Dunlop UV-1 Uni-Vibe

[VintageGear]

It's cold outside, so time to fix something that has been bothering me for a long time.. my UV-1 UniVibe (revision F, somewhere from the 90s).
The issue concerns the speed. Without an expression pedal connected, the speed goed from fast to blazing-fast. No way of getting it down to a nice slow sweep. When I connect an expression pedal (or just a TRS plug), the signal is lost completely. Also, the LED goes of which says the unit is engaged.
So I suspected the expression pedal jack input, but upon replacing it, no change!

Over on "the other forum" one has attempted to draft a schematic, but not really connecting a TRS would cause the unit to stop passing signal....

Schematic by lolbou and Dirk Hendrik: http://freestompboxes.org/members/lolbou/Dunlop%20-%20Univibe%20RevA/Dunlop%20-%20Univibe%20UV-1%20-%20Schematic%20(2-3).pdf

[Kevin Mitchell]

You should fix the schematic link and perhaps put up a video showing the problems you are having with the pedal.

[PRR]

> fix the schematic link

This forum censors that forum's links.

Here is a similar-name file, with same drawer-names, on a mystery server in Russia. I downloaded it, and my PC has not gone up in smoke, yet.
http://milas.spb.ru/~kmg/files/schematics/Dunlop/Dunlop%20-%20Univibe%20UV-1%20-%20Schematic%20(2-2).pdf

[VintageGear]

Cheers! I cannot edit my openings post unfortunately. I'll make a video after Christmas! Happy holidays 

[VintageGear]

Alright, I got to shoot a video of the problem! I attached a 1khz tone generator to the input.
Rotary speed is way too high, and when with the expression pedal input occupied, nothing comes through!
https://vid.me/96wY

Any ideas???

[VintageGear]

Maybe Q12 and Q13 problems?

[VintageGear]

It's still staring at me at my workbench...

[anotherjim]

How is the on-off switch at the expression jack implemented? Built into the jack or separate?
I suppose the FET that takes the Jack to a parallel connection from the speed pot could have been damaged if something very wrong has been plugged into the jack at some time.

[VintageGear]

Oof, I totally forgot about this UV1 -- revision F! Now, two years later, during my quarantine cleaning, I found it stuffed away in a corner. Kinda gave up on it last time after replacing the expression jack.
Currently it does work on an expression pedal (well, I used some fixed resistors which I held against a patch cable, that changed the LFO rate).

Problems that I found were: 1) In bypass, no signal comes through and 2) The LFO is too fast, say 10x too fast. No gentle vibe at the lowest speed, but towards a tremolo speed.
Poked around a bit and found the 074 suspicious. Decided to be radical: Replaced the TL074 and yep that fixed the bypass issue.
Hurray.

Now the LFO is baffling me still. Had a good look around and tried to understand which section of the PCB did what.
There is a mainboard, and a daughter board. I believe the latter contains the LFO section alongside with the lightbulb that flashes on the LDR's inside the metal housing.
I found some great gutshots online:
Mainboard: http://www-personal.umich.edu/~damont/Dunlop1.JPG
Daughterboard comp side: http://www-personal.umich.edu/~damont/Dunlop4.JPG
Daughterboard solder side: http://www-personal.umich.edu/~damont/Dunlop3.JPG

There is a TL061 which at its output (pin 6) shows a LFO signal on my scope (well, not really low frequency, rather high freq)
I replaced the two elec. capacitors at the bottom right, to no avail. Figured it was worth a try for a quick fix.
So.. back to the books... I read https://www.electrosmash.com/mxr-phase90#link6 and wanted to discuss with you if I should try to play around with different values of capacitors to see if I can get the lowest LFO setting to be slower.

What do you guys recon?

[VintageGear]

Although it is a different revision, I believe the LFO is rather similar to my unit.


Played around with the capacitors around the TL061 but to no avail. Unsoldered the light bulb to see if that had any effect, but nope.
Any ideas how the speed of LFO is generated? The 2x 680n caps around the VR1 (LDR/LED)? Around U12?

P.s. whole schematic: https://i.imgur.com/MprXsZq.png

[VintageGear]

Any ideas?

[Eb7+9]

#%€¥ me ...


somebody reactivated my account
(pesky kids)


first off, the schematic looks questionable around Q12 and Q13 - maybe you can give that a closer look ... also, the 690r to ground next to Q12 might be in series with Q14-4 instead ... that would make more sense ... the rest of the control circuit looks ok with Q15 and Q14-1 doing the final v-I conversion (same circuit block JC Morisson used in the EHX "Black Finger" sidechain) ...

the main chain of control goes like this:

Control resistance (pot) >> converter current >>
LED light >> opto-cell Resistance >> LFO frequency/ speed

as far as converter current response to the 10k control pot goes, TP2 sets the overall "average" of the current range while TP1 sets the "width" of the current range (in a Limiter setting the exact same circuit would see these two trimpots become "ratio" and "threshold" controls) ... we want to make sure these actually do what they're meant to ... bear in mind that if cell resistance is inversely proportional to LED current and so is LFO frequency to resistance, then LFO frequency is proportional to converter current ... this way we can directly associate driver current with LFO frequency

to test the function of TP1 and TP2 we can thus directly measure the drive current produced by the converter circuit by sticking an ammeter across the opto LED terminals (NOT to gnd or anywhere else) ... we can now see how the drive current responds directly to the control pot, and two trimpots, taking the guess work out of the picture

on another front, I can also see an "old" signal path mistake towards the end there // wonder if that's a mistake in the drawing or the actual design ... see if you can spot it!

[VintageGear]

Thanks for sharing your knowledge! I did as you suggested, there is 0 amps between the LED lugs. There's 2v between them. Made me wonder if the LED lights up at all, so I desoldered the three lugs on the cell-side. INF resistance (well, >2 M, my dmm doesn't go higher) when power is OFF. Measured across the outside lugs: 7 ohm, which changes to 8 ohm with speed in maximum setting.
Conclusion: Yes, the LED lights up.
Hypothesis: Perhaps too bright, totally saturating the cell?

[Eb7+9]

if you're seeing 2 volts across that LED there has to be current flowing through it ...
let's just double check this

set your meter for CURRENT sensing
ie., put your DMM in 2000uA (DC current) scale and make sure the + lead goes in the right spot on your meter ...
(test first w battery and resistor if you're not sure)

... and now, just to make things perhaps a little less confusing :

completely lift/disconnect the optocoupler's LED side from circuit ...
next, place the test leads at each circuit pad where the optocoupler LED leads went
(this effectively bridges the circuit with the meter, which now acts as a near-short)
measuring what the LED would see for current ...

(the reason why: the driver circuit is a current source and doesn't care much if there's an LED or not, ie., a short, in its output circuit // the current will be the same ... plus/minus a tiny error)

tell us what DC CURRENT you see flowing there when you're at either end of the SPEED control ...

[VintageGear]

Thanks... Turned out the fuse in my DMM had blown and therefore I did not measure any current 
The resistor trick was indeed a good idea which made thinking about the fuse inside the unit. Had to get out to get a new one so apologies for the belated reply!

Speed at minimum: aprox 2.6 mA
Speed at maximum: approx 12 mA

[Eb7+9]

Speed at minimum: aprox 2.6 mA
Speed at maximum: approx 12 mA

well done ...

ok, so we need to know two things next: cell sensitivity to current, and current driver response to trimmer adjustment

my recommendation at this point would be to ...

(i) grab yourself a fresh 9v battery and a 10k resistor and hook them up in series with the LED of the opto-coupler ... and then read the DC resistance on the cell(s) side - read both cells to check for "balance" ... and also measure the current flowing through the LED ... you don't need to disconnect the cell resistor side from the circuit to measure its resistance since one end is blocked to dc in the circuit ...

this will give us an idea of the LED current sensitivity ... it's a good place to start from as it could influence the (potential) later replacing of the cell unit if need be, and it will confirm whether the currents you measured are indeed too high or not, or the cell resistors are too un-balanced // we don't know that yet ... so, let's find out what that first data point looks like for each photo-cell

(ii) investigate the current driver circuit adjustments by leaving the circuit the way it is atm and, again with your DMM set to ammeter mode, play with T1 and T2 trimmers and see if you can get those two end-point numbers to shift up and down ... see if you can get them to lie more in the 50uA to 1mA range ... try this at first to get a sense of adjustability on the driver side

---

just to be super clear ...

we want to know at which LED current level the two photocell resistors exhibit around 100k resistance, and also at which LED current level the photocells exhibit around 1k resistance ... we expect these DC resistance numbers to vary a little from cell to cell, depending on how well they are matched ... at least, this will be our chance to "know" with certainty what these numbers are for your particular opto-coupler ... these are about the right resistance levels a dual-potentiometer would be presenting a stock Vibe LFO circuit at either end of the speed range // obviously, the opto-CELLs need to reach those numbers ...

>> we want to match the drive circuit limits to those two (intermediate) LED-current numbers ... the LED current being a direct one-to-one "go-between" SPEED circuit resistance and opto-CELL resistance

the last step would include setting the high speed current to get about 1k resistance across the individual opto-CELLs and low speed current to get about 100k each ...btw, be careful with the trimmers - go a little at a time while constantly measuring current ... once things line up you should be able to hook both opto-cells back into the circuit and have things working closely to full-range performance

if you're lucky and there is nothing busted in the drive circuitry AND the opto-CELL unit is spec'd right (ie., fairly balanced on both sides -> at both resistance end-points) you should then be able to attain your "preferred" operating speed limits by adjusting those drive current end-numbers by slight fiddling of T1/T2 once inside the final target zone ... once fully happy with the performance of the circuit you could apply your DMM ammeter across the LED one last time and post your two final current drive "limit" values for future reference (!)

---

btw, if you decide to "cut to the chase" and replace the 10k resistor by a pot in the test above be careful
// make sure to include a current limiting resistor in series, otherwise there's real risk of blowing the opto-LED by accident

[VintageGear]

7sharp9, thanks again.  I performed the 9V + 10k resistor test:
The Opto measures 265k Ohm in dark, 10k Ohm in light for one side, the other side is quite identical at 268k Ohm and 10k Ohm light.
There is only one trimmer on the board near the opto coupler, it was hot glued in place (80% from maximum position), but could easily removed it and measured as well.
On the + and - pad of the LED it measures at approx. 20% 2.4mA, at 80% it is 2.6mA. Does not come close to the 0.5mA-1mA we were aiming for.

Without the LED connected in circuit, there is a nice slow vibe audible (but is not responding to the pot setting).

[Eb7+9]

good bits of info here ...

cells are balanced and hard limited at 250k - I'm strongly guessing to help ease with manufacture-ability of already haywire design ... making LFO work at slow speed with LED left disconnected and signal path responding well, makes sense (250k cells in parallel with 250k resistors gives typical max operating resistance levels) ... however - counting on max-dark cell resistance limits is somewhat risky and I imagine prone to creating considerable variation in the slowest-speed limit from one unit to the next

otherwise, this is almost exactly as the original vintage circuit parameters with the exception of the 220r speed limiting resistors and biasing ... as for the bias of the LFO (22k/4k7 dc divider) that's probably done to limit LFO signal amplitude at the DEPTH control pot (saving a resistor in the process) ... thus setting MAX depth

the problem then is with the current drive circuit, where many things can go wrong: any of the two jFET's, the devices in V-I converter, resistors, trimpots, the speed pot,  the external jack ... (not a walk in the park)

you might consider whether there's room to replace the single speed pot by a dual C100k instead and be done with it ...
might have to do some PCB cutting as required



note: if using a dual C100k pot then the 250k paralleling resistors are removed and 220r series limiting resistors replaced by 5k trimpots - for adjusting top speed ...

good luck

[VintageGear]

Fixed!!! I bought myself a component-tester for a few euro, and decided to pull all transistors and check them. Guess what, the transistor closest to the trimpot was kaput (reading as 2x diode?!) and a quick swap gave me back the lovely vibe sound!
Thanks for all the help Eb7+9; it was greatly appreciated to understand how the LFO circuit worked! Now.. time to rock..

[Eb7+9]

happy to hear you found the problem VG,

btw, I've heard over the years that the UV-1's response sucks // I didn't know why until you posted the schematic - looks like they copied the later Dynacord signal path ... for strongest effect the last phasing cap should go to the base of the follower device instead to the bias node (just one side of the cap needs to move) ...