EL Inverter and transformer advice

[philba]

I'm planning to build an inverter to drive electro luminescent wire. Now before you all shout "buy one, they are way cheap", I am looking to experiment with frequency and voltage. In addition, I want to experiment with some ideas that require more control over the voltage going to the EL wire. I've got lots of of those cheap inverters.

So, first the requirements:
- runs off 3V
- frequency from 400hz to 4Khz
- max 50 mA on output side, probably more like 10 mA
- output voltage initially at 100V. intertested in varying up to 150V (see below)

I'll probably wind my own transformer(s) which is the main area I have questions on. I am pretty clueless about magnetics. To start with, I'm planning on using essentially a multivibrator circuit to drive alternate halves of a CT primary on the transformer. The secondary would deliver the 100-150V AC to the EL wire. Does this make sense to you all? Is there a better circuit that i should use and still be able to vary the frequency? I have seen some designs with a feedback coil tbut that seems to lock in the frequency and I have no idea how to determine that anyway. Most of the transformer pages I've found deal with 60 hz so the first question is what are the implications of my frequency range (400 to 4K)? Should I just follow the cookbook approaches for 60 hz stuff and not worry about efficiency? What are the best cores for those frequencies? what about a toroid? I see them all the time in the surplus catalogs though they seldom say the material. I did find several audio transformer winding pages that may be applicable so I will spend a bit more time there - of course it's audio focused so I need to figure out what isn't applicable.

on varying the output, is using a variac to control the voltage a hair-brained idea? I don't know if it will work at 4Khz.

Finally, is there any free simulation software for transformer design? most of what I've found via google is not free (and not cheap, either). I'd like to try different core materials and see how they might perform.

 

[Hero999]

EL stuff is basically a phosphor sandwiched in between two electrodes. It electrically behaves like a capacitor with a huge dialectric loss. The amount of current flowing is dependant on the frequency as well as the voltage, dose the datasheet specify the capacitance?

At 4kHz you're better off using a ferrite core for the transformer as laminated steel will have too high an eddy loss.

Here's an inverter circuit, I've used to power flopurescent tubes.

L1 is a about 20 turns round a small ferrite core, it mearly stops surges when Tr1 and Tr2 are on simultaniously (which happens at part of the oscillator cycle).

T1's primary is 10 or so turns of 0.5mm wire wound bifilar style, the secondary is 100 turns 0.15mm. You'll probably want to adjust this ratio to give you the voltgae you want, you'll probably want to use more tuns on the primary and secondary to lower the frequency.

C1 and C1 might be required to get the oscillator running, it it works without them then good for you. I'd start off with 1nF then adjust their value to give best results.

R1 and R2 depend on the gain for Tr1 and Tr2, and will also need adjusting, use between 1k and 10k, start with 4k7.

Tr1 and Tr2 can be any general purpose transistor, but check the power ratings first.

 

[philba]

thanks for the response. The lytec/elam datasheet doesn't specify capacitance. I measured around 7 nF on a 4 ft length of their 5 mm Red "high bright" wire but wasn't able to get consistant readings (because of the leakage?).

the circuit you posted is similar to others I've seen. I'll probably start with 10 mA which at 3V is 400 mA on the primary for a 120 V secondary.

edit: I measured a 10" length of 5mm blue EL wire and got a consistant 2 nF. So, the wire seems to have 1.75 to 2.125 nF per foot.