Bomb-proof func. gen. power amp

Where's the ground reference, did you mean +/-12V or is that single-ended?

A friend of mine has an audio amplifier design with variable output impedance, all the way from constant-voltage to constant-current.

I suggested he extend its output impedance from the positive real axis to the entire complex plane, but we decided most of the negative half-plane would be entirely unstable, and weren't certain how to implement the complex impedances (i.e., constant complex impedance isn't normal). Even so, the positive real axis should be more than overkill for you, since you say you only need two points on said line (0 and infinity).

Tim

-- Deep Friar: a very philosophical monk. Website:

Yes, single supply. 12VDC 2A. 24W.

David King

And the source and load are also single ended? So you basically need a variable power supply more than a power amplifier? (Notwithstanding the reactive current requirement, which will require a bidirectional power stage. If you can allow flyback on transitions, then this gets even easier.)

Tim

--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms

Hmm, well I was going to suggest a burr-brown (TI) power opamp. I'd have to look up the part number. It does cost ~$10.

But goes right down to DC. It's thermally protected and doesn't need to be replaced.

George H.

I have seen these, but, by the time I get them here, for qty. under 10 each would cost nearly $40. Not that I am cheap, but I need a few of these, and discrete devices are easier to upgrade later if necessary.

I suppose I was looking more at a rail-to-rail, single supply op amp (DIP) with an added constant current output stage and variable feedback loop. Seems strangely elusive as a working optimized circuit though.

I just can't seem to get it right on the bench, at least to the point where I can plug it into the range of applications as previously described.

I know it's Christmas (Merry Christmas!), but any advice would be appreciated ... or even a circuit, SPICE sim or whatever. It's just me and my soldering iron this year.

David King

At that sort of speed, you can use a complementary buffer, like this:

0 +12V | R R 200 R R | 10 uF *--------------------*---CCCCC----* | |/ | | +----------| 2N4401 GND |\ | | |> 0----|+\| | | | \ | 100 | 50 | >------*---RRRR-----*-----------*---RRRR--0 Output | / | | | *--|-/| | | | | |/ | | |< | | | +----------| 2N4403 | | GND | |\ | | CCC | | | 47pF CCC GND | * ------------+-------RRRR-------------* | 2k R R 1k R | GND

The 200 ohm/10 uF network will protect it from shorts and other misbehaviour.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

Doesn't this distord a lot at zero crossings? Or does the opamp take care of that? It is nice & simple though. I guess almost any old fashioned NPN/PNP pair (like TIP31/TIP32) will do.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------

The 100 R will bypass the transistors when bias levels get below .7 volts on the bases..This will force the Op-Amp output to deliver the signal down in the valleys and you won't detect to much of the transistors switching over between the On/Off set..

Jamie.

Sure it does, but the OP only wants it to work up to 5 kHz, so the loop gain of the op amp should handle it okay.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

Thanks for the great Christmas present. I will put one together tomorrow!

David King

Single power supply, constant current driver. Do you want the output to be unipolar? How accurate?

John

Is that better than something like this?

Cheers! Rich

Yes, 12VDC single supply is preferred. Not particularly accurate. I am more interested in successful capacitive and inductive loading within the Dc to 5KHz bandwidth, and maximiizing the voltage swing. IOW as close to 12V as practical.

Sorry for any delayed answers. I am in a different time zone..

Thank you,

David King

Nice circuit, but I was aiming for single supply. 12V swing is enough.

Thanks,

David

F

=A0 =A0 =A0 |

=A0 =A0 =A0 50

=A0 =A0 |

=A0 |

=A0|

=A0 =A0|

=A0 =A0 |

=A0|

Dont forget diodes on the output to each rail

NT

Been a long time since I learned the terminology, but are we forgetting

1. Constant current (transconductacnce) mode. Default waveform sinusoidal.

???

The collector-output approach gives you a wider output voltage range--in fact with a bit of work it can go much wider than the op amp's supplies. On the other hand, it's much more load sensitive, harder to stabilize, and more complicated.

Merry Christmas!

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

Shouldn't need them. If there's an inductive load, the emitter follower will take care of it--it'll just keep conducting. It's like driving a relay with an emitter follower--no snubber needed.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal
ElectroOptical Innovations
55 Orchard Rd
Briarcliff Manor NY 10510
845-480-2058

email: hobbs (atsign) electrooptical (period) net
http://electrooptical.net

OK, but unipolar output current drive? That could be pretty simple.

John