I'm hoping someone on here may be able to help with a problem I have. Some background first:

I have been asked to run an experimant using a bipolar waveform at 500V peak to peak. To achieve this the signal from a signal generator was passed though an audio amplifier then through a small audio transformer. A basic set up I know but worked well for the application it was required for.

The problem:

I have now been asked to run a similar experiment using a unipolar square waveform of the following parameters

Vout - 60V-120V
Frequency - 50 Hz
Duty cycle 25%

I have a custom signal generator that can produce a maximum of 10V but if I put the unipolar signal through the original set up in comes out bipolar. I have been told this is because the audio amp does not have a 0V reference so cannot distinguish between bipolar and unipolar.

Does anyone out there know a way of amplifying the signal from the signal generator whilst still keeping it unipolar?

Thanks in anticipation of any answers.

 

Is duty cycle of signal generator adjustable? Is an inverted output acceptable?

 

An audio amp does have a 0 volt reference - circuit ground. If it is DC coupled and gets fed a DC signal, the output will also be DC.

The 500 volts might be a challenge is any power is needed. Do you have a 500 volt source? If the frequency is not too large, about any IGBT can switch that level of voltage with a 0 - 10 volt drive signal.

 

For an example , Google up National Semiconductor, AN 72 pg 43 fig 94 ; some uses of LM3900 as high V amp, inverting & non-inverting.

 

If you do not have a 130 to 150V power supply, we do not recommend that you half wave rectify line voltage with 1N4004 diode and filter with 10 to 50 uF @ 250V cap.

 

Thanks for all of you replies. You'll have to excuse the comments or answers, I'm not very knowledgeable about electronics, I'm on a bit of a steep learning curve.
I have some answers to the comments as follows:

An audio amp does have a 0 volt reference - circuit ground. If it is DC coupled and gets fed a DC signal, the output will also be DC.

Not sure whether the amplifier is DC Coupled or not, model being used is a Sony TA-FE370. When the unipolar sqaure wave is applied the output looks more like a bipolar saw tooth wave.

The 500 volts might be a challenge is any power is needed.

I managed 500V for the first experiment, using the amplifier and transformer. For this experiment I only need an output of around 60-120V, the output of the signal generatir is 0-10V, hardly any power is required as signal will be used in an electrolysis type set up to determine the effect of different waveforms and parameters on the corrosion of the electrodes.

For an example , Google up National Semiconductor, AN 72 pg 43 fig 94 ; some uses of LM3900 as high V amp, inverting & non-inverting

Thanks for that, I've had a look at the non-inverting amp, looks like the sort of thing that I'm looking for as far as I can understand, although I'm not sure if I have a good enough power supply.

Hope these help clarify some points. Thanks for in adveance for any further help.

 

The problem here surely is that a 25% unipolar square waveform has a DC component. A transformer works on AC only!!

Get rid of the transformer.

You probably need a special amplifier that will give the output voltage levels directly. DC coupled.

 

Transformer has been ditched, I have just been looking at the audio amp we have.

Looks like I need a DC coupling amp then.

 

Maybe one of these high voltage op amps would help out? - http://apex.cirrus.com/en/products/pro/detail/P1156.html

 

You might be able to make this work if your load impedance is high. The sim shows a 50/60Hz power transformer used as a step-up. It must be driven from a very low impedance, hence use your audio amp to drive the low voltage winding on the transformer. Use the high voltage winding to drive your load, with a clamping diode and series capacitor as shown.

The output waveform exhibits some distortion due to the diode, but short of looks like the input waveform stepped up ...

Google "diode clamping circuit"