electromagnetic noise issue with long cables

[deuns0615]

hey Guys,

I put this in the off-topic section as it's not a MD issue, but more like a "newbie in electronics" issue

I built a percussion which the player holds in his hands, far away from the MD module which is connected to the computer backstage.
Right now it sends piezo signals through a 10m multiconductor cable (in fact a 7 pin midi cable), then each wire in the cable gets connected to a pin on MD's PCB, one wire is used for the ground

I thought it was working perfectly, we were rehearsing with it without any issue, but one day someone played with the lights in our rehearsal room, and i realized it was triggering sounds.

i shouldn't have been surprised: the electromagnetic noise from the electric circuit in the building or the room is somehow received by the 10m cable, and MD interprets it as a piezo signal, and bang! The problem occurs with both my laptop on battery and my other computer.

I have 2 questions:

1. on the cable i'm using the shield isn't connected to any pin, would it change anything to have the shield connected to the ground pin? (doing this "faradet cage" i understand nothing about?)
2. If i decide to put the MD close to the player (reducing the multiconductor cable to 1m), and use a 10m midi cable connected to MD's midi output (with MD running on batteries) and going into a midi to USB interface, could this "noise" affect the midi data travelling through the 10m 5 pins midi cable as well? (after all, i understand that a "noise" signal can be confused with a piezo signal, but "noise" in the middle of coded hexadecimal language should be pointless and thus ignored by the interface, am i wrong?)

thanks in advance for your help

[dmitri]

1. You have to connect the shield of the inputs wires to the ground to minimize electromagnetic noise pickup.
2. Using shorter inputs cables and longer MIDI or USB cables is a better idea since MIDI oand USB cables are better protected from the electromagnetic noise both electronically and through protocols checks (parity/ checksums and etc).

[deuns0615]

Once again, thanks for your kind advice Dmitri!

After a few trials, i can say that a long shielded cable with the ground connected to the shield solved the electromagnetic noise issue BUT, the length of the cable is still a problem, it adds somehow some random latency, some hits are delayed, it forms some clusters of hits sometimes, so it's definitely an issue.

The right solution is, like you advised, to have a short cable to carry the piezo signal (ground connected to the shield) and a 10m midi cable or even better: a wireless midi system (Panda Audio Midibeam in my case, which works really well so far). I had no issue with this configuration.

We had to run our MDs on batteries, put them in small boxes and hang them to our belts like our wireless headset microphone transmitters, we tried it on stage and had no issues at all, i'm really glad!

thanks again Dmitri