High Supply Voltage Op Amp

[Ripthorn]

Hi guys, in my quest for as much gain recovery as I can get, I am looking for an opamp that has really high supply voltage to allow for maximal boost.  I have looked at TL072's, but 18V seems low to me.  I have also looked at the BA10358 which has 32V max and I think the 4558 has something like 22V.  Any other suggestions out there?  I'm looking for the most clean boost I can possibly get.

[alanlan]

TL072 - you can use +/- 15V safely as with most op-amps.  It will do 27Vpp - I mean how much headroom do you want?  It's no use saying that you want the maximum possible because anything is possible.

[alanlan]

...and you could take a differential output approach which will double your 27Vpp to 54Vpp!

[Sir H C]

You are starting to hit power amp levels, check  out some chip power amplifiers.

[Brymus]

the mc3304p will handle 36V supply its a quad opamp.
But the first stage on most amps doesnt want to see more than a 2V swing what are you making?

[zyxwyvu]

Why not use a transformer to step up your output?

http://www.hollis.co.uk/john/titan_boost.jpg
(from http://www.hollis.co.uk/john/circuits.html)

You can also check out specifically high voltage opamps from various companies:

TI: http://focus.ti.com/paramsearch/docs/parametricsearch.tsp?family=analog&familyId=1482&uiTemplateId=NODE_STRY_PGE_T&paramCriteria=no
National: http://www.national.com/cat/index.cgi?i=i//4:s//92//Driver//Op Amp:n//256//15//84:m//256//0:m//700//0

[Ripthorn]

I am trying to do gain recovery in a preamp after the tonestack and my TL072 is clipping before recovering very much gain, so I wanted to see what is out there to boost the signal back up higher.  The differential amplifier approach looks really interesting.

[slacker]

I have looked at TL072's, but 18V seems low to me. 

18volts is for if you use a bipolar supply, so +18 and -18 which gives a total of 36 volts. You can run them of +36 volts and ground, with an 18 volt vref instead if you don't want a bipolar supply.

[alanlan]

18 volts is the absolute maximum in this case, so I wouldn't actually use 18/36 for that op-amp.  +/-16 would be more like it.

[R.G.]

Well, if you're serious about this -  

Look at this http://www.johnhardyco.com/pdf/990.pdf; perhaps do some searching for "990 discrete opamp" on line. It is *possible* to make one of these things capable of running from +/-50V power supplies or even more.

A simpler approach might be with a two-transistor gain pair or a ring-of-three circuit to get over 50-60V of power supply headroom.

And if you really want even more headroom, you can use a high-voltage MOSFET to cascode a small signal JFET and get literally hundreds of volts of signal swing.

[slacker]

18 volts is the absolute maximum in this case, so I wouldn't actually use 18/36 for that op-amp.  +/-16 would be more like it.

Good point.

[Cliff Schecht]

My dad was working at a (now defunct) power company working on some digitally-programmable power chip. He was telling me about an all discrete op amp that he got to play with that did some ungodly numbers (I'll ask him later to get the specs and model). I think they were somewhere above $1k a unit though..

That 990 is very cool as well!

[R.G.]

My dad was working at a (now defunct) power company working on some digitally-programmable power chip. He was telling me about an all discrete op amp that he got to play with that did some ungodly numbers (I'll ask him later to get the specs and model). I think they were somewhere above $1k a unit though..

That 990 is very cool as well!

I got to play with some Air Force surplus Philbrick opamps back at school about 6,000 years ago. They used two 12AX7s per opamps, one for a diffamp and the other for a pushpull output stage. Neat stuff. They ran from about +/-100V as I remember it. We had to get a double integrator sine-cosine generator to work as a homework assignment.

Of course, modern audio power amp practice is simply to build a discrete opamp without a lot of concern for DC accuracy. The standard Linn-format power amp works well as an opamp if the designer didn't play games with the compensation to get good test numbers. Some of them are only stable above a certain gain. Maybe the simplest single chip solution is an LM3886 at +/- 40V or a TA7293 at +/- 50V.

[Ripthorn]

What about bootstrapping the opamp?  I read a paper on it last night and it seems interesting, though not sure if it is quite as practical in this case.  I will try the higher bipolar power first to see what I get, and maybe try the +32V or thereabouts second.  Thanks for the input so far guys.

[Ripthorn]

So what would be the easiest way to get +/- 16V?  I used a TC1044 to do +/- 12V and that worked well, but I need just a little more headroom.  Any simple ideas?

Edit: in the amp, I have +12VDC, +6VDC and 12VAC, one of which I would like to use as opposed to a transformer.

[R.G.]

So what would be the easiest way to get +/- 16V?  I used a TC1044 to do +/- 12V and that worked well, but I need just a little more headroom.  Any simple ideas?
Edit: in the amp, I have +12VDC, +6VDC and 12VAC, one of which I would like to use as opposed to a transformer.

First, read here: http://geofex.com/Article_Folders/Power-supplies/powersup.htm

I would use the 12Vac, and double-half-wave rectify it to +/- 17Vdc  if one side of the 12Vac can be tied to ground. If it's centertapped and the centertap tied to ground, you'll have to double or triple on each side of the centertap.

[Cliff Schecht]

Speaking of GEOFEX, I saw a reference to it today in a Nuts and Volts. The article was about how to choose the correct part and referenced RG's "The secret life of pots".

[snap]

80V (+/-40V) supply http://www.prema.com/Application/pr2201_e.html

[stephanovitch]

Use this for audio application:
http://focus.ti.com/lit/ds/symlink/opa445.pdf

+/-45V

[Ripthorn]

Hey RG, that is exactly what I was hoping for.  I will give that a shot tonight and see how it works.  Thanks.