Op-Amp boost with power transistors.

[Don_de_marco]

Hi,
I want to build this op-amp based amplifier, using theese comonents and this circuit:

http://www.party-concept.dk/opamp_boost.gif

Can any of you help me with determining the size of the resistors, if I want the specs. to be:

PL= 3W at RL=8Ohm (When Vin=1Vpp).
IoMax<1A
Tamb<85 degrees celcius
Rin=100KOhm
Can resist eternal shortcircuit at both DC and AC.

Hope anyone has some tips for me!

 

[thespeakerguy]

Ilimit = 600µA?

Looking at the positive side only...
Suppose the load draws more than 1mA. This current flows through R5. This creates a drop of 1V across R5. This one volt actually is clamped by the base-emitter junction of Q4, and about 400µA of current flows into the base. Multiply by 100, and it's clear Q4 is in saturation, stealing all the base current from Q5.

The LF356 can supply more than 1mA unboosted. This circuit cripples the output.

 

[Don_de_marco]

Thanks thespeakerguy,

I simulated the curcuit and build it for a test, and that worked fine.

Now I just have to deal with some distortion at high output power, but I will try with some Lead-compensation, and a couple of well placed capacitors.

 

[richie00boy]

I can't work that circuit out. It looks to me as though that output stage is connected to the offset null pins of the chip

Also, a simpler and more stable way would be to just add a buffer pair of transistors to the output, biased with 3 diodes. This kind of thing is shown in most op-amp textbooks.

 

[Don_de_marco]

Right you are...something went wrong with the artist who made that schematic

Of course the transistors open when the op-amp draws a certain amount of power thru the resistor, and then again the resistor mounted on the Tip3x makes shure that no more than x amps can be drawn form the entire system.

Works for me!

 

[richie00boy]

No offence, but it's a very crude method and suffers from a bad 'kink' in the operating curve around the point when the power devices kick in. It would be less parts, lower distortion and more stable to use a single pair of devices in the normall buffer configuration

Check out a diamond buffer if you really must use more parts

 

[richie00boy]

An externally hosted image should be here but it was not working when we last tested it.

Above is an example of a neat little circuit that should do the trick. It was intended as a headphone amp or powerful buffer, but I've scaled it up to your 3W with just a few component value changes and bigger output transistors.

There is no short circuit protection. This can be added by base-clamping transistors sensing the emitter resistor voltage like below. The values of resistance forming the potential divider will need adjusting to suit your application, these are random values.

An externally hosted image should be here but it was not working when we last tested it.