Forum-Vibe . . . . no oscillation

[Fastocker]

Hi, just finished my Forum-Vibe build yesterday . . . now looking for some troubleshooting help.  I used a Classic Amplification board and followed the Forum-Vibe "Mods" version during the build.  I'm using a single power board with a 24VDC, 250mA wall wart (center +) to power everything.  Right now I can get the bulb to light up and can change the intensity using either trimmer 2 or 3.  When I plug a guitar in, everything works except the bulb is not oscillating.  I get a bypass signal and the signal through the circuit when I hit the footswitch (i.e., you can definitely hear the vibe's preamp when you click the footswitch on).  The volume pot works but I get nothing from the intensity or the speed pots.  While playing through the circuit and manually adjusting the trimmers, I can hear the vibe effect . . . the bulb is just not oscillating on its own.

I'm using a 12V, 40mA bulb with NSL-7532 photocells.  I originally installed MPS-A13s into Q11 and Q12; wasn't sure if that was a problem so I replaced them with the same 2N5088 I used throughout the rest of the build (Q10 is an MPS-A13, though).

I searched and found this thread:

http://www.diystompboxes.com/smfforum/index.php?topic=91135.0

Now, first I checked and rechecked all my wiring and connections and even replaced the speed and intensity pots and that didn't help.  Then I pulled the ground side of resistor R45 as Redhouse suggested in that thread and hooked up an LED in line --- I have the anode connected to the lifted side of R45 and the cathode going back to the ground trace on the board.  The LED lights up but does NOT oscillate . . . it simply stays fully lit as shown.  So it sounds like I have a problem with the oscillator section of the circuit.

Any suggestions on what to try next?  Here are a few photos of the board -- Thanks!

Jim

[R.G.]

1. Go read and follow "Debugging: What to do when it doesn't work". In spite of the notes that follow, you'll need to do this unless you just count on being lucky.
2. Remove the LED Mod. Does the light bulb - not the led - oscillate?
If not, you have problems with the odd phase shift oscillator that the univibe used. The problems are usually too little gain in the darlington connection in the LFO or bad wiring/wrong component/wrong orientation. These cover high-90s of the problems with the LFO. Sometimes it helps to replace the pot with two fixed resistors of about 47K to set the speed to a fixed value and eliminate pot wiring as a cause.
3. Go back to 1.

[Fastocker]

OK, should have followed the debugging thread -- everything seems hunky-dory on the build as far as I can tell.  I'm getting power, passing a signal through the vibe, etc.  I did go back and measured voltages just now -- here they are compared to the original values presented in the Forum-Vibe documents.  My measured values are in red and, although slightly lower, I didn't see anything that jumped out at me.  I measured Q11-Q14 just for the heck of it:



The LED Mod was just a test that Redhouse described in the second post of this thread: http://www.diystompboxes.com/smfforum/index.php?topic=91135.0  The bulb would not oscillate before I tried the LED test, either.  I'm going to go back and re-re-recheck all my pot and wiring connections again to see if I just simply missed something.  Also, for Q11 and Q12 I've had either 2N5088 or MPS-A13s in there . . . neither worked.

With the voltages all checking out, I just keep thinking I have one component out of whack or something wired wrong . . . although I can't seem to find it.  Again, going to go back and recheck everything.  I'd like to try replacing the speed pot with the fixed resistors as recommended . . . I'm just not sure how to do that with the dual-gang pot.

Jim

[Fastocker]

BTW, R.G. --- I built a Neovibe (GGG board) recently and it fired up on the first try and sounds beautiful.  Hoping to get the Forum-Vibe up and running here shortly as well.

[R.G.]

OK, should have followed the debugging thread -- everything seems hunky-dory on the build as far as I can tell.  I'm getting power, passing a signal through the vibe, etc.  I did go back and measured voltages just now -- here they are compared to the original values presented in the Forum-Vibe documents.  My measured values are in red and, although slightly lower, I didn't see anything that jumped out at me. 

They look reasonable. Even if the voltages all look OK, that tells you something about where not to look.

The bulb would not oscillate before I tried the LED test, either.  I'm going to go back and re-re-recheck all my pot and wiring connections again to see if I just simply missed something.  Also, for Q11 and Q12 I've had either 2N5088 or MPS-A13s in there . . . neither worked.

Try this. Fire it up. Then temporarily short the base of the first transistor in the LFO darlington to ground. When you let it go, does the bulb light a couple of times and tie out? This applies a huge kickstart to the LFO. If it's working but just too low gain, this will make it wobble a little.

I'd like to try replacing the speed pot with the fixed resistors as recommended . . . I'm just not sure how to do that with the dual-gang pot.

If you stare at the univibe schematic and think about that pot, it is wired like two variable resistors with a common end. The common end goes to ground, the variable ends connect to the capacitor chain. So a fixed resistor from each of the variable ends to the common end subs in a pair of fixed resistors.

I'd be more specific with parts numbers but I don't have the Forum Vibe schemo or documentation at hand.

[Fastocker]

Try this. Fire it up. Then temporarily short the base of the first transistor in the LFO darlington to ground. When you let it go, does the bulb light a couple of times and tie out? This applies a huge kickstart to the LFO. If it's working but just too low gain, this will make it wobble a little.

I tried grounding out the base of Q11 and the bulb dims; when I let it go, the bulb flickers and lights back up to full brightness.

[R.G.]

That's actually pretty diagnostic. C22 and C23 (if I'm looking at the right part numbers; one before and one after the depth pot) block any DC conditions from the lamp driver. If shorting Q11 base causes the lamp to dim as long as you hold the short, then something is wrong with the DC isolation that C22 and C23 should be providing. It would also account for the LFO not oscillating.

Measure all pins of the depth pot for DC. If the DC is not less than a few millivolts, something is wrong with C22 and/or C23; probably C22.

[Fastocker]

That's actually pretty diagnostic. C22 and C23 (if I'm looking at the right part numbers; one before and one after the depth pot) block any DC conditions from the lamp driver. If shorting Q11 base causes the lamp to dim as long as you hold the short, then something is wrong with the DC isolation that C22 and C23 should be providing. It would also account for the LFO not oscillating.

Measure all pins of the depth pot for DC. If the DC is not less than a few millivolts, something is wrong with C22 and/or C23; probably C22.

Got a chance to measure DC voltage at the depth pot --- a few millivolts on pins 2 and 3, no voltage on pin 1.  If I read your response correctly then those readings are OK.  Still unable to get the bulb to oscillate . . . . grrrr . . . . 

[R.G.]

Sometimes one comes along that's a little tougher.

Time for next-level debug.
1. which board rev did you use, and does the board rev actually match the schematic you think you're using?
2. just to keep my curiousity in check, are Q11/Q12 actually 2N5088?
3. Can you measure and report:
- voltage with respect to ground on both ends of resistors R38, R39, R45, R46
- voltage with respect to ground on both ends of C19, C20, C21
- voltage from base to emitter (i.e. meter leads right on those pins) for Q11 and Q12

We'll start there.

[Fastocker]

Sometimes one comes along that's a little tougher.

Time for next-level debug.
1. which board rev did you use, and does the board rev actually match the schematic you think you're using?
2. just to keep my curiousity in check, are Q11/Q12 actually 2N5088?
3. Can you measure and report:
- voltage with respect to ground on both ends of resistors R38, R39, R45, R46
- voltage with respect to ground on both ends of C19, C20, C21
- voltage from base to emitter (i.e. meter leads right on those pins) for Q11 and Q12

We'll start there.

1.  I believe the board I used is the V3 board as I bought it recently from Classic Amplification.  I'm using the Forum-Vibe "Mods" version with this schematic: http://classicamplification.net/forumvibe/fv_sch_mod.htm.  The parts layout is here: http://classicamplification.net/forumvibe/fv_mod_08.htm

2.  Yes, Q11 and Q12 are 2N5088s.  I had MPS-A13s in those spots originally (per the Forum-Vibe mod suggestions) but pulled them and tried the 2N5088s to see if they would work.

3.  Voltages (with respect to ground) for each end of the components you listed read as follows:

R38:  18.28v --- 10.75v
R39:  0.0v --- 10.74v
R45:  0.0v --- 9.71v
R46:  7.59v --- 7.59v

C19:  9.69v --- 10.75v
C20:  9.71v --- 9.71v
C21:  0.0v --- 9.71v

Q11, base to emitter:  0.44v
Q12, base to emitter:  0.60v


Let me know if you see anything in those readings -- thanks again for your help!

Jim

[R.G.]

It needs some thought, but try this: set  your meter to ohms, and measure resistance between the two ends of C20

[Fastocker]

It needs some thought, but try this: set  your meter to ohms, and measure resistance between the two ends of C20

Alrighty, then -- I'm getting a resistance of 6.82kΩ across C20.

[R.G.]

Does that change as you rotate the speed pot?

[Fastocker]

Does that change as you rotate the speed pot?

Yes it does --- I'm reading about 103kΩ with the speed control turned all the way down; about 5.11kΩ with it's turned fully up.

[R.G.]

I've spent some time thinking about this, and still haven't come up with a smoking gun.

It's possible that there's a gross error hidden in there somewhere, in the form of a circuit board flaw, soldering issue, whatever. However, the voltages don't reveal it if there is. It acts very much like the gain of Q11/Q12 is simply too low, or that there is too big a load on the emitter of Q12 through C22, or that there is a flaw with the three-capacitor stack or limiting diodes D1/D2.

It's unlikely that Q11/Q12 are themselves the problem since you've replaced them, unless you're getting them in backwards. Transistors DO work with emitter and collector swapped as long as the voltages across them are low, and this gives low gain, so that's one thing to check just to be sure. I would expect other problems from this, but maybe not. Check the pinout and connections with a datasheet for your transistors. Sorry - yes, again.

Beyond that, it's time for divide and conquer. Remove C22 to eliminate loading issues. Use your ohmmeter to set the resistance in the speed pot as measured across C20 to about mid range, about 50K maybe. Set the meter to DC volts and hook it across R45. Now power it up, watching the meter. The meter should wobble, showing you the oscillation if it's working. If it is, see if the speed pot is effective. In this case, the problem is past C22.

If the meter doesn't show oscillation, the problem is in the LFO itself. As a last attempt at the obvious suspects, remove D1 and D2, and try it again. If it now oscillates, the problem was in the diodes.

If it still doesn't oscillate, I would get very frustrated and clip all the parts in the LFO off, clear all the holes, and rebuild that section with different parts. In a world of cheap parts, remanufacturing a small section of board is faster than internet typing in many cases. It's not very intellectually satisfying, but it's fast.

[Fastocker]

Well, still no luck . . .

• I checked Q11 and Q12 and they are installed correctly -- I used my Peak Atlas DCA55 to double check the pin orientation and both transistors are correct.
• I set the speed pot to a middle setting and removed C22; I then measured DC voltage across R45 -- no wobble on the meter, it simple read a steady 9.7V.
• I removed D1 and D2 and tried again; still no oscillation.

So I guess I'm to the point of completely rebuilding the LFO section?  Would that be the entire lower left corner of the board?



Man, this thing in no fun . . . shoulda stuck with my Neovibe . . . .

[R.G.]

Let's see. You have replaced the transistors a couple of times. You've cut it loose from the load of the lamp driver. You've removed the limiting diodes. We're kind of down to one of the three 1uF caps not being right somehow, one of the resistors not being the right value, or the board or wiring containing a flaw.

Does the common connection of the two sections of the speed pot indeed contact the emitter of Q12, as measured by a meter?

If it does, I would replace the 1uF caps one at a time. If they're polarized electro, observe polarity. If it works when one was replaced, it was that cap or its solder connections. If that doesn't work trace around all the copper areas with the sharp point of an X-acto knife to cut any hair-fine copper leftovers; that tripped me up once in 30 years. If none of this helps, replace all the resistors. Yes, they're all in the lower left.

[Fastocker]

Thanks for all your help on this one, R.G.  No progress to report . . . I've done all the diagnoses up to replacing the components one by one but still no oscillation.  I got a bit disgusted with the thing and just set it aside for now . . . not giving up, just going to take a break and then go back and revisit the board (with a cooler head) a bit later.



Just as an aside, I assembled another Neovibe board last weekend and it fired right up and sounds great, LOL.  I did few interesting mods to this one:

• Replaced R47 with a 100k resistor and 250k trimmer in series
  
• 1K5 resistors for R42 and R43 (more top end speed)
  
• Used a 4M32 for R40 . . . seemed to enhance the effect at slower speeds.
  

The trimmer on R47 works well in tandom with the lamp bias trimmer . . . it sort of allows you to dial in everything from a smooth 'wah wah wah' sound to that cool uneven heartbeat sound.  I think this is similar to what SteveB on this board did on his great sounding Neovibe.  I also used a 7818 for the voltage regulator just to see if running it at 18 volts made any difference.  I did notice that this build has a bit more depth to the effect than my first Neovibe . . . I had to use a 24VDC adaptor with this one, though.  Fun stuff these vibes . . .

[lonewolf]

I built a easyvibe..It acted the same way..no wobble...turned out to be a short on a trace that killed the oscillator..it was almost impossible to see..everything checked with a meter..moved a lead..fixed the problem..after 6 hours with a magnifyer..

[Fastocker]

Hmmm . . . thanks for the tip, lonewolf.  I might grab my loupe and a meter and start checking traces.