A few years ago I started working on a DIY power supply to power EL tape and panels. Those require 100-200V at 400Hz-3kHz. I've designed a circuit in LTspice and started a breadboard but haven't gotten around to finishing it yet.

My plan was to run a 9-12 V wallwart transformer in reverse, supplying a 12V, ~1kHz square wave to the secondary and look for 100-150V at 1kHz on the primary.

There's one question to this strategy, and that's whether the wallwart core can operate at frequencies as high as 2kHz. If you've got a wallwart and a signal generator, could you answer this question?

The secondary of a wallwart in only about 2Ω of DC resistance, so your signal generator will need to be OK with that or else you could put a resistor in series to drop the current. I have no idea what the impedance response to frequency will be. It would probably make sense to put a nightlight or other small load on the high-voltage side. That would make it easy to see the brightness tail off as the frequency increases.

 

How much current (peak) will the EL stuff draw? Knowing that will inform whether or not the wall-wart/sig-gen set-up is feasible.

 

How much current (peak) will the EL stuff draw? Knowing that will inform whether or not the wall-wart/sig-gen set-up is feasible.

It varies with the size of the tape or panel. Here's some info from a manufacturer. I'd probably like to light 20-50 square inches, so up to 1W or so. Roughly 10mA.

 

There's one question to this strategy, and that's whether the wallwart core can operate at frequencies as high as 2kHz. If you've got a wallwart and a signal generator, could you answer this question?

I took a 12 VAC, 375 mA wallwart and drove it in reverse from my signal generator at 2 kHz. Worked fine, 5 Vp-p in gave 50 Vp-p out with no visible distortion.

 

Hi Wayneh,
I have found a 240 V to 12 volts 1 amp AC output wallwart. The secondary resistance is 1.2 ohms. I have driven it with a signal generator fed into an audio amp. When you say 12 volt square wave I assume you mean 24 volts peak to peak. It seems to work quite well but there is a lot of ringing on the transitions as to be expected with the secondary (Original primary) unloaded. The input on my scope has a maximum volts per division of 5 (50 with the divde by 10 probe.) so I had to drop the input to just below 20 volts peak to keep the output on the screen. The first picture shows the full waveform.


The second picture just shows the ringing on one transition.


The transformer worked better than I expected at 1Khz. I willleave things set up in case you want me to put some resistve or capacative loading on the output. (I think these EL tapes look like a capacitor.) Sorry that the pictures are over exposed and a bit blured. (Hand held and auto exposure on the camera.)

Les.

 

I will things set up in case you want me to put some resistve or capacative loadinf on the output. (I thing these EL tapes look like a capacitor.)

This and the report from @OBW0549 is great news! I'll go ahead and complete my build.

You're right, EL is a lot like a capacitor and it would be useful to test driving a capacitive load. Looking now for a useful range of values... [update] Looks like 1 cm wide tape is about 10nF per meter. So testing in the range of 10-100nF would make sense.

It looks like the target drive conditions have changed a bit from when I started this project. LightTape (perhaps the leading manufacturer) is showing 800Hz and 250V. Based on your results, the 800Hz should be no problem but my voltage may be on the low side. To some degree, voltage and frequency are interchangeable since they both are proportional to output brightness.

 

When yo say 12 volt square wave I assume you mean 24 volts peak to peak.

Actually I was thinking of 12V DC pulses. This might allow operation on a boat or in a car. I may instead use an old laptop power brick at 19V to get more output voltage.

 

Here is the output waveform with a 10 nF capacitor across the output. I had to reduce the input amplitude to get all of the output waveform on the screen. The bottom waveform is the drive waveform on the transformer low voltage winding.


With a 47nF capacitor across the output the output was almost a pure sine wave. (And even greater amplitude.) I think you will have to do some tests with the EL tape as it will probably have a resistive component which will damp the ringing.

Les.

 

Here is the output waveform with a 10 nF capacitor across the output. I had to reduce the input amplitude to get all of the output waveform on the screen. The bottom waveform is the drive waveform on the transformer low voltage winding.

With a 47nF capacitor across the output the output was almost a pure sine wave. (And even greater amplitude.) I think you will have to do some tests with the EL tape as it will probably have a resistive component which will damp the ringing.

Les.

Yup, will do. The critical question was the transformer and that question has been answered. I was also a little concerned about prematurely aging the EL tape with a square wave, because of the high frequency components. It looks like the tape won't really see those sharp edges.

 

Here is the output waveform with a 10 nF capacitor across the output. I had to reduce the input amplitude to get all of the output waveform on the screen. The bottom waveform is the drive waveform on the transformer low voltage winding.
View attachment 167696

With a 47nF capacitor across the output the output was almost a pure sine wave. (And even greater amplitude.) I think you will have to do some tests with the EL tape as it will probably have a resistive component which will damp the ringing.

Les.

If your rig is still set up, could you try one more thing? I'm using a biggish capacitor, eg. 1000µF, to couple the pulsing DC signal into the transformer. One leg of the transformer is grounded. The new secondary, formerly the primary, is floating. My breadboard build is working but I'm seeing only about 60V at the output. That may be due to current restrictions from the breadboard, the wires and such, or it could be another artifact of my meter. Just curious what you'd see in your setup.

 

I took a 12 VAC, 375 mA wallwart and drove it in reverse from my signal generator at 2 kHz. Worked fine, 5 Vp-p in gave 50 Vp-p out with no visible distortion.

See "one more thing" request in #10.

 

I was driving the low voltage winding (The original secondary which is now the primary ) with a TDA2050 amplifier IC fed with a +22 and -22 volt supply so it was a true square wave AC with no DC offset. It sounds like you are using a transistor for active pull down but you don't say if there is just a resistor to the positive or another transistor for active pull up. I have now dismantled the test setup.

Les.

 

See "one more thing" request in #10.

I already had a capacitor, 470 uF, in series between the signal generator and the transformer because I didn't want the SG injecting any DC into the transformer.

 

I already had a capacitor, 470 uF, in series between the signal generator and the transformer because I didn't want the SG injecting any DC into the transformer.

OK, I'll proceed by shortening and beefing up all the wires between my power supply and the transformer. My EL tape has arrived, so I can start testing soon!

 

OK, I'll proceed by shortening and beefing up all the wires between my power supply and the transformer. My EL tape has arrived, so I can start testing soon!

Don't forget that the eddy current losses will go up as the square of the winding voltage, so if you get 200V across the original 110V winding the core losses will be nearly 4x that of the original power transformer running at 110V (independent of load). I suspect your main power consumption will be in driving the core losses, so don't use a bigger core than you need. Copper losses are probably negligible as you are taking minimal power.

see: https://www.edn.com/design/componen...sformer-at-a-frequency-it-wasn-t-designed-for