How Not To Dim EL Panels, TRIACs!

We’ve all been there: an exciting brilliant idea, scratched onto a napkin, hastily plugged into a breadboard, all for naught.  Even the best ideas sometimes suffer from a heavy dose of reality.  [Ch00f] over at ch00ftech recently had a similar experience dimming an EL panel of his using a TRIAC and some clever waveform manipulation. Instead of tossing the parts across the room in a fit aside and moving on he goes into a detailed analysis of what went wrong.

This method differs from the way most EL drivers dim output loads, instead of chopping the output like a PWM controlled LED the TRIAC snips the ends of the waveform and makes an ugly but less powerful output. The issue with this method is that when you cut the waveform during non-zero crossings it causes massive current spikes. These can wreak havoc on a cheap EL inverter and generally cause headaches all around.  [Ch00f ] speculates that his woes may be due to the fact that EL wire is a capacitive load, causing voltage to fall out of phase with the current. This is one of those engineering problems with a thousand and one answers, we can’t wait to see what he comes up with.

Check out the writeup for all the “deets” (as [ch00f] would say) as it is a pretty good primer on TRIAC operation. If there isn’t enough glowy wire in this post you can also check out this sound reactive panel or an informative guide on EL or even more from [ch00f] in general.

15 thoughts on “How Not To Dim EL Panels, TRIACs!

  1. First comment. Sparkfun had an EL inverter shield and arduino that use this method. They sucked phenomenally.

    I got the sequencer, the arduino based one. It sucks at dimming, it always reset itself, and just is terrible in general. I returned the first one as defective, and the second one (that wasn’t much better) I decided to try to fix myself, so I lifted the reset pin from the board and wired in my own reset button and wire so the AC wouldn’t couple with the reset trace and reset the atmega.

    I was gonna write the comment to complain, but it seems they finally pulled both of those products, and have a newer shield available. I will be contacting them to see if they can do any sort of upgrade deal for those that have the shitty hardware.

      1. I didn’t want to reference the thing outright but you got it, the sparkfun sequencer suffered from sharing a ground with non-zero cutoff triacs and it wold destroy the DC electronics. They added a zero crossing isolated triac (with a second load switching triac) to fix that. [ch00f] was actually trying to manipulate the wave, not flip it on and off (like the sparkfun stuff).

  2. They used to have lots of places where they sold EL wire in my neck of the woods, then they fell out of grace and now you might as well try to get blank cassette tapes or floppies or something.

  3. This is what I think is going on. AC mains dimmer used to control a tiny AC to DC to HFAC (inverter). Just to run cap-lighting. Sounds like this is supposed to happen. Just using a cap in series with the incoming AC would be enough. Dwino should control the AC coming out of the inverter (converter) PWM instead of messing with an intermediate voltage.
    There is only one dimmer I use or approve of, a Variac. Triac’s made AM radio a mess, and every guitarist will pick them up live buzzzzzzzzzzzzzzzzz! Plus a free light-show with every power factor change.

  4. Anyone fix the fact that EL wire burns out fast. Loses 1/2 it’s brightness in the first 3 months, and the last 1/2 of it in the next 9 months.

    12 months on lifespan = pretty crappy product.

  5. Wouldn’t the right way to do it be via controlling the frequency of a resonant inverter, bringing it in and out of resonance, with the EL wire in place of a traditional capacitor?

  6. I really dig how near the end of the video, you played the PWM through a speaker to demonstrate the dimming in relation to frequency. What I really liked was that as the output got close to the scanning speed of your camera, you could see the old ‘Apollo’ effect, funky bars on the monitors in mission control. For years I thought all CCTV had ugly bars ruining their picture. (Palm to the forehead)…

  7. If the issue is large spikes when turning off mid cycle, why not turn on mid-cycle and turn off at zero, this is how most digital AC dimmer trac circuits operate, or at least the one that i built does. A snubber circuit can be used to reduce the current spikes, im not an expert on them though, mroe research required

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.