My Neovibe won't vibe... please help!

[DeeBug]

Just recently completed it.  The bulb is on but doesn't pulse.  The bulb is easily adjustable with the trim pot, from virtually completely off to fairly bright.  The volume knob works, but the speed and depth knobs have no effect on the bulb or sound.  Also, there's a bit of unpleasant static when I strum the guitar with the pedal engaged.  The sound quality changes a little between "vibrato" and "chorus", but that's about it, and the static is there in both modes.

It's the GGG Neovibe.  All parts from Mouser and Small Bear.  Transistors are all 2N5088 except for the lamp driver which is 2N3904.  No modifications or part substitutions.

It's powered up with 18v from a Dunlop Brick power supply - negative ground.

Here is a report on all the voltages.

Q1
C = 1.79V
B = 1.31V
E = 0.91V

Q2
C= 5.08V
B= 1.78V
E= 1.41V

Q3
C= 10.25V
B= 5.07V
E= 4.48V

Q4
C= 14.89V
B= 4.66V
E= 4.51V

Q5
C= 11.25V
B= 4.52V
E= 3.71V

Q6
C= 14.89V
B= 4.69V
E= 4.54V

Q7
C= 11.13V
B= 4.53V
E= 3.76V

Q8
C= 14.89V
B= 4.67V
E= 4.51V

Q9
C= 11.20V
B= 4.53V
E= 3.75V

Q10
C= 14.89V
B= 5.92V
E= 5.41V

Q11
C= 17.45V
B= 8.37V
E= 9.70V

Q12
C= 17.45V
B= 9.70V
E= 9.09V

Q13
C= 14.53V
B= 2.83V
E= 2.18V

L7815ACV
L= 14.89V
M= 0.6mV (middle pin, connected to ground)
R= 17.45V (the pin connected to the 1000uf caps)


DB102 (Diode Bridge)
P1= 17.45V (+pin)
P2= 0.5mV (–pin;  connected to ground)
P3= -5.0mV (opposite the –pin)
P4= 18.18V (opposite the +pin)

D1
A (anode, the non-band end) = 9.03V
K (cathode, the banded end) = 9.00V

D2
A = 9.00V
K = 9.03V


Does anything appear amiss here?  Any idea why the bulb doesn't pulse?

[R.G.]

Q11
C= 17.45v
B= 8.37v
E= 9.70v

Q12
C= 17.45v
B= 9.70v
E= 9.09v

Your problem is in there.

Q11's base should be higher than the emitter by about 0.6-.7V, and it's actually lower than the emitter. That means that Q11 can't be amplifying, and so it does not let the whole circuit around it oscillate. Ignoring the two collectors, which are properly at the power supply voltage, the pin voltages of Q11/Q12 should be about as shown in the voltages on the schematic page at GGG.

The problem is in the parts surrounding Q11/Q12.

- Check the transistor pinout. Yes, again.
- Check all the part values, especially R39/40/41.
- Look for soldering problems with a magnifying lens.

Something is wrong in that one part. Tell us what you find.

[DeeBug]

R.G.- thanks for your reply!
And thanks for helping me narrow down where the problem is.  Also thanks to the two others who responded but had their posts deleted due to the site problems.

As you suggested, I once again checked the transistor pinout.  It's oriented in the same way as all the other transistors, and the pins are all in the right holes.  I also swapped out the transistor for another of the same kind, no difference.

The part values checkout.  I measured them each before soldering them in, and I've rechecked them visually by verifying the color codes:
R39= 3.3K
R40= 2.2M
R41, R42, R43, R46 = 4.7K
R44, R45= 220K
R47= 100K
These are the same values given at GGG.

One thing I left out earlier is that I did in fact make one substitution, which is that the lamp is actually 12V/40mA, as opposed to the 25mA that is called for.

There doesn't seem to be any problem with the soldering.

Further studying the area, I've noticed that C19 and C20 read higher on the minus side than the plus side:
C19= 8.38V (plus side)/9.02V (minus side) 
C20= 9.02V (plus side)/9.05V (minus side)

So, do you think it could it be the 40mA lamp that's causing this?  If not, might tweaking any of the resistor values help rectify the votage values?

[R.G.]

You know what? I think I told you wrong. I'm sorry I missed that.

I think that the DC voltages are probably OK, but the voltage at the base of Q11 is being loaded down by your voltmeter in measuring.

If you have V+ = 17.45V and R39= 3.3k, R4=4.7K, then the junction of the two should be
V = 17.45*(4.7/(4.7+3.3))= 10.25V.
The base of Q11 will only pull a few fractions of a uA through the 2.2M resistor, so it will be very near 10.25. If the base emitters only need 0.5V (which would be normal for the low currents involved) then the emitter of Q11, the base of Q12, and the emitter of Q12 would be very much like they show in your measurements.

Which puts us back to "what the devil is going on?"

Given that the DC voltages are close to correct, it is possible that there could be a problem with C19/C20/C21, although I'd first guess that the same thing with the transistors is getting you with C19 - your meter is loading down that very high impedance base. Next best guess - the pots might be wired wrong. Try subbing in two equal valued fixed resistors of 22K to 51K for the two speed pot sections, and remove connection G to eliminate possible depth pot problems. It might just oscillate. You would detect this by putting your voltmeter on the emitter of Q12, and seeing a varying voltage, instead of DC.

To answer your question, no, the different bulb type would not have any effect on the oscillator oscillating or not.

If that's not it, we'll dig deeper.

[DeeBug]

OK, I tried subbing two 47K resistors for the two speed pot sections, although I'm not sure I did it right (see pic).  I also removed connection G to eliminate possible depth pot problems.  There was no change, no detectable oscillation or voltage fluctuation at the emitter of Q12.  Here's a picture showing how I put the resistors in:
http://i25.photobucket.com/albums/c92/mcbn1234/Two47K.jpg

Did I substitute those two fixed resistors for the two speed pot sections correctly?  If so, what would be your next suggestion?

[DeeBug]

bttt

[R.G.]

Sorry - for some reason I couldn't post for a while.

I see one problem. The wiring diagram for the pot is wrong in the Neovibe documentation. The common terminal on the pot should be C, not E.

And your resistors don't match either one if I am looking at them correctly. The two-legs common end should be in "C", the other resistors ends in D and E.

Try that, and yell back.

Sorry for the doco bug, folks.

[DeeBug]

Thanks again, R.G.!  Woohoo!  One of the problems solved.  I'm learning a lot here as I go.

Turns out it was, as you noted, the wiring of the speed pot.  I'm not 100% schematic literate yet, so I'd followed the diagram with the different wiring.

Now, that it's oscillating as it should, it's just a matter of getting rid of the static/crackle problem.  The sound is perfectly clean with the effect off.  With the effect on, however, I get static-y crackle when I strum the guitar strings.

Except for that unwanted distortion, this is a beautiful sounding effect.  So, where do I go from here?

[R.G.]

With the effect on, however, I get static-y crackle when I strum the guitar strings.

Can you describe the crackle more? For instance, does it only happen with light picking, strong picking, bass strings, etc.?

Is it a single snap or does it persist as long as the note sounds?

[DeeBug]

It happens with medium to heavier picking.  Bass strings and chords.  Not a single snap.  It persists with the note/chord.  It happens at all volume levels.  Especially pronounced when I roll off the treble on the guitar, but it's always there regardless.

[R.G.]

That's a characteristic of a problem with a DC blocking capacitor. Heavy bass notes have a large amplitude, as do chords. Sometimes that's enough to break over a part that's kind of on the edge. Examine your coupling capacitors carefully for correct orientation first, or soldering problems around them.

[DeeBug]

Which caps are the coupling caps?

[DeeBug]

Examine your coupling capacitors carefully for correct orientation first

All of the electrolytics are oriented in the same direction.  Here's a pic:

http://i25.photobucket.com/albums/c92/mcbn1234/Boardbeforelamp.jpg

or soldering problems around them.

I've gone over the board several times, and I've scraped between solder pads to ensure they weren't being bridged by flux residue.  Here's a pic of the solder side of the fully wired and completed board:

http://i25.photobucket.com/albums/c92/mcbn1234/Soldersideofboard.jpg

That's a characteristic of a problem with a DC blocking capacitor. Heavy bass notes have a large amplitude, as do chords. Sometimes that's enough to break over a part that's kind of on the edge.

Would it be possible to 'beef up' the values on the DC blocking capacitor or something to stop this "breaking over the edge" distortion without otherwise affecting the effect itself?

[R.G.]

Would it be possible to 'beef up' the values on the DC blocking capacitor or something to stop this "breaking over the edge" distortion without otherwise affecting the effect itself?

No, the values specified are plenty for the necessary blocking.

I think that from here on you'll need to use an audio probe to figure out exactly which stage is doing the breakover. Probe the collector and emitter of Q3, 5, 7, and 9, and the emitter of Q10, listening for the first stage where you hear the breakover. You will have to use a capacitor-isolated probe.

[DeeBug]

OK, I'll build one and see what I can find out.  Thanks!!

[DeeBug]

Hi R.G., I finally got round to building an audio probe and doing what you suggested last April.

To recap, my Neovibe does vibe now (it was the perennial "speed pot wiring" problem), except that there's still a "breaking over the edge" distortion problem.  I can hear the effect real well.  It would sound fantastic except for that static-y sounding breakover (by the way, there's no problem in bypass mode).  Problem is worse with a LP than a Strat, though still bad with my single coil Strat.

I think that from here on you'll need to use an audio probe to figure out exactly which stage is doing the breakover. Probe the collector and emitter of Q3, 5, 7, and 9, and the emitter of Q10, listening for the first stage where you hear the breakover. You will have to use a capacitor-isolated probe.

Here are the results of my probing with the audio probe.  These results were obtained by strumming hard on the open D and G strings of a Strat through a SF Princeton amp.  When I tried the same thing with an LP, the results were far more pronounced.  I wanted to make sure none of the breakup was caused by the LP, so I used only the Strat results.

Q3 - collector: substantial static-y breakover with some fuzziness to it
Q3 - emittor: slightly less loud than Q3c, but clean.  The cleanest of all the points I tested.

Q5 - collector: breakover, but a bit less than Q3c
Q5 - emittor: very, very ratty breakover, sounds pretty bad

Q7 - collector: some breakover, about the same as Q5c
Q7 - emittor: similar to Q7c, but rattier (though not as ratty as Q5e)

Q9 - collector: some breakover, about the same as Q5c and Q7c
Q9 - emittor: fairly louder than Q9c and much rattier breakover (somewhere in between Q5e and Q7e in terms of "rattiness")

Q10 - emittor: breakover, much like Q3c

So, to sum up:
Q3c, Q5c, Q7c, Q9c and Q10e all have breakover noise
Q5e, Q7e and Q9e all have a dirtier, rattier, nastier breakover, with Q5e being the worst
Q3e is pretty much clean (at least with the Strat it was)


Does this give you any ideas what might be going on?  It'll be great to finally figure this out and get a nice clean sound.  The effect itself sounds amazing.