Please help, can you review this wiring diagram?

[MrRowe]

Hey Guys,
   I'm new to the forum, and was hoping for a little bit of help from people with a little more experience than myself.  I'm a live sound engineer, but have never built a pedal before, and just want to make sure I'm not overlooking anything before I start soldering.  This wiring diagram is for a "switching" pedal i am attempting to build for a friend.  He plays a Korg SV-1, into a Motion Sound keyboard amp.  The amp has three channels, channel one is a Leslie horn, channel 2 and three are standard keyboard amp channels.  He was looking for a way to utilize all three channels, while only playing the one keyboard.  He has not been using the Leslie horn, because he doesn't like the way the electronic piano and piano patches sound through it. 
    So, this pedal would take a left and a right from the keyboard, and a mono line from a synthesizer being triggered via midi by the keyboard.  the intended result would be the 4PDT would either: by pass the left and right input all together, allowing the "M" input for the synthesizer to pass through to the "M" output, and thus only sending the synthesizers output to channel 3 of the amp.  OR bypass the M input, and enable the 3PDT switch to toggle the left and right inputs between outputs A and B, output A would feed the left and right to Channel 1 on the amp (Leslie) and output B would feed the left and right to Channel 2 on the amp for piano sounds.  I hope that I'm explaining this correctly.  Basically a stereo A/B, that can be bypassed all together for a third mono channel.

My questions are as follows,

A) Does the wiring diagram look correct?
B) Do I have the grounding correct to complete the circuit and power the LEDs only when something is connected to the L input?
C) What should I do with the poles on the 4PDT switch that are left empty to bypass the signal (Lugs 3, 6, and 7) do I leave them connected to nothing? or should they be connected to ground?

Thank you in advance, I hope I have this right, I am very excited about this project.


free image uploader

[GGBB]

That will work, but it has issues. The main problem is that all of your amp inputs will be hanging when not in use - like leaving an unattached cable plugged in. This can cause noise to be picked up and amplified. You can solve part of that problem on the 4PDT by flipping the M, R, and L ins and outs around, and then grounding pins 3, 6, and 7. But the amp that is not selected on the 3PDT will still be hanging. Another problem will be switching noise - a bit like having an amp on when you insert the cable. You can help both problems by tying the outputs to ground with a large resistor, but I'm not sure that will be 100% effective. It will work fine, but it could be noisy.

[MrRowe]

OK, so I should ground the unused lugs on the 4PDT, to eliminate amp noise, but your statement about switching the inputs on the 4PDT around has me a little confused. I apologize.  As far as the switching noise goes, could you elaborate a little more on how to go about tying to ground with a resistor? Thank you

[GGBB]

Swap pins 1&2, 4&5, 8&9. This moves the 4PDT outputs from the outside lugs to the center lugs, so that the ones that are switched 'off' can be tied to ground through lugs 3, 6, & 7.

For the resistors, just connect a 1M or larger resistor between the tip and sleeve of each output jack.

Also, if you use stereo jacks for all three inputs, you can have ANY input jack turn on the battery.

[MrRowe]

To make sure I understand correctly, in other words, in its current configuration, grounding lugs 3, 6, and 7 would be grounding the input side of the switch, and I need to be grounding the output side? Correct?

[MrRowe]

Such as this?


upload an image

[MrRowe]

Can you explain why this "circuit" would cause switching noise where others don't? is it the lack of grounding the unused output on the 3PDT?

[GGBB]

Last diagram is correct.

I am not saying it *will* have switching noise that you will notice, I am saying it might or probably will.    Any switching scheme is susceptible to switching noise unless it is specifically designed to avoid it. When mechanical switches are used, some noise is pretty much unavoidable, but can usually be kept to a bare (perhaps imperceptible) minimum by applying certain techniques. What you are basically doing is plugging/unplugging an instrument in to/from an amp that is turned on. By keeping grounds connected, the biggest problem is avoided (ground lifting), but if you have ever had a flimsy jack and wiggled the cord you know that the connect and disconnect of the tip still makes noise. The same thing will happen here. Sometimes this is related to mechanical bounce and sometimes to a small but sudden change in the voltage level at the amp input. Making sure the amp inputs are grounded (by the switch or the tip-sleeve resistors) when not in use helps that.

Another potential source of noise is the "unconnected cable" when the outputs to the amp are disconnected. This acts like an antenna that can pick up AC hum or RF etc. The tip-sleeve resistors also help this by shunting it to ground.

You *might* get away without any of the improvements I suggested and have it work perfectly fine and silent. Until one day it doesn't, because of a room's especially noisy AC wiring, or an RF transmitter in the area, or the stomp switch has worn down just enough to become more noisy. EDIT: or fluorescent lighting, computer equipment, ...

[MrRowe]

I get what your saying. Can you explain to me why the resistors tying the tip and sleeve on the outputs don't create a "short circuit" if you know what I mean. I'm sorry, just very new to all of this, but very interested. Thank you.

[GGBB]

Not a problem. Because the resistors are large, they will impede current flow far more than the circuit input, so more signal current will flow into the amp than to ground through the resistors. Some signal will flow to ground, but not enough to cause any problem except maybe very slight padding/attenuating of the signal. Many circuit input stages like in instrument amp inputs have a resistor from input to ground anyway, so adding one in parallel that is large will not significantly change the input characteristics of the amp. You will have to check this for yourself though - if you notice signal loss or tonal change, use a bigger resistor. 1M is probably a bare minimum - 2.2M or 4.7M wouldn't be overkill. If you have the technical specs for the amp, its input impedances may be listed. Keyboard amps are typically in the 10k-100k range, which when paralleled by 1M-10M doesn't change significantly.

[MrRowe]

I figured that was the reason, path of least resistance kind of thing, thanks for clearing that up. I had originally intended to only run this off of a battery with no external power, but I see the T-Rex polyswitch, which is of similar construction from what I can tell, has an expected battery life of 8 hours. Is that about what you could expect from a 9V running LED's alone? If so I think I better put a jack in there.

[GGBB]

You should be able to get far far longer than 8 hours. The only thing drawing power here is one LED at a time. If you use ultra bright LEDs and a suitably high value resistor (I've been using 6.8k with my LEDs), the current draw is under 1mA. A typical 9V alkaline is rated at roughly 500mAH so that gives theoretically 500+ hours. You can use a bigger resistor to get the draw down even further - ultra brights are still plenty bright with very little current. You should be fine with battery only, especially considering that the switching itself doesn't require power so it can't fail.

[MrRowe]

So the input impedance of the amp is 100k, what would you recommend for a resistor value that would give me the most bang for the buck without effecting the output?

[GGBB]

2.2M would probably be my choice. That will drop 100k down to >95k, less than 5%, which is insignificant.

[MrRowe]

OK,
   I've decided to rework this kind of from the ground up, after receiving feedback from a couple people, I decided I needed to accomplish my goal, while trying to keep any of the inputs or outputs floating.  I am now looking at using three switches, each output having an off/on switch.  This wiring diagram is either going to work, or its a disaster. I'm not sure which, anyone?


free upload image

[GGBB]

I'm guessing the overall plan is to have an on-off switch for each output, and when off, the output is shunted to ground. Correct?

The way you have it drawn, left and right are joined/mono-ed because they are both connected to the same lug in order to ground them. You would need a 4PDT to have separate ground lugs for each channel.

However, both stereo output switches share the same inputs, so you can't ground the inputs when an output is off because the other output may need to be on. This is probably okay - a hanging input is not a big deal since it isn't going to generate any noise because it's disconnected.

So you need to put the inputs on the outside and the outputs in the middle, with the grounds on the other outside row, in order to toggle the outputs from the inputs to ground. Each stereo out switch should be:

Code Select Expand


[In  L]   [In  R]   [LED]
[Out L]   [Out R]   [Gnd]
[ Gnd ]   [ Gnd ]   [   ]


The mono out is essentially the same but only one channel, so you can use a DPDT. And in this case you can ground the input because there is only one mono output. So your wiring works, but to make it consistent with how the stereo switches need to be wired, it would be:

Code Select Expand


[In ]   [LED]
[Out]   [Gnd]
[Gnd]   [In ]


(Edited for typos)

[MrRowe]

Yes, that is the desired end result. I get what your saying about turning the stereo to mono, separate lugs on the 4pdt, Etc. But assuming I don't mind floating inputs, like you said, reconfiguring the lugs as you have shown on the 3pdt's, would get me there, correct? On your 2pdt you show two input lugs, is that a mistake? Or am I missing something? Or is that for grounding the input when not in use? In addition I was told to add a separate resistor for each led, but I have seen it both ways, is there a right way? Or is that personal preference?  Thank you so very much for all of your insight.

[GGBB]

reconfiguring the lugs as you have shown on the 3pdt's, would get me there, correct?

Yes.

Or is that for grounding the input when not in use?

Yes - so it's just a jumper wire from one lug to the other.

I was told to add a separate resistor for each led, but I have seen it both ways, is there a right way? Or is that personal preference?

You need a separate resistor for each LED that is on at the same time. In your case, all three could be on, so you want to have three resistors. You can share the resistor between LEDs if only one of them is on at a time, but even then it is sometimes useful to use a separate resistor for each LED. If you have different types of LEDs (e.g. different colours), they may have different brightness levels, so each resistor can be a different value in order to even out the brightness across all the LEDs.

Thank you so very much for all of your insight.

Glad to have helped.

[MrRowe]

They way you have laid out the 3pdt's, all the grounds can be linked together via jumper and then grounded as well, correct?

[GGBB]

Yes.