Debugging a Neo Vibe (Understanding Audio Probing and Transistors)

[Kevin Mitchell]

I hope the title was clear enough. I've made another thread for debugging this same circuit months ago but figured it would be smart to start a new one with a different approach. The lamp pulsates just fine and works well with the speed pot. The issue is the sound gets lost somewhere resulting in no signal to output.

I put together an audio probe to see where the signal is getting lost. My understanding of circuits and it's components are poor and I am hoping someone can explain to me what is happening here.

So I followed the input signal with the probe, not getting far I've noticed the signal ends at "B" of the Q1 transistor. My guess would be bad solder or a bad transistor. But I don't know whether the signal should be at "E", "C" or both. Hope someone can clear that up for me. I will be posting voltages later today if probing alone doesn't suffice although I would like to know what to look for while reading the transistors and how to go about it.

I would also like to mention according to the most recent wire diagram on Geofex.com I've noticed there is no connection to the ground for the output jack. Pretty sure that's not right so I made the connection.

The 2 bulb transistors are 2N3904 and the rest are 2N5088. I'm powering it with a 24V 100MA electro-harmonix AC/DC adapter.

[PRR]

Without a link to the schematic it can be hard to tell if signal "should" be at an Emitter.

In "many" transistor amplifiers, there's small signal at Base and large signal at Collector, but small or no signal at Emitter.

There are many exceptions.

It helps to stare at the road-map (schematic) and try to figure out how signal gets from here to there. Not a deep understanding, just a good guess of the highlights of the tour.

You stopped at Q1? Go further. If there's good signal into Q2 then we are not too concerned about oddness at Q1, especially since you are looking for a NO-signal problem rather than a garbled signal problem.

[Kevin Mitchell]

I didn't stop at Q1, the signal does.

Here's the PDF from Geofex;
http://www.geofex.com/FX_images/vibeschm.pdf

[R.G.]

OK, same advice applies as with your first posting. But I'll expand the advice so it's clearer.

Go here: http://www.diystompboxes.com/smfforum/index.php?topic=29816.0
Read it and follow the instructions. It's a reliable way to find out what's wrong, and that's why it's titled "Debugging: what to do when it doesn't work" and why it's stuck to the top of the list of posts, so that you actually have to actively skip over it to the most recent posts.

The audio probe is telling you that once the signal hits the input to the first active device, it disappears. That's an important clue, but that's all the audio probe can do for you - tell you where the signal is and isn't. After that, you have to go digging into where it disappeared to figure out why.

This is what the debugging procedure does for you. It leads you into measuring the things that would let someone sitting across the continent look at the schematic (which is why the request for the schematic is there) and reason about why the voltages might be as they are, and what likely flaws could cause the observed voltages. This is a much faster and surer way than people simply opining that it might be that the wires are the wrong color, or other things that are equally useful.   

So, get out your voltmeter and start measuring voltages on those first three transistors. Something in there is not working right, and that's what the audio probe told you. That's where to start digging. You'll save yourself and us a lot of typing if you go ahead and measure the pin voltages of all the transistors while you're at it.

[Kevin Mitchell]

Thanks R.G. I already figure as much. But a question I had was if the signal should be present in E or C so I could understand transistors and probing a little better.

I'm sure this is normal, the voltages drop and go back up so I took the higher end of the voltage. Please let me know if you think I should go back and remeasure anything.

Q1;
E: 23.9
B: 23.8
C: 12

Q2;
E: 28.2
B: 27.6
C: 26.4

Q3;
E: 23.8
B: 27.6
C: 26.4

Q4;
E: 28.1
B: 27.6
C: 27.2

Q5;
E: 28.6
B: 28.1
C: 27.1

Q6;
E: 28.1
B: 27.7
C: 27.3

Q7;
E: 28.6
B: 28.1
C: 27.3

Q8;
E: 18.7
B: 26.8
C: 27.4

Q9;
E: 27.8
B: 18.3
C: 26.7

Q10;
E: 27.3
B: 27
C: 26.8

Q11;
E: 7.7
B: 3.1
C: .6

Q12;
E: 15.4
B: 7.7
C: .6

I measured the power and for some reason it says 32.5 but it should be 24. This is as far as I got last time I tried to get R.G. some readings and that alone is very discouraging for me.

[R.G.]

All right, let's start unwrapping the onion.
From the debugging thread:

2.Links to the source of the project web site with the schematic or project, layout and wiring diagram. Example : Neovibe
3.Some candid admissions about how closely you followed the suggested layout, if any. If you just perfboarded it, say that, and if you can, include a picture of the board. Scanners do a GREAT job of making pictures of PCBs, by the way.

The biggie here is this: did you buy a PCB, or did you make your own PCB, or did you perfboard or build it some other way. Those requests were not just to make you jump through hoops. There are different things to check based on which way you did this. You gotta tell us what you did before we can tell you which piece of it is fouling you up.

4.Any parts substitutions or modifications you made to the original. If you didn't have any 10uF caps and used 22uF, say that. If you had to use a 2.7K resistor instead of a 2.2K, say that. If you used a BC109 instead of a 2N3904, say that.

Did you use the original parts suggested in the parts list? And if so *which* parts list?

By the way, if you substituted any transistors, it's a sure bet that the first question you get asked will be "Did you check the pinout of that transistor?" Delight the questioner by saying (truthfullyâ€) "Yes, I did a google search on that part number and check the pinout against how I soldered it in." If you can't say that, chances are good that you've already found your problem, all by yourself.

So, did you look up the proper pin out for the transistors you actually used? Do you know the pins are in the right holes?

That's preparatory to looking at the pin voltages. But the voltages do tell a cautionary tale. The voltages on all the transistors are way high, much beyond what they should be. So the question then is *really* which version of building did you choose, because it looks like either the measurements are way off, or the power supply setup is severely wrong, or the pinouts are wrong, or some or all of those.

Let's do power supply first, because if the power supply isn't right, nothing else will work.

[Kevin Mitchell]

R.G., I etched my own board using your most recent layout provided and I also used the suggested parts. As for what parts list, your parts list is what I used. Nothing funky. I also mentioned the type of transistors I used. "The 2 bulb transistors are 2N3904 and the rest are 2N5088. I'm powering it with a 24V 100MA electro-harmonix AC/DC adapter." The adapter was brand new I have no idea why it's pushing out 32.5 volts. No subs, no mods. (Other than connecting the ground of the output which I have previously mentioned.)

As for the transistor pins, I didn't look into that. I just assumed how they went according to the layout you have made along with the parts you have suggested. It provides the pins for 2N3904 which I assumed was the same for the 2N5088. E, B, C being bottom, middle, top for how it is laid out on the board.

Thanks by the way.

As for wiring, I wired it just as it is shown in the diagram here; http://www.geofex.com/FX_images/vibeschm.pdf
Maybe I shouldn't have?

Looks like I need to find another adapter huh?

[Kevin Mitchell]

Let's do power supply first, because if the power supply isn't right, nothing else will work.

The bulb works, speed works, depth works and the volume seems to work. A faint hum is there when the volume is crank up and nothing at all when it is down.

[R.G.]

R.G., I etched my own board using your most recent layout provided

OK. Just to be sure we're on the same page, that's this one, right?
http://www.geofex.com/Article_Folders/univibe/vibeupdate.pdf

"The 2 bulb transistors are 2N3904 and the rest are 2N5088

OK.

I'm powering it with a 24V 100MA electro-harmonix AC/DC adapter." The adapter was brand new I have no idea why it's pushing out 32.5 volts. No subs, no mods. (Other than connecting the ground of the output which I have previously mentioned.)

OK, unplug the power supply from the univibe clone, and probe the output of the power supply with your meter, but test it for  both DC and AC levels. I'm not familiar with that adapter, so I don't know whether it puts out AC, rectified but not filtered DC, filtered DC, or filtered and regulated DC. If it's AC or rectified but not filtered DC, the increase in DC level is understandable. AC outputs are specified in RMS voltage, but the peak of an RMS sine wave is 1.414 times the RMS. So a "24V" AC output would be 24*1.414 = 33.9V peak minus one or two diode drops, or about 31-33Vdc, which is remarkably like what you see. If it's rectified but not filtered, it would have the same peak value. Your meter might read it as the average of half- or full-wave rectified AC, or might read a "corrected" value. Depends on the meter. But test what you get right out of the adapter.

With that number in hand, plug it back into the unit and clip your meter black/(-) lead to the ground contact of the input jack. Then measure the DC voltage across C24 and C24. This should be positive on the + pin with the meter (-) lead on the minus pin. I'm guessing you read 30+ volts here. Leave the black lead on the input jack ground/bushing contact for all further measurements, and leave it on DC settings.

Then measure the voltage on all three pins of the 7815 regulator. This ought to read the same as the C24/C25 + voltage on one pin, 0.00000000 V on the middle pin and 15V on the third pin. If not, this is a big hint that something is wrong here.

As for the transistor pins, I didn't look into that. I just assumed how they went according to the layout you have made along with the parts you have suggested. It provides the pins for 2N3904 which I assumed was the same for the 2N5088. E, B, C being bottom, middle, top for how it is laid out on the board.

If you used that layout and observed the direction of the flat face of the transistor, it should be correct.

As an aside, people have been building neovibe clones since the mid 1990s from this or it's ancestor layouts, and they've used just about everything that could be possibly called a transistor, so I have to ask these questions.

As for wiring, I wired it just as it is shown in the diagram here; http://www.geofex.com/FX_images/vibeschm.pdf
Maybe I shouldn't have?

That wiring does work.

L

ooks like I need to find another adapter huh?

Not yet. Gotta do the due dilligence measurements first.

[Kevin Mitchell]

OK. Just to be sure we're on the same page, that's this one, right?
http://www.geofex.com/Article_Folders/univibe/vibeupdate.pdf

Yes with the updated layout and diagram; http://www.geofex.com/FX_images/vibeschm.pdf

I have just noticed that this adapter reads backwards. Usually the black lead of the MM goes on the inside and the red to the outer layer. I'm getting a reading of -33.3 DC and 0 AC. But when I switch them around I get +33.3 DC and +72.6 AC.

Also these readings aren't consistent and I have no idea why. They go up or down around .2 volts every time I measure it.
Here's a picture of what I'm pretty much using. Same model # and specs. Please excuse the rather large image;


I suppose it wouldn't be as simple as switching the + and - wires from adapter to diode bridge?

Why is the lamp working fine under these circumstances?

I did not use a 7815 regulator. Should I have? I thought the 2N3904 would be fine for my case. But perhaps not with this power adapter?

[Kevin Mitchell]

I would guess the issue (other than this confusing adapter) is I'm using a 2N3904 instead of something like a 78L15? When I built this thing I was under the impression the two unlabeled transistor pads were an optional thing depending on if you're using AC or DC? I can't recall completely why I put the two 2n3904s there (the spots above C23 and C24).

Are Q11 and Q12 where I meant to put the 2N3904? Looking at the layout that would make more sense based on what I read months ago. But as I said the lamp is working fine... All transistor readings I have posted earlier (Q-1 to Q 12) are 2N5088.

[R.G.]

Yes with the updated layout and diagram; http://www.geofex.com/FX_images/vibeschm.pdf

OK, we are now in lockstep.

I have just noticed that this adapter reads backwards. Usually the black lead of the MM goes on the inside and the red to the outer layer. I'm getting a reading of -33.3 DC and 0 AC. But when I switch them around I get +33.3 DC and +72.6 AC.

I'm a little confused by that. I would expect you to get either +33.3 and -33.3 depending on the way you had the probes inserted, and much less than 72V AC either way. It's an odd circumstance, but for the moment let's just leave that, and I'll stay suspicious of it.

I suppose it wouldn't be as simple as switching the + and - wires from adapter to diode bridge?

Sadly, no. One of the beauties of the rectifier bridge that's on the PCB - um, you did use a diode bridge ("DB1"), didn't you? - is that the diodes route the incoming voltage, AC or DC around to the right way on the PCB. You pay for this in the 1.4V lost to the diodes in the bridge, but you can't get the incoming polarity wrong with that thing in there.

Why is the lamp working fine under these circumstances?

Because the design of that section of the circuit is tolerant of higher voltages than the rest of it (the audio section...) in the original design.

I did not use a 7815 regulator. Should I have? I thought the 2N3904 would be fine for my case. But perhaps not with this power adapter?

Ah, my circuit-antennae just twitched. Should you have used a 7815? Well, it's in there for a reason, so probably yes.

And this is one of those little things that the debugging procedure is intended to flush out - not using a 7815 means the voltages in the audio section will be off, perhaps a LOT off, which is one of things you're finding. That may not be THE problem, but it certainly could cause problems. Not using a voltage regulator is a modification to the original circuit.

The twitching of my circuit-antennae is from one of the possible meanings of what I read you as saying. When you say " I thought the 2N3904 would be fine for my case" did you mean that you put in a 2N3904 *instead* of a 7815 in the spot reserved for the 7815 on the PCB?

[Kevin Mitchell]

The twitching of my circuit-antennae is from one of the possible meanings of what I read you as saying. When you say " I thought the 2N3904 would be fine for my case" did you mean that you put in a 2N3904 *instead* of a 7815 in the spot reserved for the 7815 on the PCB?

I just updated my last post;

I would guess the issue (other than this confusing adapter) is I'm using a 2N3904 instead of something like a 78L15? When I built this thing I was under the impression the two unlabeled transistor pads were an optional thing depending on if you're using AC or DC? I can't recall completely why I put the two 2n3904s there (the spots above C23 and C24).

Are Q11 and Q12 where I meant to put the 2N3904? Looking at the layout that would make more sense based on what I read months ago. But as I said the lamp is working fine... All transistor readings I have posted earlier (Q-1 to Q 12) are 2N5088.

Okay time to order some regulators lol. And yes I am using a diode bridge. Same one in the parts list I believe.

[R.G.]

I would guess the issue (other than this confusing adapter) is I'm using a 2N3904 instead of something like a 78L15? When I built this thing I was under the impression the two unlabeled transistor pads were an optional thing depending on if you're using AC or DC? I can't recall completely why I put the two 2n3904s there (the spots above C23 and C24).

I posted just as you did, apparently.

No, two 2N3904s are not going to act the same as a 78L15. Although the 78L15 is the same package as the 2N3904, it is an entire integrated circuit that just happens to have the same number of pins and plastic case as the 2N3904. The 2N3904 is a single transistor. The two are NOT interchangeable, which is one of the reasons the debugging thread goes on and on about looking up pinouts on datasheets. Two things with different type numbers that have the same package may be very, very different inside.

There are two versions of the 7815: a low powered one (78L15) and a high powered one (7815), either of which will work in this circuit. The extra pads are so you can use either the low or high powered version of the 7815.

the 2N3904s are definitely a problem, and they need to come out and a 7815 of one of those flavors needs to go in. You could also use a 7818 or 78L18 as these are compatible with the 7815, just putting out 18V instead of 15V, as the datasheet says.

Are Q11 and Q12 where I meant to put the 2N3904? Looking at the layout that would make more sense based on what I read months ago.

No. Q11 and Q12 should be 2N5088. The 2N3904 should go immediately above C23.

[R.G.]

Okay time to order some regulators lol. And yes I am using a diode bridge. Same one in the parts list I believe.

OK. We have peeled back one layer of the onion. Post back here when the regulators are in and you've measured the DC voltage on all three pins of the inserted regulator.

Notice the pinout of the regulator you put in. It would be best to get the actual datasheet of the one you get - and you might as well get the high-power version, they're about the same price as the low-power version -  and carefully check how it goes on the board so that the input voltage pin is connected to the top of C23/C24.

[Kevin Mitchell]

Thank you so much R.G. you're a great fella. I usually do these things in a tired post-work daze so I guess a mistake like this is understandable. Especially when the one above Q23 can be a 2N3904 but also has pin for something else. Let's call it an honest mistake.

Do you think I should find a better adapter as well? Any suggestions?

Notice the pinout of the regulator you put in. It would be best to get the actual datasheet of the one you get - and you might as well get the high-power version, they're about the same price as the low-power version -  and carefully check how it goes on the board so that the input voltage pin is connected to the top of C23/C24.

No. Q11 and Q12 should be 2N5088. The 2N3904 should go immediately above C23.

But you said the 2N3904 goes above C23 how can I put a 7815 there as well? And I will definitely make sure all connections are correct.

[Kevin Mitchell]

Thank you so much R.G. you're a great fella. I usually do these things in a tired post-work daze so I guess a mistake like this is understandable. Especially when the one above Q23 can be a 2N3904 but also has pin for something else. Let's call it an honest mistake.

Do you think I should find a better adapter as well? Any suggestions?

Notice the pinout of the regulator you put in. It would be best to get the actual datasheet of the one you get - and you might as well get the high-power version, they're about the same price as the low-power version -  and carefully check how it goes on the board so that the input voltage pin is connected to the top of C23/C24.

No. Q11 and Q12 should be 2N5088. The 2N3904 should go immediately above C23.

But you said the 2N3904 goes above C23 how can I put a 7815 there as well? And I will definitely make sure all connections are correct.

So should I get the 7815 or the 7818?

[R.G.]

Notice the pinout of the regulator you put in. It would be best to get the actual datasheet of the one you get - and you might as well get the high-power version, they're about the same price as the low-power version -  and carefully check how it goes on the board so that the input voltage pin is connected to the top of C23/C24.

No. Q11 and Q12 should be 2N5088. The 2N3904 should go immediately above C23.

But you said the 2N3904 goes above C23 how can I put a 7815 there as well? And I will definitely make sure all connections are correct.

DOH!!!
Well, I make mistakes too. I meant C24/C25. 7815 is over by C24/25, 2N3904 is above C23. That's my story and I'm sticking to it.

Thank you so much R.G. you're a great fella. I usually do these things in a tired post-work daze so I guess a mistake like this is understandable. Especially when the one above Q23 can be a 2N3904 but also has pins for something else. Let's call it an honest mistake.

I can't tell you how many times I've done similar things. No worries. Mistakes are a way to learn if you let them be.

Well, if you can make them non-fatal mistakes, anyway.  

Do you think I should find a better adapter as well? Any suggestions?

I don't know if that's needed or not. Frankly, if you have time, drop by a Goodwill store and pilfer through their rack of appliances and electronics, looking on the plug in adapters for something that says "18V", either AC or DC. One Goodwill store I went to had a basket of entangled adapters for $0.50 each. I must have had a dozen of them as I walked out. But the 24V DC adapter may be just fine. The only issue is whether that "73VAC" thing I said I was going to keep worrying about rises up as a flaw in the adapter leaking AC wall voltage through it. I think it was a measurement error, but I'm going to continue to worry until it's proven nothing. We can tell when we get the regulator in.

[Kevin Mitchell]

Okay thank you!

Sorry I keep going back to edit my posts. I'm a very indecisive guy. Added another question;
Should I get the 7815 or the 7818?

[R.G.]

It doesn't matter much. If you keep the 24V adapter, which as we know can put out nearly 32V, the 7818 is probably better, as this makes the voltage regulator chip not get as hot. But the 7815 will probably not be too hot anyway, especially if you get the non-low-power version - 7815, not 78L15.

So - if you're going to keep that adapter, get the 7818. If you get a lower voltage one, the 7815. I picked the 7815 as being closest to the "vintage" voltage in the original univibe back in the mid 90s when I first did this thing.

OK,  I talked myself through it. Get the 7818. Make sure it's the "TO-220" package, flat plastic rectangle with three beefy pins sticking out.