BMP Creamy Dreamer problems (Aion Electronics PCB)

[doof]

Hi everyone, first time posting here.  I've been building pedals for a while, so i'm not completely new to this, but lately i've bought a few of the Halo Fuzz (BMP) boards from Aion Electronics to try some builds.  The Green Russian version worked perfectly the first time, but then i tried a Creamy Dreamer, and it was perfect for a couple days, then started to sputter really bad and sound weak.  i found a couple of lifted traces from when i had to desolder a bunch of capacitors that i found out too late were waaaay out of spec.  I fixed those traces but still no luck, so i thought maybe i just didn't do it quite right.

I decided to rebuild it from scratch with a new board rather than try troublshooting something so iffy.  Again, it worked perfectly for about 2 days, and now today it's sputtering out and getting really weak. 

I methodically tested every single component before soldering it into the board, and didn't make any value substitutions.  I used 1% metal film resistors, and all caps were within +/- 5% of the values listen on their build documentation - https://www.dropbox.com/s/h2oqzwwf1ue4seo/aion-halo-bmp-documentation-v1.5.pdf?dl=1.  I also confirmed the pinouts of each transistor before placing them.

Voltages at the transistors are:

Q1
C: 3.606
B: 0.588
E: 0.002

Q2
C: 3.578
B: 0.585
E: 0.000

Q3
C: 0.028
B: 0.614
E: 0.000

Q4
C: 3.846
B: 1.832
E: 1.24

Seems like Q3 is the big glaring problem.  Checking the schematic, seems like R17 and R15 bias that transistor,  +9v to R17 has good continuity, and voltage at R17 is 9.45v.  Immediately after R17 it drops to 0.028.  There is good continuity between R17 and R15.  Voltage on that side of R15 is 0.028, and voltage on the other side of R15 is 0.622, which is basically what is expected since that connects to the base of Q3. 

Both R15 and R17 measure at the correct ohms.  This second build went extremely smoothly, with nothing getting desoldered.  all the solder joints look good and i reflowed them all just for good measure. 

Anybody have any guesses as to why i'd get the exact same problem on two separate builds?  Thanks in advance for any help you can provide!

[PRR]

Welcome.

> that side of R15 is 0.028, and voltage on the other side of R15 is 0.622, which is basically what is expected since that connects to the base of Q3.

Yeah but... Base does not "make" voltage. You have to feed current AT the Base, then it tends to sit at 0.6V as you say.

Where is that current coming from?

It won't flow uphill through R15 from Collector.

That kinda leads to leaky caps C7 C8 C9. I'd be inclined to lift one leg of each and see if the DC changes.

BUT you have 9V across R17 22K, and zero across R16 820r. You would think, if Q3 were conducting hard, R16 would show about 0.3V. Zero suggests Q3 is not conducting C-E at all.

It's not yielding to my poor logic. I hate to shot-gun new parts at it, and suspect a short or open rather than a bad part. Maybe in process of starting a shot-gun attack you will find a connection fault.

[doof]

Thanks for your response!  As you can tell, even though i've built quite a few pedals, i still have a lot to learn about how the circuits actually work.  i'll have to try what you suggested regarding those caps.  i'm just surprised i'm getting the exact same results from two different builds.  On the first one i used ceramic disc caps and on this second one i used polyester film box caps.  Very frustrating!  i'm tempted to do a build on veroboard and see what happens. 

[robthequiet]

The problem occurred on the second board, too. That's weird. Then it takes time to develop. Does it get better if the pedal stays off for a day or does it just deteriorate from the start and not recover?

Teaching myself to fish here... 

Tried wiggling the components a bit? I've had a transistor with a cracked leg.

[doof]

it does not recover, just stays bad.  i have tried wiggling, but everything is solid and the problem remains the same. 

[robthequiet]

OK, noted. Paul probably gave the best advice as far as the shotgun approach. Second the motion on leaky caps. A cap gets reset during assembly then can't dump charge or something, so it degrades over time. You could test this by placing a resistor across each cap to see if it restores at all.

[doof]

Ok, i think you boys solved the problem.  i connected one leg of C7 to ground which silenced the pedal, but once i let go, it immediately came back to life sounding much healthier for a few seconds before degrading again.  C8 and C9 didn't really have a noticeable effect.  So i have replaced C7 and so far it's sounding good and Q3 voltages are substantially similar to Q1 and Q2 now.  I can't thank you guys enough!  I owe each of you a coke

hopefully it stays good for longer than a couple days!

[robthequiet]

Good catch! Don't know how many times we've gone around in circles in a circuit until we've no hair left from scratching. Enjoy! h/t Paul, cheers.