CD80-AE and Dimmable LEDs

EdSavoie

Well-Known Member
We have this Bulletproof CD80 Advanced Electronics Rack filled with the 2.4KW modules, and life is just dandy. Travel to a few years ago and part of an "eco friendly" initiative has our house lights replaced with LEDs.

I'm sure most of you can see where this is going now, but the LEDs aren't liking it very much.
The lamps in question are GE PAR30s model LED26DP38S840/40

Now, these lamps are listed as dimmable from 10-100% and have a rather long life expectancy, yet we see them fail in one of two modes, being either Off, or Strobing once per second. Sadly, most of them fail in the second method, which as you can imagine, is the visual equivalent of water torture.

I'm wondering if there are any quirks in how the CD80 Dims circuits that causes it to do this, and please note that some bulbs have failed during a period of time where they were either off, or at 100%, and thus never dimmed during it's life.
 
It's not so much a quirk as a case of using the wrong tool for the job. LEDs are not really intended to be voltage dimmable so the retrofit lamps are going to have issues. The nature of the issue depends on the dimmer design and the power supply circuitry in the lamp. Your dimmer modules are most likely forward firing. LED power supplies don't always get along.

You may have better luck if you replace at least one of the LED lamps on each circuit with an incandescent. It will change the load characteristics and maybe fix some of the issues that the dimmer is facing trying to work with an incompatible load.
 
Hmm, alright. I'll throw some incan bulbs into the end of each row, and see what happens!

To be perfectly honest, i'm surprised they actually bought these rather nice bulbs, I would have thought they were installing cheap capacitive-dropper bulbs...

Looking again at the bulb, the PSU on it is only rated for 90-265V
With you bringing up voltage specifically, I wonder if these are intended to dim via some kind of Controlled Current supply?
 
Last edited:
Again, try some ghost loading. There is also a minimum LED load needed to work correctly when using a theater style dimmer. Usually about 75 watts of load. Below that, the leakage in the firing circuit in the dimmers can be a problem. Still, the core issue is that the LED lamps each contain a DC power supply and in some cases it is a half-wave supply (one diode) that actually end up with the two (dimmer and lamp) interacting with each other, thus the oscillation. The phantom load of a conventional lamp will insure that the voltage in the circuit returns to 0 at the zero-voltage-cross-point. Start with a 100 watt phantom load lamp, then lower it and see how little you can get away with.
 
Hmm, alright. I'll throw some incan bulbs into the end of each row, and see what happens!

To be perfectly honest, i'm surprised they actually bought these rather nice bulbs, I would have thought they were installing cheap capacitive-dropper bulbs...

Looking again at the bulb, the PSU on it is only rated for 90-265V
With you bringing up voltage specifically, I wonder if these are intended to dim via some kind of Controlled Current supply?

I have some LED lamps on SV racks, and would also second the ghost load idea. You might also want to look at the manual on page 86, which explains how to build a custom profile to attach to a dimmer, so that you can get a smoother and more linear fade curve. See http://www.strandlighting.com/clientuploads/directory/downloads/CD80SV.pdf

Rob Montgomery
 
The specific behavior described sounds like those LED lamps are causing the SCR in the dimmer to get stuck on. That can happen either for random individual AC half-cycles (causing strobing) or until the dimmer is turned off (stuck at full). This sort of problem usually gets worse as you add more LEDs, so experiment with finding a ghost load that fixes things on whichever circuit has the most lamps on it.

Also, note that a lamp specified to dim from 10% to 100% is going to appear very bright at its minimum level--much brighter than what you would expect from an incandescent lamp set to a control level of 10%. Picture a 10W lamp next to a 100W lamp and you'll have a better idea of what that 10% minimum actually means.
 
I'll have a chance on Monday or Tuesday to test a ghost load.

Just for reference, my understanding (quite possibly incorrect) is that the SV is a fairly different beast from the original and AE models.

The smallest number of lamps in one of the house circuits is 11, and each lamp is a 26W model.

I'll get right to testing when i get back to the auditorium (presuming i don't get distracted by the new instrument DHL is delivering. New thread for that one though ;) )
 
It appears I wasn't the only one working to solve this. An ancient work ticket was finally filled to fix the house lights, and the GE bulb I mentioned turns out to be the replacement...

Turns out they got to the bulbs right before I went and looked for a model number.

So, yes those GE bulbs are working flawlessly, aside from the contractor buying COLD WHITE! ARGH!
 
Again, try some ghost loading. There is also a minimum LED load needed to work correctly when using a theater style dimmer. Usually about 75 watts of load. Below that, the leakage in the firing circuit in the dimmers can be a problem. Still, the core issue is that the LED lamps each contain a DC power supply and in some cases it is a half-wave supply (one diode) that actually end up with the two (dimmer and lamp) interacting with each other, thus the oscillation. The phantom load of a conventional lamp will insure that the voltage in the circuit returns to 0 at the zero-voltage-cross-point. Start with a 100 watt phantom load lamp, then lower it and see how little you can get away with.
I had a show once with a quantity of the little flickering neon flame lamps needing to be individually controlled from our CD80 AMX-192 racks. I found I could get away with a 6 Watt 120S6 lamp in parallel with each neon flame lamp. The designer didn't need to dim the neons he just wanted to be able to cue each lamp individually in his LXQ's. We started with more substantial dummy loads but found we could get away with the physically tiny 120S6 lamps secreted in black vinyl electrical tape right below the neon flame lamps.
Toodleoo!
Ron Hebbard.
 

Users who are viewing this thread

Back