designing an op amp with complementary push pull driver?

[bobcat1]

complementary op amp circuit

Hi All

I would like to design a simple current boster op amp with complementary push pull driver who can work from a single supply(mandatory)

It should drive up to 150 mA of current so 2N3904 & 2N3906 should serve best as a complementary peir in the design

The circuit should work from +9V and have a Gain of 5V/V max, the op amp should be LM358 or compatible single supply IC.

I would appreciate any kind of help and direction

Thanks

Bobi

 

[Audioguru]

push pull driver

The lousy old LM358 has 3% of crossover distortion because it was the first low power opamp and they eliminated the bias on its output transistors to reduce its current. Also its respose fails above only 2kHz.

Any opamp (even the good ones) will operate with a single supply if its input is biased near half the supply voltage.

I would use a low noise, very low distortion, wide bandwidth and inexpensive TL071 audio opamp.
The TL072 is a dual and the TL074 is a quad.

 

[FvM]

There are a lot of electronic applications, that don't require low THD and audio or even higher full power bandwidth. Thus LM358 still serves it purpose. Unfortunately, no respective specifications have been given for the driver.

But I agree, that a single supply specification is meaningless, if you don't tell the required minimum amd maximum output voltage. In addition, I don't understand the "5V/V max" gain specification related to a current booster.

 

[bobcat1]

HI

Sorry for miss understanding

I am looking for a low cost solution that involved op amp and a driver that can work using singl supply

The design can have any low cost op amp - the LM358 was just an example.

The amplifire should have gain of 5

See attached schematic - i need the same but working from a single supply

Can you help please?

Thanks

Bobi

 

[Audioguru]

Your (a) push-pull output stage has no bias for the output transistors so they remain turned off for plus and minus 0.7V of the output swing creating horrible crossover distortion. If the opamp is very fast (but the LM358 is very slow) then negative feedback from the output to the input of the opmp will reduce some of the distortion. The 470 ohm resistor does nothing but waste voltage swing.

Your (b) push-pull output stage again has no bias for the output transistors but the opamp by itself can drive the load through the 220 ohm resistor until the signal swing reaches about plus and minus 0.7V when the output transistors begin working. the opam's gain changes drastically between the high gain when it tries to drive the load and the low gain when it drives the output transistors which causes noticeable distortion, especially with the very slow LM358 opamp.

A single TL072 dual low noise, very low distortion wide bandwidth opamp costs $.56US at Digikey today. A lousy old LM358 costs pennies less.

It is easy to bias an opamp at half the supply voltage in a single polarity supply amplifier like this:

 

[FvM]

If you write GND for V-, you have single supply. I guess, you also know how to add the two resistor feedback network to set a gain of 5.

But the driver may be unable to achieve an required minimum output voltage, unfortunately it isn't specified.

Audioguru is basically correct regarding the crossover distortion problem of a class B output stage. When driving this output stage with a LM358, you have two class B stages cascaded, further increasing crossover distortions. But as said, it may be still sufficient for some applications, e.g. a servo motor driver.

 

[Audioguru]

Bobcat,
What is this circuit driving and over what range of frequencies?

If it is an audio amplifier driving an 8 ohm speaker then it will sound awful and will have low output power.

 

[bobcat1]

Hi

Thanks for your replay

The circuit should drive a 1:1 transformer with a signal of 81.952 KHz
and a 9V ptp signal.

circuit distortion should be kept to less then 1%.

The source is a sin wave DDS generator (AD9850) with 200mV/ptp output

I am looking for a basic schematic (in any format) that can be simuleted using SPICE program

Thanks

Bobi

 

[Audioguru]

Hi Bobcat,
You need a voltage gain of 9V/0.2V= 45.
But with a low supply voltage the lousy old LM324 or LM358 has a voltage gain at 85khz of about 6 then will have no extra gain to reduce distortion.
At 85kHz its max output is only 1.25V p-p.

A fairly good opamp like the OPA134 will have a voltage gain of about 105 at 85kHz so it can have a little negative feedback. The output level can be almost at the supply voltage.

You will need to bias the output transistors with a Vbe multiplier transistor to reduce their crossover distortion.

 

[bobcat1]

Hi

Yes you are right - the LM358 has low gain @ 85KHz and I will replace it.

Can you please upload a schematic diargam who show your proposal?

All the best

Bobi

 

[FvM]

The distortion spec clarifies, that a class B output stage is not recommend (at least, it would require excessive loop gain to reduce crossover distortions below the 1 % level). Providing a GBW of 10 MHz and more can also achieve the intended specification with a class B output stage.

You will need to bias the output transistors with a Vbe multiplier transistor to reduce their crossover distortion.

Right, if the output stage shall be operated class AB. But this involves also emitter resistors to prevent thermal runaway of discrete output transistors and at least a current source. Thus I wonder, if an integrated audio amp or power OP would do the job with less effort.

 

[Audioguru]

Can you please upload a schematic diargam who show your proposal?

It is very simple and you can do it:
1) Good opamp.
2) Vbe multiplier.
3) Complimentary emitter-followers with emitter resistors.
4) Overall negative feedback.