Neovibe debug (voltages included)

[Auke Haarsma]

I had my shot at the Neovibe. No dry or wet sound coming through, no vibing. The bulb flashes only if I bypass R48 to ground (it flashes a bit...not very bright) Speed pot also seems to affect the brightness of the bulb.

I've used the updated layout. My voltages (R48 is NOT bypassed with these voltages):

78L15
in: 16.40 (de BR gets 18v DC)
puts out: 14.73

Q1
C:1.20 V (not stable. it fluctuates from 1.19 to 1.22, due to vibing?)
B:0.55
E:0.01

Q2
C:10.43
B:1.20
E:0.6

Q3
C:14.73
B:10.42
E:10.0

Q4
C:14.73
B:4.64
E:4.20

Q5
C:11.25
B:4.20
E:3.52

Q6
C:14.73
B:4.70
E:4.25

Q7
C:11.14
B:4.25
E:3.60

Q8
C:14.73
B:4.68
E:4.22

Q9
C:11.20
B:4.25
E:3.57

Q10
C:14.73
B:5.82
E:5.26

Q11
C:16.53
B:7.0 - 8.6V
E:4.0 - 14.4V

Q12
C:16.4
B:5.0-13V (I'd expected these to be the same as E Q11, probably me or my DMM? )
E:3.3-14V

Q13 (un-labelled)
C:16.4
B:3.9 -6.3V
E:3.3 - 5.5 V

Any help much appreciated!

[R.G.]

I had my shot at the Neovibe. No dry or wet sound coming through, no vibing.

No signal at all coming through is equal to saying "Something is wrong in the audio path as opposed to the LFO." That localizes one fault to the row of transistor circuits starting with Q1.

The bulb flashes only if I bypass R48 to ground (it flashes a bit...not very bright) Speed pot also seems to affect the brightness of the bulb.

"Speed pot seems to affect the brightness of the bulb." is equal to saying "there's a DC fault somewhere with the speed pot".

78L15
in: 16.40 (de BR gets 18v DC)
puts out: 14.73

That's low, but probably OK. If you're sure you'll only ever use 18V DC, you can rewire and remove the diodes, replacing TWO AND ONLY TWO of them with wire jumpers to get the +18V and ground connected where they're supposed to go. That will boost your regulator input to 18V and get you better regulation. Although this is NOT your problem now. It should work with 14.73Vdc.

Q1
C:1.20 V (not stable. it fluctuates from 1.19 to 1.22, due to vibing?)
B:0.55
E:0.01

That's because your regulator is not regulating, see above. It's also not a problem that would prevent audio from coming through.

Q2
C:10.43
B:1.20
E:0.6

This is OK.

Q3
C:14.73
B:10.42
E:10.0

This is not OK, and is at least one of the reasons you don't get any audio. With the collector of Q3 at 14.73, that is equal to saying "there is no voltage across R11, so Q3 is not pulling any DC current. Inspect Q3 for the usual problems: wrong pinout/orientation in the holes, solder shorts, solder opens, open/cracked traces, problems with R8, R9, or C18. Q3 will have to have its collector near 10-11V and its emitter near 3-4V to work right here.

Q4
C:14.73
B:4.64
E:4.20
Q5
C:11.25
B:4.20
E:3.52

Looks OK.

Q6
C:14.73
B:4.70
E:4.25

Q7
C:11.14
B:4.25
E:3.60

Looks OK.

Q8
C:14.73
B:4.68
E:4.22

Q9
C:11.20
B:4.25
E:3.57

Looks OK.

Q10
C:14.73
B:5.82
E:5.26

Looks OK.

Q11
C:16.53
B:7.0 - 8.6V
E:4.0 - 14.4V

Q12
C:16.4
B:5.0-13V (I'd expected these to be the same as E Q11, probably me or my DMM? )
E:3.3-14V

Looks OK. It's oscillating big time.

Q13 (un-labelled)
C:16.4
B:3.9 -6.3V
E:3.3 - 5.5 V

OK-ish. But something about the depth pot is messed up, either C22, C23, or the wiring has a problem, I think.

Probably not C22 or the LFO would not be oscillating.  Look at C23 and wiring.

[Auke Haarsma]

R.G.... you are simpy amazing. Your comment was SPOT ON.

A cold solderspot at the bridge from Q3 to C18 caused the audio-muteness. Reflowed the solder.. solved!

I'm getting a vibe. Flipping the chorus/vibe switch gives me quite a big difference in output volume. Also I haven't looked at what causes the dim bulb yet (it's still dim unless I bypass R48). The speed pot is exactly wired as shown on the pdf.

[R.G.]

R.G.... you are simpy amazing. Your comment was SPOT ON.

Well, remember I've been remote-diagnosing that circuit for over 15 years now. It's like Martin Mull's classic song "Licks Off of Records":
You oughta know the fact is
I'm this good cause I practice.

A cold solderspot at the bridge from Q3 to C18 caused the audio-muteness. Reflowed the solder.. solved!

Kewl!

I'm getting a vibe. Flipping the chorus/vibe switch gives me quite a big difference in output volume.

That's a bit odd. Check the values of the resistors leading to the chorus/vibe switch.

Also I haven't looked at what causes the dim bulb yet (it's still dim unless I bypass R48). The speed pot is exactly wired as shown on the pdf.

Some of these things, depending on the driver transistor and bulb need more current than others. What value do you have for R48?

[Auke Haarsma]

78L15
in: 16.40 (de BR gets 18v DC)
puts out: 14.73

That's low, but probably OK. If you're sure you'll only ever use 18V DC, you can rewire and remove the diodes, replacing TWO AND ONLY TWO of them with wire jumpers to get the +18V and ground connected where they're supposed to go. That will boost your regulator input to 18V and get you better regulation. Although this is NOT your problem now. It should work with 14.73Vdc.

Could you explain why I can't remove all 4 the diodes? I'm already feeding it DC, so is it for polarity protection? (but in that case 1 diode would do).

Some of these things, depending on the driver transistor and bulb need more current than others. What value do you have for R48?

R48 is 68 Ohm. I took it out and measured it just to be sure and it is 68 Ohm as it should be. Also C23 is 10uF (63V cap). Depth pot seems wired correctly. Middle lug goes to pad H. Outer lugs just affect the way the pot works (CW or CCW), right?

Should I try a different transistor in Q13? (all Q's are 2n3904's btw)?

[R.G.]

Could you explain why I can't remove all 4 the diodes? I'm already feeding it DC, so is it for polarity protection? (but in that case 1 diode would do).

Actually that's what I meant - remove all 4, and then put in wires to the two places which would make your 18V source come out the right polarity into the circuit. When I said "remove the diodes" it would have been clearer if I'd said "remove all four diodes". Although removing 3 and leaving one in for polarity protection is not a bad idea.

R48 is 68 Ohm. I took it out and measured it just to be sure and it is 68 Ohm as it should be. Also C23 is 10uF (63V cap). Depth pot seems wired correctly. Middle lug goes to pad H. Outer lugs just affect the way the pot works (CW or CCW), right?

Should I try a different transistor in Q13? (all Q's are 2n3904's btw)?

Hmmm...
(a) what's the value of your trimmer pot?
(b) Can you measure the DC voltages on all pins of the Depth pot and report it back here?
(c) Only if we can't do this some other way should you replace Q13. It could be the problem, but let's settle the other issues first.

[Auke Haarsma]

a) I'm not sure about the trimpot. The value is stamped to small to be sure. I measure 8k at max (and ofcourse 0k at min). Seems a 10k pot too me, though that would be way to big. My intention was to put in a 1k (couldn't get a lower value), but even with the current pot at min setting the bulb stays dim.

b) depth pot Voltages
Fully CW
lug 1: -0.8 to +0.8 V
lug 2: 0V
lug 3: 0V

Fully CCW
lug 1: -0.9V to +0.9V
lug 2: -1.5V to 1.5V
lug 3: 0V

And also the resistance to ground: (just because I first misread your reply... I thought you needed these)
Fully CW
lug 1:48k Ohm
lug 2: 0
lug 3: 0

Fully CCW
lug 1: 48k
lug 2: 48k
lug 3: 0k

[R.G.]

b) depth pot Voltages
Fully CW
lug 1: -0.8 to +0.8 V
lug 2: 0V
lug 3: 0V

Fully CCW
lug 1: -0.9V to +0.9V
lug 2: -1.5V to 1.5V
lug 3: 0V

The reason I had you do that is that the depth pot should never have any DC voltage on it. Those should all be 0V.

This shows that DC voltage is leaking into the pot through either C22 or C23 (or both!).

Possible causes:
C22 or C23 is inserted backwards.
C22 or C23 is shorted by solder between its pads.
C22 or C23 is defective (this is rare with new parts!)
There is some other random short/solder fault letting DC get to the Depth pot.

From the voltages, I suspect C23. If it were mine, I would, in this order:
(a) inspect C22 and C23 to see if one is inserted backwards. The + terminals should both be closest to the wiring pads for the depth pot, the (-) leads to the interior of the boards; the + lead also goes into the hole which has a square pad.
(b) Unsolder and remove C23. Measure all the terminals of the Depth pot for any DC voltage. It should be 0V on all terminals. If it is not all 0Vdc, put C23 back in and remove C22. Test for DC voltage again. If removing a capcitor makes the DC voltages on the pot go to 0V, that cap was letting DC through somehow and needs to be fixed/replaced.

I really think this will fix the issues with the bulb.

[Auke Haarsma]

The voltage on the depth-pot is gone if I remove C23. I replaced that cap, but still I have voltages on the pot. I checked the traces going to C23, they all seem ok. Resoldered several parts around that spot.

Could be of interest, I find Q13 getting pretty hot. Let's assume I'm not a sissy, but I don't wanna keep my finger on it longer than a couple of seconds. The other Q's don't get warm at all. To me this doesn't look good.

[rhdwave]

Not to butt in  but i think q13 is normal to get a little hot.  This happens with mine as well.  Unless there is a problem on my q13 too, which is of course possible.  Didn't mean to confuse things, just wanted to add my observations.

[R.G.]

The voltage on the depth-pot is gone if I remove C23. I replaced that cap, but still I have voltages on the pot. I checked the traces going to C23, they all seem ok. Resoldered several parts around that spot.

You'll need to dig back into that area. Something's still wrong there, and it vanishes when C23's removed. Just out of curiousity,
(1) is c23 a new part - not salvaged from another board?
(2) what are the capacitance and voltage ratings for C23?
(3) With C23 pulled out, can you vary the bulb's brightness with the trimpot?
[quote author=Auke HaarsmaCould be of interest, I find Q13 getting pretty hot. Let's assume I'm not a sissy, but I don't wanna keep my finger on it longer than a couple of seconds. The other Q's don't get warm at all. To me this doesn't look good. [/quote]
It's normal. Q13 is right out on the edge of what you can do with a TO-92 package. It works most of the time, gets a little hot, but works OK.

But Q13 is what fails most often in real univibes, too.

You can fix that by using a TO-126 or TO-220 packaged transistor in the board. That extra pad above Q13 marked (B) happens to be in the correct place for the base lead of a TO-126/220 power transistor package to solder right in. A TIP111/112/113 darlington would work, as would a BD675 or a 2SD985. I recommend darlingtons for the bigger transistors because power transistors tend to be sub-100 current gains. You may have to dink the biasing resistors a little with one of these.

I recommend leaving the TO-92 in there until and unless it fails.

[Auke Haarsma]

I'm at work now, but the caps in C23 where twice new parts. One 10uF cap with 35V rating and one with 63V (iirc).

I have to check (3) at home. What I do recall is that the light stayed dim/unlit with C23 pulled.

I'll report back asap when I get back home.

@rhdwave: thanks for chipping in. With your and RG's comment I think Q13 is not the problem right now.

Thanks for all the help!

[Auke Haarsma]

C21 is btw a 400V rated cap (1uF). At that time it was the only rating I had in stock. Should I replace it with a lower rated cap?

[rhdwave]

Anytime i can contribute here makes me feel better about mooching off of the generous help i get from so many on this board.  So, you are extremely welcome   Good luck!

[R.G.]

I'm at work now, but the caps in C23 where twice new parts. One 10uF cap with 35V rating and one with 63V (iirc).

OK, nothing obviously wrong there.

So I think there is a problem with the PCB (solder threads, invisible copper threads perhaps). SOMETHING about C23 is letting DC get from the Q13 circuitry back into the Depth control. You're going to have to find that.

[Auke Haarsma]

I understand, it has to be something in that area. Let's start Trouble shooting again

Troubleshoot-log:
-16.26 into regulator, 14.70V out.
-depth pot fully CW, lug 1 has voltages from -0.83 to +0.83 V.
-removed C22. 0V voltage on all lugs of depth-pot.  Q13 C:15.45 B:2.24 E:1.59 (static voltages!).
-put C22 back in. Voltage on lug 1 of depth pot is back...
-just to be sure: removed C23. Voltages are still on the Depth-pot. I think we can qualify C23 as being OK.
-put C23 back in.
-replaced Q13, same issue.
-tested it in a dark room, so I could see the bulb very clearly flash. It seems touching a potmeter or switch affects the brightness of the bulb slightly. Must be something wrong with my grounding on this pcb (?).

Let's have a look at the power-section of the PCB:

-removed the 4 diodes in the diode bridge.
-put back 1 diode as polarity protection.
-connect 17.5v DC on board. Voltage regulator now in: 17.5V, Out: 14.56V.
-checked depth-pot, still voltage on pot.

Back to the Caps

-desoldered C22 again. Put testleads from the depth and speed pot to C22. Teastleads to breadboard. C22 is now effectively placed off-board, to exclude PCB-etching/soldering errors around the C22-area. -same issue, still voltage on depth pot.
-if I replace C22 with a red LED, the led pulses. Brightness and speed of LED can be controlled with depth and speed pot. It seems to lit up when the bulb dims and vice versa.
-with C22 removed no flasing of the bulb
-With C22 removed, if I put a bridge where C22 would normally sit, I get ONE bright flash of the bulb and then it's back to it's dim state. Am I correct in thinking that that single bright flash drains C23?


Gotta take a break. Does the above give any new clues to where I may find my error?

[R.G.]

Actually it does give me what I needed - a good swift kick in the seat of the pants. 

Somehow I read what you clearly labeled as voltages alternating between positive and negative as ... DC ... voltages, not the AC voltages they are.

There definitely should be AC voltages on the depth pot. That's where the LFO signal is. And has to be. Once I got it into my head that you were describing DC voltages, they always read that way to me. Duh...

So put C22 and C23 back in and see if you can adjust the bulb to a dim orange glow with the Depth pot fully CCW. If so, does advancing the depth pot to full CW make the bulb flash on and off? If so, it's fine, and you need to move to a listening test. Otherwise, there may be some problem with Q13.

My apologies - I just KEPT misreading that on.

[Auke Haarsma]

No Problem R.G.! Knowing that the voltages on the pot are ok I put back C22 and C23, and vibe-ed away. I've already had so much fun playing around with this circuit!

The bulb needs finetuning. It only works good if I bypass R48, but I can live with that. The phases need some finetuning too, but I can do that by ear.

Thanks for all the help and info!

[Auke Haarsma]

Final update on this one:
-removed trimpot, put a bridge there instead.
-removed R48 (68Ohm) and replaced it with a 10Ohm resistor.

Sounds wonderfull! Sweet vibing sound, subtle to wobbling! This is a really sweet effect!

[rhdwave]

Final update on this one:
-removed trimpot, put a bridge there instead.
-removed R48 (68Ohm) and replaced it with a 10Ohm resistor.

Sounds wonderfull! Sweet vibing sound, subtle to wobbling! This is a really sweet effect!

That's great! Congrats! Just wondering how you have your ldrs positioned? Facing the lamp or straight up?