Troubleshooting Boss switching circuit

[disorder]

I have a Boss pedal that will not switch between bypass/effect. I have eliminated the switch as an issue and I have also tried replacing both 2SC945 transistors with 2N3904 with no luck. The pedal starts out in the "on" state with the LED lit and I can't get it to switch even if I directly ground either transistors base or directly connect the east end of R55. Q9 has 0V at collector, 0.67V at base, and emitter is correctly at 0V. This base voltage looks correct to me as it is the base-emitter drop meaning it is on. Q8 has 6V at collector, 0V at base, and 0V at emitter. So far everything seems normal to me.

My next step is to replace C21 and C18. All relevant resistors measure correctly according to the schematic below. I am starting to wonder if one or a number of the JFET switches are bad and somehow keeping the BJTs from flip flopping.

http://www.hobby-hour.com/electronics/s/schematics/boss-hm2-heavy-metal-schematic.png

[R.G.]

Give a read here:
http://geofex.com/Article_Folders/bosstech.pdf
for a view of how they work.

If you're that deep into this, open the resistors that lead away from the collectors of the flipflop to isolate the flipflop from possible external shorts.

If the flipflop works without the JFET loads, then connect the 1M coupling resistors one at a time and find out what stops it working.

It makes a whole lot of sense to do some probing of the base, emitter, and collector voltages to see what the transistors are doing. A shorted cap would give you heartburn too.

[PRR]

> wonder if one or a number of the JFET switches are bad and somehow keeping the BJTs from flip flopping.

The JFETs are 1Meg away from the flipflop. Hard to see how a fault on the far end of 1meg would really upset a 56K flop.

C21 and C18 are VERY unlikely to fail. However your path so far hasn't found any better suspects. Perhaps someone has been shooting big zaps of static electricity at the switch (which might also kill C19). I'd lift the Base end of C21 and C18, then do your thing of shorting one base to ground to flip the flop.

Next lift the Base ends of C17 C20. If these have gone short (grasping at straws here) the flop will stay stuck. (However I'd expect a collector sitting at 0.6V, not zero or 6V.)

[disorder]

Lifting the 1M resistors leading to JFET switchers did not change anything. I am still getting the LED turned on when I apply power, and now all three terminals on each switching BJT reads 0V. 

[R.G.]

Lifting the 1M resistors leading to JFET switchers did not change anything. I am still getting the LED turned on when I apply power, and now all three terminals on each switching BJT reads 0V. 

0.00 V or just "way less than 1V but bigger than 0V"? Can your meter tell the difference?

If it's really 0.000V, then there are gross problems, like collectors shorted to the emitters, either by transistor failure (unlikely) or by solder shorts (more likely).

[disorder]

I spoke too soon. Here are my voltages I am currently dealing with... I have swapped known good transistors in and the problem persists. I compared voltages with a DS-1 that switches correctly and what I find strange is the R32/R55/C19 node that connects to C21/C18. On the DS-1 this node measured as the 9V supply. On this broken HM-2 it starts at around 7V and creeps up to about 8V and doesn't go higher even though the supply is 9.4V. Not sure if its slowly increasing with time or increasing because I'm probing there with my meter. I should note the momentary switch is GOOD (brand new) so it's not that.

[Fender3D]

When you swapped 2SCs with "known good transistors" did you check the pinout?
2SCxxx are ECB, while 2Nxxxx are usually EBC...

[disorder]

Yes confirmed pinouts matched with my Atlas tester.

[amptramp]

The R32/R55/C19 junction should be actually at battery voltage unless the switch is actuated, but the input impedance of the meter may be what is dragging this voltage down to 7.4 volts.  Temporarily disconnect C19 and see if the voltage changes.  If it does, replace C19.  When the switch is actuated, this point should go down to 0.9 millivolts, which you may not be able to measure.  A new switch is not necessarily a good switch and both connections to the switch are always suspect.

Your measurement that is 6.25 for the Q8 collector should nominally be 6.62, but the measured value is not out of line.

Just out of curiosity, disconnect the LED or the series diode or the resistor.  If it operates normally after that, then you have another data point.

Check that C23 and C24 are not shorted.  It will not affect the flip-flop but it will affect the signal path.

[R.G.]

I spoke too soon. Here are my voltages I am currently dealing with...

Open R40 by raising one leg. You should see the base of Q9 go to zero, its collector go high, and the base of Q8 go to 0.6V and the collector of Q8 go low.

If that does not happen, and particularly if the base of Q8 stays low, then there is a base-emitter short on Q8, or the path from the collector of Q9 through R39 to the base of Q8 is open.