n8, etc for C6. Even dropping straight to 4n7 is only changing the cutoff by an octave. For -6dB/oct filters like we're talking about, that's not a huge difference.

Personally, I'd drop the value in octaves (10n, 4n7, 2n2, 1n), and if I found I'd gone a bit too bright, I might go back up by a half-octave. Tastes differ, I guess. Maybe it's just a question of how much time and patience you have! A bit of experimentation is definitely a good idea though - we'd all agree about that.

Sorry to be a dick and back track on what I've said.   I shouldn't work off my memory of the schematics. 

The cap in that position could be moved to 1n.    After that play with another cap across the LEVEL pot, maybe in the 4n7 to 12n zone.

[holio cornolio]

I think Rob is being very cautious in his recommendation of going from 10n to 8n2, 6n8, etc for C6. Even dropping straight to 4n7 is only changing the cutoff by an octave. For -6dB/oct filters like we're talking about, that's not a huge difference.

Personally, I'd drop the value in octaves (10n, 4n7, 2n2, 1n), and if I found I'd gone a bit too bright, I might go back up by a half-octave. Tastes differ, I guess. Maybe it's just a question of how much time and patience you have! A bit of experimentation is definitely a good idea though - we'd all agree about that.

Sorry to be a dick and back track on what I've said.   I shouldn't work off my memory of the schematics. 

The cap in that position could be moved to 1n.    After that play with another cap across the LEVEL pot, maybe in the 4n7 to 12n zone.

I don't think that's being dickish at all. It's incredibly helpful. Frankly though, if I'd dropped in stages as you originally suggested, I doubt if have noticed the very subtle and progressive brightening without rechecking the original value between swaps. I'm pretty happy with what I have in there now, it's not a huge step away from the stock tone, just a hair of extra brightness in a range that tweaking the tone pot clockwise can't achieve. Also, I only have a limited range of cap values. I may revert to stock at some point. We will see...

[Electron Tornado]

Just my 2 centavos:

One thing though that almost every mod thread suggests, is to remove C6. C6 is a 0.01uf cap which runs in parallel to a 10k resistor in the second stage (tone section?) of the IC. Mods suggest that removing this cap introduces 'clarity'. What does that even mean?

Keeping C6 in makes the amp sound like it's under a blanket, compared with removing it. That's where the "clarity" description comes from. I liked the sound with C6 removed completely. If the pedal starts to sound harsh, try adjusting the Tone knob, or increase the values of C2 and/or C3. (Referencing schematic here:  https://www.hobby-hour.com/electronics/s/sd1-super-overdrive.php  )

I know you don't plan on any diode changes, but the stock clipping of the SD-1 is asymmetric, while the Tube Screamer is symmetric. It's a nice mod to add a switch so you can select between the two.

Changing the IC. Watch this video and decide:

[ElectricDruid]

Keeping C6 in makes the amp sound like it's under a blanket, compared with removing it. That's where the "clarity" description comes from. I liked the sound with C6 removed completely.

While it's easier to remove it completely, like that the op-amp is also amplifying any ultrasonics or radio frequency noise that you've picked up in your signal chain. C6 acts to limit the upper bandwidth. To reduce noise, it's usually recommended to limit bandwidth only to what is required - in our case that's audio, so we can certainly roll off everything above 20-25KHz since we can't even hear it.

C6=1n gives 15.9KHz, which is very close to full audio and probably still plenty for a guitar.
C6=680pF gives 23.4KHz which is definitely full audio.

Going lower than that, or removing C6 altogether, is just going to amplify stuff you can't hear and cause noise problems. That said, the tone control isn't boosting things enormously so it's not a massive problem, just a question of best practice.

C2/18n forms a highpass filter with R4/100K. These values give a rolloff below 88Hz. Increasing C2 pushes that rolloff lower and might marginally increase bass if a guitar actually had any frequencies down there, which it probably doesn't. You might notice on a Bass guitar.

C3/47n and R6/4K7 form another highpass rolloff (a shelf, technically, since the gain can't go below one for a non-inverting op-amp). These values give the classic tube screamer 720Hz bass roll-off. This definitely does affect the sound, and increasing it will give a thicker, boomier sound. It won't remove harshness, but it might provide some heavier bottom end to balance it out.

[Electron Tornado]

While it's easier to remove it completely, like that the op-amp is also amplifying any ultrasonics or radio frequency noise that you've picked up in your signal chain. C6 acts to limit the upper bandwidth. To reduce noise, it's usually recommended to limit bandwidth only to what is required - in our case that's audio, so we can certainly roll off everything above 20-25KHz since we can't even hear it.

C6=1n gives 15.9KHz, which is very close to full audio and probably still plenty for a guitar.
C6=680pF gives 23.4KHz which is definitely full audio.

I'll have to try some different values for C6 if I can get some time in a noisier environment. I haven't noticed that it became any more noisy without C6. Unless the first op amp circuit is where the noise is actually picked up, I wonder why the RF noise wasn't filtered earlier.

That brings up an interesting question - what is a good way to test a pedal's susceptibility to picking up ambient electromagnetic noise?

[holio cornolio]

Just my 2 centavos:

One thing though that almost every mod thread suggests, is to remove C6. C6 is a 0.01uf cap which runs in parallel to a 10k resistor in the second stage (tone section?) of the IC. Mods suggest that removing this cap introduces 'clarity'. What does that even mean?

I know you don't plan on any diode changes, but the stock clipping of the SD-1 is asymmetric, while the Tube Screamer is symmetric. It's a nice mod to add a switch so you can select between the two.

So I have been a/b ing my now modded sd-1 against my od-1 clone. It's based on the byoc kit so is the quad opamp circuit. The 2 pedals are quite close in tone but there is something about the drive on the od-1 that appeals to me a bit more. It seems a bit less compressed and a little more raw sounding than the sd-1. I don't think that would be opamp related (maybe I'm wrong) bit I am now toying with the idea of shoving in some diode options. I have some bat62 Schottky diodes in the box that I e never used in a build. Anyone any thoughts on using Schottkys in a sd-1? Looking for a looser, less compressed drive sound.

[Rob Strand]

You might be a bit surprised at these findings: 15n SD1 treble cap closer match
to OD1 than 10n or 4n7.

I think one key difference is the input filter OD1 4n7 vs SD1 18n.


https://www.hobby-hour.com/electronics/s/od1-overdrive.php
https://www.hobby-hour.com/electronics/s/sd1-super-overdrive.php

Key points:
- the tone control messes with the response a bit
  even in the flat position.
- the 10n cap tries to compensate
- it actually turns out 15n on the SD1 is closer
  to the OD1 than 10n.  (below and above 500Hz)

Other differences:
- The input filter or the OD1 cuts the bass more
   than the SD1.   This gives the OD1 more clarity
   and the SD1 more thickness.
   OD1 input cap 4n7
   SD1 input cap 18n



Edit: schematic links added.

[holio cornolio]

Thanks. That is interesting. The input cap would be c2 right? I will need to check my I've actually bumped this up to 0.047uf in my sd-1 (although mod suggested bringing it up to 0.1uf to flatten the mod hump, but I actually like the mod hump. I might try that swap. I think I have those cap values in the drawer.

[Rob Strand]

Thanks. That is interesting. The input cap would be c2 right? I will need to check my I've actually bumped this up to 0.047uf in my sd-1 (although mod suggested bringing it up to 0.1uf to flatten the mod hump, but I actually like the mod hump. I might try that swap. I think I have those cap values in the drawer.

Yes C2 (on both the SD1 and OD1 schematics).

As far as the low end of the hump goes you can also play with C3 (on both the SD1 and OD1 schematics).

C2 cuts the bass before the clipper and C3 cuts the bass within the clipper.  If one of those cap is largish and the other is used to cut the bass you get a slightly different sound when C2 cuts the bass than C3.  After spending many hours playing with this in my younger days it seems using C3 to cut the bass is slightly better.   It's when you have bass cut in both positions you can over do it.   Nonetheless before I knew what what actually going on inside the OD-1 sounded fine  .  There's always a trade-off between mud and clarity, and also Mid-hump vs transparency.  I don't think you ever win that game.

Yesterday I just spent 10 mins comparing the two circuit and came up with the 15n thing.   This morning I had a bit more of a critical view.   So it's now obvious making the cap 18n makes response of the the OD-1 and SD-1 the same when the SD-1 tone pot is centrered.    It's just symmetry, the 10k + 18n on the feedback matches the 10k + 18n on the input and that helps remove the secondary effect of the tone control.      The side-effect of that matching is when you advance the tone pot to the boost side the 18n feedback cap will make the pedal darker - good bad who knows.

Another way to match the two pedals is to leave the 10nF in the SD-1 and use the Tone control position to get the best match.   It turns out with the existing part values it can't be done.  Backing off the Tone can match the highs of the SD-1 to the OD-1 but the SD1 then has some more attenuation in the mids.

If you wanted to have the SD-1 match the OD-1 *and* not make the feedback cap a low value like 15n or 18n then it's going to take more work.  Probably needs some careful choice of component values and adding some more parts.

[holio cornolio]

I didn't start out with the intention of matching the sd-1 to the od-1 to be honest, but when playing them side by side, setting aside the eq, there is something in the quality of the drive of the sd-1 that is different. It's not just eq, although that is part of it, but there's a compression and a control to the sd-1 whereas the od-1 has, as you say, more clarity but also a looser, less compressed feel. I don't have great ears, and it's hard for me to pick out the subtleties (the c6 removal was audible but not breathtakingly so for me) and the 2 pedals do sound very close - definitely related, but there's something wilder and more ragged about the od-1 that I would like to transfer to the sd-1.

[Rob Strand]

I didn't start out with the intention of matching the sd-1 to the od-1 to be honest, but when playing them side by side, setting aside the eq, there is something in the quality of the drive of the sd-1 that is different. It's not just eq, although that is part of it, but there's a compression and a control to the sd-1 whereas the od-1 has, as you say, more clarity but also a looser, less compressed feel. I don't have great ears, and it's hard for me to pick out the subtleties (the c6 removal was audible but not breathtakingly so for me) and the 2 pedals do sound very close - definitely related, but there's something wilder and more ragged about the od-1 that I would like to transfer to the sd-1.

Just the same, by putting pedals side by side you do start to pickup on small differences.  It's pretty normal to like characteristics of one and not the other.    To me OD-1's are always clear and SD-1's are muddy even though I know there's not much difference in the circuit.   The TS-9's are similar but different, even if you throw asymetrical diodes in the TS-9.

Working out what *really* causes the difference is the hard part.

If you want to see the effect of the different C2 values between the pedals, what you can do is lift one lead of R8 (470 ohm) on the SD-1, or, desolder the centre wire on the SD-1 Tone pot.   What that does is force the filter on the SD-1 to be the same as the OD-1.  It should completely removing the difference from the messy tone control section.   That leaves only the C2 value and if you still hear a difference then it's likely to be C2.  So then as an experiment make the C2 value on SD-1 the same as the OD-1.   Does that reduce the difference and get you where you want to be?