Op Amp (OPA134) driven Power Amplifier

[lineup]

Hello!

Op-Amps would be perfect to drive and control power amplifiers
if it wasn't for one thing:
- they work at too low voltage when it comes to higher power.

With the method I will show you, you can drive a power amp
even from an opamp working from +5/-5 Volt supply.

First, opamps can often put out 20-40 mA.
OPA134, which I have used for my SPICE, puts out 35-40 mA.
This is more than enough to drive one power output stage.
A triple Darlington needs no more than like 0.1 mA for plenty of Watt out.
One ordinary Darlington (driver + power transistor)
should do well on a couple of mA. Maybe max 5 mA in worst case.

This tells us, that there is no need to amplify the current from Op Amp.
We can let the output current, as it is, drive the power amplifier output.


If you look at my diagram, you see the opamp output is split in 2 halves
and via foldback transistors fed to the output.
The blue lines show the path.

I have done SPICE for 100 W/8 Ohm 200W/4 Ohm with great results and performance.
OPA134 is good enough for this job.
I think you should have a look att this method if you are intesrested.

Regards Lineup

[stratus46]

Originally Posted by lineup

Hello!

Op-Amps would be perfect to drive and control power amplifiers
if it wasn't for one thing:
- they work at too low voltage when it comes to higher power.

With the method I will show you, you can drive a power amp
even from an opamp working from +5/-5 Volt supply.

First, opamps can often put out 20-40 mA.
OPA134, which I have used for my SPICE, puts out 35-40 mA.
This is more than enough to drive one power output stage.
A triple Darlington needs no more than like 0.1 mA for plenty of Watt out.
One ordinary Darlington (driver + power transistor)
should do well on a couple of mA. Maybe max 5 mA in worst case.

This tells us, that there is no need to amplify the current from Op Amp.
We can let the output current, as it is, drive the power amplifier output.


If you look at my diagram, you see the opamp output is split in 2 halves
and via foldback transistors fed to the output.
The blue lines show the path.

I have done SPICE for 100 W/8 Ohm 200W/4 Ohm with great results and performance.
OPA134 is good enough for this job.
I think you should have a look att this method if you are intesrested.

Regards Lineup

Nothing new about a power amp with more voltage gain after the VAS stage. Dan Meyer of SWTPC used that technique in the Universal Tiger back in the '70s. It did not use an IC opamp like yours but a discrete one. It had an annoying tendency to blow up.

G²

[DCPreamp]

Hi There,

There are a number of variations of the op-amp-front-end circuit you designed in use out there in the real world. Carver uses a similar topology in most of their amps with a TL072 front-end, but the rest of the circuit is fairly complex implementing their class-G/low heat design. Alesis also uses a similar design using an NE5532, but the output stage is inverting, so the negative feedback is fed into the non-inverting input.

Attached is a simple op-amp/power amp circuit using an NE5532 and darlington outputs. It's very basic, but offers great sound and good stability into varying loads.

Paul

[mlloyd1]

also, lets not forget before Boulder went down the "cost no object" path of the current equipment, they made the 250 and 500 series power amps that were basically an op amp (albeit the 990 discrete op amp) driving an output stage with voltage gain.

mlloyd1

[jaycee]

Isn't this the same topology that Jan Dupont's LYNX amplifier uses?

[teemuk]

- they work at too low voltage when it comes to higher power.

And designers worked around that decades ago. I don't see an amp design, merely a sketch of a basic topology that you can very often find from OpAmp driven amps. I've seen this scheme in dozens of guitar amps, for example. If you want something else, you could e.g. bootstrap the OpAmp's supply rails or just use the QSC-style scheme where the output stage itself additionally amplifies voltage.

[lineup]

Here is one of my latest simulations setup.

Some points:
Shows a run with 1Vrms input and 200 Watt out into 4 Ohm.
The output peak voltage is +-40 Vp. From 45V supply.
Distortion according to Fourier is 0.00066%
The AC current from op-amp shows 723 uA rms. Which is like +-1 mA peak!
For 200 Watt ....

The BC639/640 and MJE340/350 runs both 10 mA.
The output stage is biased in Class AB, like 75 mA per Power Device.
MJE15030/31 are drivers for 2 parallelled pairs MJL3281A/1302A

There are 3 Capacitors used to tailor the AC behavior and ensure stability.
C3 - can often be 5 - 22pF, depending on what transistors you use
C1, C2 - can be 100 - 680 pF to make a nice frequency/phase curve


What more can I say?
Only that I wish more people would try this.
Even in real life circuits. Because that is what counts.
Regards Lineup

* I have used andy_c own models for MJL3281A / MJL1302A
You can find them here:
Improved SPICE Models for MJL3281A and MJL1302A - Section 1

[homemodder]

Lineup out of my curiosity could you show us the THD spectrum of the sim, maybe at 10w, half power and full power.

[jez]

Many musical fidelity amps from the '80s and 90's used a similar circuit but with output mosfets... then there is the Audio Analogue puccini et al... and various others including certain Fender guitar amps.... nothing new under the sun
The biggest problem with this type of amp is that the added phase shift from driver and output stages can make them awkward to stabilize without ending up with a slow, limited bandwidth amp. Musical Fidelity actually rolled off the noise gain of the op-amp in order to re-gain stability margin.

[gootee]

See aplication notes AN-272 at national.com, AN18 (and AN21) at linear.com, and AN-211 at analog.com.