Audio Matrix (Analog Mux/Demux selection)

Hi guys,

I'm trying to make a switching matrix for my guitar, so I can wire up all my pickups and switch between many different 'virtual' wiring schematics a.k.a. presets.

It would need to be 8x8, and I'm having trouble deciding which route I should go.

Route 1. I could use a Max7219 to control 64 optocouplers, only I can't seem to find any resistive type ones that aren't expensive at this quantity. Optocouplers that I did find for a good price and size are the LTV-816S chips, although these have a phototransistor instead of photoresistor. Would the transistor portion cause much distortion?

Route 2. I could make an 8x8 matrix of transistors (FET, MOSFET, JFET?) and shift registors.

I need this project to be cheap and small, and I only need on/off style routing. I have no need for fading/interpolation between presets.

Thanks for any insight, and I'm open to other suggestions as well.

What you need is an analog multiplexer:

And, a demultiplexer if you also truly need 8 outputs.

If I used a mux/demux, would I be able to select more than one channel to output to per chip? I have a bunch of CD74HC4051s lying around somewhere, but they only allow the input to go to one of the 8 outputs.

I need something that can route one input to multiple outputs. If I can find something just like this chip (+5v logic and +/-5v signal switching) that allows multiple connections then that will be the way I will go.

Thanks guys

The 4051 can be used as 8 to 1, or as 1 to 8 switch.
Multiple chips can be stacked.
There is also an enable input that opens all the switches.

IMHO, not wise to use anything like this on a passive guitar pickup.
Guitar pickups work with high impedance (1Megohm) inputs.
And more capacitance than the guitar lead (~500pF) kills the tone.
Leo…

See, if you can find them:

MT8806A 8 x 4 Analog Switch IC MT8816A 8 x 16 Analog Switch Array IC

I found the MT8816 and MT8808 to have -0.3v to Vee +0.3v. I'd prefer to not have to bias the signal as it adds more components, but I will as a last resort. Do you know of any crosspoint switch that is 8x8 that has a larger analog voltage swing? I just don't want to take a chance of damaging anything, as I would use this for my live rig and it would really suck if anything broke while performing.

EDIT: I think I will just try the MT8808 and bias the signal to keep it in the correct range. Maybe bias it at 2.5 volts.

Questions: 1. To unbias the signal as it leaves the guitar, would I just use a dc blocking capacitor? If I use a dc blocking cap, wouldn't I need to create an RC filter (high pass)? 2. If I have to make an RC filter, wouldn't the varying impedance of switching the pickups in and out of the signal path screw with the filter?

  1. To unbias the signal as it leaves the guitar, would I just use a dc blocking capacitor? If I use a dc blocking cap, wouldn’t I need to create an RC filter (high pass)?

Naturally, capacitor and input impedance of the following circuit IS a first degree HPF.

1.2. If I have to make an RC filter, wouldn’t the varying impedance of switching the pickups in and out of the signal path screw with the filter?

The value of the capacitor must be chosen in accordance with minimum loading resistance, in order to get low freq. end range to pass via HPF.

I’m thinking that I could do something like the following:

Where R is 1 Mohm and C is 150 nf, as according to this calculator my cutoff frequency is about 1hz.

I chose 1Mohm as it doesn’t bleed treble noticeably and the minimum load from the pickups is capable of 0 ohms (nothing connected).

Does this sound about right or am I going about this completely wrong?

Thanks for your help Magician

What you draw is LPF.
HPF you may see here:

Figure-1 on a right side. “C” is supposed to be your DC blocking cap, and “R” is input impedance following circuitry.

Sorry, I posted the wrong photo. The R and C were supposed to be switched like this:

For the calculation I had in mind a correct HPF.

As for the values, does that seem about right?

Wiki agrees with you on values, there is a formula in a link I've posted. 1M and 0.15uF indeed provide 1 Hz cut-off freq.