Troubleshooting Madbean Glam chorus: Very high voltages observed?

[D.C.]

I've been trying to build a Madbean Glam chorus circuit using this layout from TagboardEffects, and I'm running into some issues. (For reference, see also this nearly-identical post I wrote on Reddit.)

First of all, subtitutions: I didn't have 2N5457, so I used a BF244A, based on this thread here on DIYStompboxes. The pinout differs from the layout, which I accounted for by twisting the G and S pins (see the datasheet here). Also, my 78L05 is this component from Tayda: L78L05. (Not sure if that "L" at the beginning matters.) All other parts are as listed.

According to Madbean's instructions, you're supposed to adjust the trimpot until the JFET's drain measures 6 V. (See the last page for the actual schematic.) Very strange things happened the first two times I attempted this, but then it began to behave more normally.

When I first applied power to the circuit with a 9 V battery, the measured drain voltage was about 35 V. (Yes, thirty-five.) When I adjusted the trimpot, it changed from maybe 31 to 37 V. After removing power, I watched the voltage gradually decrease towards zero. I then fiddled with some connections and reapplied power, and the measured voltage was the same. After adjusting the trimpot a bit (and possibly bumping some wires in the process -- not really sure...), the voltage soared. Into the hundreds of volts, peaking in the 300-400 V range!! Again, I disconnected power, and I watched the drain voltage drop to 200... 100... 10... etc.

You may be thinking: "Are you sure it wasn't millivolts?" I'm pretty certain I wasn't misreading my multimeter. I've had it for a few years and am pretty familiar with it. And I definitely watched the voltage rise and had to adjust the range upwards.

Given the large voltages, I assumed the board was fried, but a few minutes later, I decided to just give it one last try. Sure enough ... the circuit started acting normally. As I adjusted the trimpot, the drain voltage was in the neighborhood of 5 to 7 V.

Before I continue, I have to ask: Is it really possible that the voltage was in the hundreds of volts? If so, those values exceed the voltage ratings of probably every component on the board. I'm afraid the board is toast.

If the board is, in fact, toast, these voltage measurements may be useless, but here they are anyway. (See second-to-last page of the Madbean document for expected values. Discrepancies of note: PT2399 pins 5 and 12; pins S and G of BF244A.)

Code Select Expand

    PT2399
    1 4.91
    2 2.45
    3 0.00
    4 0.00
    5 4.84
    6 2.87
    7
    8
    9 2.45
    10 2.45
    11 2.46
    12 3.63
    13 2.45
    14 2.45
    15 2.45
    16 2.45

    TL062
    1
    2
    3
    4 0.00
    5
    6
    7
    8 8.67

    78L05
    I 8.37
    G 0.00
    O 4.92

    BF244A
    D 5.45
    S 4.66
    G 1.71

[GibsonGM]

Hi D.C., and welcome. 

No, those high voltages aren't possible, at least in the universe I live in.  AFAIK, the only way you could see something like that would involve an inductor dumping its magnetic field rapidly, and there isn't one on the parts list.  So, I would chalk that up to "what the hey, that was weird!"   Also, and inductor wouldn't be able to sustain high voltages - they dump their charge instantaneously (usually).

I haven't used the PT2399 so can't help you there, but someone should come along shortly who will know how to interpret your voltages! 

Don't panic, we'll help you sort it out! 

[PRR]

Change the battery in your meter. Please. We have seen such "impossible strange" displays when the battery is on the last gasp, the internal references are gone wacko in different directions, or something. (Maybe the CMOS forgets how to count.)

Or mV, but you knew that.

Very remotely possible you lost your Ground clip, and 400V is sneaking over from your tube amp. (There should not be 400V in the walls of any house or small shop I know.)

[duck_arse]

also, you are not allowed empty opamp pins, we want all the voltages please. even if they are 200V and up.

[vigilante397]

Change the battery in your meter.

+1 This is a thing.

[D.C.]

We have seen such "impossible strange" displays when the battery is on the last gasp

That was likely it. Honestly, it's been a long time since I've replaced my batteries. I figured the display would dim, or just not turn on at all; I didn't imagine it would generate crazy measurements. Good to know.

you are not allowed empty opamp pins

That's rather aggressive..... Anyway, they were just quick and dirty measurements. At first I wasn't even going to post them at all. I felt there was a high chance the measurements were worthless, which turned out to be not far from the truth, given what I've learned about my multimeter.

I've since gone back and re-measured the voltages. Turns out the voltages were actually pretty reliable, once the 300 V nonsense died out. Here's a complete table:

Code Select Expand

PT2399     Expected   Actual
    1      5.03       4.92
    2      2.51       2.45
    3      0          0.00
    4      0          0.00
    5      2.64       4.81
    6      2.44       2.81
    7      n/a        1.84
    8      n/a        0.68
    9      2.51       2.45
    10     2.51       2.45
    11     2.51       2.46
    12     2.51       3.64
    13     2.51       2.45
    14     2.51       2.45
    15     2.51       2.45
    16     2.51       2.45
           
TL062      Expected   Actual           
    1      n/a        oscillating (4.2 to 4.9)
    2      n/a        oscillating (4.2 to 4.9)
    3      n/a        oscillating (4.2 to 4.9)
    4      0          0.00
    5      n/a        oscillating (4.2 to 4.9)
    6      n/a        oscillating (4.2 to 4.9)
    7      n/a        oscillating (2.5 to 6.5)
    8      9.13       8.93
           
JFET       Expected   Actual
    D      5.94       5.76
    S      2.9        4.63
    G      0          1.72
           
78L05      Expected   Actual           
    I      9.13       8.78
    G      0          0.00
    O      5.03       4.92

[D.C.]

Now, my current problem: the circuit passes the clean signal, but with somewhat reduced volume (maybe 1/3 less than bypassed).

I tried swapping the PT2399 with a different one. Same result. I also changed the BF244A with an actual 2N5457. I had to adjust the trimpot to get the 6 V bias, but again, the result was the same.

I am suspicious that I may have briefly accidentally shorted out a couple of the 2N5457 pins by trying to measure with a probe. Could I have blown the JFET?

A couple of the pins have suspicious measurements, compared to the expected values. I've done some cursory examination around the pins with the most suspicious measurements, but I don't see anything obviously wrong. I thought I may have found a solder bridge, but I think that was a false alarm.

It's strange that I'm having so much trouble. No one else in the TagboardEffects comments seemed to have major issues like this. Wondering if I should junk this board and start from scratch...

[D.C.]

Sorry for so many multiple posts. Just discovered something new.

I swapped in a new BF244A and PT2399. Pin 12, whose voltage disagreed with the expected value earlier, is now as expected... but pin 5 is still bad, and pins 9 and 10 (which were good before) are now bad.

I actually noticed this earlier while I was swapping out JFETs and PT2399s, but when I checked it again, the numbers were as I reported initially, so I didn't mention it. Now that I've seen it again, I know this is really happening.

Code Select Expand

PT2399     Expected   Before   After
    1      5.03       4.92    4.92
    2      2.51       2.45    2.45
    3      0          0.00    0.00
    4      0          0.00    0.00
    5      2.64       4.81    4.91
    6      2.44       2.81    2.38
    7      n/a        1.84    0.00
    8      n/a        0.68    4.51
    9      2.51       2.45    0.13
    10     2.51       2.45    4.16
    11     2.51       2.46    2.45
    12     2.51       3.64    2.45
    13     2.51       2.45    2.45
    14     2.51       2.45    2.45
    15     2.51       2.45    2.45
    16     2.51       2.45    2.45
           
JFET       Expected   Before   After
    D      5.94       5.76    5.84
    S      2.9        4.63    4.30
    G      0          1.72    1.66


[PRR]

> noticed this earlier ..., but when I checked it again, ... I've seen it again

Different numbers with "no other change" is bad joints.

Same as the danged light in my bathroom. It is on, it is off, touching it may or may not get it on again. Some loose connection inside.

[D.C.]

Different numbers with "no other change" is bad joints.

You seem to be under the impression that I took a measurement, made no change, took the measurement again, and got different numbers. That is not the case. I've been swapping out various JFETs and PT2399s and getting various different measurements along the way. Still, point taken: perhaps I have some bad soldering.

In any case, I've discovered a couple rookie mistakes, partially thanks to Reddit user midwayfair. I had an extra cut on my veroboard, so a whole section of my circuit wasn't getting ground. I also had a couple misplaced capacitors that were shifted by one row.

After fixing both of those mistakes, pin 5 is now oscillating around 4.5 V or so, instead of the fixed 2.64 V that it's supposed to be. I'm afraid all of these mistakes may have toasted all of my PT2399 chips. Maybe I'll order some replacements; Tayda has a sale going at the moment...