Op Amps driving 10w+ output stages

[rcollege]

I hope I'm not the 1000'th person asking this...

but it seems no one in electronic stores around the San Jose area knows much about audio design. (is it too much to ask when living in the heart of Sillicon Valley?)

Is it feesible to use op-amps to drive output transistors/ mosfets in amplifiers 10 watts (RMS) or more? Or is doing this begging the op amp to sizzle and fry?

I'm not referring to chip amps such as LM1875, TDA 20xx or other all in one amps. More LF353, TL082 and the like. I know there are better ones out there...but I'd rather play with the cheaper common ones first.

Also...if my Denon Blu Ray states that its audio output is 2V RMS...wouldn't you use this figure for your signal supply in simulation programs? I see many simulations using 0.7V pk.
I ask because it make a HUGE difference when it comes to clipping.

[Mooly]

Of course it's possible to use an op amp as you describe... and in different ways to. The supply voltage of the opamp needn't be a limiting factor either.

As to the clipping... well the amp will clip at a given output, usually within a couple of volts of the rails, and that for that output to be reached requires a given input voltage. How much voltage... well that depends on the gain. The gain doesn't alter the clip point though, it just determines how much is needed to get there

[multisync]

back in the early 70's there were several designs using 741's or 709's or 301's driving an output stage. I'll look around and see if I can find one in the basement.

[Mooly]

First out of the bag,
Opamp Based Power Amp

[johnr66]

The problem is that the output current is too low and you would need a gain stage ahead of the output pair and that is just one issue...

Mosfets are transconductance devices, so it might be a little easier.

You can't expect the "dude" behind the counter to know what could be considered an advanced electronic engineering topic. Open a good audio power amp design book. It ain't no cakewalk.

[multisync]

here is an example don't know where it came from, if there is a copyright infringement please delete.

[johnr66]

Originally Posted by multisync

here is an example don't know where it came from, if there is a copyright infringement please delete.

Can anyone else else the glaring problem in that schematic?


Also, to me at least, it sounds like the OP wanted to drive the two output transistors directly from the OP amp.

[rcollege]

Thank you all for you comments and answers- I really appreciate it.

Johnr66- I never thought of audio amplifiers as such an engineering feat. So it makes me feel a little better about feeling so naive on some of this.

As far as driving 2 transistors...you are correct! I was even thinking of a class a output using a darlington...but it takes a lot of amperage that = heat. I've simulated as high as 11amps to get 1 amp to the speaker.

However, you have me thinking about the current supply to the output devices.
I would go entirely discreet, but the op amps just seem superior in specs. I'm even thinking of increasing the current to 100mA peak using high current transistors in chips such as CA8013. This may give the accurate transistor matching of an op amp while allowing higher current output?

Thank you everyone for the schematics. I am going for simplistic circuitry, the reason for the op amps is I've read that op Amps will most often be more accurate than discreet components.
I will keep you all updated during the upcoming designs (and disaters)

[jcx]

for 10 Wrms into 8 Ohms there are fast CFA op amps that can do it with only 2-3 packages worth of chips paralleled

LT1210 come in a TO-220 for easy heatsinking at 1 A outpu current you'd only need 2 parallel

TPA6120 are harder to heatsink but 2 packages with 4 amps in parallel would (just) drive 8 Ohms to 10W


simple bjt buffers driven by op amps can work too - Mosfets require too much gate V drive for most op amps to drive them to the ~13 V peak you'd need


for more Vswing you can bridge op amp driven transistor outputs


there's no need to stick with SE resistor output stage bias - Class A push-pull can be 50% efficient

[Mooly]

If you really want to use just opamps to drive the speaker then try,
Doug Selfs NE5532 Power Amp. Thoughts anyone !