I'm in the planning stage of a project to light about 8m of EL tape. I need some ideas about the overall strategy for powering it.

As I understand it, EL wire or tape acts like a capacitor, about 10nF per meter of 1cm wide tape. To light it up, it needs over 100V at 400Hz or more, up to say 2500Hz. Let's say the target is 120V at 400Hz (for longer life of the tape). I believe this works out to ~100mA for the 8m of tape.

So the project is to make an inverter producing 100mA at 120V AC at 400Hz. It will be battery powered and I'm thinking 12v SLA but weight and size is a concern, so I'm open to other options. Run time up to a few hours will be an issue, as 12W will deplete any battery at a good clip.

I'm thinking of just running a wall wart transformer in reverse to step up the voltage, and just powering it with a 555 timer at 12V to set the frequency, and a MOSFET to switch current to the secondary. I'll need over 1 amp at 12v to step up to 120V at 0.1A.
Something like this?

Does this make sense? Should I be concerned about supplying a square wave to the transformer? What else might be required? I understand that there are boost converters that could give me the voltage without a big transformer. Would it make sense to pursue that route?

 

Well just try it and see.

I've learned all the electronics i know today by trial and error over the years.

 

I think they need AC, so the circuit you have might be just the ticket. Let us know how the transformer works at the higher frequencies.

 

That tape works like a capacitor with a lossy dielectric, the more DV-DT the brighter, you will find the brightness fairly proportional to the frequency.

Beware of the pathetic life span of these things, 2K hours or so- terrible.
The harder you drive it, the faster it will die.

 

Any boost converter recommendations? Ditching the transformer would be nice but I've never worked with a boost so I'm not sure how I might control the frequency and amplitude of the output.

 

I think the problem with a boost type is since the thing looks like a cap and the inductor releases it's energy very quickly the current in the strip is very high but very short.

 

How about another angle: Could a commercial 12VDC to 120VAC inverter like you'd use in your car, be hacked to run at 400Hz and above? Those obviously don't use big transformers and yet can handle quite a bit of power. It'd be perfect for my project if only the frequency were higher. Or are there aviation inverters out there that might work (I think they run at 400Hz?)

 

You don't need that amount of power for driving EL strips/wire.

Commercial drivers are just the size of like a CCFL inverter in like a flatbed scanner and such.

 

Here is one:

http://www.ebay.com/itm/12V-DC-to-A...709?pt=LH_DefaultDomain_0&hash=item35d1f0ef35

 

Interesting. I wish it allowed turning down the frequency, since the life of the EL device is related to frequency. But that thing is cheap enough and rated high enough to be exactly what I need.

Oh, wait, it's only rated for 400 cm2 of EL tape. I need 8000.

 

Every commercial EL driver I can find, while impressive, cannot handle the current of 8 meters of EL tape (800cm^2). While I could just use several of them in parallel, instead I'm proceeding with building my own inverter based on using an h-bridge to drive a big old wall-wart in reverse. A nagging concern is whether the transformer core will sustain frequencies up to 2.5kHz or so. We'll see.

In my thread on the inverter design, RichardO provide a very nice reference regarding EL panels, excerpted below.

The folks at adafruit quote these values for the EL tape.
Size: 100 cm x 1.5 cm about 1mm thick
Panel lifetime: >25000 hours
Operating voltage: 60-250VAC
Operating Frequency: 50-5000HZ
Current Draw: 0.14mA/cm2 (max) @ 110V / 400Hz [14mA for 100cm^2]
Initial Brightness: 40 cd/m2
Operating Temperature: -50 C / 65 C
~10 nF per meter [0.645nF/in^2 this is much lower than the 45nF per 100cm^2. An error?]

For EL Panel squares they provide this:
Glow Size: 10 cm x 10 cm (3.95" x 3.95")
Plastic Size: 10.4 cm x 10.4 cm (4.1" x 4.1"), about 0.5mm thick
Panel lifetime: >25000 hours
Operating voltage: 60-250VAC
Operating Frequency: 50-5000HZ
Current Draw: 0.14mA/cm2 (max) @ 110V / 400Hz [14mA for 100cm^2]
Initial Brightness: 85 cd/m2
Operating Temperature: -50 C / 65 C
45 nF per panel [2.9nF/in^2 45nF per 100cm^2]

 

An audio transformer might be better than a 60 Hz one. Also , do you mean 800 sq cm?

 

The circuit looks all there to me.

 

I have a 400 Hz filament transformer, 120 V, 6.3 V CT @ 1A, 6.3 V ct @ 1A, good to 10,000 ft, never used WWII surplus that you may have for shipping $.

 

That's an awesome option. I may take you up on it, but let me first nail down the details of the EL tape and the project. My initial estimate was 800 cm^2 (yes, square cm) of lit tape (8m at 1cm lit width). I've found a U.S. supplier that makes a higher quality tape and they offer it in a thinner, 1/4" lit tape. So I'm still figuring things out.

 

On my boat I have a typical 12V system, so I may use a commercial inverter to get a higher voltage for my EL tape, instead of building my own inverter. I'll still need to generate a higher frequency, but I have some old TV transistors that should easily handle the voltage.

Here is some data that may come in handy: Output of an inexpensive inverter

Device tested: Black & Decker inverter model PI100AESB rated at 100W AC ouput, with a USB port rated to 350mA.

This was bought at Walmart some 5-10 years ago and I don't think it was more than $30. More or less the cheapest I could find with enough wattage for my laptop power brick.

It has a "Caution" on the back, that the output is "non-sinusoidal". I became curious about just what this thing puts out. To make the story short - it's a modified sine wave. It's a square wave with a short flat spot at zero volts between each up or down transition. The flat spot is much narrower than the up or down portion, maybe 1/4th or 1/3rd.

My multimeter readings on the inverter attached to a ~13V battery:
No-load voltage - 92.8V
~10W load - 112V (a small lightbulb)
Peak detector - 154V (bridge rectifier plus 470µF electrolytic, under load)

My oscilloscope is software on an old Mac. I divided the rectified inverter voltage down 1000X with a 1M and 1kΩ resistor. If I can figure out how to get a screen shot off of this ancient computer, I'll post the waveform. Nothing modern I have can read a floppy!

 

An option might be to rectify the high voltage AC, then use high voltage transistors to generate the pulses from that. Then you don't have to worry about the frequency capability of transformers.

 

wayneh:
I just remembered something. The EL power supplies I have for LCD backlights claim that they compensate for EL panel aging. They say they do this by adjusting the voltage and frequency.

Here is a link to the data:
http://www.electronicsdatasheets.com/pdf-datasheets/endicott-research-group--erg/lps1212/


It can't drive your large EL strip but, for reference: about $19 -- if you could find one.

 

An option might be to rectify the high voltage AC, then use high voltage transistors to generate the pulses from that. Then you don't have to worry about the frequency capability of transformers.

That's exactly the plan I'm looking at now. Using a commercial inverter solves a lot of problems for cheap.

 

How do I choose the component (R & C) values to use a 12V push-pull signal to control a higher voltage pulse? The Class B amp circuit attached seems to do a nice job but even modeling it with LTspice I cannot settle on a set of coupling capacitors and current-limiting resistors that seem "optimal". Goal is maximum power into C3 with minimal losses. (Not smoking the components should come along with efficient design.) If it matters, C3 is the EL tape, and has a DC resistance of 10-20Ω.

I've been reading here about Class B and Class AB amplifiers. It seems I want a class B but I'm not sure how to pass the signal and bias the bases of the power transistors.

In the schematic, the feed to R2 is the output of a 555 astable in the range of 100-2kHz. The higher voltage rail at R3 is 100-160VDC. I have a matched pair of power transistors from a TV to handle the pulses.