[solved] Vari-Mid IIIB preamp - figuring out where voltage offset comes from

[frozenfrogz]

Dear all,

the small preamp in my Ibanez BTB-405QMTK bass guitar is giving me a hard time. The problem is, that I get a +9V offset on the output where it should be 0V.
The preamp circuit is a Vari-Mid 3B (IIIB), Ibanez and Meinl who are the official distributor and service center for Ibanez here in Germany refuse to send out schematics for this board.
They play the same f’d up game like Apple and Ersa do: No repair for you, just switch out the complete circuit board at your own cost.
The strange thing is, that the circuit itself is working without problems. Just the massive offset on the output is driving me nuts -> you can guess what happens if you plug it into an amp before turning down the main volume :/ The same is true for switching effect pedals with the bass connected of course.

The circuit is powered with 18VDC from two 9V batteries wired in series, however I am seeing more or less exactly half the supply voltage on the signal output (pin C1). This is true with everything wired in place (some voltage drop through the volume pot) and also with only the bare board connected to power.

Can anyone give me some pointers on where to start looking? Currently I am trying to RE the schematic, but it is quite a task with all the traces running under the SMD parts.

The SOT-23 transistor seems to be a KTC3875, marked ALY and checks out OK.
The two SO8 op-amps are TL062C.

Pin C5 is VCC (+18V)
Pin C4, C2 GND
Pin C3 Signal In
Pin C1 Signal Out

Every tiny pointer is highly appreciated!

[Armadillo]

could that be conductive?

Firstly, the two 9v batteries are forming a balance power supply. Are they the same voltage?

If above didn't work, then the simplest is to unsolder one by one the capacitors and check it off the board. Only 8 of them. No need to crack the brain. 

Those are the 2 most likely causes of voltage offset.

[frozenfrogz]

No, that is some residue from the isolation foam pad that was covering the components. It is definitely non-conductive. I checked the foam and even with the probes almost touching (less than 1mm on the surface between them) I do not get significant readings in the nanosiemens range of my Fluke 8050A.
But thanks anyway

Edit: Saw that you had further suggestions.

You mean the bigger tantalum caps I guess?

[Armadillo]

Only the coupling caps.
When it leak, offset problem occur.

[Armadillo]

You can unsolder the caps quickly and easily with 2 small size soldering iron like a Chinese chopsticks but using 2 hands to pick it.
The expensive method will be the desoldering tweezer of course. 

[Andy Watson]

The circuit is powered with 18VDC from two 9V batteries wired in series, however I am seeing more or less exactly half the supply voltage on the signal output (pin C1).

I would expect (though I could be wrong) half the supply voltage to appear at the output - because this would be the obvious reference point with the availability of a split-supply. My question would be why is your reference at -9v - i.e. the most negative of the two batteries?

Has the unit become faulty or has some "work" been recently performed on the unit?

[frozenfrogz]

OK I see, thanks for clarifying
I have an el-cheapo hot-air rework station and suitable pliers to do the job.

[floobydust]

I see a tantalum 10uF 16V output coupling capacitor from terminal "C1". I would suspect it is shorted or leaky.

The DC offset should change when you turn the level potentiometer, if it's coming through from the last op-amp. Otherwise things are wired wrong.

Running +18V through the output jack as an on/off switch is not the best move by Ibanez, because it sends live DC spikes to the output and the main amp, when you jack in. Ouch.

[Armadillo]

*** CORRECTION *****
Yeah, I know I made a mistake somewhere.

I double checked the battery, it is in-fact in series and not balance power supply. So it is using the tantalum capacitor and resistor to divide the 18volt by 2 to form balance power supply.

Please check the tantalum caps and resistors [along C1 to C5] first before you check the coupling caps. For the resistors along C1 to C5, just use ohms to measure and make sure its not opened.

This supposed "balance power supply" should be the priority check. You can take the attached as reference.   

[frozenfrogz]

I see a tantalum 10uF 16V output coupling capacitor from terminal "C1". I would suspect it is shorted or leaky.

The DC offset should change when you turn the level potentiometer, if it's coming through from the last op-amp. Otherwise things are wired wrong.

Running +18V through the output jack as an on/off switch is not the best move by Ibanez, because it sends live DC spikes to the output and the main amp, when you jack in. Ouch.

After desoldering the cap my multimeter does not even want to take a guess (nF range) and the transistor checker is having a hard time, returning a value of 27pF.

Hooray We’re on to something!

Need to check the drawers for a suitable part...

Btw. DC offset does change on turning the volume pot. I checked all the wiring beforehand to eliminate that source of error.
I agree that switching through the jack/sleeve is not actually a nice thing to do. All active electronics in guitars are wired that way because musicians are lazy and it is more convenient to remember in what order to plug into the amp than to turn up to a gig and find the battery was drained because you forgot to switch of the electronics inside the guitar. However, I am using cables with mute switch on the input jack and this usually works fine for all the instruments I have.

[frozenfrogz]

Of course I did not have the right value SMD tantals, but found a salvaged 10uF 35V through hole part. Looks a little funky on that board
While I was at it I desoldered all the other 1u / 10u caps, they check out fine though. Next time I order some parts the orange blob will be replaced with an SMD part.

Guess what: Offset is gone now! 

Thanks to all of you! Not only does my beloved bass work flawlessly again, but I also learned a lot tonight from reading up on keywords and concepts you mentioned.

Wishing you all the best,
Frederik

[frozenfrogz]

Guess what, the other 10u/16V cap died just days ago.
Seems like EOL-timer is pretty much "programmed" into them xD
I changed out all three 10u/16V caps for 10u/20V Kemet tantalums - hope this will keep things quiet for a while.

I should remember to not bother and simply replace all similar caps if one already failed in the future, just like with electrolytics on old hifi gear.