Jump to content

DMX ground problems?


ceecrb1

Recommended Posts

Long story, but I have a monthly maintenance contract at a cinema complex, they have 2 highend cyberlights there.

The basic setup is that about 200mtrs away from the fixtures there is a DMX splitter, which also feeds many other studio duo mini cities...

All mini cities and also some space cannon focus´s which are fed from that splitter/ universe work faultlessly (wrt signal).

 

Both cyberlights are addressed as 1 (ie same address).

Both have their own DMX cable run from the splitter.

The job of these lights is to scan the logo (gobo) from left to right slowly.

 

Now the interesting part..

One of these lights is really "unstable". It will run the show but offset by 45 degrees, or will pause and insert come colours, change gobo or any other random effect as and it chooses. For about 50% of the time it will do as is expected.

Now I have done endless tests..

I´ve tried terminators, I´ve run the DMX signal from the output of the "good" light to the 2nd bad one, rather than use its feed from the splitter.

I have even sent the fixture to highend to be serviced and fault finded..

It spent a WEEK running every working day in my workshop, working faultlessly.

 

So. Today I bought myself a pocket oscilloscope on ebay. (I know its hard to trigger DMX but I can pause the show... I want to look for noise, voltage drops, that the 0v has not risen up etc etc.)

I´m also wondering about ground loops.. BUT the fixture and splitters are all patched into the same DMX controlled relay panel (etc unison dimmer racks), and so power feed.. so a ground loop is REALLY unlikely I admit (but still possible if the shield of the cable is touching some metal somewhere in its run, the installation is 15 years old).

 

I did have them working perfectly for a month with the companies (only) DMX recorder (elc ministore) where the show was recorded directly from the back of one of the HOGs.

 

So, there you are.. the problem is out there in the world.. Feel free to send some ideas back my way.

 

This has been a VERY long term problem (years).. Why has it not been fixed sooner? well honestly there have been other problems more important and the client isnt that bothered about it not working as there are other parts which are far more important. I´m only there 2 days a month and I have a LOT of kit to work on. We have even looked into replacing both fixtures with two 2k profiles with gobos.. but having problems finding ones which are "all-

weather".

Link to comment
Share on other sites

Couple of suggestions:

 

Try swopping JUST the units over, but keeping everything else (including power leads, DMX cables etc) where they are. If the problem goes with the unit, it is definately something specific to that unit. If it stays in the same position (ie. manifests itself in the previously good unit) then you'll know it is some external factor.

 

It would seem odd to me that something like a DMX glitch, however caused, would give an effect such as a pan/tilt attribute being offset by 45 degrees (or whatever), or a colour/gobo value changing and not immediately changing back. Rogue DMX values should not 'latch' in the fixture, unless it has some obscure mode setting (I'm not familiar with the model in question).

 

If I were a betting man, I'd say the most likely scenario is that the unit has a fault with its DMX input circuit - either an IC fault or a related circuit board problem, but when fed with good strong signal (ie. on a short cable direct from a desk, as on the test bench) it manages to 'blast through' the problem and the unit appears to work fine. Put it on the end of a long line and it starts to struggle. Do you have a spare DMX buffer/booster/splitter that you could place in the line immediately before the problem fixture (so it gets the strongest possible signal)? Also, can you try swopping to a different desk, but still using the existing DMX distribution? While all desks should output 'standard' DMX, in my experience some are closer to the standard than others! Again, a faulty DMX input circuit might respond okay to a 'good' DMX signal but struggle with a 'borderline' one, even if other fixtures cope fine with it. One of these suggestions might well be the case as you say it worked okay with the DMX recorder... was this placed close to the units or at the normal desk location?

 

Good luck!

 

Ben.

Link to comment
Share on other sites

Yeah I´ve been thinking along the same lines as you.. a "tired" component inside rather than broken..

 

The DMX recodeder was recoded at the desk but installed by the fixture..

So I can no be sure that the "solution" (I admit the fault might be the actual fixture) will be cleaning up the DMX signal.

This is why I bought the scope.. I want to SEE the signal and how much noise etc there is... if there is a voltage drop in a long, old, dried out, horrible nasty cable.. etc.

 

Actually yes there is a spare splitter!! I will try that in this months trip.. some how that obvious test had escaped me.. however this is why I´m putting the problem out there.. my mind has gone mad with this one.

Link to comment
Share on other sites

If you've got a scope, then its worth knowing that the Artistic Licence MicroScope can has a scope trigger output so you get a consistant triggering point. The normal DMX update rate is fast enough that persitance of vision makes the image continuous, anbd you can see noise on the trace.
Link to comment
Share on other sites

Well I'm on site (lunch time woop woop!) checked out the entire signal path with my new pocket scope.. Clean as a whistle..

So its clearly the light..

 

So had a brain wave..

As rhe only thing the fixture does is pan left and right, I am going to find a programmable stepper motor driver.. Rip out all the internal system and use that to do the pan.. Block off all other effects and disconnect their cables. Had enough of this and the end result will be the same!

Link to comment
Share on other sites

Well I'm on site (lunch time woop woop!) checked out the entire signal path with my new pocket scope.. Clean as a whistle..

So its clearly the light..

 

So had a brain wave..

As rhe only thing the fixture does is pan left and right, I am going to find a programmable stepper motor driver.. Rip out all the internal system and use that to do the pan.. Block off all other effects and disconnect their cables. Had enough of this and the end result will be the same!

 

Errr... that doesn't sound like such a great idea.

1-You will have difficulty finding a microstepping driver to give smooth panning.

2-You will have to create a way of zeroing the mirror position at power up or it will be different every time you turn it on.

3-You will need to mechanically lock off every other effect inside the unit, and the mirror tilt, at exactly the right position. Just unplugging the motors will leave them free to move and they will vibrate out of position.

Link to comment
Share on other sites

1. Aww poo didnt realise that.

2. If I create a mechanical stop at each extreme of the pan. It will hit it then work with that as a reference point. Same as the fixture already has, just need to move it.

3. Zip ties and high temp silicon can work wonders :)

 

I know.. Not a pretty way to do it but the fixture has been bad 2 years now. Loosing so many man hours on it every month its just silly.

 

Truth is I think your right.. Not all brainwaves are good.. Back to the drawing board.

Link to comment
Share on other sites

I share your pain, we used to have 8 elderly cyberlights and it was a challenge keeping them going.

But there's a lot of engineering goes into getting a smooth and repeatable pan and tilt. You won't get it from an off the shelf stepper driver board.

 

Question: do the cybers get powered on and off every day?

One of ours used to power up in a funny mood sometimes. We did a job where we had 4 of them projecting angels going up and down a tower block, they were on a time switch. 3 of them worked fine, the 4th one would sometimes not start up right and point its beam somewhere completely wrong. We changed the programming so that they stayed powered on all the time and just blacked out during the day. No more problems from the dodgy unit. The problem turned out to be when the lamp struck, it zapped the electronics in some way causing all the motors to "let go" for a few seconds. This resulted in all the functions losing calibration.

 

So, possibly try swapping the lamps between the units?

Link to comment
Share on other sites

Wow thats very interesting.

 

There are however some drawbacks that mean I cant do the same.

First the desks here are Hog 2.. And I dont know how to program the hogs... plus the dimmers which power the units are controlled via an AMX interface system which I have NO idea how to program.

 

Finally.. NO NEED!! I FIXED IT WOOOOOOOO!!!!

 

at 8pm tonight. By Pure, complete and utter LUCK!

I climbed down from the roof.. went to the control room to look at the desks, because for the first time ever, BOTH fixtures were displaying the same problem, ie doing "whatever they wanted". BUT that "whatever" was identical....

So, clearly a desk problem.

got back to the control room.. swapped from the main hog to the backup.. and started reboot the hog, clean it and reload the show. in the middle of all this, the desk's furman filter "rebooted" but for a millisecond. It wasnt enough of a power cut to reboot the fixtures but clearly it made something unstable in the desk... but at EXACTLY that moment, my workmate told me by radio that the fixtures had "gone mad" again.

 

So, swapped each desk to another filter.. reloaded the shows and sat down to have a beer.. no errors in over an hour!!! (they were every 2 or 3 mins max.) Also the gobos were far more focused and the pan far smoother.

 

 

Its taken 2 years to find this! and I only found it as BY chance I was looking at it and noticed the 3 delay leds turn off and reappear at that moment.. imagine the number of times it has happened and I wasnt there to see it!!!

Real case of right place, right time!!!

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.