What to look for on datasheets to get low power op amps?

[Djentronio]

Im looking at 500uV, 1mV, and 10nV style op-amps, but which one is best for low consumption amplification of guitar signals? What should I be looking at specifically?

[Hatredman]

You want to build onboard stuff on your guitar, don't you?

.

[PRR]

Well, you want a Supply Voltage that works for guitar-level and the batteries you can get small/cheap.

Then, to take the question as-asked, you sort on Quiescent Current, Iq. Here is TI's parametric chip-push page, opamps like TL072, 5V to 12V supply, sorted by Iq.



BUT you also want to "pass audio". There are MANY sides to this: bandwidth, speed, hiss.....

At a glance, those Slew Rate numbers are bad for audio. I would expect to want near 1V/uSec. The first handful are 1000 times slower than this!! While not the same thing, I would want GBW to be a large fraction of a MHz. And while hiss figuring is complicated, I would not want anything worse than TL072.

There's hundreds of choices at TI, and dozens more unique to other suppliers.

And just because it has specs does not mean you can buy one or two cheaply.

[Djentronio]

Thanks. I was worried that a very low quiescent current would also indicate an inability to drive another amplifier for a guitar based system. Looks like tiny power also comes with bad slew rate.

[reddesert]

I would look at commonly used audio op-amps and read the datasheet for I_cc, supply current, in the "Electrical Characteristics" section.  Thus, some values from TI datasheets for standard op-amps in "typical" current draw, not max:

TL062/TL064 - 200 uA, per amplifier
TL072/TL074 - 1.4 mA, per amplifier
RC4558 - 2.5 mA, both amps
LM324 - 1.4 mA, all four amps

(these are tabulated for operation at +/-15 V or so, so they may be different when operated off a 9V battery).

[merlinb]

Looks like tiny power also comes with bad slew rate.

You may not have to worry about this as much as you think. If you're talking about a 9V guitar system then even if you needed to produce 10kHz (you don't) at the full 4.5Vpeak (you don't), you would only need a slew rate of:
2 * pi * 10kHz * 4.5V = 282743 volts per second, or 0.28 V/microsec.

[Hatredman]

Looks like tiny power also comes with bad slew rate.

You may not have to worry about this as much as you think. If you're talking about a 9V guitar system then even if you needed to produce 10kHz (you don't) at the full 4.5Vpeak (you don't), you would only need a slew rate of:
2 * pi * 10kHz * 4.5V = 282743 volts per second, or 0.28 V/microsec.

Low slew rate is what makes The RAT sound like a Rat, isn't it?

.

[R.G.]

A lot depends on what you're trying to do. If you're doing active filtering and such, opamps are about the only choice because they are less complicated (ie. fewer parts) and make the job simpler, and because they don't necessarily have to drive cable capacitance. If what you're trying to do is just to drive cables cleanly to avoid high frequency loss, you might want to consider this:
http://www.geofex.com/FX_images/Onboard_Preamp.pdf
This was discussed here on this forum back in 2013. You might like to read that thread.

I'm generally a proponent of opamps over discretes in most cases, but electronics inside a guitar are a special case. The need for low power consumption and long battery life while keeping good audio performance makes it a challenge for opamps - at least the ones available up through 2013 when I wrote that.

Low power opamps sacrifice some things to get low power. In many cases the sacrifices are in the areas of speed and current capability, and also in overall gain-bandwidth. The opamps available up through 2013 had issues, so I went off and designed that onboard buffer.

With opamps, you are a little disadvantaged by the very high gain and the need to cut gain at quite low frequencies by the compensation networks to keep them stable.  They keep getting better all the time, but you can still do a better job with discretes in special situations.

[PRR]

> 0.28 V/microsec.

At that point the NFB is nearly gone. "Good Practice" suggests taking a good margin over the signal's slew rate. Hence my wild-pitch at 1uV/uS.

Yes, 10KHz at 4.5V is very extreme for guitar, so 0.2V/uS may sound fine.

Note that the VERY low iQ chips are showing numbers like point oh oh V/uS. We can certainly buy a chip too slow to pass guitar cleanly.

[Hatredman]

RG, I have an old question I've always wanted to ask About that onboard preamp, and I think It would be of interest for the OP, só I'll Shoot It here.

You Said:

"If your onboard tone andvolume controls add cable-drive issues back in, one solution is to use one of these buffers for the pickup coils, and another right at the guitar's output jack."

Why not only one at The Jack? Why we would need two?

I'd like to build that for an A/B test (with/without), but this "two circuits needed" business puzzles me.

.

[R.G.]

Guitar pickups are very fussy about exactly how they're loaded, and guitarists are very fussy about what they hear. For most people other than the "I love brown-sound" crowd, more sparkly treble is better.

The pickups in a guitar are loaded by the on-guitar controls, Some people would prefer all the treble they can get, so buffering right after the pickup would keep the treble that would be lost to guitar controls. Then another buffer would drive the cable, although one could argue that the buffer before the controls has already wiped out the treble loss. But some people ...like... what their controls do to their sound by loading. So -one buffer? two buffers? Season to taste.

[Djentronio]

Curious question: If a chip is too slow to pass a guitar signal cleanly what use is it? For really, really tiny signals? For something else?

[Hatredman]

.
So - one buffer? two buffers? Season to taste.

Hummmmmm, haven't thought of that, thank you!

Makes a Lot of sense, and now I'll have to experiment a Lot with It.



.

[Hatredman]

Curious question: If a chip is too slow to pass a guitar signal cleanly what use is it? For really, really tiny signals? For something else?

For a Rat

.

[highwater]

Curious question: If a chip is too slow to pass a guitar signal cleanly what use is it? For really, really tiny signals? For something else?

Looking at a few datasheets, buffering/amplifying sensors seems to be a big one. Working in the audio field as we are, it's easy to forget just how many opamps get used for DC-only applications.

For guitar pedals, they are quite likely to be useful for LFOs - low power consumption would help with ticking, and low-voltage chips tend to swing closer to the rails.

For a Rat

LOL. It happens that I'm in the middle of building a Rat... I might just have to throw a few ultra-low-power chips into my next Mouser order to try them out.