Univibe repair advice

[soggybag]

A friend of mine has a broken Univibe. Everything seems like it is working fine but it doesn't vibe. I opened up the cse and took a look at the lamp under the little aluminum box on the PCB. It lights up when I turn the vibe on but it just glows orange. It doesn't pulse and it's not super bright but it lit.

Any suggestions? I'm thinking the speed pedal cable or pot might be bad?

[spudulike]

LFO is shot, usually the bulb driver tranny. Replace it with a 2N3904.

Read this http://www.lynx.bc.ca/~jc/pedalsUnivibe.html

[soggybag]

Thanks for the fast reply. Let me ask another question. Is there a test you could recommend to confirm that the driver transistor is the problem? Or does the dim bulb point to this? I have never owned one of these, so I'm not sure how bright the bulb should be normally.

thanks again.

[R.G.]

The 2n3904 is OK if marginal as a bulb driver.

Does this one have a trimmer pot on the emitter of the driver transistor or a fixed resistor?

Measure the DC voltages of the transistors and diodes in the LFO section and pop them back here. We can probably tell you what is broken. We can certainly tell you if it's the driver transistor.

For reference, if you haven't yet, read "The Technology of the Univibe" at GEO - http://www.geofex.com

[soggybag]

There is a trim pot on the lamp driver transistor.

It looks like these transistors are ECB.

The Darlington pair in the LFO read:
E 11.7v
C 22.8v
B 12.4v

The lamp driver reads:
E 3.6v
C 17.5v
B 4.2v

Thanks for the help. I just took this on to help out my friend. It also looked like it might be a small challenge. I read the Univibe article at GEO and used the schematic there for reference. This looks to be the definitve guide on the subject. JC's site had some good information also. We'll see how the repair goes, this is inspiring me to build my own Vibe. I had looked at this project before and it looked like to much work and too many parts. Now that I have one in front of me it doesn't look as complex.

[R.G.]

The Darlington pair in the LFO read:
E 11.7v
C 22.8v
B 12.4v

Hmmm.... that's suspicious. If you are measuring across two base-emitters, both transistors, from the base of one to the emitter of the second, you'd expect to see at least a volt, maybe a bit more, as the first junction would be barely turned on, maybe 0.5V, and the second would be the standard 0.5 to 0.7V for silicon. You're showing only 0.7V total. So it could be that one of the B-E junctions is shorted or has a short around it. It could also be that you only measured one junction and the other is actually OK. Check that again, and measure each transistor B-E separately. If one of them has under 0.5V base to emitter, you have found your problem. Or ...A... problem, anyway.

The lamp driver reads:
E 3.6v
C 17.5v
B 4.2v

That looks OK. Since you have the trimmer pot, see if wiggling the trimmer pot changes the brightness. If it does, chances are the driver/bulb circuit is OK. The lamp is supposed to sit there at a dim orange glow normally. The trimmer is to let you adjust it to that level. In actual operation, the bulb rests at a dim orange with the depth control fully down. When you increase depth fully, the bulb oscillates from a bright flash to off.

[soggybag]

Thanks for the reply.

I measured the EB of both transistors in the darlington pair the first measures .7 and the second .6v.

I took the cover off the lamp again and checked the trimmer. Adjusting it changes the brightness of the lamp.

I notice that as I turn ont he device the lamp gets brighter dims then settles into it's steady dim glow.

The owner explained to me that the vibe had periodically stopped working. He said when this happened he would push the pedal all the way forward and that would get it started again. I'm not sure if that points to anything. This is what initially made me think there was something wrong with the pedal or it's cable.

[R.G.]

OK. That's a common malady in Univibes - a pulse or two right at turn on, then nothing. The gain of the darlington is too low to sustain oscillation. It may also be that there is a problem with the pots being open or intermittent or a cabling short or open. Get out your ohmmeter and ohm out the cable, and then leave your ohmmeter on the cable to measure the pot resistances as you sweep the rocker slowly. You should see nice, smooth resistance changes, no droputs or funny stuff, on both pot sections.

If the cable and pots check out OK, replace the first darlington pair transistor with a new high gain transistor. Try a 2N5088 if you have it, minding the different pinout.

The LFO in a univibe needs all the gain it can get. Sometimes the darlington gets tired.

[gez]

Sometimes the darlington gets tired.

How/why does this happen RG?  I don't doubt you...just curious.

[R.G.]

How/why does this happen RG?  I don't doubt you...just curious.

I don't know. I wish I did.

I would have speculated that it did not happen except that I've repaired half a dozen Univibes with exactly that symptom and the solution in every case was to put a new high gain transistor in the first transistor of the darlington. I concluded from this that the transistor used to be OK or the unit would have failed ship test or have been junked long ago, and if it was once good and is now bad, something happened along the way. That's the sum total of my info on this.

I do know that a discrete darlington made from two same-type-number transistors is not nearly as high a gain as a monolithic darlington The gain of a transistor is exponential with the collector current over a large range of currents (that's how multipliers work!) and the first transistor in the discrete darlington is running at 1/beta of the current of the second one, so the gain is quite low. But that doesn't explain how they get worse - if they do.

There is a wear-out mechanism that involves reverse breaking the base-emitter junction of a silicon bipolar. In addition to making the transistor noisier, it lowers the gain. I would propose that in this case because the three caps in series there could get up to over 12V (enough to reverse break two BEs) and if the power supply were turned off when the voltage was high, the caps could supply current to break the junctions. But - there's a 2.2M resistor in the way, and that would limit the current to incredibly small, which ought to keep the damage down.

Maybe it keeps the damage down so it only takes forty to fifty years to happen. If that's the problem, the fix is a reverse biased diode from emitter to base to clamp the reverse voltage to less than breakover.

Right now, darlington tiredness is a disease of Univibe LFOs until proved otherwise - but I have seen it there.

[soggybag]

I replaced the one of the transistors in the darlington pair with a 5088 and now the lamp does not light at all.

I measure .48v EB on one and .62v on the other.

I think I replaced the wrong transistor. I replaced the one closest to the 10u cap. I think I try replacing the the other and see what happens.

[soggybag]

Ack! After a closer look I found that one of the leads coming off the lamp is broken! I must have bumped it as I was working.

There is a tiny stub of a lead coming out of the bulb. I'm not sure it looks long enough to solder something to it. I might be bale to crimp a short stretch of wire to it.

Is it possible to replace the bulb? Or will all the mojo be forever and irrevocably lost?

I reinstalled the first transistor and replaced the second with a 5088. I'll have to get the sorted out before I can test it.

[gez]

the first transistor in the discrete darlington is running at 1/beta of the current of the second one, so the gain is quite low.

Although that's really obvious (now that I think about it), it had never occured to me before.  Thanks for mentioning it.

Right now, darlington tiredness is a disease of Univibe LFOs until proved otherwise - but I have seen it there.

Stange indeed!  Thanks for you reply.

[R.G.]

Ack! After a closer look I found that one of the leads coming off the lamp is broken! I must have bumped it as I was working.
There is a tiny stub of a lead coming out of the bulb. I'm not sure it looks long enough to solder something to it. I might be bale to crimp a short stretch of wire to it.

Bad juju there.
There are two possibilities - (a) you can solder a wire onto it, in which case that bulb has a chance; and (b) you can't, and the bulb is a throwaway.

Is it possible to replace the bulb? Or will all the mojo be forever and irrevocably lost?

Put it this way - whatever bulb you put back in there, original, replacement, or whatever, will have a thousand times the mojo it has now. Right now it's only a somewhat interesting piece of sculpture.   

[Paul Perry (Frostwave)]

Another possible long-term 'wear out' mechanism, is gradual diffusion of water vapor into the transistor (maybe along the legs). A long shot, but stranger things have happened.

[soggybag]

Is there a recommended replacement bulb? JC's site talks about a 12v 25ma bulb from RS.

[alderbody]

Is there a recommended replacement bulb? JC's site talks about a 12v 25ma bulb from RS.

Try Small Bear. Steve has a great bulb type.

btw, regarding the Darlington pair in the LFO, i finally settled with two BD139's in my Clone and it really works! (power @24V)
Of course those two are connected off board...

If i remember well the bulb driver i ended up using is a BD137.

a bit too large for the application, but i got the best ehaviour of all i tried (3904, 5088, 5089, FET's, Daringtons, etc...)

Very stubborn pedal!

...but if you get it right.... 

[soggybag]

I looked at Small bear, Steve has dual ganged reverse log 100K pots, which I thought would impossible to get, but I could not find a lamp. It looks like he has LEDs, what category did you look in?

[soggybag]

I went to RS and found a some bulbs. They are listed at 12v 50ma and 12v 60ma. I looked for the 12v 25ma recommended on JC's site, but couldn't find them. These were the closest they had.

I think I will first try and repair the old bulb. And if this doesn't work give one of these new bulbs a try.

[Eric H]

Right now it's only a somewhat interesting piece of sculpture.