Simplest 10W mono audio amplifier?

Yes, you can use an LM317 but consider that your 'no modulation' condition requires 50% of the maximum voltage at it's output pin. That means that the other 50% plus the ~3V overhead of the LM317 are across the device.

Example: you have a 28V supply. Lets say you lose 3V in the ovehead needed by the LM317, that leaves 25V for the PA stage. For AM you set the average carrier level to 50% so you want half the supply to feed the PA, thats now 12.5V. This will rise and fall according to the modulation of course. If the PA draws say 1A, the average heat dissipation in the LM317 is (12.5 + 3) * I = 15.5W.

So it isn't very efficient. If you uses a class AB or class B modulator and transformer you could significantly increase the overall circuit efficiency.

Brian.

BradtheRad said:

I've seen articles about making an LM317 into a power amplifier at audio frequencies. So your idea is an atypical role, although not totally farfetched. You'll have to test to find out how it behaves at higher frequencies.

Click to expand...

I do not know if this is exactly like you mention. I have seen lm317 work as an audio power amplifier. However, I think the way it is done in gunn modulators is a bot different.

For AM modulation in my voltage-output-power-sensitive ransmitter, you only need to vary the vcc at a speed of about 6khz max (AM bandwidth).
The lm317 can be used as a variable voltage source and it's voltage can be varies from about 1v or so to 30v or more.

The whole purpose is to have a single stage do the job of a multi-stage power amplifier and output transformer. In other words slam signal line audio can be injected directly to the lm317 and vary it's voltage.
A way I think is to replace the adjust resistor to ground with an LDR that is facing onto a LED, which is driven by line audio signal.
A small variation of the LED brightness can cause a big variation of output voltage, hence the TX can be AM modulated.

How do you find this idea?

- - - Updated - - -

betwixt said:

Yes, you can use an LM317 but consider that your 'no modulation' condition requires 50% of the maximum voltage at it's output pin. That means that the other 50% plus the ~3V overhead of the LM317 are across the device.

Example: you have a 28V supply. Lets say you lose 3V in the ovehead needed by the LM317, that leaves 25V for the PA stage. For AM you set the average carrier level to 50% so you want half the supply to feed the PA, thats now 12.5V. This will rise and fall according to the modulation of course. If the PA draws say 1A, the average heat dissipation in the LM317 is (12.5 + 3) * I = 15.5W.

So it isn't very efficient. If you uses a class AB or class B modulator and transformer you could significantly increase the overall circuit efficiency.

Brian.

Click to expand...

All right, I see... it is a matter of power efficiency, like in the series transistor/mosfet case in previous posts.

Please see my previous post #20 of how I am thinking of doing it. This would hopefully require only an lm317 without additional active stages for full modulation.

Now I wonder about the efficiency issue you mention, why is the transformer more efficient? I mean the audio amplifier itself could require 15W or so (tda7360 in bridge mode), so the total system efficiency might be worst, despite the modulation efficiency is better with the transformer.
Do I have a correct point here?

A class-A amplifier has high current all the time, even when there is no audio. But a class-AB amplifier draws current when there is audio but draws almost no current when there is no audio or when the audio level is low.

An LM317 used as an amplifier is class-A so it heats all the time, even with no audio.

Audioguru said:

A class-A amplifier has high current all the time, even when there is no audio. But a class-AB amplifier draws current when there is audio but draws almost no current when there is no audio or when the audio level is low.

An LM317 used as an amplifier is class-A so it heats all the time, even with no audio.

Click to expand...

Ah yes, I see your point, absolutely understood.
However,
In my design I am going to use an LM317 eitherway, as a variable power supply, to vary the output power of the transmitter, since the amount of output power depends of the Vcc value.
As you would guess, when the Vcc is set too low, almost all the power is consumed in the LM317, but this is the only simple and cheap way I can think of, in order to vary the Vcc from a fixed higher voltage transformer.

Given this fact, it migh be more convenient to avoid extra harder to find components (like the transformer). Indeed if a kit is to be produced, one might want to avoid the transformer at all. And if the regulator can AM modulate as well, this doubles it's job.

So in the one hand I have efficiency and on the other cheaper components.

Which way would you go in my case?

Well, look at it this way - if you use a transformer, the secondary (PA side) will only have a resistance of a few Ohms so you can still pass maximum current at maximum voltage to the PA stage. The signal fed in at the primary will be added to the PA voltage. For example if you are using a 28V supply and the secondary produces 10V of modulation voltage, the PA will see a nominal 28V with peaks at 38V and dips of 18V.

As Audioguru and I pointed out, the modulation signal can be produced from an energy efficient class AB or class B amplifier.

Brian.