Boss SD-1 questions

[snk]

Hello,
I have an original Boss SD1, and on most instruments i try it on, i think it eats up too much bass (i am a synth player). But, on a couple synths, it can also sound really good.
I got interested into some TubeScreamer clones recently, until i read that the SD-1 was actually very close to the TubeScreamer.
I searched for mods on the SD1 (in fact, keeping mine intact, and building a new one on veroboard), and found many
I would have a couple questions in order to get a better understanding of the circuit (and not following blindly random advices on the web).

My sources were :
http://www.musicgarage.fi/mods-repairs/boss-sd-1-bass-mods/
https://electric-safari.com/2018/08/28/boss-sd-1-super-overdrive-mod/
https://www.talkbass.com/threads/boss-sd-1-mod-for-bass.588792/
https://modyourownpedal.com/blogs/news/a-few-easy-boss-sd-1-super-overdrive-mods



I understand that i could just use bigger values for C2, C3 and C8, and also lower R6.

But, from what i read :

C3 and R6 together set gain as well as dictate the frequencies that are boosted and clipped, in this case 723hz and above;

I guess this is a RC Hi-pass Filter, right ?
But, from a technical point of view, what will increasing C3 and decreasing R6 will do ? Lower the side-chain cutoff point ?

Changing R6 to a 2K2 will increase the amount of available overdrive.

What does it mean ? Does it mean that i will get bigger distorsion even at minimum gain, or does it mean that i will be able to reach a wider range of gain (= more distorsion at higher settings) ?

I also read that

"Changing C2, C3, and C8 to 0.1µF (100nF) will increase bass response".

Bt what does C8 do in the circuit ? If it is a RC filter, what resistor is his "buddy" ?

I also read several different things about C4 (18nF) :

The signal goes through R7 and passes a cap (C4) to ground which is a low pass filter reducing highs at about 720hz and above;

Some websites suggest to increase C4 up to 220nF (!), while other suggests to lower it to 4.7nF for « for more clarity » (what does that mean ? more treble ?)
What will increasing C4 do ?

Other mod suggestions are about C5 and R8 :

Another low pass filter is formed from the tone control which connects at this same spot and goes through C5 and R8 to ground.

Suggestions are made to increase C5 to 150nF, and decrease R8 to 220R : what would it actually do ? Damping the output further ? Or brightening it ?

I am also wondering about C6 : some say it will give more bass, but elsewhere i have read that it is in fact dampening the highs...

Last, it the choice of opamp important in these pedal ?

Thank you in advance for any help (i'm mainly trying to understand here the inners of the circuit)

[snk]

... Ah, also :
- is the SD-1 really that close to the TubeScreamer, sound-wise ? From what i remember, the TubeScreamer really was not transparent at all, with a huge bass roll-off and scooping...
- Does anyone have any experience with the Chirality Splinter ? The demos i have heard were really right up my alley

[snk]

... and one more thing : i forgot to tell what  am looking for 
- I want the pedal to let pass through more bass...
- But without changing its character, and without more distorsion (so, maybe not changing the side-chain circuit too much. I want the output to be more bass-heavy, but maybe not allowing the bass to trigger the clipping too much);
- So, I may consider a pot to tweak the sidechain response (a bit like the Ruetz mod on the Rat).
- I might also try with leds instead of silicon diodes...

[PRR]

Two main bass-cuts:

[Steben]

... and one more thing : i forgot to tell what  am looking for 
- I want the pedal to let pass through more bass...
- But without changing its character, and without more distorsion (so, maybe not changing the side-chain circuit too much. I want the output to be more bass-heavy, but maybe not allowing the bass to trigger the clipping too much);
- So, I may consider a pot to tweak the sidechain response (a bit like the Ruetz mod on the Rat).
- I might also try with leds instead of silicon diodes...

Bass cut before clipping is in the guitar world very crucial in defining the sound. Much more than clipping diode type.

[snk]

Thank you, PRR and Steben.

Bass cut before clipping is in the guitar world very crucial in defining the sound. Much more than clipping diode type.

Yes, i am aware about that, this is why I asked about which capacitors/resistors were responsible for the clipping detection circuit
In this case, this is C3 and R6, right ?

[Rob Strand]

 

I want the pedal to let pass through more bass...

It's a common goal.

Some options:
1) Less LF cut.   see AMZ's fat mods
2) Variable LF cut.  see the way it's done in Timmy, Zen Drive and many others
3  Post distortion LF boost.  Check out the aion's version of the Xotic soul driven and the Barber drives.  To some degree Boss BD-2 and OD-3 - not so easy to transfer completely the to TS-9. compared to previous examples.

All move towards the goal.  Something is gained something is lost.

[Mark Hammer]

There is always greater amplitude from bass frequencies than from mids and highs, if only because larger ferromagnetic materials moving over the pickups yield greater voltages from the pickups.

Diodes are level-sensitive devices.  They conduct as a signal approaches and exceeds their forward voltage, and do NOT conduct if that voltage is well below the forward voltage.

So, trimming back the bass content in that circuit attempts to make the amplitude of the lows mids and highs approximately equal, such that they have a similar likelihood of clipping.  Higher notes on the unwound strings should clip roughly the same amount as low notes on the wound strings.

However, clipping results in an increase in harmonic content.  As such the SD-1, TS-9, and similar circuits do several things to offset this.

• They use a capacitor in the feedback loop of the clipping op-amp, and set gain via the feedback resistance.  As that feedback resistance is increased (to increase gain) the combination of feedback-R and C rolls off more highs.  Not enough to make the impact of clipping disappear, but enough to make it sound relatively "smooth".
• They use lowpass filtering after the clipping stage to further smooth out the impact of the clipping on HF content.

The simplest and most straightforward mod to achieve a fuller sound is to increase the value of C3.  Values from 100-220nf are sufficient to do this.  Depending on the guitar, pickups, and desired tonal goals, a more pleasing result may be found with the less or greater-value within that range.  For guitar, there may be little point in going beyond 220nf.

Why THAT capacitor and not any of the others in the signal path?  In principle, any capacitor in series with the signal can have an impact on the bandwidth.  But C3 dictates where gain will be applied in the clipping stage, and so has a greater impact on the resulting sound than any of the others.

[snk]

1) Less LF cut.   see AMZ's fat mods

Thanks, i didn't know that AMZ article, and it confirms what i suspected

PS : Thank you, Mark for your knowledge, I only noticed your post afterwards.


I have a question about the C3(47nF)/R6(4.7k) configuration : entering their values in the RC filter calculator gives me the 720Hz cutoff point as expected.
1- AMZ is suggesting to change the capacitor value in order to get a different, lower cutoff point. So, i could use 2 capacitors on a switch.
2- But would it be any different to use a pot as a variable resistor instead ? Would there have any other side effect ? If I was going for a 100nF cap and a 10K pot, it would allow me go go as low as 160Hz, and as high as 1000Hz (if i limit the pot range using a 1.5K resistor, and the "original" 720Hz value using a 2k resistor).

[Mark Hammer]

Your logic is sound.  The key question is "How easily do you want to be able to replicate a given tone?".  If the amount of bass content is something you want to make instantly available, in a predictable way, then I would recommend a 3-position toggle switch.  If it is something you want to be able to fine tune to the context, and don't really care too much about replicability, then a rotary control (pot) might be the way to go.

For the toggle strategy, the circuit would depend on whether it's something you want to be able to switch in a live performance context, or merely something for futzing around in your basement or studio.  Different rolloffs can be achieved by adding capacitors in parallel, but there may be an audible pop when doing so.  That's why I distinguish between the live and noodling-around context.  Such pops won't be audible when the pedal is bypassed, but folks don't always remember to do that.  While different effective capacitances can be achieved by placing several caps in series and bypassing some selectively, the math may not work out as easily as with parallel caps.  Not a tremendous hardship, but may not concur with what you have in your parts bin.

The variable control approach has two different variations.  You are correct in assuming that changing the value R6 changes the rolloff, when in conjunction with C3.  So if we increase R6 to 10k, that lowers the rolloff to 338hz.  But, it also reduces the maximum amount of gain in that stage from 221x to 104x.  And if 104x still gives you the degree of grit you want, great.  But it IS a tradeoff.

The other variation on this approach is to do something like what is done in the Rat pedal, and many others: use a second RC network in parallel with R6/C3.  For instance, if you used 6k8 and 100nf, that would give equivalent gain down to around 234hz, as you increased the gain knob.  R6/C3 would still give more gain to content above 720hz as you turn the gain control up, but it would not be a steady 6db/oct rolloff from 720hz on down.  In some contexts, that can be the preferred option since it retains the basic character of the drive without having to completely sacrifice low end.

Your call.

[Ben N]

If the amount of bass content is something you want to make instantly available, in a predictable way, then I would recommend a 3-position toggle switch.  If it is something you want to be able to fine tune to the context, and don't really care too much about replicability, then a rotary control (pot) might be the way to go.

When modding, there are also practical constraints. A micro-toggle is a lot easier to find space for on a Boss enclosure than an added pot, so even if you'd rather have a rotary control, a switch may be a practical compromise.

[snk]

Thank you all for your insight. There are plenty of food for thought.

1- About the context

When modding, there are also practical constraints. A micro-toggle is a lot easier to find space for on a Boss enclosure

This is not a mod : i will keep my SD1 as it is, and i may build a "variation" of it on veroboard (more bass-friendly, and maybe with a 4th knob to fine tune the clipping detection circuit).

Your logic is sound.  The key question is "How easily do you want to be able to replicate a given tone?".  If the amount of bass content is something you want to make instantly available, in a predictable way, then I would recommend a 3-position toggle switch.  If it is something you want to be able to fine tune to the context, and don't really care too much about replicability, then a rotary control (pot) might be the way to go.

It is for studio use, and i am a tweaker, so i'd rather go with an added pot, if possible

2- About the sound.

You are correct in assuming that changing the value R6 changes the rolloff, when in conjunction with C3.  So if we increase R6 to 10k, that lowers the rolloff to 338hz.  But, it also reduces the maximum amount of gain in that stage from 221x to 104x.  And if 104x still gives you the degree of grit you want, great.  But it IS a tradeoff.

Ok, this is very useful to know. I wasn't aware about the gain/frequency relationship.
So, R6 defines the GAIN, right ? Increasing its value will decrease the maximum gain, and decreasing its value will increase the maximum gain available, is that correct ?

If so, this is my thoughts :
1- I have played further with my old SD1, and it would be acceptable if it had a tad less gain (in fact, less gain at minimum setting would be even good). I like the way it sounds at full gain, but i have many other distorsion & fuzz pedals, so i am planning to use that modded SD1 for making audio more full and dirty but not over the top (Rat, Mxr Dist+, Barbershop overdrive and several fuzz pedals will take in charge the heavy duty ). If possible, the same gain factor at maximum value would be best, but i can live with a little less gain (if that means that the sound is tweaked to my taste).

2-But going from 221X to 104X seems quite a lot, isn't it ? I made some calculations :
-If R6 defines gain, and if increasing its value to 10K causes a big gain loss, then i could try with 5K, right ?
- According to CR-Tool, using a 220nF cap as R6 and a 5K pot (followed by a 680R to limit the range), i would get a 127Hz cutoff point at minimum value, and a 1kHz a maximum, which seems a perfact range to me
- How much of gain loss would we get, if using a 5K pot ?

[Steben]

The Bluesbreaker has a very "in between" tone spectrum with a resistor/cap in the first gain stage set worth checking, especially the low gain version:

[Mark Hammer]

Ok, this is very useful to know. I wasn't aware about the gain/frequency relationship.
So, R6 defines the GAIN, right ? Increasing its value will decrease the maximum gain, and decreasing its value will increase the maximum gain available, is that correct ?

If so, this is my thoughts :
1- I have played further with my old SD1, and it would be acceptable if it had a tad less gain (in fact, less gain at minimum setting would be even good). I like the way it sounds at full gain, but i have many other distorsion & fuzz pedals, so i am planning to use that modded SD1 for making audio more full and dirty but not over the top (Rat, Mxr Dist+, Barbershop overdrive and several fuzz pedals will take in charge the heavy duty ). If possible, the same gain factor at maximum value would be best, but i can live with a little less gain (if that means that the sound is tweaked to my taste).

2-But going from 221X to 104X seems quite a lot, isn't it ? I made some calculations :
-If R6 defines gain, and if increasing its value to 10K causes a big gain loss, then i could try with 5K, right ?
- According to CR-Tool, using a 220nF cap as R6 and a 5K pot (followed by a 680R to limit the range), i would get a 127Hz cutoff point at minimum value, and a 1kHz a maximum, which seems a perfact range to me
- How much of gain loss would we get, if using a 5K pot ?

R6 doesn't define gain, but plays a role in it, along with R5 and VR1.  Bear in mind that the natural tendency of an op-amp is to "want" to achieve its maximum possible gain.  What results in gains less than that is negative feedback from the output; the more negative feedback, the lower the gain arrived at.  The total feedback resistance partly determines how much negative feedback occurs, but so does the "ground leg" resistance (R6 in this case).  Think of the feedback resistance and ground leg resistance functioning like a pot -a voltage divider - with the inverting pin of the op-amp serving as the wiper of that imaginary pot.  When R6 is MUCH less than VR1+R5, then more feedback signal is bled off to ground.

If you make R6 larger, or replace it with a fixed resistor in series with a pot in order to make it larger, you not only affect the maximum gain, but also the minimum gain as well.  With VR1 set to minimum, R5 = 33k, and R6 = 4k7, that minimum gain = 8x,  Not normally enough to clip unless you pick hard.  If R6's value is increased, yes you would get more bass content, but that minimum gain would change as well as the maximum gain.  If R6 = 10k, not only is the maximum gain 104x, but the minimum gain is now 4.3x.  To get back to a minimum gain of 8x you'd want to increase R5 from 33k up to 68k or 75k.  That would also increase maximum gain, but not so much that you'd notice.

If you went with your suggestion of 220nf + 5k + 680R, you would certainly be able to dial in more and much less bass.  But by making the ground leg much smaller when the pot is at minimum (680R), you will have bled off most of the negative feedback and could risk unreasonable gain levels.  At the very least, changing the "bass" amount, would always require adjusting the gain level, sometimes substantially.

In light of this,my gut says that, if there sufficient space to install a SPDT 3-position toggle, that would be the best option.  If you use small ceramic disc caps, it wouldn't need to take up much room.  I made a couple of Menatone Red Snapper clones in 1590A enclosures.  They normally have a bass control like you describe, but I simply went with switching caps via a 3-way toggle to get a trio of rolloffs.  Works just fine.

[snk]

I really liked the idea of a potentiometer to tweak the sidechain (clipping detection), but from reading what you wrote i understand it is much wiser (and safer) to go with a switch.
It also seems that is what the Chirality SPlinter is doing.
I will think about it further



If you don't mind, I am considering other things (but my mind is not made up yet) :

- I played further with the SD1's gain : I think I would be happy with a bit less minimum gain, and also the last part of the gain pot doesn't do much. In order to get less minimum gain, should use a lower value for R5 (33k) ? And a 500K pot for a more useful range ?

- I am also considering getting a wider bandwidth (less hicut) : should I try to tweak values for C5/R8 or C4/R10 ? Or maybe that reducing C6 might be enough ?

[edit]
I think i found my own answers :
- There is a mod called "C6 mod" which should allow to let more treble pass through. It is an easy to try mod, so i will test it.
- I have found suggestions to lower C5 down to 10K for less minimum gain.

[snk]

Hi,
I am now considering making a veroboard, with a couple more mods, to check it against the original SD1

Two more questions, before building it :

1- About the BASS : I have read quite often, when searching for mods advices on the web, that C2 (input cap) and C8 (output cap) should be upped to 100nF (some people even suggest 1µF!). But when i check the values in a RC filter calculator, it tells me that the original value for C2 (18nF) makes a hipass value of 90Hz. So, it seems that using 33nF would go down to 45Hz, which is already quite good. And 47nF would allow all the bass to pass : am I right? If so, isn't the 100nF cap a bit overkill in this context, or is there something else to consider ?


2- Two questions about the TREBLE : I want to experiment with some values to allow the DS1 to reach a wider bandwidth, and let more treble pass if i want to.
- I have read that C6 can be removed (or reduced) to get some treble back. I might put it on a switch (so i can get the original tone, as well as increased bandwidth).
- But I have also read that C4/R7 was a fixed LPF set at 880Hz. Using a 10nF cap instead would put the corner frequency up to 1.6kHz (which I think will be better for electronic instruments with higher frequency content than a guitar).
- Then, C5 & R8 are another LPF, working in conjunction with the tone pot. The max value is set at 12.5kHz, and i see that using a 330R resistor and a 27nF cap would set the higher frequency around 17kHz, allowing for a wider bandwidth.
Now, my question is the following : do you pedal gurus think that all the 3 mods would work well together, or would there have a very "cumulative" effect, one single good mod being better than 3 average ones (if that makes sense) ?
I know that i can try and experiment (i will use sockets), but guidelines from trusted people are always a good help

[Mark Hammer]

The cap values in the input and output buffers are quite sufficient.  Remember that they are intended to provide for an unaffected bypass signal, so why would they shave off bandwidth with no attention to what comes before or after?

The TS-9 includes a feedback cap in the clipping stage that removes a little more high end as gain is increased, in order to compensate for the added upper harmonic content.  The  lowpass filter after the clipping stage is fixed, set around 720hz, and aims for providing the same sort of tone, independent of where the gain is set.  Together, they also eliminate any of the "fizz" that can often occur with drive pedals.  I have often touted the Klon Centaur as being one of the first boosters/drive pedals deliberately designed to shape tone in order to push amps into a more pleasing overdrive, as opposed to providing an overdrive tone of its own.  To a limited extent, the TS-9/808 was also designed with the purpose of pushing an amp to provide a more pleasing overdrive sound, although it clearly has an overdriven tone of its own that is readily identifiable, no matter what amp you play into.

In the case of the SD-1, there is no feedback cap in parallel with the diodes to "de-fizz", but then the use of a diode trio with a little higher headroom (so that clippping isn't arrived at quite as soon) offers the opportunity to have something approximating an unclipped-and-bright boost.  I gather the slightly higher lowpass filter, provided by C4/R7 supports that.  Still, once the gain gets high enough, you won't want so much bandwidth.  Boss went with more lowpass filtering downstream, instead of upstream, via C6.  C6/R9 provide a rolloff beginning around 1600hz.

How much bandwidth does a "clean" guitar need?  Well, most guitar-amp speakers begin to seriously roll off above 6khz.  So, unless it's an acoustic guitar, extending pedal bandwidth past that is as pointless as increasing C2 and C8.  If you'd like the option for a brighter-but-fizzless sound, I'd suggest replacing C6 with a smaller value, like 3300pf, and using a toggle to add in 6800pf in parallel.  3300pf gets you a rolloff around 4800hz, which is more than bright enough, and the additional 6800pf brings things back very close to stock.  Keep in mind that all of these so-called rolloffs are 6db/oct, so quite gentle.  4800hz may seem low and restricted, but trust me you'll still have gobs of treble.

[snk]

Thank you for your reply.

How much bandwidth does a "clean" guitar need?

One (unusual) thing to have in mind is that i am not a guitar player : I am a synth player, and my instruments have a very high bandwidth. So, i am often looking for a trade off between keeping the effect own's character, and the instrument bandwidth (which can go both deep down and very high).
I also process quite a lot of drum machines with pedals, and snares and cymbals have their sweet spot much above 880Hz. When i play with the SD1 and a drum machine, i have to use an eq to cut a bit below 1k, and boost around 2.5-3.5khz and 15-17khz.
So yes, as you wrote "4800hz is more than bright enough" for an analog bass synth (and the DS1 sound good on bass sounds of a Yamaha CS10), but I wish it could also keep the sizzling sound of a TR-606 cymbals, if possible

In fact, most of the time i am using a RAT or an MXR dist+ or a TC MojoMojo on synths and drum machines, and always found the Tubescreamer too mid-rangey. But this video of the Chirality Splinter changed my mind; I realized that with some mods, the TS can sound great for beefing up drum machines without the signature mid bump. This is why I decided to experiment with a circuit which i can compare side by side with the original version.

[Mark Hammer]

If you want sizzle for cymbals, then you do not want an overdrive box; or at least not one optimized for string instruments.  What you want is an "exciter".  In principal, they aren't all that much different from a guitar overdrive, with a few basic exceptions.  First, the overdrive signal is blended in with clean/dry and not heard on its own.  Second, what is overdriven is not the full bandwidth but rather a seriously highpassed signal.  The result is that any existing harmonic content in the original signal is exaggerated.  Probably the best DIY example is Jules Ryckebusch's "Harmonic Sweetener" project that appeared in Electronic Musician magazine some 35 years back ( https://www.experimentalistsanonymous.com/diy/Schematics/Tone%20Control%20and%20EQs/Harmonic%20Sweetener.pdf )

Jules uses 4-pole highpass filters to remove the fundamental.  The result is boosted by a gain of 100x, and clipped with some red LEDs, then mixed back in with the clean signal.  There is nothing sacred about the filter corner frequency.  I built the circuit a few times and made some of the filter variable.  Similarly, the clipping can be varied and the gain of the clipping circuit as well.

The only caveat is that some of what gets boosted is hiss, so some form of noise control may be called for.  One possibility is to stick a back-to-back pair of diodes in series with the highpass signal before the mix-level control, to serve as a crude noise gate.

[snk]

Thank you for the hint, Mark.
I am not after an exciter : I already have an Aphex Aural Exciter (and most of the time I prefer to (ab)use my console eq, some HPF in parallel, or some saturation box of some kind run in parallel). It is just that sometimes, i find some guitar pedals too "narrow sounding", and i wish they could process a quite wider frequency range (without losing their character).
I know Baja made a clone of the BBE Sonic Maximizer, also. i will give the Harmonic Sweetenera look.

Anyway, all the informations, cap/frequencies values and directions you suggested are more than enough for me to play around with a DS-1 inspired design suiting my tastes
I fear that changing C4/R7 would change the "signature sound" of the DS-1 a bit too much. I will socket them anyway, but I will start by decreasing C6, and then maybe tweak C5/R8 a tiny bit.
For C6, putting a 10nF and a 1nF cap on a switch should allow me to keep the "factory setting", as well as a fullband signal (16khz). If there are too much treble, I will use the tone pot