Help!!! Looking for Transformer for DC AC Inverter

[shenlufei2005]

Hello everybody!

I am designing a DC-AC Inverter for an EL-Lamp. The DC Input is 12 VDC. And the AC output should be about 120 or 130 VAC at the frequency above 20 KHz. The whole circuit should have a small size lower than 2cm X 4cm, since the EL-Lamp ist about 20 cm2. Because we don't have signal generator with large Outputvoltage like 120 V, or available Amplifier , I have done the Measurement and Characterization of this EL-Lamp by using Keithley sourcemeter, which can provide approximately through programming 100Hz 120 Vrms Sine Wave Output. That means, the modelling of this EL-Lamp is under low frequency. According to the general equavalent circuit of an EL-Lamp, I think it should consume about 1 W Power.

My idea is to use a DC-DC converter to generate about 36 VDC and after that the 36 VDC will be converted to 36 VAC and by using a high ratio transformer, I can transform the 36 VAC to about 180 VAC. Of course, the Waveform is still Rectangular. So use a low pass filter. I may obtain a pure sine wave output voltage with Vrms=130V at frequency above 20 KHz.


However, I cannot find the transformer, which is suitable for this application, namely high frequency, high voltage , reasonable small size.


Can anybody help me? Or has somebody found the mistakes of my design idea? The DC to DC Converter Chip, which I am going to use is MC34063 of On Semiconductor.

Thanks


Lufei

 

[WimRFP]

Hello,

Why do you need the additional DC-DC conversion?

I think you can go directly from 12V DC to the requried output voltage with a push-pull type oscillator (2 transistors with center tapped primary). One example is the "Royer oscillator". It is not the most effective one with regards to requried magnetic volume, but it is easy and provides a sine wave directly. It was frequently used in CCFL converters.

Regarding the transformer, I think you have to calculate this yourself so you have dive into magnetics. As your requireements are just 1W, and voltage is not that high, it can be a small one when using modern ferrite materials.

for application information look at EPCOS (former Siemens) or Ferroxcube. Digikey, Farnell, RS, Schuricht, etc have small cores in stock.

 

[shenlufei2005]

WimRFP,

thanks for the reply.

Sorry I forgot to mention that the efficiency is also an very important issue if the power is from Battery. The Dc-DC Chip can usually provide better about 90% efficiency. But the oscillator, I think, normally 70% , isn't?

 

[WimRFP]

Hello,

Such switching oscillators at these power levels have efficiencies generally over 80%. The loss is mostly in the copper or the magnetic core. copper and core loss is also present in your two step approach. So I think the single step approach will have better overall efficiency. As it will also provide you a sine wave, you do not have to add additional filtering.

When costs are not the dominant issue, you can use core material intended for over 100 kHz at a lower frequency. This will give good overall efficiency.

The Royer oscillator is zero current switching oscillator, so switching losses (even with general purpose bipolar transistors) are negligible. In most descent designs, the transformer has a an air gap as it has to store the resonant energy. The air gap also provides defined inductance, with low depencend on variation in permeability.

One thing that people may forget is that the maximum Vce is about 3.5*Vsupply.

Off course you can also use other topology (for example half bridge via controller IC), but the Royer type works very fine with (cheap) BJTs and requries few other components.