Debugging help: Swollen Pickle (Original version)

[splice]

Hey all,

I recently tried to build a clone of the Way Huge Swollen Pickle, the original version without the scoop and crunch pots. The vero layout I used is at http://tagboardeffects.blogspot.ca/2014/06/way-huge-swollen-pickle.html and the schematic I think it's based on is at http://www.bigmuffpage.com/images/schematics/Swollen%20Pickle%20Schematic.jpg . I used 4x 2n3904 transistors instead of the resistor array, which should be pretty much the same thing. I used 470k resistors since I didn't have 475k ones, I measured those I had and put the highest values into the board (may have been about 471-472k). The vero layout omits the D5, C14 and R27 power filtering on the schematic and instead uses a 100uF cap, which is what I built. It also omits the link from Tone 3 to the base of Q1 on the schematic (Q4 on vero).
I thin it also reverses the diodes and cap on the second clipping stage. The schematic shows the signal going into the D1/D2 diodes, then C7 cap, then out, but the vero has it going to the cap, then diodes, then out. My LTSpice sim seems to show the order of the diode and cap there shouldn't make a difference in the signal.

I've had a few issues but where I'm at right now is that the sound isn't quite right. It sounds like the notes cut off midway, and higher pitched notes often don't get fuzzed at all. Playing softly makes it worse, and rolling the sustain pot back as well.

I originally built it and forgot the 10k resistor at the top right of the vero, R6 in the schematic. Not sure what effect this could have had. So far, I replaced all the electro caps and also replaced Q2, Q3 and Q4 on the schematic (annoyingly, Q1, Q2 and Q3 on the vero layout).

My measured voltages are (going by the schematic, invert for the vero numbering):
Q1: C 3.3V, B 1.3V, E 0.7V
Q2: C 2.9V, B 0.69V, E 0.037V
Q3: C 2.9V, B 0.69V, E 0.04V
Q4: C 4.6V, B 0.67V, E 0.029V
D1 anode/D2 cathode: 2.9V
D2 cathode/D2 anode: 2.6V
D3 anode/D4 cathode: 2.9V
D3 cathode/D4 anode: 2.6V

9V rail/battery: 8.78V

I'm pretty much a beginner in electronics, I know some stuff but it's mostly a jumble. The voltages don't look immediately wrong to me, but... I built up the schematic in LTSpice IV and looked at the voltages I should be getting. They're a bit off. The simulated voltages (off a 9V DC source with no input voltage) should apparently be:

Q1: C 3.77V, B 1.42V, E 0.79V
Q2: C 4.12V, B 0.65V, E 0.027V
Q3: C 4.24V, B 0.66V, E 0.031V
Q4: C 4.24V, B 0.66V, E 0.031V

This sim is with all pots set at noon. I built the vero and sim with a 25k linear tone/filter pot, although the schematic (and I suspect the correct value) is 100k linear for tone/filter. I had a 100k pot lying around and tested my board with it but the problem didn't seem to go away, I just had more variance in the tone.

I have an audio probe built up and can use it for troubleshooting, but the results so far haven't led me to any valuable insight. I see the voltages are off from the sim, but I'm not sure what could be the issue. I re-verified all resistor and cap values and they seem right, all links and cuts are as well. I re-re-re-verified the orientation of transistors against schematic, vero and datasheet, and they're right. I have an hFE tester on my meter, all transistors were tested at between 145-160hFE (@ 1.5V, 10mA). Continuity testing with my meter didn't reveal any obvious shorts. I can't spot any obvious cold solder joints or cuts in the tracks.

Could anyone let me know if I missed something obvious or what I might do to troubleshoot further/perhaps fix whatever's wrong with it? I've got some 5 pedals waiting to be built off PCBs, parts are coming in, so if I can't figure this one out soon I'll probably just take it apart entirely and try another schematic/vero some other time. I'd love to get it working though, considering the time I already spent on it... Thanks.

[duck_arse]

hello and welcome, splice.

with regard to the circuit diagram, Q2, 3, and 4 all have the same value resistors at emitter, collector, C-B, and B-gnd. allowing for the slight variation in your 470k's, the b, e, and c volts on those three should all be the same, very nearly. your Q4 volts match ltspice, so Q2 and Q3 are way off. if you are using sockets, swap Q4 with Q3, then Q2, and see what volts result. if still off, measure all resistor values around Q2 and 3.

and leaving off R4 would mean no output signal, no current to the transistor collector. do you have any connection between C3 and the Q1 base resistors? can you show photos of what you built?

[splice]

I didn't socket my transistors (likely will in the future), but I did unsolder Q2 and Q3 and swap them for 2 other tested 2n3904, with the same result. I'm fairly confident the transistors are fine right now.

The more I look at it, the more the identical voltages for Q2 and Q3 seem strange to me. If a part failed or I made a mistake, I'd expect one of them to be off, not both. That leads me to believe that despite my double and triple-checking, I must have missed something obvious and made some mistake in resistors I put in those stages. I'll take a look this week, remove all resistors from one stage and test them against the schematic values.

I'll take a pic before I do that and check for a connection between C3 and Q1 base. Will report back, thanks.

[splice]

Here's what my board looks like. Colours still look right to me on the clipping stage resistors, but what else could it be that would affect both clipping stages' collector voltage in the same way?

Notes on the board: the ceramic dipped caps have one leg bent back under them since they span two adjacent rows, even if at first sight it looks like they skip a row because of their size. They all read 471 (470p). The film cap which reflects the flash reads 332 which is 3n3. The black marks indicate a cut on the other side of the vero.

The back looks cleaner live, the flash makes it look pretty poor. I've measured for continuity and there doesn't seem to be any bridging, but I can retest easily enough.

Front: http://i.imgur.com/uNsaMd4.jpg
Back: http://i.imgur.com/8keXBG7.jpg

I'm about to pull the resistors on the lower left clipping stage to measure them, even if the colours look right. Maybe the exact measurements will give me a clue.

[duck_arse]

well, I went over your topside, and since that 15k has one leg folded under, it all looks right. underneath, you should maybe run a sharp knife point along between coppers, just to pick-up/off any micro-dags. and have a real, REAL close look at the edges of your track cuts. I always do the drill-cut first, then go back with the exacto and maggy, and trimm the sharp points off the sides of the cuts, so they end up flat/straight.

[jobrien781]

I built versions from tag board effects, the ic version and transistor version. Built them exact to the layout. They both sound the same but the notes get chocked out I would say half as soon as it should. I double checked my placement of links cuts and components and all seems in order. I haven't seen any corrections for the layout. Maybe a componet value in the layout is wrong? Just a guess. Anyone out there go through this and found a fix? Thanks for the help.

[duck_arse]

the procedure:
http://www.diystompboxes.com/smfforum/index.php?topic=29816.0

if you follow those instructions on both builds, you'll either findd/fixxe the problem yourself, or have all the info you'll need to post here again. that's when the experts will stop in to help.

[splice]

I'm really curious about this given that it's pretty much the exact same issue I have (the high pitched whine turned out to be my power supply, replaced with battery, much cleaner but notes get cut off, sustain is whacky). There's one guy claiming that it's working perfectly for him and all the knobs work as they should, but somehow he missed that the tone/filter knob is listed at 25k when it should be 100k. Makes me doubt his claim that everything is fine...

At any rate I've pretty much given up on my build. I've built 4-5 pedals since with no issue. I'm starting to doubt the layout, and if I try again, I'll likely go for another layout instead.

[jobrien781]

Ok. I got it! Change the filter pot to 100k linear! But otherwise followed the vero exact. Sounds fantastic! Consider it verified! Thanks for the direction on trouble shooting. I'm new to the pedal building world. Off to make another!

[aron]

What was the pot value before? Was it 25K?

[splice]

Yes, it was 25k. I reported this on tagboardeffects over a month ago now, you can see my post on the swollen pickle page. Strange how no one picked up or commented on my report of the pot value then. Oh well, jobrien781/Jim took my advice, posted that it worked and got credit for the correction at TBE.

My build's still faulty so I'm going to take it apart, but it's at least nice to know that my concern about the 25k filter pot was well founded and the reason no one made any comment on it wasn't that I was obviously wrong.

[jobrien781]

I went back through the comments on TBE and saw your post about the 100k pot. But not till I saw it on the schematic. The 25k pot made the voltages quite different compared to the 100k. That is where I learned something, that the pot values can greatly effect how the circuit works. I wasn't quite sure if I was going to be able to get it to work myself, but was happy I got it with some help from you guys. This was the first pedal I checked voltages. It was a good learning experience. And I learn something from each pedal I make. If you need I can can post a picture. I don't know if it makes a difference but I also used monolithic caps for the three 470p caps. They blend in with the background. Almost missed them. I hope you don't take your pedal apart Splice. It awesome once it's up and running.