Massive thunk when putting input jack in.

[Kesh]

I've built many pedals, but suddenly I've one that makes a very loud pop/thump when the input jack is inserted. Normally it's there, but not at this crazy level.

It's the standard 3PDT true bypass circuit. Obviously this connects the battery negative to ground when I insert the jack.

And it makes no difference if the effect is engaged or on bypass.

Any idea what's going on, and any fixes?

[R.G.]

Two possibilities:
- the output of this pedal/circuit is very effective in transferring its output capacitor's charging current to the signal line, perhaps as a result of an oddity of the pull down resistor, if any
- bad/reversed (electrolytic) output cap

The more I deal with input jack power switching, the more I don't like it. Not that going to another solution would fix this. I just had to vent. 

[Kesh]

Thanks, but the thing is, it does it when the board output is not connected, when the effect is bypassed. So it can't be putting the output cap to signal. And it's a 1u poly.

It's actually worse when the effect is bypassed, so perhaps input to ground interacting with battery negative being attached is doing something odd.

Circuit is here but it doesn't show switching or LED. That all follows the normal thing of grounding input when effect is bypassed and illuminating the LED by connecting it to ground.



Maybe I have a really aggressive battery

[Kesh]

Actually, the more I play with it, I think it's actually the jack tip hitting signal in that causes it, rather than sleeve connecting the battery's neg.

[Pyr0]

I wouldn't worry about it too much.  It's not like anyone would normally unplug the input jack while playing 

[Kesh]

OK, I figured out what is wrong, but not sure about solution.

For a moment the input jack's pin is connecting the negative battery to signal in, which when the effect is bypassed means negative battery is going to signal out. For that moment the ground is not yet connected to negative battery, but is, through the circuit, connected to POSITIVE battery. The ground is at about +7V for that moment. So you are getting about -7V across the output.

I wonder if it's because the circuit is such a large load for a pedal at 60mA. Then the board ground can swing very positive when not grounded, as it were.

RG?

[R.G.]

Sounds like a reasonable guess at the situation. I suspect the pulling up to 7V is being done by the 7805. Just a guess.

I believe you could do much better by putting a PNP in the battery (+) line and pulling the base of the PNP down to ground through a resistor by inserting the mono plug and grounding the ring contact. I use this all the time for higher-current pedals and any place where I can afford the $0.05 or so it costs.

In this situation, the tip would momentarily contact the 10K or so to battery, and might or might not pull it hard enough to turn on the PNP through the inductance/resistance of the guitar pickups. But since the power and ground are intact from circuit to power source, the current would not be going through the ring contact and the circuit's input and output voltages would remain normal -- well, at least what they normally do in a power-up.

This trick is remarkably effective on curing a number of ills associated with the input-jack grounding power switching setup. I really don't know why it doesn't get used more often. Maybe it's just hard enough for people to wire a jack, much less adding another transistor and resistor.

[Kesh]

Thanks RG. Next high current pedal that solution is going to happen, but this pedal is being delivered right this minute. I will just have to tell him to plug the input jack first and unplug it last.

Luckily high current means it will mostly be run on external power, which stops this happening.

[R.G.]

... but this pedal is being delivered right this minute. I will just have to tell him to plug the input jack first and unplug it last.

Fixing issues in the documentation instead of the actual hardware or software is a time-honored computer industry practice.   

Microsoft didn't invent it, but they sure made letting customers do final test a lot more pervasive.    You're well ahead of that!

[Muthauzem]

RG.. Do you have more info on your proposed solution? Like a schematic or something?

It seems interesting.

[R.G.]

See: http://geofex.com/FX_images/PNP_power_switching.pdf

[electrosonic]

It's not the industry standard, but why not switch the power using the output jack? The input jack can use a switching jack that shorts the tip to ground when no jack is inserted.

Andrew.

[R.G.]

It's not the industry standard, but why not switch the power using the output jack? The input jack can use a switching jack that shorts the tip to ground when no jack is inserted.

I've suggested exactly that here and in other places so many times I give up. Notice that the jack in the diagram is not labeled input or output. Works either way.

[Seljer]

See: http://geofex.com/FX_images/PNP_power_switching.pdf

Is the PNP transistor just like that with other components also adequate as polarity protection?

(I actually did use this circuit for power switching when I used a stereo 1/4" jack for something. I used the PNP, then on the base along with the resistor a NPN with it's base connected to ground by the NC contact of the ground lug of the switching jack. But I didn't bother to see how hard I had to try to fry it.)

[Muthauzem]

See: http://geofex.com/FX_images/PNP_power_switching.pdf

That's so simple and seems so much better that I really wonder how I haven't stumbled upon it before.. Definitely gonna try on my next build.
One other think that it also seems to resolve is the case of inverting the power supply. Would the PNP BJT also protect the circuit or a diode/mosfet still be necessary?

[R.G.]

That's so simple and seems so much better that I really wonder how I haven't stumbled upon it before..

There is a lot of stuff at geofex that is never looked at. This circuit is actually a snippet of a larger version which turns on *both* a +9V and a -9V battery, one each, in a bipolar supply, at the same time, from one contact to ground. It's in one of the "hum free" schematics.

One other think that it also seems to resolve is the case of inverting the power supply. Would the PNP BJT also protect the circuit or a diode/mosfet still be necessary?

It's a snippet of a circuit. As it sits, it won't help all that much, as the base-emitter junction of the PNP will break over at 5-8V. Every time this comes up, I think "hey, I really ought to do design an end-to-end power supply protection/switching/etc. circuit for geofex." But I get lazy, and the question only comes up a couple of times a year.

There is a different version of this with a perverted version of the MOSFET and "cheap - and good" protection circuits.

[EATyourGuitar]

It's not the industry standard, but why not switch the power using the output jack? The input jack can use a switching jack that shorts the tip to ground when no jack is inserted.

I've suggested exactly that here and in other places so many times I give up. Notice that the jack in the diagram is not labeled input or output. Works either way.

the russian big muff has the footswitch wired to ground the signal after Q1,Q2 but before Q3,Q4 when in bypass. you can also do this with a switched ground jack. they are rare but they do exist. switchcraft enclosed has them.