high voltage "opamp"

[mw]

Is there a standard circuit connection that can be used to raise the
output swing of an opamp to a higher voltage ? For example, a TL084
opamp with a +18V / -18V supply can give an output swing of about +12V
to -12V. But what I am looking for would permit a voltage swing of +36V
to -36V. I would of course expect to provide higher voltage power
rails, and add a few discrete transistors, etc., to massage the signal.

It seems like this circuit has been used as an example in an app note or
text book somewhere, but I haven't located it.

Ideas, anyone?

 

[Pooh Bear]

mw wrote:

Is there a standard circuit connection that can be used to raise the
output swing of an opamp to a higher voltage ? For example, a TL084
opamp with a +18V / -18V supply can give an output swing of about +12V
to -12V. But what I am looking for would permit a voltage swing of +36V
to -36V. I would of course expect to provide higher voltage power
rails, and add a few discrete transistors, etc., to massage the signal.

It seems like this circuit has been used as an example in an app note or
text book somewhere, but I haven't located it.

Ideas, anyone?

One method is to drive the op-amp's power pins ( only works for a single
device ) with emitter ( or source ) followers driven by voltage dividers
between the op-amp output and the high voltage supply rails.

You can source more current if required by adding emitter followers to the
output pin directly.


Graham.

 

[mw]

Chris wrote:

Possibly you're thinking about National Semiconductor AN-272, "Op Amp
Booster Designs". Look particularly at Figure 6.

http://www.national.com/an/AN/AN-272.pdf

Yes. That's the one.

There are several subtle and not-so-subtle difficulties with output
voltage boosters for op amps that degrade performance. And you have to
pay some extra attention to op amp input protection, particularly
transients. You'll almost always end up looking better and getting
better performance with a single chip solution if it's available.

I just now simulated various versions of the National circuit and did
find that they do not perform well at high frequencies (>20kHz). I
think it is because impedances are too high. It's the trade off between
speed and power.

Thanks for yours (and Pooh Bear's) help.