Boss SD1 bleed through fix mod

[rosscocean]

I've been playing around with modding my SD1 for a while, but I'm just starting to look into the bleed through fix mod.

I know the mod can be done using a J201, cap and a diode, and that would sort it out but I came across something on Analogman's site where he mentions a much easier method, involving like parts etc. Does anyone know what it is?

Here's a link to the Analogman page http://www.analogman.com/boss.htm

Cheers,
Ross

[rosscocean]

Anyone?

[MrStab]

after having stared at the circuit for about 10 minutes trying to think of various ways to ground the gain stage or repurpose Q1 & Q2, the best i can suggest is cutting the leg of R14 on the side it meets C8, connecting a 1-10uF cap on the then-free end of R14, and wiring the negative terminal of that straight to the output jack. possibly with a shielded wire. the input/clean signal is already buffered by that point and maybe you'd be avoiding crosstalk with a high-gain area that way.

i can't guarantee that'd work, but if Q5 seems pretty clean with an audio probe, it's a fair bet. you may wanna put more thought into things like output impedance and pulldown resistor than i have, though. bottom line is, though, it's essentially just moving the same out-put switching operation to a different part of the circuit.

(i'm going by this schem, btw: http://www.hobby-hour.com/electronics/s/schematics/boss-sd1-super-overdrive.gif )

[MrStab]

i have to say, even though it may work, this solution is not ideal. it leaves inputs hanging and i think there are better solutions involving more parts. nonetheless, here's what i was on about:



or some variation thereof. maybe someone else can see issues with the polarity of the cap i added (in relation to Q2 in off state), and maybe any potential problems caused by the Q6 buffer.

emphasis is on avoiding the clean Q5 signal not picking up signal from the area round IC1, so would be pointless if you did all this only to let a wire fall on that region. it's possible you could maybe even join the signals before the output buffer as intended, depending on how long and where the problematic traces are.

[rosscocean]

Thanks for your reply, I'll have a look at that.

[MrStab]

no problem, Ross.

can anyone pitch in before Ross tries it? my gut & tiny brain says it'd work, but it's all theoretical and a second opinion could help

[Groovenut]

I still have a sneaky suspicion this bleed issue is one cause by layout. The tracks of the R11/R14 nodes are too close together and when the gain of the circuit is increased you get bleed. Strange that all of them dont do it though.

[MrStab]

the gist of what i'm proposing is to basically cut that trace out and reroute further away. just not sure whether avoiding the output buffer is overkill or not.

[wampcat1]

Here's a super easy mod to fix it. http://brianwampler.com/downloads/boss-sd1-bleed-fix.pdf

[VanUsen]

Hi Everyone,

I'm new to the forum but have been lurking for a while. I've recently been playing around modding my SD-1 with a Monte Allums kit and have changed a few things here and there with the info that I learned. In regards to bleed fix, I have an Analogman SD-1 and I think this is what he does:

There is a resistor with one leg connected to one of the lugs on the input jack and the other leg (via a wire) is connected to R1 (R1 has been completely removed). I don't know the value of the resistor but it's colors are brown, green, red and yellow (I think! It's hard to make out in its location). I hope this helps!

[Silvio55]

Hi Everyone,

I'm new to the forum but have been lurking for a while. I've recently been playing around modding my SD-1 with a Monte Allums kit and have changed a few things here and there with the info that I learned. In regards to bleed fix, I have an Analogman SD-1 and I think this is what he does:

There is a resistor with one leg connected to one of the lugs on the input jack and the other leg (via a wire) is connected to R1 (R1 has been completely removed). I don't know the value of the resistor but it's colors are brown, green, red and yellow (I think! It's hard to make out in its location). I hope this helps!

So the mod consists on removing the input resistor R1 (10K) from the board and replace it with a 1K5 resistor soldered directly to the jack?
I can imagine this solves a problem with the layout that leads to the bleeding but wonder why changing the value, or what effect could it have.
Thanks!

[Rob Strand]

I still have a sneaky suspicion this bleed issue is one cause by layout. The tracks of the R11/R14 nodes are too close together and when the gain of the circuit is increased you get bleed. Strange that all of them dont do it though.

Quite possible it's adding to the problem.

It's a while since I've looked at this issue but I seemed to recall it was related to the switch Q1 not turning off.
I suspect replacing R14 with 0R or 100R would help a great deal.

If you look at the OD-1 it doesn't have this problem and the only difference is the R11/R14 thing.

On the other hand if you look at higher gain effects like the OD-2 they have  an extra switch at the input to completely stop any buzz propagating throughout the circuit.
The wampler link achieves a similar goal but I if was designing from scratch I'd go with the OD-2 structure.
It is possible to do the OD-2 mod on the SD-1 but you need to add a couple more parts compared to the wampler.

[Groovenut]

I still have a sneaky suspicion this bleed issue is one cause by layout. The tracks of the R11/R14 nodes are too close together and when the gain of the circuit is increased you get bleed. Strange that all of them dont do it though.

Quite possible it's adding to the problem.

It's a while since I've looked at this issue but I seemed to recall it was related to the switch Q1 not turning off.
I suspect replacing R14 with 0R or 100R would help a great deal.

If you look at the OD-1 it doesn't have this problem and the only difference is the R11/R14 thing.

On the other hand if you look at higher gain effects like the OD-2 they have  an extra switch at the input to completely stop any buzz propagating throughout the circuit.
The wampler link achieves a similar goal but I if was designing from scratch I'd go with the OD-2 structure.
It is possible to do the OD-2 mod on the SD-1 but you need to add a couple more parts compared to the wampler.

Yes, however the OD-1 dual opamp version (as well as the quad version) is a completely different layout than the SD-1 even though it has the same switching. This is part of what leads me to believe it's a layout issue and not merely a FET switch issue.

The easiest fix of course is the input cutoff fet, but it's not a very elegant fix for the existing SD-1s with this issue.

I am curious as to why the 1k5 on the input jack would work to fix this at all. But that remains to be seen I suppose.

[Rob Strand]

This is part of what leads me to believe it's a layout issue and not merely a FET switch issue.

Perhaps lifting the C8 end of R11 would help isolate this to a layout issue.

I am curious as to why the 1k5 on the input jack would work to fix this at all. But that remains to be seen I suppose.

Same here.   The only way I could see this doing anything is if the input impedance of the first opamp (pins 4,5,6) was going low impedance due to clipping - which I doubt is the case.   If this was the mechanism then the wampler mod wouldn't work because is has no affect on that opamp's input impedance or the degree of clipping.
If the wamper mod works then it points to the layout and/or Q1 switching issue.

[Silvio55]

VanUsen mentioned that he owns an Analogman modded SD-1, and analogman mention in his website that there's a simple fix for the bleeding that consists in "moving just one component and one wire", it seems to be it, maybe the value of the resistor is not changed, or it doesn't matter.
There's another thread in this forum that describes the same mod, moving R1 to the jack, the guy says it worked fine:
http://www.diystompboxes.com/smfforum/index.php?topic=101879.0
But there's seems to be no confirmation if this is indeed the simplier mod or why does it work!

[Rob Strand]

What if it something simple like keep the Drive, Tone and Level pot wires away from the input jack wire?
I don't have a unit to check the routing of the wires.

Maybe when people mess with R1 they are just moving the wires!

Edit:
Refer to the original schematic, on the PCB layout  the input terminal PCB terminal 3 is right next to the output of the overdrive stage PCB terminal 4!    So it's possibly a layout issue.  The R1 mod will help here, however keeping the wires on PCB terminals 3 and 4 away from each other will probably help a great deal.    Taking R1 off the board and not using PCB terminal 4 is better than replacing R1 with a 0R as it avoids coupling on the PCB tracks.  (Also the other side of the drive pot is on PCB terminal pin 2).

So now I'm puzzled how the wampler mod  helps the problem.

[Groovenut]

What if it something simple like keep the Drive, Tone and Level pot wires away from the input jack wire?
I don't have a unit to check the routing of the wires.

Maybe when people mess with R1 they are just moving the wires!

Edit:
Refer to the original schematic, on the PCB layout  the input terminal PCB terminal 3 is right next to the output of the overdrive stage PCB terminal 4!    So it's possibly a layout issue.  The R1 mod will help here, however keeping the wires on PCB terminals 3 and 4 away from each other will probably help a great deal.    Taking R1 off the board and not using PCB terminal 4 is better than replacing R1 with a 0R as it avoids coupling on the PCB tracks.  (Also the other side of the drive pot is on PCB terminal pin 2).

So now I'm puzzled how the wampler mod  helps the problem.

I suppose it's always possible there is capacative coupling through the wiring of the input jack and the gain or tone pots. That makes the most sense if there is bleed through even when Q1 is removed.

The Wampler mod appears to help this by reducing the clipping opamps gain to 1 when switched to bypass mode by inserting a very low resistance (~10-100R ish) between pins 6 & 7 via the extra fet (according to the posted link any way, thats how I read it).

[monkeyxx]

I just moved the 10K R1 resistor to the input jack, and moved the brown wire.  In fact, it did not work.  I am still hearing some bleed through at high gain settings, on the clean bypass sound.

I guess I'm going to have to go for the Wampler mod since that's the only well documented one I can find to date.  I would be interested if some of the new mods could be figured out.

[Rob Strand]

The Wampler mod appears to help this by reducing the clipping opamps gain to 1 when switched to bypass mode by inserting a very low resistance (~10-100R ish) between pins 6 & 7 via the extra fet (according to the posted link any way, thats how I read it).

That's how I saw it too.

I haven't look at this issue for many years.    One school of thought is some sort of capacitive coupling the other is leakage through the JFET Q1, the one that disconnects the effect signal.

I haven't seen anyone comment on whether the bleed through is dependent on the LEVEL (volume) control position.  If the bleed only occurs with the LEVEL with the level below the 25% position it may be capacitive coupling.  If it only occurs with the LEVEL on a high setting then that would point to the JFET switch turning on due to the signal level.

The think that puzzles me is you never hear people complaining about TS9's doing this - even ones with diode mods.  For the modded case the direction of the two series clipping diode may have an effect.

[Groovenut]

The Wampler mod appears to help this by reducing the clipping opamps gain to 1 when switched to bypass mode by inserting a very low resistance (~10-100R ish) between pins 6 & 7 via the extra fet (according to the posted link any way, thats how I read it).

The think that puzzles me is you never hear people complaining about TS9's doing this - even ones with diode mods.  For the modded case the direction of the two series clipping diode may have an effect.

It's for this very reason that I tend to believe it's a pcb layout issue. I believe (but have not yet confirmed) that there must be two traces that are too close to one another and at higher signal levels coupling happens

my 2 cents