Measuring for DC offset: the saga continues

[onboard]

I think there's an unwanted DC offset in this circuit that's causing big switching pop, but I don't even know how to measure for it.

Here's the schem...I've since changed R5 to 22k and R6 to 1.5k. Now the jfet channel is popping huge when switched, and the gain wasn't increased by that much. Now I'm thinking it's because the jfet input isn't capacitively coupled.

[KarbonHed]

It might be the effect or whatever following your unit. Try a 1M pull down resistor on the output of SW1b. Possibly one on the input side too.

[onboard]

Unfortunately, the pedal was going straight to the amp    Otherwise good advice, I think.

There's a 1M pulldown resistor at the input and ouput of each channel, so my guess is there is a problem with the jfet channel's circuit itself. I only mentioned capacitively coupling the jfet's input because it seems to me that it came up in another thread discussing emulations of different amp input stages.

Neither the Fetzer Valve or Don Tillman's preamp include a cap up front, but the poster mentioned keeping DC from appearing at the jfet's output. Actually, shouldn't the +10uf cap at C7 keep DC off the output?

My breadboard bypass switch is terrible, so any time I get popping there it get's chalked up to the cheap-o toggle. But now this is built and in a box with a 3PDT and the pop is huge.

[R.G.]

To measure for DC offset:
1) get out your digital multimeter (DMM)
2) turn it on, set it to DC 2V or 20V scale to get a good reading
3) connect the black lead to the circuit's signal ground somewhere
4) probe all lugs of SW1 and write down the resulting voltages
5) post them back here.

It's possible you have an electrolytic backwards - that always leaks voltage.

[onboard]

Here's the entire switching arrangement. I'm using 3PDT's and there's LED supply voltage on the extra lugs.


SW1a & b lugs are all reading 0.0 ~ 0.1mV at rest.

SW2a & b are reading -01.6mV at in/out poles and bypass throws, 0.0mV at the circuit pole.

I measured SW1b pole while switching. +/-2.2mV when bypassing the opamp channel, +/-15mV when bypassing the JFET channel. Those are peak numbers, the meter settles back to 0.0 ~ 0.1 in a second or two.

If I accidentaly overheated the 3PDT's, would LED supply voltage be interfereing?

Hopefully I'm presenting this clearly, thank you for helping.

[Joe Kramer]

Hi!

That JFET input with no cap is giving you the pop.  It would be all right without the cap if you didn't have a switch, but as it is, every time you leave that loose end hanging in space, a DC charge builds up and then discharges when you switch back again.  Put  your favorite flavor cap on there and hang another 1M resistor on the input side and the pop should stop.

Joe

[onboard]

the pop should stop

There's a slogan in there!  A call to arms - "Stop the POP!"  

I thought the problem could be the open pole opposite the LED  power supply pole, but grounding that didn't stop the pop. Good practice to ground that anyway?

Gazing around in the box, it looked like the lugs on the jfet channel's bright switch (mounted directly under the board) could be touching a power rail - I got excited. Removed the spacer from the switch shank and no luck.

I'm sure rooting for the missing jfet input cap solution.

edit really ought to use the preview button

[onboard]

oops, double submit...

[onboard]

Joe, you were right. I added an input cap to the jfet channel and moved the pulldown resistor in front of it to keep my input Z high. Almost all of the pop is gone. I was guessing in that direction, but a second opinion nudged me to do it. Thanks man!

So for anyone else interested in multi-channel preamps that include a jfet type emu - Fat Bastard, Fetzer Valve, Umble, Thunderchief, Alembic Stratoblastor, Uno for example - add that missing input cap after the pulldown resistor.

I should change this thread title... but I still would like to learn how to go a through a circuit front to back and eliminate any unwanted dc offset...

[Joe Kramer]

I added an input cap to the jfet channel and moved the pulldown resistor in front of it to keep my input Z high. Almost all of the pop is gone. I was guessing in that direction, but a second opinion nudged me to do it. Thanks man!

So for anyone else interested in multi-channel preamps that include a jfet type emu - Fat Bastard, Fetzer Valve, Umble, Thunderchief, Alembic Stratoblastor, Uno for example - add that missing input cap after the pulldown resistor.

Great!  Glad it worked!  One caveat to doing this on circuits other than the one you posted: sometimes the transistor's gate (or base) is getting a bias current from the same resistor that's setting the impedance--it would look just the same as the circuit above, but the 1M would be coming from the positive supply instead of ground.  In that case. you'd need to put the cap on the input side, not the transistor side, of that resistor, and then add  another resistor on the input side of the cap.  Otherwise the cap will block the bias current, and you'll have to call the plumber or something. . .

Joe