max op amp voltage?

[TimWaldvogel]

i am designing a mic preamp based around a dual op amp but i am curious how i go about raising the voltage for headroom?

whats the max voltage a opamp can take? i was think +18 volt guitar power supply around the whole thing. but i am curious if 24 volts would work and what i gotta do to prevent the op amp from frying lol

[R.G.]

Datasheet. lol.

[TimWaldvogel]

Everytime I look at it I get lost in technical terms and jumbo jumbo I don't understand yet

[JKowalski]

Datasheet. lol.

Never thought I would see you "lol"  

Learning how to read datasheets is a very important thing to do. Most of the basic information is stated fairly clearly. Just look for "Max Operating Voltage" or Vcc min/typ/max.

Vcc is the power supply voltage, and also can be found as Vdd. The ground/negative power supply voltage terminal is often called Vss or Vee.

Vcc/Vee (corresponding to bipolar collector/emitter) is used for analog IC's, Vdd/Vss (corresponding to FET drain/source) is used for digital IC's - these conventions are pretty strong but there are exceptions.

Datasheets usually have all the major info in a big table with min/typ/max for each value that they can apply to. If you are looking for a basic parameter (max voltage, max power dissipation, max output voltage swing, etc.) then this is where you'd look. If you are looking for how the chip responds to a variable you change (eg. frequency, voltage on terminal, ambient temp, whatever applies to the chip), look at the graph section. If you just want to know what the chip does, look at the summary at the start. If you want the chip pinout, well, if you can't find that then I don't know what to say  

[Gurner]

There are heaps of opamps that will go to 18V.

In this instance, my first port of call would actually not be a datasheet, but a search facility - like the one Digikey has......

http://search.digikey.com/scripts/DkSearch/dksus.dll?Cat=2556125&k=opamp

look at the column "Voltage - Supply, Single/Dual (±)" ....if you are just going with 0V & 18V (vs +18V & -18V), then scroll down until you get to the that don't start with a ± in front (albeit confusingly you can still use one ±9V for a single 18V supply!)  - it's the latter number that tells you how high you can take its supply, so for example if you see "1.8V ~ 30V" , it means you can put up to 30V on the supply pin etc

Of course, once you've slimmed the list down, then you're gonna need to look at the datasheet to see if it meets all your main requirements.

[R.G.]

Everytime I look at it I get lost in technical terms and jumbo jumbo I don't understand yet

Then the right question is "Where do I find the maximum power supply voltages for an opamp on a datasheet?"

That does two things: first, it tells you what you need to know right now, and second it tells you how to start understanding the technical terms. One very important section of the datasheet is cryptically marked "Absolute Maximums". That means that if you go beyond these values, the part may very well go up in smoke. This section is probably the first place to look on ever datasheet after reading the few paragraphs of overview under the titles.

There really is only one way to learn to speak technical - you start with one term, learn it, then go to the next. I'm still learning the occasional technical definition on a datasheet after decades. And once you learn it, you refresh yourself all the time by using what you've learned.

I encourage you to dig into datasheets. You'll need the knowledge if you keep up any kind of technical work.

Never thought I would see you "lol"   

Me either.    I just had to be inspired by the moment I guess. 

[teemuk]

As a side note, even the maximum power supply limits really arent' a problem that would limit headroom because you can always bootstrap the power supply. Basically, with proper design you could easily get even a +- 100V output voltage swings from an opamp while still retaining the OpAmp's rail-to-rail voltage within the margin.



All you need is a high voltage power supply and few additional components for the bootstrapping circuit. See:
http://www.mikrocontroller.net/attachment/23492/45890.pdf


At that point you'll actually be caring more about a parameter called "input common mode voltage range". It's less than the actual maximum rail voltage that an OpAmp tolerates.

[PRR]

> raising the voltage for headroom?

Don't raise the supply, lower the gain.

Whole systems can run great at 0.3V nominal, peaks not over 2V. With a bit of fudge, the system can work at 5V total. Indeed today we are usually going to an ADC which won't take over 5V peak to peak. We no longer dump +30dBm through a 6dB pad to drive a long telco line, passive equalizer, and passive distribution.

As an aid to reading: 741, 4558, TL072, LM324(ugh) are all 36V max. We usually run them on +/-15V supply so that static bias is near zero DC and some interconnects don't need DC block caps. They can run on +/-18V to "use the full 36V rating", but 36V/30V is barely over 1dB higher level while OTOH your "+/-18V" has tolerances which could put you over 36V total and ZAP.

5532 is rated 44V total, I think a 604 is also. This suits legacy systems with +/-18V or 20V rails.

If you are a bad boy: I have worked 36V chips on an unregulated supply which probably touched 40V a few times, they live. (Internally this doping process is called 40V; the 36V public rating covers their butt against our sins.) I have worked 28V chips at 35V and they lived a few years, failed, the replacements  lived a few years, failed. ROFL (not).

Really, if +/-12V is "not enough", you are doing something very-odd or just wrong.

[Gurner]

> raising the voltage for headroom?

Don't raise the supply, lower the gain.

I second that motion - I don't know what all the obsession with headroom is!

A non active (typical) guitar only puts out a few hundred millivolts (a typical dynamic mic much less) - even running an opamp at single supply of 5V gives absolutely heaps of 'headroom' (I think that term gets way over-used, as if taking on some mythical, magical place where apparently sonic fidelity is raised into another plateau). Sure, nobody wants to 'clip' a mic signal, (or a guitar signal if 'clean' is the end goal), but really, in most cases 5V is fine, 9V way sufficient etc etc

[Mark Hammer]

When people talk about "gain" someties I get that feeling you have when you've somehow strayed into the wrong washroom.

Is this the ladies or the mens?  Are you aiming for the sort of clean gain people use for sample mic-ing or are you aiming for the sort of gain that provides desired coloration?  Which "universe" am I working in here?







...and because it seems appropriate, and I desperately want to feel like I belong...  LOL

[TimWaldvogel]

Mark I mean both. I am talking clean gain with clean sound until I push it harder and then it might add a little
Coloration

[ashcat_lt]

But (at least theoretically), you don't get coloration until you've run out of headroom.

I think what Mark's getting at is that a lot of folks - especially guitarists - use the term "high gain" when they mean lots of distortion.  Technically, lots of distortion comes from having lots more gain than the circuit can handle with a given input voltage.  You can get a "high gain" sound without any actual gain, if you've got components which clip at a low enough voltage.

Now in the case of trying to capture a truly clean sound in the era of 24-bit digital recording with its enormous dynamic range, and considering that every stage through which the signal passes is going to add noise, there is something to be said for getting as much clean gain as possible at the front of the chain, and having enough headroom to pass it through the chain, even if that means attenuating at the end before hitting the ADC.

[PRR]

> clean gain with clean sound until I push it harder and then it might add a little
Coloration

Run a reasonable supply voltage. Turn gain up only until you see meter-motion. That's clean.

Now -reduce- supply voltage until it dirts-up.

In general, opamps don't so much "color" as "fuzz". You may be thinking of sloppier amplifiers.

> But (at least theoretically), you don't get coloration until you've run out of headroom.

A naked vacuum tube tends toward 6% THD at clipping, 3% at 6dB below clipping, 1% at 15dB down. These THD levels are not obvious, but subtly audible.

High NFB amps tend to <0.05% THD from noise up to 2dB short of clipping, then THD skyrockets in a very typical way.

But we learn by doing, not by talking. Solder and opamps are cheap.

[stringsthings]

... But we learn by doing, not by talking ...

well done ! 

BTW, very much enjoyed the discussion on impedance

... and, for more guitar-effect projects that run on higher supply voltages, i recommend taking a look at Anderton's EPFM projects .... i'm doing his pre-amp at the moment and the extra headroom adds some "sparkle" to my ears ....