Twin face puzzle

[comradehoser]

On a first fire-up building streak, and making this one for my brother in law, the streak abruptly ended

Built everything to spec as far as i know except the 20uf caps (used 22uf). Have a barely audible signal. The issue i think is the DC is blocked. Only have fractional voltages after the 5817 (including to pins 1 and 8 of the charge pump, the resistors after D1, and all transistor legs) which i noticed immediately while going to trim the transistor bias before fire up.

Reflowed all points, rechecked wiring, continuity (including socket to board), components and cap orientations. Subbed the 1044scpa, tried different transistors, same deal.

At this point, I'm kind of stumped as nothing is coming up for me in the components or soldering. Bad trace after the schottky is all i can think of.

But maybe you all can see something i can't.

Gutshots

 

[music6000]

On a first fire-up building streak, and making this one for my brother in law, the streak abruptly ended

Built everything to spec as far as i know except the 20uf caps (used 22uf). Have a barely audible signal. The issue i think is the DC is blocked. Only have fractional voltages after the 5817 (including to pins 1 and 8 of the charge pump, the resistors after D1, and all transistor legs) which i noticed immediately while going to trim the transistor bias before fire up.

Reflowed all points, rechecked wiring, continuity (including socket to board), components and cap orientations. Subbed the 1044scpa, tried different transistors, same deal.

At this point, I'm kind of stumped as nothing is coming up for me in the components or soldering. Bad trace after the schottky is all i can think of.

But maybe you all can see something i can't.

Gutshots

When you biased the trimmer, did you switch the toggle for each circuit?

 

[comradehoser]

I did. They are both equally dead. Ran the full range of the trimmers while testing, and can't get anything but the teeny-tiniest signal through the entire range

 

[Coda]

Test the power at the 9v jack and where it connects to the pcb, before the diode. Also, measure the collector voltage of the q2’s while adjusting the trimpots. Does the voltage change?…

 

[comradehoser]

No change. It's why I noticed the problem, because i was trying to bias before plugging in the signal.

Sorry, i forgot that part of the troubleshoot measures. 9.4v DC all the way into the board, and on the diode.

I guess i will try to bridge the point between the diode cathode/stripe and the next component. (learned that one from Music6000)

[edit for report:] On the npn side, I have an amplified fuzzy signal (!) when bridging the diode cathode to either of what is I assume are the shared ends of R5 or R7. Nothing on the pnp side, either when bridging to any part of R1/R3, or to either or both of pin 1 or 8 on the charge pump, although I didn't sweep the trimpots, just left them both maxed--is it a very precise "sweet spot" in the trim?

Very puzzling to me.

SCHEMATIC

 

[Coda]

Is the diode oriented correctly? It looked correct in the photo, but I would double-check just incase. Kinda sounds like what would happen if it was in backwards...

Also, what are the hfe measurements for the transistors?...

 

[comradehoser]

That was my immediate first thought, but yeah, the diode is installed as indicated.

If you can tell me how to measure the hfe, I'll do it, but I don't have a transistor measurement device, just a multi-meter. In any case, they are the specced transistors--AC128s and the BC108s. I don't think that's the issue, though, since the BC108's work with the jumper from the 5817 cathode to R5 or R7. How optimally they work, I don't know, but I do have a loud fuzz-facey signal.

The ac128s remain inoperative

 

[comradehoser]

Soooo... just trash the board?

 

[szukalski]

Where did you get those AC128s? Not from a Chinese parts seller? They look identical to the ones offered on ebay from places which have had examples of counterfeit chips.

Not saying that's the case, but it's my first thought.

 

[Coda]

Put your multimeter in DC mode, and place the positive lead on the lecture oc the collector of Ge Q2. The negative lead goes to any ground point. Can you tell us the voltage it reads with the pedal turned on?…

 

[szukalski]

That was my immediate first thought, but yeah, the diode is installed as indicated.

If you can tell me how to measure the hfe, I'll do it, but I don't have a transistor measurement device, just a multi-meter. In any case, they are the specced transistors--AC128s and the BC108s. I don't think that's the issue, though, since the BC108's work with the jumper from the 5817 cathode to R5 or R7. How optimally they work, I don't know, but I do have a loud fuzz-facey signal.

The ac128s remain inoperative

When you say they're specced.. you've measured them and they are in spec? Or they have the right letters on the outside of the package?

 

[comradehoser]

Coda: yes, absolutely.

Szukalski: specced as in, installed what was required according to the letters on the outside of the package. I'm pretty sure I ordered them from Banzai music, but who knows--could be they are supplied by bogus chinese manufacturers.

I also had the NOS 2n169 Ge black top hat transistors that are NPN, recommended in another discussion somebody, and put them in first without the charge pump, but nothing doing.

Will measure and report back ASAP this evening.

 

[comradehoser]

The measures plugged in, grounded, and switch-selected:

NPN 1: 0, NPN 2: 0, PNP 1: 0, PNP 2: 0

The measures plugged in, grounded, switch-selected, and jumpering from cathode of diode to R3 for PNP and R7 for NPN:

BC108 NPN1: 1.265 when NPN 2 trimmed to 4.558, distorted amplified signal at r7, not r5.
"AC128" PNP1: 4.8 when PNP 2 trimmed to 5.0, no sound.

THEN THE PLOT THICKENS.

Decided to pull the "AC128s" and the charge pump and put in the 2N169s. Just for the hell of it. Aaaaand, we have distorted PNP-side amplified sound (even though they are Ge NPN) when jumpering to R1 or R3. V reading of 0 without jumper, and no sound, obviously.

2N169 PNP1: 0.914 when PNP2 is biased to 4.407

There is something incorrect in the board traces, right? Could it be a switch issue? I triple checked the diode orientation on the board after pulling the charge pump. Should I ask Banzai for a refund, or at least alert them to a possible problem?

 

[music6000]

Did you check to see if the toggle switch is shorting out on the pins underneath the PCB due too to much Solder?

 

[comradehoser]

Yes, if anything i was stingy with the solder. No visible bridges and the continuity checks out.

 

[comradehoser]

Vf on the diode is 0.107. I can't find a continuity reading or anything around that from the stripe/cathode end of the diode to any other point on the board or the switch. So... bad traces?

 

[comradehoser]

Just jumper it and call it a day? Kind of unsure where to look for the fault.

 

[comradehoser]

Update:

This won't be marked solved and we'll have to pour one out for the poor little dude.

Went in to make the jumpers permanent per observed behaviors above. Reflowed everything, added a bit of solder to low spots and the 4pdt switch. The pedal just quit working entirely. Jumpered several different ways, and was getting 9.3 volts straight up in the transistors with nothing doing on the trimpots, and no signal except 9VDC whine. Decided to desolder the 4pdt and the charge pump socket to see if there was some bridging going on, ended up destroying both and pulling pads to boot (also my first time doing that ever).

Such a weird experience--just very odd with the inconsistency of the behavior in troubleshooting, and the board itself seemed more fragile than I've dealt with before (now something like 60 pedals in, I think). Off the first solder, the diode trace to the charge pump pin was DOA.

It's my very first and only time binning a board, and I've built some complex ones.

Thanks for the help folks, this one is going down as a loss.

RIP Twin Face.

Hopefully I will get to build you again. { :,^( ,,,, ,, ,,,

 

[comradehoser]

Update: ordered a new board, moved everything over, and after a lot of more effing hard work, I have a functional-ish pedal.

Got the board, and tested all pads for stated continuity, all was good. Transferred everything from the old board with a new switch and diode, since the other components seemed to be in decent condition. No love.

Worked on the NPN side first since I had two spare 2n169 transistors that I knew worked and swapped in. Nothing. Measured voltage and getting 9v straight into the NPN collectors and 0v in the PNP. Focusing on the NPN, I swapped trim pot, and the two resistors first in line to the 9v in (R5 and R7) resistors since voltage modulation didn't really seem to be happening. Tried the old BC108s, nothing. Suspected the sockets and soldered the spare 2N169s back in, and ta-daa, had fuzz. Reinstalled new sockets, and it still worked.

After a long time and subbing in a new charge pump, I got the PNP side to work with the same subs (R1 and R3 resistors). The 2N169 NPNs didn't work with charge pump removed, so I'm guessing there's a connection to make even though the schematic and continuity check seems to indicate that the 9V should flow regardless of the 1044SCPA being there or not

So, somehow I killed both sets of my NPN transistors (BC108 and 2N169) in my jumper noodling, and probably killed resistors somehow even though they looked fine, and the trimpots and sockets were probably wrecked in the desoldering.

Love the 2N169s. The fake AC128 PNPs actually worked, but they sound like total ass.

Now I just have to find some PNP and silicon NPN combinations that I actually like before I box this up and send it to my brother in law.

For a "simple" build, this one is effing aggravating.

 

[STEVE]

I’m having the exact same issues with my two twin face boards , Npn works sounds fuzzy but flick the switch and nothing happens