Some dimming problems

[fatbarry_2000]

Hi there,

 

We're having a spot of bother here at the Phil with a few dimmers. Just wondered if anyone might have any suggestions/thoughts!

 

#1. We have approx 900 40W GLS lamps making up our houselights, run from two LD90 dimmers. For some time now, one single cove (that's

three channels) has been brighter than the other 8 during dimming - not in 'full' intensity, but at 20% it is noticeably brighter than the rest.

This doesn't appear to be a problem with the desk, as it happens on all 3 of our desks, as well as our Artistic Licence remote fader unit.

 

Also, according to the dimmer, it's not set on a different curve to the rest.... Should I assume the dimmer is lying?

 

#2. In the last week, our Source 4 19deg profiles fitted with scrollers have started to flash - very short duration 'off then on' type

behaviour. It occurs with a single lamp, or a few of them, or all 12. It seems to occur at any intensity.

I have assumed that it isn't a DMX issue - again it occurs with the 2 desks I've tried - and it would have to be a very specific problem

to affect just 12 out of 512 channels.

Also, it doesn't seem like a dimmer problem - I would expect such a problem to affect the entire dimmer, and in any case, the 12 lanterns

are spread across two LD90s (due to some odd installation). They only connection between the channels as far as dimmers are concerned is

that they are softpatched at the dimmer.

 

#3. About 2 weeks ago an Alphapack 2 we have for small gigs has started flickering - this ocurs when a fader on the pack is brought up, and the

flickering extends to the overload LED as well. We are using small Selecon 500w lanterns, and even using a single lamp (ie. obviously not an overloading risk)

the lantern output varied between about 50 and 30 percent. I've had a quick look inside but can't see any obvious problem. Any help appreciated on this one!

 

Please contribute any thoughts you might have!

 

Thanks,

Richard

 

Richard Curtis

Production Technician

Liverpool Philharmonic Hall

[woody74]

Everything you describe (2 of the 3 are flickering issues) sounds like an infrastructure data transmission issue (which others here are more capable to address than me), so I'll throw out my two bits:

 

#1...one single cove (that's three channels) has been brighter than the other 8 during dimming

Any chance that cove had all its lamps replaced at once, and they are from a different manufacturer, or maybe a bad batch?

 

#2...started to flash - very short duration 'off then on' type behaviour.

Don't know if this would have any bearing on you, but this happens with Etc Sensor Racks, wherein the internal temperature gets too high, and the dimmers go into "safe mode", dimming themselves randomly in order to decrease the load and thus drop the temperature.

 

The other thought was: are you routing your DMX through a switch, hub, or bridge? If so, have you tried resetting it?

 

#3. About 2 weeks ago an Alphapack 2 we have for small gigs has started flickering

Maybe a power-input issue, but again, if it's not the whole pack but just one or two channels, data transmission seems more suspect.

 

HTH,

-w

[Mexi]

could it be voltage drop on item 1? Cabling may take a longer run to get there, which could cause the reduced brightness. Probably not so noticable during the last few dimming steps up but more at reduced illumination?

[Dmills]

Ok, have you checked that both legs of the DMX line have continuity and that neither has ended up grounded somewhere, that can cause all sorts of weird problems (BTDT). an easy way to check is to first ensure the line is terminated, then meter across pins 2 and 3 at the desk end, if it shows 100 to 150 ohms (with better then a kohm or so to pin 1) then fine, if not it is a DMX wiring problem....

 

Next thought is common mode voltage, RS485 is only good for about 12V common mode and a missing earth connection can cause problem that way. Check the earth to the desk location.

 

A really weird one that did this to me and had me tearing my hair was eventually traced to a dried out PSU smoothing cap in a DMX splitter, so you might want to stick a scope on the DMX and see if anything untoward is happening.

 

The above three relate to flicker but will not cause the dimmer law weirdness.

 

Check if there is any difference in 'bottom trim' on those channels, also that the bulbs are actually 240V and not 230V, which would maybe do it.

 

Final thought is power harmonics playing hock with the dimmers (Possibly exaggerated by neutral impedance problems, they have been digging in that area recently (The Hardman street corner), if you have the kit, then checking voltage harmonics might be informative. Also check phase to neutral voltages, stranger things have happened.

 

If you need more help with this PM me, I live literally around the corner.

[Biskit]

I have assumed that it isn't a DMX issue

 

From what you describe you are probably right about this... however it does seem like a 'clutching at straws' situation so it might well be worth double checking all your DMX cables including any short links, adaptors etc.

 

In a recent new-install we had 5 new dimmer racks installed: a Strand Act6 Digital right next to the control room (supplying some fixed FOH outlets) and four Zero88 Betapack 3s (in a conventional location backstage).

 

DMX came out of the desk, into the Act6, then a long run (70m ish) to the first Betapack, with short (under 1m) links daisy-chaining the remaining Betapacks. No termination. Initially we were using an old Strand GSX desk (pending a decision on which new desk to buy!) and had no problems whatsoever with it. Occasionally, hirers would bring their own desk, ranging from 'Showtec Scenesetters' to Behringer to Frogs and even an Avo Pearl.

 

With every desk apart from the GSX and the Pearl, we had random flicker problems. Mainly on the Act6 dimmer but occasionally on the Betapacks, any channels, at random. We found that unplugging the long link between the Act6 and the first Betapack (at either end) removed the problem, however because everything worked ok with the usual GSX desk, we put it down to an issue on the Betapacks rather than a DMX line issue.

 

Eventually it turned out to be a faulty link between Betapack 1 and 2. One of the data lines was shorted to ground in the XLR at one end. We fixed it and all has worked perfectly ever since, with any desk. We still can't explain why the problem didn't occur with the Strand or Avo desks... perhaps they send out a 'more powerful' DMX singal or something?

 

Anyway the moral is... don't rule anything out until you are 100% sure!

 

Ben.

[boatman]

DMX came out of the desk, into the Act6, then a long run (70m ish) to the first Betapack, with short (under 1m) links daisy-chaining the remaining Betapacks. No termination. Initially we were using an old Strand GSX desk (pending a decision on which new desk to buy!) and had no problems whatsoever with it. Occasionally, hirers would bring their own desk, ranging from 'Showtec Scenesetters' to Behringer to Frogs and even an Avo Pearl.

My bold.

 

Eventually it turned out to be a faulty link between Betapack 1 and 2. One of the data lines was shorted to ground in the XLR at one end. We fixed it and all has worked perfectly ever since, with any desk. We still can't explain why the problem didn't occur with the Strand or Avo desks... perhaps they send out a 'more powerful' DMX singal or something?

 

Anyway the moral is... don't rule anything out until you are 100% sure!

There's another moral here too. Had you put in a termination the faulty cable would very likely have shown up from day one. Then it could have been located very quickly by breaking the line at different points and inserting the termination.

[Dmills]

There's another moral here too. Had you put in a termination the faulty cable would very likely have shown up from day one. Then it could have been located very quickly by breaking the line at different points and inserting the termination.

 

I never have understood why people think that leaving DMX lines unterminated is a good thing, it hides faults that then bite you at the worst possible time.

Yea, sure if the cable is faulty leaving it unterminated might allow things to sort of work, and if you have a cable fault during the overture then pulling the termination to see if it will get you to the interval is possibly a reasonable thing to try, but this does not mean that running without a terminator routinely is good practise.

 

You occasionally get some (normally) moving light type insisting on no termination "Because it doesn't work with one", IME you normally find it doesn't work reliably without one either (And almost always there are one or more faulty cables involved)!

 

Regards, Dan.

[Biskit]

There's another moral here too. Had you put in a termination the faulty cable would very likely have shown up from day one. Then it could have been located very quickly by breaking the line at different points and inserting the termination.

 

I knew that would be picked up! :o Rest assured it is now terminated. We had assumed (clearly unwisely!) that the professional installers would have correctly installed the line, including termination (and non-faulty links!) and tested it thoroughly. Seeing no physical terminating connector, we thought the sparkly new Betapack 3s (of which ours were some of the first installed in the UK) must have built-in termination.

 

Anyway, it was basically by the means you describe that we did ultimately track the fault down. We have had other, shall we say, quality control related issues with this install, which is as much as I can really say about it.

 

My point still holds though that DMX problems shouldn't be completely rulled out in the OP's situation.

[fatbarry_2000]

Thanks for the replies so far. Do keep the suggestions coming in - eventually there'll be enough

free time at work to work through them all.

 

I have another quick question:

If I were to meter the various parts of a DMX chain, what would be sensible things to look for?

I've read somewhere that with a terminated line you should see resistance of 120 ohms and anything

else equals a problem, but is it worth checking voltage from + and - to ground, or

continuity between the same? If so, what should I expect from a 'healthy' system?

 

Thanks.

[Dmills]

If so, what should I expect from a 'healthy' system?

 

120 ohms or thereabouts between pins 2 and 3 (with the desk unplugged and a terminator fitted), and at least 1K (and could be much higher) between pin 1 and both pin 2 and 3.

 

With the desk unplugged there should be much less then 7V between pin one and the local earth at the desk position (Common mode range of RS-485). If everything is opto isolated there should be no continuity from any pin of the DMX line to earth (well, only leakage).

DMX out should be a 5V differential square wave between pins 2 and 3, but you will need a scope to see it as it runs at a few hundred Khz,

 

Note: For desk read DMX source - Splitter outputs each count as a 'desk' for this purpose.

 

Do you have a two LED tester of the type described here?

http://www.emanator.demon.co.uk/bigclive/dmxtest.htm

A seriously useful bit of kit for tracing these problems.

 

Regards, Dan.

[Kazeja]

LD90 dimmer racks have a facility to control dimmer intensity locally (Levels button) which bypasses any other control input to the dimmer.

 

Do the problems occur when you control the dimmer locally via the dimmer(s) keypad? If yes then problem is not DMX related.

 

If no, turn off everything that is connected to the DMX chain and measure between pins 2 and 3 at the source of the DMX chain. As mentioned in other posts, you should get a reading of about 120 ohms. Anyting in the Kilohm reading would indicate an unterminated or broken line. Anything below 70ish ohms would indicate an over terminated line.

 

Please advise status when turning dimmers on locally.