GOTCHA!! Data and Noise

G

garyjac1

Apprentice elf
Joined
Dec 26, 2012
Messages
51
Location
Dalby
G'Day All,

I'll post this experience in the hope that some other poor bastard doesn't have to go through what I've been through...but then, probably not anyone else is as dumb as me :)

I was testing PPR-A receivers with a PPD-C unit this morning. In fact, I'd been doing this for a week, off and on. This morning I decided to use longer Cat-5 cables (20 and 30 metre). I use these for data only and power the pixels/strips from a separate 12 volt supply.

Swapped out a 15 metre cable that I'd been using for a week with no probs and introduced a 20 metre Cat-5. Some pixels didn't light, others were "off" colour and so forth. WHaaaaa!!!!!

Here's what caused it.

I thought: "Why unroll that 30 metre/ 20 metre cable? Just unroll a metre at each end. No need to unroll the whole thing until it goes into the display outside." So, one meter from the ECG-P12R to a large coil of cable with the twist-ties still on it from the shop. Then about a metre run to the PPR-A unit and pixel string.

Madness ensued until I thought about noise. I uncoiled a bit more cable, like about 5 metres each end. Things improved. Lights came on! I uncoiled most of the cable. It was in a tangled mess all over the room, BUT, all the lights were doing precisely what they were supposed to do!

Aha! I carefully recoiled the Cat-5 and, sure enough, after a certain length was nice and neatly coiled as it came from the shop I started to see colour aberrations and then entire pixels turning off/on where they shouldn't.

ALL cables uncoiled and it ALL works fine. Induced noise and crosstalk (possibly) in the tightly coiled cables.

I hope everyone has a good lauch at this one. I certainly did :))

Regards,
Gary
 
Can you post a link to these PPR-A and PPD-C thingys as I have no idea what they are?
 
This is an issue that has had me in the past as well, everything tested on the bench but once you lay all that cable out they become big antennas to absorb interference and some will have some nasty surprises this year, So setup early as to give yourself enough time to deal with any potential surprises.
Ive had situations where just moving a cable a couple of feet across stopped issues. There are many causes of this and nailing it down to what it is can be very difficult and time consuming at times.
 
They ahve differnetial receivers / trasnmitters each end so they should be fine.


Something that could be problematic is that if your line is not terminated/matched correctly, and this would . with Cat5 cable, you really should have a 100R resistor on each end of the cable, near the tx and rx. I'm not entirely sure of the design of these units, i would have thoght they woudl have been included on each end.
 
Yes, the resistors are there on J1Sys PPR-A rcvr board. It is just occurs when you have it coiled tightly up and the whole thing is picking up noise, or, indeed, as Eddy has pointed out ambient EM of any kind. In situ, I have had no issues -- so far :))
 
Something is oddly wrong then. I use the same technique as the J1sys boards and i can push 1000' on a reel of Cat5 thats still in its box. ( its a bit of amission to find the end, but some brands tuck the end inside the inner drum. )
 
Professional stage lighting and audio gurus have long known this little tidbit of info. Data lines need to always be uncoiled, and should always be kept away from mains voltage lines as well - eg: the microphone cable (the same as a standard DMX cable) should never cross power supply lines - unless at right angles. Where possible, keep at least 12 inches apart, cross at right angles where you have to, and never ever ever use a cable (data or power) while coiled - even for testing.


The manufacturers of extension leads don't tell you to fully unwind it for the fun of it....
 
You must log in or register to reply here.
Back
Top