Stereo jack switching +9v and -9v question

[tempus]

Hey all;

I've been looking at RGs article on +9v and -9v switching here: http://geofex.com/circuits/+9_to_-9.htm

In the lower diagram, how does connecting the base of the transistor to the tip of the input jack turn on the transistor, thus providing a path for the 9v from the battery to get to the IC? Does the resistance of the pickup provide a path to ground for the base?

[B Tremblay]

While the diagram makes it appear that the tip is connected to the transistor via the 6k8 resistor, I believe it is the ring lug - based on the other contact being marked as "signal".

[syntaxera]

same topic...  http://www.diystompboxes.com/smfforum/index.php?topic=98687.0   started a few hours ago.   Maybe together we can both figure this thing out.

[syntaxera]

another thought,  since the output ring is already connected to the 3PDT make the ground connection when the cable is plugged in... 

could I add a stereo jack to the input side and use THAT ring to connect the power / transistor / resistor configuration in the GEO diagram? 

[tempus]

While the diagram makes it appear that the tip is connected to the transistor via the 6k8 resistor, I believe it is the ring lug - based on the other contact being marked as "signal".

I was reading the ring lug as "signal ground" vs the sleeve which would be the power ground. Yours makes sense though, but I don't see how this isolates the power and signal grounds.

Syn - it was your thread that got me looking into this, so I thank you for that. After some study, I think I have an answer to your question as to how to hook up the PNP. I would say connect the battery + to the emitter, the ring lug to the base, and the +9v to your circuitry to the collector. This way the PNP will be turned on when you plug in the guitar (because the base will now be grounded) and current will flow from the battery to the circuitry. Now, this is based on the possibility that RG's diagram has an error in it (with the ring and tip connections reversed) as per BTremblay's suggestion.

[R.G.]

I should have marked that better. The contact that signal comes off is the tip, and the power connector is indeed the ring. They're just not marked and the schemo symbol is misleading. That's my story and I'm sticking to it.   

I did point syn to that schemo for his question about how to hook up his PNP. It's not a real mystery.

[tempus]

Thanks for that clarification RG.

How does this design keep unwanted noise out of the signal path?

[syntaxera]

I did point syn to that schemo for his question about how to hook up his PNP. It's not a real mystery.

hi R.G.

the mystery for me was...

a) was it connected to tip or sleeve (that has been cleared up). lol
b) my output sleeve is acting as a cutoff for the ground on the 3PDT...  so that's leaves the question I had in the other thread..   do I rewire so that I can use the PNP with the output sleeve?   OR should I change the MONO input to a STEREO and hook the PNP up over there? 

thank you both for the tips...  I think I might be close to getting it going on....  but will it work, that is another question all together.

[R.G.]

How does this design keep unwanted noise out of the signal path?

The fundamental thing to remember is that all the conductors we get to play with are really resistors. Every wire has a resistance, and so the only way to have it have zero volts across it is to have the current through it be zero.

The wires to the input jack (and output jack, and controls, and all other wires) are really resistors, albeit small ones. The stereo jack switching trick connects the battery minus (usually) to the ring contact, and the input plug "shorts" the ring to the ground wire leading to the circuit board. So 100% of all the ground current in the pedal is funneled through the input jack ground wire and two possibly dirty contacts with the sleeve of the input plug. This travels with the input signal ground, so it causes a voltage to appear at the INPUT of the pedal equal to the total current of the pedal, LED, whatever is connected to that ground connection, times the wire and contact resistances.

Usually we get away with this. Sometimes, the contacts are dirty, the wires are small, the input impedance is really high, and the gain of the circuit is really high. How many posts have you read that start with "Can anyone tell me how to get more gain out of ..." ? The currents in the pedal vary with the signal. Sometimes we're lucky and they semi-cancel. Sometimes not. When that happens, you can get oscillation, almost oscillation, whatever. It may be at audio or a few megahertz and sound like loud, angry hiss. If the LED current goes through that path, the LED current's instant change with switching is impressed in the input as well. 

It makes hugely more sense to use the output jack to switch power. Well, frankly, in today's era of DC power adapters, power switching with jacks may not make sense any more at all. And also, most people have pedalboards, not single pedals. Who unlaces all the input jacks on their pedalboard when they quit playing? It solves a problem we pretty much no longer have.

But, like center-negative DC jacks and single ended cables, we're stuck with input jack switching by history. If that's true, then we ought to make the best of it and make it not noisy and oscillation prone. The PNP power switch does this by making the current through the input jack ring contact (1) smaller by the gain of the transistor than it would otherwise be and (2) very nearly constant, not varying with the signal in the pedal. This effectively prevents signal feedback from getting into the input jack circuits and still switches off the power when the input plug is pulled out, whether we really need that or not.

b) my output sleeve is acting as a cutoff for the ground on the 3PDT... 

Why is that? This is pure curiosity. There are lots of small decisions in design work. What does this achieve for you?

so that's leaves the question I had in the other thread..   do I rewire so that I can use the PNP with the output sleeve?   OR should I change the MONO input to a STEREO and hook the PNP up over there? 

Either way. The PNP acts like a series switch in the + power lead. It is off (not conducting current) when its base is open. It is closed (conducting current) when its base is grounded through a resistor. It doesn't care where the ground comes from. I frankly prefer using the output jack to switch power, even when it's the much-less-trouble-prone transistor switching. But history has pre-overruled me. Most all pedal users expect power switching on the input jack, whether it's good or not.

Since you are presumably your own customer, you can use the mono-plug-in-a-stereo-jack trick on either input or output. The PNP doesn't care.

Notice that you are not limited to only one PNP. You can switch two different power lines with two different PNPs and maybe eliminate positive power side ticks and pops from the LEDs. You could even use slowed-down transistor switches on the LEDs so the transitions in current are not as fast and therefore don't cause as much audible ticking.

There are many ways.

[syntaxera]

b) my output sleeve is acting as a cutoff for the ground on the 3PDT... 

Why is that? This is pure curiosity. There are lots of small decisions in design work. What does this achieve for you?

I thought it was what you do with the sleeve in a 3PDT true bypass situation... I switched it from the input to see if it helped at all with the pop.  I can easily switch it back if needed.    So I would then have the ground cut on the input jack sleeve...   and the PNP configuration on the Output sleeve...   this makes more sense to you I am guessing? 

man i hope this works..  hahaa

[syntaxera]

How does this design keep unwanted noise out of the signal path?

The fundamental thing to remember is that all the conductors we get to play with are really resistors. Every wire has a resistance, and so the only way to have it have zero volts across it is to have the current through it be zero.

The wires to the input jack (and output jack, and controls, and all other wires) are really resistors, albeit small ones. The stereo jack switching trick connects the battery minus (usually) to the ring contact, and the input plug "shorts" the ring to the ground wire leading to the circuit board. So 100% of all the ground current in the pedal is funneled through the input jack ground wire and two possibly dirty contacts with the sleeve of the input plug. This travels with the input signal ground, so it causes a voltage to appear at the INPUT of the pedal equal to the total current of the pedal, LED, whatever is connected to that ground connection, times the wire and contact resistances.

Usually we get away with this. Sometimes, the contacts are dirty, the wires are small, the input impedance is really high, and the gain of the circuit is really high. How many posts have you read that start with "Can anyone tell me how to get more gain out of ..." ? The currents in the pedal vary with the signal. Sometimes we're lucky and they semi-cancel. Sometimes not. When that happens, you can get oscillation, almost oscillation, whatever. It may be at audio or a few megahertz and sound like loud, angry hiss. If the LED current goes through that path, the LED current's instant change with switching is impressed in the input as well. 

It makes hugely more sense to use the output jack to switch power. Well, frankly, in today's era of DC power adapters, power switching with jacks may not make sense any more at all. And also, most people have pedalboards, not single pedals. Who unlaces all the input jacks on their pedalboard when they quit playing? It solves a problem we pretty much no longer have.

But, like center-negative DC jacks and single ended cables, we're stuck with input jack switching by history. If that's true, then we ought to make the best of it and make it not noisy and oscillation prone. The PNP power switch does this by making the current through the input jack ring contact (1) smaller by the gain of the transistor than it would otherwise be and (2) very nearly constant, not varying with the signal in the pedal. This effectively prevents signal feedback from getting into the input jack circuits and still switches off the power when the input plug is pulled out, whether we really need that or not.

I finally had time later in the night to read this in detail, what a great explanation... thank you.  I think I might have a grasp on what's going on here now.  I am going to try a bunch of these things tomorrow (hopefully) and I will let you know how it went.  I have a feeling it might be the only thing that does the trick.  Nothing else has worked at all...  but this makes a lot of sense... and it seems to have worked really well for you.   Thanks for sharing!

[tempus]

Thanks for that explanation RG.

So I wonder if this little trick might also help with switching pops in other applications. I've got a looper that switches 3 of my effects in and out of the signal path that (even without any LEDs connected) still pops a little when I switch and I've never been able to nail down where the noise is coming from. Switching is achieved with a PIC and JFETs, and I wouldn't have thought that a lot of current would be getting switched but maybe I'm wrong. Do you think it might help in this application? Not sure where I would put the PNP, but I was thinking between either the 9v coming from the DC power wall wart and the 7809, or between the 7809 (and/or 7805) and the circuitry.

Any thoughts?

[syntaxera]

Thanks for that explanation RG.

So I wonder if this little trick might also help with switching pops in other applications. I've got a looper that switches 3 of my effects in and out of the signal path that (even without any LEDs connected) still pops a little when I switch and I've never been able to nail down where the noise is coming from. Switching is achieved with a PIC and JFETs, and I wouldn't have thought that a lot of current would be getting switched but maybe I'm wrong. Do you think it might help in this application? Not sure where I would put the PNP, but I was thinking between either the 9v coming from the DC power wall wart and the 7809, or between the 7809 (and/or 7805) and the circuitry.

Any thoughts?

hey Tempus, have you tried the trick on any of your circuits yet?  Curious to know how it worked for you.....

[tempus]

Not yet. The only thing I have is my looper switching matrix, as described above, and it would be a huge pain to experiment with it as it's already built, assembled, and in use. If it might help, though, I would definitely try the modification.

If I get a chance I may try it to see if it gets rid of the high frequency hash on a switching power supply...

[syntaxera]

Not yet. The only thing I have is my looper switching matrix, as described above, and it would be a huge pain to experiment with it as it's already built, assembled, and in use. If it might help, though, I would definitely try the modification.

If I get a chance I may try it to see if it gets rid of the high frequency hash on a switching power supply...

that's kinda my issue here...  I had to send a video of my current project to some educators in NY, and was simply out of experimentation time.  Once I am done working this thing out with them I will have more time to build it again with this in mind.  I will probably try to figure out these problems on the breadboard before building it again.  After this last build I now know ALL the issues I will encounter and can plan better to resolve them.

fun fun fun !   lol