Audio modulate fluoro inverter

A square wave, by definition, has two equal excursions - both time and deviation are equal, so there is no "modulation."

What you want to try is a pulse width modulation scheme where the positive transition is modulated (made longer as sound, modulating signal, gets louder).

PWM is fairly easy to implement - you feed a triangle wave (or the charge/discharge on the cap of a 555 astable) that continually moves between two voltages into one input of a comparator. You feed your audio signal into the other (properly massaged for voltage levels) and what comes out is a rectangular wave with varying on versus off ratio.

An "off the shelf" lamp will operate at some frequency if it is very high with respect to the modulating signal no problems. If it is low enough you will experience aliasing - the clock of the PWM is in and out of sync with the natural frequency of the lamp oscillator and that is likely to be visible.

You are dealing with an unknown lamp - they don't all behave the same. Some come on at a reduced level for a time before coming on fully. It should be a 100% cold cathode lamp with a fairly high driving voltage if you want modulation to work well.

Thank you for your reply. PWM is more like FM. There are existing "dimmer" circuits to do this with a TL494.

Sorry if I was unclear, but what I am looking for is amplitude modulation. For example, most commercial 12V fluorescent inverters run at around 20KHz. How can I best adapt such a circuit to modulate, or gate, the 20KHz with a lower audio frequency squarewave?

Barry

The TL494 will work for PWM, but I don't think you want the feedback, boost inductor, etc they show in the application notes - it is designed or optimized to provide a constant vantage in a boost or buck regulated power supply. Fixed output voltage - not what you are doing

- but it will work,or can be adapted for PWM "modulation."

Amplitude modulation is probably not the best idea. That would be, for instance, connecting the speaker output of an amp into a full wave bridge rectifier, then feeding that to your lamp. I don't think you'd like it.

But you did mention PWM and that is the way to modulate a fluorescent light. So I think we have a communication problem.

With pure amplitude modulation you don't have the "strike" voltage for the fluorescent tube. The tube really wants to see several hundred volts to start and warm filaments. Once it is ionized it may only require 50 volts with a current limiter.

Straight amplitude modulation will be well below the strike voltage half of the time or more.

Enter PWM - the audio signal is converted to PWM and gives the tube the full 12 volts (several hundreds after the inverter) and turns it on and off fast enough that you won't see it with your eyes. The important thing is that even when the "on" pulse is very short in duration, it still supplies the full 12 volts so may work for modulating a fluorescent tube inverter.

Now, you say the tube is working at 20 KHZ - your modulation can't also go up to 20 KHZ (the oscillator in the pwm) it has to be slower to look good. The audio modulation source should probably be below

You have three things that can "alias" with each other. The tube should be ~200 KHZ, the oscillator in the PWM at 20 KHZ, and the audio no higher than 3KHZ for things to be happy.

High frequency is more effective at ionizing lamps than low frequency at the same voltage (or at some given voltage). So, Ideally you'd want the lamp to be running at a much higher frequency than 20 KHZ for audio - and not waiting for a heater to help the gas ionize.

What you want to do, can probably be done. Lamps are moving to higher and higher frequencies as the switching parts get better and cheaper - they don't have to spend as much on the inductors and efficiency goes up.

BUT you may have problems just taking some off the shelf 12v lamp and just adding pwm to it and expecting it to track an audio signal.

"Barry Hughs"

The light output will persist even when the AC drive is gated off for a short time - the phosphors have persistence.

The tube may not ionise and re-light immediately if the drive is removed for a short time.

If you gate the 12 volt DC power, the inverter will have a time delay before it kick starts.

You have not thought this through much - have you ?

....... Phil