Dimmer Doldrums

tdtastic

Active Member
Our dimming system has been doing funny stuff this summer; seeking advice for things to look for and maintenance techniques that we could or should be doing on the regular even despite our current issues.

We have three SR48 racks with 2.4k dual D20's -- whole system installed about eighteen years ago with almost no troubles ever since, aside from the occasional module needing swapped out or the rack reset after a power outage.

We've been running RAGTIME for nearly two months now and literally EVERY night I discover some oddity in the system. Random fixtures will flicker during the show but be totally fine at dimmer check even when parked at various intensities. It's never the lamp or circuiting. I always end up back staring down my rack with my mean face on: I breaker off, I breaker on. I pull out the module and re-seat it. Fixed. Then the next night it's the same problem somewhere else. Or a fixture(s) simply doesn't get power at all. I know I have several modules that need repair (and of course we don't utilize those) but not this many! These are my good modules that keep misbehaving.

Speaking of misbehaving, we HAVE experienced an inordinately high number of outages with building power all together lately which I'm afraid is not helping. With all the rain and bad weather the deep south has gotten this summer, we experience at least one or two total losses of building power a week. I keep our console and other equipment shut down and unplugged as much as I can lately, but the racks have been booting back up seemingly with no problems...

What can I do to help my dimmers get their crap together, and SOON!?!?

it's worth mentioning that our theatre does have a ghost...

P.S.
When a D20 does "go bad," what is usually the specific nature/diagnosis of that? I have about nine modules that need servicing -- Would it be exorbitant to ship those off and have them overhauled? Or is this something I should/could try and repair at the shop myself? How old does a module need to be before it might be time to replace the whole thing all together?
 
Dimmer modules are pretty dumb devices. They tend to either work completely or not at all. The only common failure method I've run into on D20's is that after a number of years some of the power bricks fail, at which point either that circuit is stuck on at full or stuck off.

I'm inclined to say that this sounds much more like a control issue than a dimmer fault. Either a gak'd up CEM or a faulty DMX chain to the dimmer racks. Some of the weirdest issues I've had to trace out in a system were actually because of a UPS battery backup in front of the Unison processor where no one had changed the battery in 10 years. The UPS ended up spewing out a pretty nasty waveform and bypassing it until the battery could be replaced seemed to solve the problem.

Have you tried contacting ETC tech support directly? They're very helpful and often times can point you in a pretty good direction right away.

If there does seem to be some coincidence with the weather I would investigate the DMX runs. Make sure all of your DMX cable doesn't run from your control position free air near the ceiling all the way back to your dimmer racks, etc. It'll be much better shielded from electromagnetic interference during thunderstorms in metallic conduit.

In terms of servicing your existing modules, if the issues are indeed with the power bricks, ETC sells that part individually. It's a straightforward repair you can make yourself.

By the way, it kind of sounds like you're pulling and reseating modules with the racks energized. If that's the case, I cannot urge you strongly enough against that. You really should be deenergizing the racks before pulling any modules.
 
Last edited:
With the constant power failures I would also say the CEM or brick failure.

The other thing to look at is heat. Dimmer check is great but it's not a show. Do a dry run of the show if you got time and sit in the dimmer room and watch the racks.

You could possibly just be overloading the rack. As 18 years is a long time. Personally I would use this as a way to upgrade your CEMs if possible.
 
Thanks guys! MNicolai, yes sometimes I do pull out modules while the racks are on. I've heard to never ever do that, and I've heard it's no big deal if you do as long as they are switched off. Can you elaborate on why that is a bad thing? I've always been confused about that...
 
Thanks guys! MNicolai, yes sometimes I do pull out modules while the racks are on. I've heard to never ever do that, and I've heard it's no big deal if you do as long as they are switched off. Can you elaborate on why that is a bad thing? I've always been confused about that...

It's just good practice when dealing with that much power. The chances that you put the D20 in crooked and it arcs and blows apart are slim. The chances you put the D20 back in and accidentally flip the breaker as you put it back in which throws power to the dimmer and arcs the light fixture attached shattering the lamp and sending glass or starting a fire ( worse case scenario) is possible too.

It takes 5 seconds to flip breakers to the whole rack but it only takes a split second to get zapped and die.
 
I have seen individual dimmer modules have intermittent flicker problems. The problem will only occur in the circuit that the module connected to and will move if the module is put into another slot. A not-too-uncommon problem in Sensor Classic was that airflow sensor in the CEM would start to fail. This can cause random flicking in multiple circuits. Sometimes a rise in the ambient temperature would make this problem worse. As was mentioned previously I think a call to ETC tech service would be recommended.
 
Air flow can be an issue on any Sensor.

When was the last time you cleaned filters, cleaned air flow sensors, etc....

Is the problem with a particular rack, or random ?. Random and across racks might point to bad DMX, cable, etc...

The power shutdowns are a concern. Have you done a back up of the CEM config' and a reset ?.
 
Random acts of weirdness can usually be traced down to one of three problems:
1) Flaky CEM (rack brain)
2) DMX cabling issue.
3) Board
We haven't talked much about how the system is being controlled and that can be a real source of random problems.
 
Good timing! This years ETC Cue conference is recording many of the classes. Lots of tech oriented tips, tricks and one on maintenance.
https://www.youtube.com/watch?v=8AIltDWER0k
More are happening this weekend so check back later too!
Well, hats the {expletive-deleted} off to ETC for realizing that this generally does not cannibalize your in-person attendance, and the benefits outweigh the costs.
 
It takes 5 seconds to flip breakers to the whole rack but it only takes a split second to get zapped and die.
Yeah, google arc-flash and that'll be an education.

Is the CEM throwing any errors? I've gone through the wars with our racks on a few different types of CEM faults and on occasion you'll get one that only displays intermittently when the fault is actually occurring. I also plug the ETC helpline, I spent many, many hours on the phone with them in this past year and those guys working the helpdesk are total rockstars.

The first thing you really should do is some tear-down troubleshooting. Eliminate everything from your system except the console and the rack. Try a difference console, if it still flickers then it's probably not your console doing it. Try and find what actually correlates with the flickers to isolate it to a specific component by eliminating everything and adding or replacing one piece at a time. Once you replicate the correct circumstances for the failure to occur, you'll have a good idea of what is causing it. ETC's helpdesk will walk you through this too!

One thing I've found that's a common failure in lightning heavy areas is the failure of a 75176 transceiver chips. This is the little guy that takes your DMX data and turns it into the RS-485 compatible differential signal. It's also the first thing in the box that the lightning/static/ESD tends to zap since it's pretty much the last component before you hit the DMX cabling. Fortunately, those crazy, wonderful kids at ETC made this a socketed chip so it's pretty easy to swap.

You have two in your CEM toward the back, and you also will have two optocouplers (6N137), one for each DMX input into the CEM (see the pic). The optos can also get zapped. If you check one of the electronic component supply houses you'll find these for pretty cheap, I picked up a tube of them for less than $10 last time. There's a few different models of the 75176, so your best bet is to just take the whole number off the chip and google it. Check the spec sheet to make sure the pinout is the same and you should be good to go.

In anything connected to your DMX line there will be some flavor of RS-485 transceiver and usually an optocoupler (including optosplitters!). ETC likes the TI ones, but there's also a MAXIM version that is virtually identical, but more expensive I think. I'd check your arch controller as well and your console if swapping these chips don't make much difference.

That all said, DMX is pretty funky and the RS-485 standard is very robust, but prone to EM interference when not properly installed. I'd go through your system with a fine-toothed comb. Check polarity on cabling, check terminations. I threw out the chip replacement since I've had that happen a few times after bad storms and it often fixes bugs here and it's a cheap and easy swap, but we still don't really know what part of your system is getting the errant data.

proxy.php
 
It's just good practice when dealing with that much power. The chances that you put the D20 in crooked and it arcs and blows apart are slim. The chances you put the D20 back in and accidentally flip the breaker as you put it back in which throws power to the dimmer and arcs the light fixture attached shattering the lamp and sending glass or starting a fire ( worse case scenario) is possible too.

It takes 5 seconds to flip breakers to the whole rack but it only takes a split second to get zapped and die.

On that note, I should probably highlight that I've seen racks where the only immediate "breaker" visible to shut power to the whole unit was actually an isolation switch not designed to be flipped under load. The chances of this blowing apart are also rather slim, but it's certainly worth reading the unit to check beforehand.
 
Yeah, google arc-flash and that'll be an education.

Is the CEM throwing any errors? I've gone through the wars with our racks on a few different types of CEM faults and on occasion you'll get one that only displays intermittently when the fault is actually occurring. I also plug the ETC helpline, I spent many, many hours on the phone with them in this past year and those guys working the helpdesk are total rockstars.

The first thing you really should do is some tear-down troubleshooting. Eliminate everything from your system except the console and the rack. Try a difference console, if it still flickers then it's probably not your console doing it. Try and find what actually correlates with the flickers to isolate it to a specific component by eliminating everything and adding or replacing one piece at a time. Once you replicate the correct circumstances for the failure to occur, you'll have a good idea of what is causing it. ETC's helpdesk will walk you through this too!

One thing I've found that's a common failure in lightning heavy areas is the failure of a 75176 transceiver chips. This is the little guy that takes your DMX data and turns it into the RS-485 compatible differential signal. It's also the first thing in the box that the lightning/static/ESD tends to zap since it's pretty much the last component before you hit the DMX cabling. Fortunately, those crazy, wonderful kids at ETC made this a socketed chip so it's pretty easy to swap.

You have two in your CEM toward the back, and you also will have two optocouplers (6N137), one for each DMX input into the CEM (see the pic). The optos can also get zapped. If you check one of the electronic component supply houses you'll find these for pretty cheap, I picked up a tube of them for less than $10 last time. There's a few different models of the 75176, so your best bet is to just take the whole number off the chip and google it. Check the spec sheet to make sure the pinout is the same and you should be good to go.

In anything connected to your DMX line there will be some flavor of RS-485 transceiver and usually an optocoupler (including optosplitters!). ETC likes the TI ones, but there's also a MAXIM version that is virtually identical, but more expensive I think. I'd check your arch controller as well and your console if swapping these chips don't make much difference.

That all said, DMX is pretty funky and the RS-485 standard is very robust, but prone to EM interference when not properly installed. I'd go through your system with a fine-toothed comb. Check polarity on cabling, check terminations. I threw out the chip replacement since I've had that happen a few times after bad storms and it often fixes bugs here and it's a cheap and easy swap, but we still don't really know what part of your system is getting the errant data.

proxy.php
@StradivariusBone An excellent read you wrote Sir and, in return, a dumb little tale of ESD.
Some decades ago, I'd gone with an amateur group to spend a long day in a province wide competitive theatre festival here in Canada. The host group had an Express, driving a Unison driving a Sensor rack plus a collection of analogue zero to ten house and lobby light dimmers controlled by wall mounted architectural stations in at least three locations. The competing group I was visiting with were carrying their own Strand MX48. Their Strand board had all of their LXQ's in it and normally outputted AMX-192 to their dimmers but would readily put out DMX 512 in this competitive festival situation. The host group's board was in a decent booth at the rear of the balcony and spoke DMX via a short 5 contact cable into a wall mounted 5 contact input to the Unison located in the basement below the stage and adjacent to the Sensor stage lighting rack. The visiting group who'd competed the previous day / night had used the host's Express and everything was functioning normally. The group I was visiting with began their day with their MX48 on a production table on the lower level for the mutual convenience of the director, LD, SM and board op'. They were also travelling with all aspects of their own sound, task lamps and intercom and had no trouble lashing everyone together for an early afternoon tech, followed by a full run for everyone's benefit. The visiting group had routed their own 100' DMX cable from the booth's DMX input, out the window, through the balcony, over the rail and down to their MX48. The host group's technicians had been asked to please remove their Express from the booth counter over the course of the day so the competing group could place their MX48 in its place during the dinner break in preparation for the evening's competitive performance with sold out audience.
So far, so good. "Peachy keen!" Everything's going swimmingly.
The set went up in the morning and a minion ran the MX48's sliders for focusing and then everyone broke for lunch.
After lunch the tech' ran smoothly followed by a full run through with LXQ's being quietly / seamlessly touched up / down on the fly with nary a disruption.
The happy director, cast and SM broke for dinner with the LX and sound crew expecting to join them shortly.
You can probably see this coming. It didn't happen quite like that.
I'll boringly plod through this step by step.
The sound crew struck their ClearCom and relocated it to the booth and both sides of the stage.
The LX crew took their MX48 up to the booth and set it up in the host's booth.
The MX48's DMX512 output was plugged into the host's wall mounted input receptacle with the host's 15' DMX cable and the MX48 which had just worked flawlessly all morning and afternoon no longer worked. It looked like it was working but nary a stage light was responding.
Huh? What the heck? What's going on?
Maybe the host's 15' DMX cable which had worked the previous day had suddenly failed? Nope!
The host's techs tested their cable with their Express and now that wouldn't work either.
Downstairs, the lobby was filling up with anxious patrons.
Upstairs in the booth, fingers were being pointed and tempers flaring.
O.K. Step by step.
The day / night before all of the host's systems had functioned perfectly.
The visiting group had run a 100' DMX cable, DMX not mic cable, to the main floor and all had worked.
The 100' cable was struck and the visiting group's MX48 relocated to the booth and things were no longer talking.
A check with a Goddard Mini DMX'ter revealed the following.
Both boards', the Express and the MX, were spitting out workable DMX.
The host group's 15' five contact cable was functioning on contacts one through three as expected, no problem.
Sending DMX from the Mini DMX'ter produced nary a glimmer from the Unison / Sensor combination. The wall stations controlling the house and lobby lights were still functioning normally.
The host theatre's manager phoned his 'hot line number' and Chris Mentis [Who was then commisioning and servicing ETC systems across all of eastern Canada responded.] Chris wisely recommended powering down and re-booting the Unison but the theatre's manager refused to do it as he had a lobby and bar full of happy patron's sucking back their favorite beverages.
Next step, the Mini DMX'ter went to the basement dimmer room, bypassed the Unison by unplugging one of its outputs and connecting directly into the Sensor rack which woke up immediately wondering why the work lights were on in the normally dark dimmer room.
Next the MX48 and the 100' DMX cable were rushed from the booth to a dark USR corner and the cable routed downstairs to the Sensor rack. LX organized power for the MX and a task lamp for the board op'. Sound organized ClearCom for the theatre manager who agreed to operate houselights from a DSR location and a headset for the board op'. A relieved ASM organized an icy cold (non alcoholic) beverage for the LX board op'. Doors were opened and gleeful patrons were allowed to enter only moments past the half.
After the performance, and public adjudication, once the house and lobby had cleared. The venue manager agreed to reboot the Unison to ascertain the situation for the next day's group. The Unison was happy upon its reboot and life went on with tempers somewhat frayed.
Here's my best guess as to where it all went wrong due to the eager impatience of an overly eager willing helpful hand.
Had I struck the 100' DMX cable from the main floor myself, I'd have first unplugged it from the Unison's input PRIOR to dragging it across the lower level's carpeted aisle and neatly over and undering it in the booth. The overly eager willing helpful hand chose to wrap the 100' cable as he would've his garden hose leaving it still connected to his garden tap. I strongly suspect dragging the cable's jacket the length of the lower level's dry carpet generated enough static electricity to totally confuse the poor Unison. It probably was scared enough in its dark room in the basement without someone giving its input the 'cattle fence' treatment.
That's my best suspicion and I'm standing by it.
To this day, I've never regretted purchasing my Goddard Mini DMX'tr. At the time, Mr. Goddard was the only one marketing such a product.
Toodleoo!
Ron Hebbard.
 
Last edited:
Like always Sir, an excellent read.

The amount of times we've run around like headless chickens trying to make DMX work once again before a show (incidently, also with the MX 48) is quite relatable. During our Musical production of the year, we utilized a narrow beam moving head in order to freeze actors out of time during dialogue being directed towards the audience. Being a 1960s facility with no renovations in sight at the time, our DMX run was essentially one giant loop that started at the FOH boards, travelled to the main snake box, crossed stage right, up to the fly gallery, draped between the two tips of the 'C' shaped walkway going upstage, down into the Chauvet Geyser, crossing all the way stage left into the wing, then heading downstage, draping up to the first electrical, and running into the moving head.

Anyone familiar with the standard will have realized about three commas in that this is a *bit* out of spec for DMX. Of course, we were unfazed as we were getting away with it without any issues. Of course, it wasn't until 15 minutes to curtain that we tried doing a test using the laptop assigned for controlling the intelligent effects which of course decided then to not register commands, and instead treat us to a display of convulsions, strobes and puffs of smoke from the Geyser.

In comes me, frantically running to stage setting up the laptop with a direct connection to the moving head, setting up a new ClearCom line and fetching the IR remote so the second Lighting tech could periodically run the geyser when needed.
 
Like always Sir, an excellent read.

The amount of times we've run around like headless chickens trying to make DMX work once again before a show (incidently, also with the MX 48) is quite relatable. During our Musical production of the year, we utilized a narrow beam moving head in order to freeze actors out of time during dialogue being directed towards the audience. Being a 1960s facility with no renovations in sight at the time, our DMX run was essentially one giant loop that started at the FOH boards, travelled to the main snake box, crossed stage right, up to the fly gallery, draped between the two tips of the 'C' shaped walkway going upstage, down into the Chauvet Geyser, crossing all the way stage left into the wing, then heading downstage, draping up to the first electrical, and running into the moving head.

Anyone familiar with the standard will have realized about three commas in that this is a *bit* out of spec for DMX. Of course, we were unfazed as we were getting away with it without any issues. Of course, it wasn't until 15 minutes to curtain that we tried doing a test using the laptop assigned for controlling the intelligent effects which of course decided then to not register commands, and instead treat us to a display of convulsions, strobes and puffs of smoke from the Geyser.

In comes me, frantically running to stage setting up the laptop with a direct connection to the moving head, setting up a new ClearCom line and fetching the IR remote so the second Lighting tech could periodically run the geyser when needed.
@EdSavoie Gee! None of us have EVER done anything like that when we were your age.
[Honestly, I strongly suspect many of us have, we're just too old now to remember.]
(Besides, when we were your age, we were wicking wale oil from our orchestra pits up to our footlights.)
Toodleoo!
Ron Hebbard.
 
Autoformer dimmers with circuit patch panels never had these problems.
@FMEng And then there were Strand's British built system LC saturable core dimmers. I only operated two of the three system LC's in my area. The third was in our local TV station's largest studio where I was allowed to see it but never to touch it. The LC I was most familiar with was hidden in the basement of a historic 18,000 square foot castle built in 1835 and controlled the interior and exterior lighting used in a 'sound and light' attraction which ran at dusk several nights per week in our summer months. The third system LC was in a high school gymna / cafe - torium where it was robust enough to defend itself from students and hurtling volley, basket and medicine balls. This third LC was the one I enjoyed playing 'tricks' with.
I had a small, homemade task light consisting of a 12" Lumiline lamp housed in a small home built mount / reflector / shroud.
The Lumiline, (Trade mark) was basically a low wattage incandescent with the physical size of roughly a T6 or T8 x 12" long fluorescent. When the LC was installed, the control console was immediately adjacent to the dimmers but there were no utility / convenience receptacles any where nearby. I rarely visited this LC as it was a couple of cities away but occasionally, maybe once every two or three years, someone would drag me out to light a one-nighter in the venue and I'd bring my Lumiline to clamp onto a music stand to hold my program / running order / cue sheets.
Here's my LC / saturable core trick.
I'd manage to save one dimmer for my little ~12 watt Lumiline. Without a dimmer, my reading light would've been too bright.
This was in a time long before TRIAC's and SCR's and each saturable core was rated for something like 3 Kw or 5 Kw's depending upon which of the two sizes.
Similar to resistance dimmers, if your load was insufficient, your light would barely dim whatsoever, not even coming close to going out. With my ~12 Watt load, the LC's dimmers had no dimming effect at all.
I'd plug in my 12" Lumiline and briefly parallel a 500 Watt Pattern 23 as a momentary load to build the magnetic core's inductive reactive field.
Set the LC's channel at 20 or 30%, wait a couple of seconds for the LC to settle with the 512 Watt load, then disconnect the comparatively bright 500 Watt lamp and the LC's core would maintain my ~12 Watt load at a workable level for the remainder of the evening, or at least until interval when I could stand in front of the Pattern 23 and momentarily flash it to rebuild the saturable core's magnetic field.
Memories. Times were different then. Not necessarily better but definitely different. Asbestos was still in vogue.
@JonCarter may remember Strand's system LC's. Heck! @JonCarter may even remember 12" and 18" Lumilines.
@BillConnerFASTC may remember system LC's as there was one in regular use at the Stratford Festival's Avon proscenium house when I moved to Stratford in 1977. I believe the LC at the Avon remained in regular service running in rep' until it was replaced in 1979 or 80.
Toodleoo!
Ron Hebbard.
 
@FMEng And then there were Strand's British built system LC saturable core dimmers. I only operated two of the three system LC's in my area. The third was in our local TV station's largest studio where I was allowed to see it but never to touch it. The LC I was most familiar with was hidden in the basement of a historic 18,000 square foot castle built in 1835 and controlled the interior and exterior lighting used in a 'sound and light' attraction which ran at dusk several nights per week in our summer months. The third system LC was in a high school gymna / cafe - torium where it was robust enough to defend itself from students and hurtling volley, basket and medicine balls. This third LC was the one I enjoyed playing 'tricks' with.
I had a small, homemade task light consisting of a 12" Lumiline lamp housed in a small home built mount / reflector / shroud.
The Lumiline, (Trade mark) was basically a low wattage incandescent with the physical size of roughly a T6 or T8 x 12" long fluorescent. When the LC was installed, the control console was immediately adjacent to the dimmers but there were no utility / convenience receptacles any where nearby. I rarely visited this LC as it was a couple of cities away but occasionally, maybe once every two or three years, someone would drag me out to light a one-nighter in the venue and I'd bring my Lumiline to clamp onto a music stand to hold my program / running order / cue sheets.
Here's my LC / saturable core trick.
I'd manage to save one dimmer for my little ~12 watt Lumiline. Without a dimmer, my reading light would've been too bright.
This was in a time long before TRIAC's and SCR's and each saturable core was rated for something like 3 Kw or 5 Kw's depending upon which of the two sizes.
Similar to resistance dimmers, if your load was insufficient, your light would barely dim whatsoever, not even coming close to going out. With my ~12 Watt load, the LC's dimmers had no dimming effect at all.
I'd plug in my 12" Lumiline and briefly parallel a 500 Watt Pattern 23 as a momentary load to build the magnetic core's inductive reactive field.
Set the LC's channel at 20 or 30%, wait a couple of seconds for the LC to settle with the 512 Watt load, then disconnect the comparatively bright 500 Watt lamp and the LC's core would maintain my ~12 Watt load at a workable level for the remainder of the evening, or at least until interval when I could stand in front of the Pattern 23 and momentarily flash it to rebuild the saturable core's magnetic field.
Memories. Times were different then. Not necessarily better but definitely different. Asbestos was still in vogue.
@JonCarter may remember Strand's system LC's. Heck! @JonCarter may even remember 12" and 18" Lumilines.
@BillConnerFASTC may remember system LC's as there was one in regular use at the Stratford Festival's Avon proscenium house when I moved to Stratford in 1977. I believe the LC at the Avon remained in regular service running in rep' until it was replaced in 1979 or 80.
Toodleoo!
Ron Hebbard.


Do you ever sleep lol.
 
Do you ever sleep lol?
Heck no @Amiers it's highly over rated. I'll sleep when I'm dead I think is how the famous quotation goes.
My Dad died when I was 13 and just starting grade nine in high school. My Mom had to go to work to keep a roof over our heads and I held several jobs all through high school. My favorite job was working midnight to 6:00 a.m. in commercial AM broadcast while going to highschool. The long and the short of it is I've been a night person ever since. Getting through days is hard when the world insists upon imposing a nine to five world on me but three to four a.m. is when I'm alert and cranking if left to my own devices.
Thanks for inquiring.
You want a giggle? You wouldn't believe how long I'd been reading CB and naively believing you were a 'hot babe' who performed bare back for your hobby or livelihood. One day it sank in when I learned more about avatars.
For ages, I thought you and another regular were twins.
Yeah, I'm that old, gullible and naive.
Toodleoo! @Amiers
Ron Hebbard.
 

Users who are viewing this thread

Back