Audio, data, and +5v DC???

[swinginguitar]

Is it OK to send a (guitar) audio signale down the same multi-wire cable as +5v power and digital (midi) data? is there a "right" way to do it?

the application would be a footcontroller that sends midi data remotely, but would also receive power on the same cable as it trasmits midi data...wondering if i could patch my guitar in as well and have that sent via the same cable (midi/cat 5/whatever)

[g_u_e_s_t]

if the guitar signal is straight off the pickups, it will probably have crosstalk with the midi and power.  if the guitar signal is amplified before going down the cable, for example with active pickups, it will probably be ok.  if it is low level, there are 2 things you can do to make it better.  the first is use a cable that has an internal shield for one of the conductors.  you can then put the guitar signal on an internally shielded conductor.  the other is to use twisted pair, and put a differential signal on a set of pairs.  you can either do this by keeping the guitar and midi grounds seperate, and have the guitar ground and signal make the differential pair, or by putting a small amplifier that makes it differential.

[swinginguitar]

My initial thought was to use something like a 5 pin DIN connector (you know the heavy screw on type like they use for power supplies) with one conductor shielded. If the guitar ground was the shield, would that be workable/helpful/sufficient?

Otherwise, it sounds like i could separate the grounds. would i need separate grounds for the guitar, MIDI data, and +5vDC? that's a lot of pins....

or i cuold buffer the signal?

[g_u_e_s_t]

look for audio snake cables.  they tend to be multi-pair, and shielded.

[swinginguitar]

So maybe the 2 pair version of this:

http://www.gepco.com/products/proav_cable/analog_audio/multi_gf22_M.htm

separate shielded pair for the guitar signal/ground, and the remaining pair for MIDI and +5vDC?

[Johan]

Remember, a cable is two or more conductors running parallel, forming a tiny capacitor.the risk of unwanted noise creeping into the guitar signal is big. But since youre also running 5 volts through the same cable, why not build a buffer into the instrument. A simple fet buffer with a 1k source resistor would let you run cables long enough so you won't hear your amp anymore without interference
Just a thought
J

[g_u_e_s_t]

you will want somthing more like this:
http://www.gepco.com/products/proav_cable/runone/runone_profpa_M.htm

a FET amp might be good, but i would be worried about noise in the 5V line getting into the audio signal.  using a low voltage opamp would probably be better.

[swinginguitar]

Remember, a cable is two or more conductors running parallel, forming a tiny capacitor.the risk of unwanted noise creeping into the guitar signal is big. But since youre also running 5 volts through the same cable, why not build a buffer into the instrument. A simple fet buffer with a 1k source resistor would let you run cables long enough so you won't hear your amp anymore without interference
Just a thought
J

IOW, send a buffered low impedance signal?

I'm thinking maybe an opamp buffer (would have done one anyway back at the rack)

[Johan]

Remember, a cable is two or more conductors running parallel, forming a tiny capacitor.the risk of unwanted noise creeping into the guitar signal is big. But since youre also running 5 volts through the same cable, why not build a buffer into the instrument. A simple fet buffer with a 1k source resistor would let you run cables long enough so you won't hear your amp anymore without interference
Just a thought
J

IOW, send a buffered low impedance signal?

I'm thinking maybe an opamp buffer (would have done one anyway back at the rack)

It might be hard finding an opamp thats happy at 5 volt, but do a google search for bo hansen buffer for an easy, good fetbuffer
J
Edit. Bo called his buffer "impedance converter"...same thing different name,but easier to find if you ou use the right se as rchword

[Gurner]

I'd be inclined to convert the guitar signal to a balanced signal in this instance...any induced noise from the cable run gets cancelled out when converted from balanced to unbalanced at the far end.

[swinginguitar]

starting to think the best way is a separate cable for (buffered) audio and midi...perhaps just zip tie them together (should reduce the capacitance and crosstalk sufficiently, yes?)

don't really want to get into balanced line amps etc, and as someone mentioned, 5v will be a limiting factor.