Boss OC-2/3 Clone getting faint static crackle and thinking the schematic is bad

[bushidov]

Hi All,

So, a while ago, I tried looking at schematics on the internet in regards to the Boss OC-2 to see what it would take to make a clone and learn from it. Due to some parts no longer being used, and a lot of differing schematics that left me confused, I got a working, used & beat-up Behringer UO300 and tried to reverse engineer it. It most used surface mount components, but that really wasn't much of a problem for me. The 0805 resistors were able to be read with a magnifying glass and I was able to use my fluke to verify resistance. I have a capacitance meter that goes down to the 1pF for resolution, so I was able to desolder each and every 0805 capacitor and measure those. I desoldered all the ICs and then traced out all the top and bottom layer traces. Minus the slide switch, which gave me 3 filtering options, I went with just two with a DPDT switch, but kept everything exact for my clone (well, I made everything through-hole instead of surface mount, too)

So, below is my schematic from my traces:


And for the most part, it actually does work well. The blend works like volume knob when both Octaves are at 0. Octave 1 and 2 both perform well and track well clean. The only problem is there is a semi-faint static sounding crackle. Kind of like the crackle coming off an active taser. It appears even if I have both octaves in the "off" / zero pot positions. Actually, even if I turn the blend off too, it still comes through. It only crackles when a note is plucked on my guitar or bass. If I let it ring out, it will stop before the notes fully comes to a rest, but does last for a while. If I mute the string, the crackle dies immediately too. It is kind of hard to hear when set to bed-room volumes, but turned up, it becomes more and more obvious.

Here is a video of it doing what I am talking about:
https://www.youtube.com/watch?v=qm45kYLHTKk

Upon probing it, it seems to be appearing around U1A on my schematic. It is really crackly around R25 where it touches pin 2 of the U1A op-amp. However at the center pins of all three pots, which by only 1 resistor a piece, connects to R25 and pin 2 of that op-amp, I can't seem to hear the crackle. I thought maybe my op-amp was bad. Swapped. No go. Tried a TL074 instead. No go. I think I also removed R19, R8, and R6, so nothing was going into pin 2 of U1A, but I still got crackle on pin 2, which made no sense to me what-so-ever.

So what am I doing wrong?

Did I draw my schematic incorrectly?

It wouldn't be the first time I missed one small detail, which in turn completely pooched the product.

Thanks for any and all help you throw at me!

[Rob Strand]

The unconnected J & K inputs on the 4027 look like a strong candidate for such a problem.

There was an original schematic around.     

The original doesn't actually use a 4027.  That's one of the main "mods" you see on the web.  The problem is a lot of those schematics claim to be OC-2 but they don't say it's modded.  Those same schematics often have inputs hanging on the 4027.   

IC6 on the original uses a small SIL package part, with a BAxxx number, instead of the 4027.  You can see it as IC6 in the pics on this page,

https://gottarocks.com/boss-oc2/

You can get the 4027 to work.  It's digital so there should be no impact on the performance.

[Slowpoke101]

Connect the J and K inputs to +9V (pin 16 of the 4027 would be suitable ).
Unused inputs on CMOS logic chips should not be left floating. This may not get rid of your noise problem but it's a good place to start. You may need to have a some more filtering capacitors (decoupling ) on the 4013 and 4027 flip-flops - they can be a bit noisy when changing state.

[Rob Strand]

FWIW, this is the original,


I also found this later version of the OC-2, which uses 2x4013's (bottom right),

https://mirosol.kapsi.fi/2014/08/



Completely forgot about that one.  You could trace-out the 4013 version from the PCB but the 4027 you have should work fine.

[bushidov]

The unconnected J & K inputs on the 4027 look like a strong candidate for such a problem.

Connect the J and K inputs to +9V (pin 16 of the 4027 would be suitable ).

You may need to have a some more filtering capacitors (decoupling ) on the 4013 and 4027 flip-flops - they can be a bit noisy when changing state.

Alright, so I did those following changes to my PCB. Schematic updated here:


It still has the exact same problem. Basically, with all three pots at 0, there really shouldn't be any noise. But I still hear a faint, clean signal and that crackle/static. Furthermore, with all 3 knobs at 0, if I take my output probe and bring it over to U1A, pin 2, I get nothing, as would be expected. Even if I crank my amp to 10. However, if I bring my probe to U1A, pin 1, the crackle appears, but none of the faint clean signal.

Thanks for the help, thus far.

[Slowpoke101]

Remove the following components:

R24
C16
C17

These components are part of the original electronic bypass switching and are not needed here.

[Eb7+9]

... notice your schematic omits the (non-negligible) D10/D11 diodes //

[bushidov]

Remove the following components:

R24
C16
C17

These components are part of the original electronic bypass switching and are not needed here.

So, I did so, update to the schematic here:


It still is doing it the static fuzz. But, I think Slowpoke101 might be on to something. I might have not gotten off all the "switching bypass stuff off the circuit" from the Behringer UO-300, which uses the same circuit as the BOSS OC-2. After removing those components, it also still works as before, meaning they didn't need to be there. Is there anything else that I need to remove?

notice your schematic omits the (non-negligible) D10/D11 diodes

. I have the UO300 here too. It doesn't use them either and has no problem with it.

Here is the UO300 schematic that I traced a while ago:

[willienillie]

I also found this later version of the OC-2...

I find it a bit odd that the PCB has the Matsushita logo screened on it.  Did they make pedals for Boss?  Or PCBs?  I would have assumed Roland was fully capable of doing everything in-house, maybe that changed for the Taiwan-made stuff.

[Rob Strand]

What I suspect is going on is oscillations and I suspect it could be occurring in the rectifiers U2C, U2D.   If that's true then perhaps adding some small caps, 10pF to 47pF region, across diodes D1 and D3.

If it's not that, it's possible another part of the circuit is oscillating.

Another difference came to mind.  The original OC-2 uses a divider at the input of of the JFET gates, parts R47, R48, R50, R51.   Also note the "vcc/2" divider voltage has been tweaked.   I believe this all done deliberately to prevent the JFET gates forward biasing and injecting ticks onto the vcc/2 rail.   *However*  at this point I don't think that's your problem as you have a buffer bias.   If the buffer impedance isn't low enough you might get problems.  You can try adding a cap from the output of the bias opamp to ground but adding caps to the output of opamps can cause oscillation.   Getting around that is messy (RG and I have discussed this a few times on the Forum). 

You could also try adding a cap 100nF to 100uF (have a play with the value) across the power rails of U2.    Again I don't suspect the problem is there but it's something to try.

Basically, with all three pots at 0, there really shouldn't be any noise. But I still hear a faint, clean signal and that crackle/static. Furthermore, with all 3 knobs at 0, if I take my output probe and bring it over to U1A, pin 2, I get nothing, as would be expected. Even if I crank my amp to 10. However, if I bring my probe to U1A, pin 1, the crackle appears, but none of the faint clean signal.

That would point to crap getting onto the vcc rail or vcc/2 rail, or oscillations (which can get into other parts of the circuit).

R25 where it touches pin 2 of the U1A op-amp

It's weird it is crackly there when all the pots are off.    If you want to try something try adding 47pF to 100pF across R25.

I find it a bit odd that the PCB has the Matsushita logo screened on it.  Did they make pedals for Boss?  Or PCBs?  I would have assumed Roland was fully capable of doing everything in-house, maybe that changed for the Taiwan-made stuff.

Boss is  Japanese so it's not unreasonable.   I can't say I've noticed that before.  The other markings don't look un-boss-like.

[Rob Strand]

FWIW, it's a little odd the 4013 has it's own filtered supply but the 4027 doesn't use that supply.

[bushidov]

If that's true then perhaps adding some small caps, 10pF to 47pF region, across diodes D1 and D3

I tried this. Still crackling with all the knobs off and volume up

You could also try adding a cap 100nF to 100uF (have a play with the value) across the power rails of U2.

I tried this. Still crackling with all the knobs off and volume up

If you want to try something try adding 47pF to 100pF across R25.

Tried this as well. Still no go.

The crackling does kind of remind me of a splatty fuzz effect.

I am just confused as to what the differences are. It might be something that was inherent with the Boss-like switching circuit that I need to still remove part of to make work as Slowpoke pointed out. I pulled the parts he mentioned, but are there more, and maybe that's what I am goofing on?

[Rob Strand]

I tried this. Still crackling with all the knobs off and volume up

The crackling does kind of remind me of a splatty fuzz effect.

I am just confused as to what the differences are. It might be something that was inherent with the Boss-like switching circuit that I need to still remove part of to make work as Slowpoke pointed out. I pulled the parts he mentioned, but are there more, and maybe that's what I am goofing on?

Well, that pretty much writes off any of the short-cut solutions.   If nothing weird is going on, the parts Slowpoke mentioned shouldn't affect things either.

There are quite a few difference between your circuit and the original OC-2.  *However*, it might not be a matter of the differences but more a matter that there's a problem.   This task is to isolate that problem.

So one sledge-hammer test is to see if the problem is caused by the Rectifier / Digital Divider circuit or it is something to do with the analog stuff.

Try lifting the input side of R2 and R5 and wire them back to VBIAS. That should disable that entire part of the circuit.

If that *did* fix things then,
- re connect R2 and R5
- lift C22
otherwise stop  (Unlikely to be caused by the rectifier/dividers.)

If that *does* fix things then,
- reconnect C22
- lift R37 and R41
otherwise stop

If that *does* fix things then,
- reconnect R37 and R41
- lift the 4013/4027 side (not the JFET side) of R20 and R22
- Just let R20 and R22 float
otherwise stop

If that *doesn't* fix things,
- try wiring the hanging ends of R20 and R22 to +9V
otherwise stop

If you hit a 'stop' just come back.

The whole point of the above is to pin-point what part of the rectifier/divider circuit is causing the problem.

[Rob Strand]

BTW, a couple of major difference are:
- R11 330k not 27k.     This high value could cause clipping and make it splattery.
- The circuit around Q2 is completely different.

But the fact you get the problem with all pots on zero means it's a bigger problem.

[bushidov]

Try lifting the input side of R2 and R5 and wire them back to VBIAS. That should disable that entire part of the circuit.

This did remove the crackle with all knobs at zero but amplified with amp.

reconnect R2 and R5
- lift C22

It kept quiet, so far so good

If that *does* fix things then,
- reconnect C22
- lift R37 and R41
otherwise stop

It kept quiet, so far so good

- reconnect R37 and R41
- lift the 4013/4027 side (not the JFET side) of R20 and R22
- Just let R20 and R22 float

The crackle returned with all knobs at zero

If that *doesn't* fix things,
- try wiring the hanging ends of R20 and R22 to +9V
otherwise stop

Crackle still remains.

I am seeing that you are slowly isolating it down the circuit. It is looking like something wrong in the wiring of CD4013BE and CD4027BE?

[Rob Strand]

I am seeing that you are slowly isolating it down the circuit. It is looking like something wrong in the wiring of CD4013BE and CD4027BE?

Yes, it's looking like a problem in that area.    It could also be the outputs of U2A and U2B switching off and on in bursts when they change state.

Are you sure you have the vdd and ground pins of both the 4013 and the 4027 connected to the power and ground?

Perhaps the next step is to check the voltage on the power rail for the 4013.   If there's a problem with the 4013 it could be pulling too much current and causing a voltage drop across R46 (1k).   If that happens the 4013 will try to power-up by drawing current  from R37/R41 through pins 4 and 6 of the 4013.  The inputs of the 4013 have diodes to vdd (and ground) making it possible for power to flow through the input pins when the actual power rails are low.

You could also check for any unintended shorts on the 4013 and 4027 output pins (check all pins even the unused ones).

[bushidov]

Are you sure you have the vdd and ground pins of both the 4013 and the 4027 connected to the power and ground?

Verified

Perhaps the next step is to check the voltage on the power rail for the 4013.

Voltage Supply reads 8.93V
Pin 7 of the CD4013BE reads 8.90V

You could also check for any unintended shorts on the 4013 and 4027 output pins (check all pins even the unused ones).

No unintended shorts, I believe. All the grounding is based off the schematic, which I have verified it matches.

Looking at other schematics though, I did notice something odd. From the BOSS OC-2, to these schematics for the "Chopped OC-2", I noticed that the section before the CD4013 and CD4027 don't have any "biasing". Yet on mine, which I thought I copied off a working Behringer UO300, does via R33, R35, and R39, not to mention the C23 cap going to bias. None of these appear on the BOSS OC-2 or Chopped. Are these my problem?

Also, I know that the BOSS had those 2 Germanium Diodes (D10 and D11). I am certain they are not on the Behringer, but I am also not sure what those were for. Is this also contributing to my problem?

[Rob Strand]

Verified

No unintended shorts, I believe. All the grounding is based off the schematic, which I have verified it matches.

OK cool.  It pays to check stuff like that.

It does mean the problem is going to get a bit trickier to debug.

Looking at other schematics though, I did notice something odd. From the BOSS OC-2, to these schematics for the "Chopped OC-2", I noticed that the section before the CD4013 and CD4027 don't have any "biasing". Yet on mine, which I thought I copied off a working Behringer UO300, does via R33, R35, and R39, not to mention the C23 cap going to bias. None of these appear on the BOSS OC-2 or Chopped. Are these my problem?

Hard to know if they would contribute.   I always work on the idea that a change is always a potential for trouble - not unlike the "quick" fix in software, it always creates more bugs than it fixes.   It wouldn't hurt to try pulling those bias resistors.

Also, I know that the BOSS had those 2 Germanium Diodes (D10 and D11). I am certain they are not on the Behringer, but I am also not sure what those were for. Is this also contributing to my problem?

These won't cause the problem.  However, I do know they contribute to the tone of the octave signals.  That's what makes it sound different to other pedals.

Your schematic has a lot of differences to the original.   Some will affect tone.   On the whole I don't think many of the differences should cause the problems you are seeing.   The problem you have is more of a fundamental problem.

A couple of things that look like trouble to me are:

The 4027 is not powered from same power rail as the 4013.    Unfortunately with a PCB it's not a simple matter to wire all the 4027 signals back to the 4013 supply.    So it's a lot of work for a 'test'.   I suppose you could check the layout.   

The second is the way the VBIAS circuit is set-up.     Circuits which use two caps C25 and C26 can have noise problems.  In your case the addition of C26.   With only C25 present the circuit looks like a normal Vcc/2 circuit.   In this case C25 filters junk off the supply.  When you add C26 it puts junk back onto VBIAS and that affects the *whole circuit*.  It's probably worth while pulling C26.    With large cap values of 220uF and large bias resistor values pulling C26 means the circuit is going to take a long time to start-up so you might want to pull C26 and replace C25 with 10uF.

It could be both of the above are contributing to the problem but changing only one might fix it!

[Rob Strand]

Oh, one more thing about the C25/C26 mod.     Wire one 220uF across +9V and 0V.

[Eb7+9]

can you look with a scope and see if you've got spikes at the output of U1B and U1D ?