Ibanez SD-9 Sonic Distortion Bypass Issue?

[sevenisthenumber]

I have an OLD SD9 that will not bypass... It just mutes when off and if you play super hard, it will have a very low volume gated tone of sorts. It works fine when on?? Any ideas?

[deadastronaut]

dodgy switch?....loose wire on switch....maybe..

[Kindly Killer]

I have the same problem with the same pedal. The switch in mine works fine; it's the bypass circuit, which I don't understand.

[anchovie]

It's audio probe time!

[Kindly Killer]

'kay I found this

how about some info on how the switching circuit works? or what do i search for other than "switching" to find info?

i.e. what am i looking for? i

[R.G.]

I found this, which looks accurate enough for our purposes here: http://www.aronnelson.com/gallery/main.php/v/Schematics-etc/sd_9.jpg.html

Things to note about bypassing in this pedal:
1. It does not bypass, not in the sense most people think. It immediately buffers the input jack (first transistor from input jack, upper left in the pic).
2. There is a SPDT switch implemented as two JFETs (amber background circles).
3. The switch works by turning on one or the other but not both JFETs at a time.
4. The gate signal to turn on JFETs is generated by the two transistors drawn with cross-coupled collectors and bases. This makes a digital flipflop, which has either the transistor on one side on (i.e. collector pulled to ground) or the other, but not both at the same time. Pressing the momentary "bypass" switch causes the flipflop to flip.
5. The two JFETs are N-channel, so they're turned OFF by their gate being pulled below their source by enough voltage. In this case "enough" is half the power supply.
6. The two flipflop collectors are connected to the JFET gates by 1M resistors and diodes. There is a 47nF cap to ground from the junction of the 1M and diode to slow the transition down a bit.

That's really all you need to know to debug this. If it were me, I'd do the following:
1. Test the power supplies. Always, always, always check to see that the board is getting good power; if you don't, nothing else you "measure" is correct. So check for "9V" being 8.0-10.0V. And then measure Vr for being half that. If these are not true figure out why not. Nothing works right if it's power supply is wrong. (Note that this applies to **everything**: plants, animals, humans in particular, stars, etc.)
2. Think. Does the flipflop work? Probably. You said that "It just mutes when off", which kind of implies that the flipflop is flipping and the LED is going on and off. If the flipflop itself was messed up, that would not be true. Therefore, you can concentrate on the JFETs and the controls to them.
3. Test the voltages on the flipflop collectors. Do these go a major fraction of the power supply high (above Vr at least) and then down to very close to ground when you flip the flop? This is a kind of warmup, as it would not be flipping if this was not true. But it gets you up close and personal with the measurements.
4. When the collectors are verified as going above Vr and nearly to ground, start looking at the JFETs. Both drain and source of each JFET is pulled to Vr by a resistor or an opamp output. Probe both JFETs source and drain (... oops! About now you're wishing you had looked up the JFET datasheet on line and found out which pin is drain, gate, and source, yes?  ) and verify that both of these pins of both JFETs are sitting with in a few millivolts of Vr. If this is not true, it's likely to be your problem.
5. Now probe the voltage at the junction of the 1M from the flipflop and the diode and 0.047uF caps. Does this voltage go above Vr and down to ground?
I suspect that one of them does not. Failures can be the JFET, the 1M, the 0.047, the diode, the solder joints, a crack in the PCB copper, anything that keeps the right voltages from getting to the right place.

From your description (just mutes when off) is it very likely that the gate circuit has a problem for the lower JFET, the one connected to pin 7 of the opamp. If it's not the gating signal, and all is well there, start suspecting the opamp and its connections. For opamps acting as amplifiers, the + input, the (-) input, and output all need to be within a few millivolts of the same voltage (Vr in this case) or the opamp is not working linearly. This is true for 99+% of opamp circuits used in effects.

Go ye, measure and report.

[Kindly Killer]

I found this, which looks accurate enough for our purposes here: http://www.aronnelson.com/gallery/main.php/v/Schematics-etc/sd_9.jpg.html

Things to note about bypassing in this pedal:
1. It does not bypass, not in the sense most people think. It immediately buffers the input jack (first transistor from input jack, upper left in the pic).
2. There is a SPDT switch implemented as two JFETs (amber background circles).
3. The switch works by turning on one or the other but not both JFETs at a time.
4. The gate signal to turn on JFETs is generated by the two transistors drawn with cross-coupled collectors and bases. This makes a digital flipflop, which has either the transistor on one side on (i.e. collector pulled to ground) or the other, but not both at the same time. Pressing the momentary "bypass" switch causes the flipflop to flip.
5. The two JFETs are N-channel, so they're turned OFF by their gate being pulled below their source by enough voltage. In this case "enough" is half the power supply.
6. The two flipflop collectors are connected to the JFET gates by 1M resistors and diodes. There is a 47nF cap to ground from the junction of the 1M and diode to slow the transition down a bit.

That's really all you need to know to debug this. If it were me, I'd do the following:
1. Test the power supplies. Always, always, always check to see that the board is getting good power; if you don't, nothing else you "measure" is correct. So check for "9V" being 8.0-10.0V. And then measure Vr for being half that. If these are not true figure out why not. Nothing works right if it's power supply is wrong. (Note that this applies to **everything**: plants, animals, humans in particular, stars, etc.)
2. Think. Does the flipflop work? Probably. You said that "It just mutes when off", which kind of implies that the flipflop is flipping and the LED is going on and off. If the flipflop itself was messed up, that would not be true. Therefore, you can concentrate on the JFETs and the controls to them.
3. Test the voltages on the flipflop collectors. Do these go a major fraction of the power supply high (above Vr at least) and then down to very close to ground when you flip the flop? This is a kind of warmup, as it would not be flipping if this was not true. But it gets you up close and personal with the measurements.
4. When the collectors are verified as going above Vr and nearly to ground, start looking at the JFETs. Both drain and source of each JFET is pulled to Vr by a resistor or an opamp output. Probe both JFETs source and drain (... oops! About now you're wishing you had looked up the JFET datasheet on line and found out which pin is drain, gate, and source, yes?  ) and verify that both of these pins of both JFETs are sitting with in a few millivolts of Vr. If this is not true, it's likely to be your problem.
5. Now probe the voltage at the junction of the 1M from the flipflop and the diode and 0.047uF caps. Does this voltage go above Vr and down to ground?
I suspect that one of them does not. Failures can be the JFET, the 1M, the 0.047, the diode, the solder joints, a crack in the PCB copper, anything that keeps the right voltages from getting to the right place.

From your description (just mutes when off) is it very likely that the gate circuit has a problem for the lower JFET, the one connected to pin 7 of the opamp. If it's not the gating signal, and all is well there, start suspecting the opamp and its connections. For opamps acting as amplifiers, the + input, the (-) input, and output all need to be within a few millivolts of the same voltage (Vr in this case) or the opamp is not working linearly. This is true for 99+% of opamp circuits used in effects.

Go ye, measure and report.

WOW great information - thanks!

A "microburst" tore thru my neighborhood Sunday, which is my normal guitar tech/tinker day. My power just came back on. I will report back when I can carve out a little time to tinker with this. Thanks so much for info.

[sevenisthenumber]

Okay...  The pedal I had to repair originally, i fixed thanks to this post.. It turned out to be a lot of corrosion on the PCB that was doing it.
I have another, same issue..
I cant figure this one out...
Someone had tried to fix broken trace and I cant verify if they did it correctly... can anyone lead me by these pics? It is a distorted bypass signal and the led does turn on and off properly.. What transistors should I try? What replacement transistors will work for these if that is the case.

[Kindly Killer]

Hey R.G. I just wanted to thank you for your help with this. I never did pinpoint my problem, but two good things came as a result. 1) I became sufficiently frustrated that I started studying electronics so I can finally understand a simple dirt box, and 2) you provided enough information that I was able to get the pedal back on the road:



Anyway, thanks.

[chauncey0303]

I know this is an old post but Sevenisthnumber who put those pictures up Thank you... I now have a working Ibanez SD9   I suspected that the fix would look like this but It made it easy to look at a picture and do it right...