Batteries, bipolar supplies, and offset ground voltage

[Eddododo]

Is there any reason one could not take two 9-volt batteries, and use them for a +/-9v bipolar supply?
If the batteries age differently, is the only issue the unequal positive and negative headroom, or would there be issues other than the theoretical max swing being something like +8 to -7.5 or whatever..

 I know in the datasheet world, op-amps are designed with bipolar supplies as the standard, and with that in mind, it brings to mind that specs like 'offset voltage' have a significance beyond the world of .5Vref in stompboxes..

The most prominent appeal of a bipolar supply is the idea that a true 0v ground would allow one to remove or at least minimize coupling caps between stages; I assume that you'd still want a cap on the output and input to be safe.. but am I oversimplifying things (or just wrong) to assume that a 0v ref can behave a little better and remove the need for as many series caps?

Assume that I'm aware of alternatives, that I'm not looking for cork-sniffing tonal purity, and otherwise don't want to waste anyone's time talking me out of it, beyond trying to understand the actual mechanics and limitations. Thanks!



 

[PRR]

> Is there any reason......

Collect and READ every schematic you can find.

(Not "just audio"; voltmeters do it different and thinking on that can be revealing.)

Why do they do this? Or that? Yes, if you Hoover-up everything you will find two-9V batt audio things. I saw one last week. To what advantage, if any?

Why do you have eustachian tubes? (pharyngotympanic tube)

[antonis]

Batteries are expensive..

[ElectricDruid]

The most prominent appeal of a bipolar supply is the idea that a true 0v ground would allow one to remove or at least minimize coupling caps between stages; I assume that you'd still want a cap on the output and input to be safe.. but am I oversimplifying things (or just wrong) to assume that a 0v ref can behave a little better and remove the need for as many series caps?

You've got it exactly. It's often simpler to design op-amp circuits with a genuine bipolar supply, since the biasing gets a lot simpler and you're not constantly having to worry about a Vref and coupling caps. Single-supply design with op-amps is a pain in the neck, really, but we do it so often we've stopped noticing!

[antonis]

It's often simpler to design op-amp circuits with a genuine bipolar supply

Transistor circuits, too..

[amptramp]

Eliminating coupling capacitors may be good in some circuits but in high-gain circuits, the offset voltage of the op amp is multiplied by the gain.  This may cause you to run out of headroom.  Also, the coupling capacitor from the output would have to be bipolar electrolytic caps or back-to-back caps.  But you would save the expense of the Vcc/2 supply splitter and the possibility of coupling between stages if you have more than one stage using this supply.

As for 9 volt batteries, this is a bit of 1960's technology that should have been superseded long ago.  We have rechargeable batteries and chargers are not all that difficult to find or design.

[antonis]

in high-gain circuits, the offset voltage of the op amp is multiplied by the gain.

That's the reason for inter-stages coupling capacitors presence in "random" points..
(they settle up DC offsets accumulation..)

[PRR]

As for 9 volt batteries, this is a bit of 1960's technology

You got me wondering. Of course off-the-shelf cells are as old as telephones, and crow-foot cells even older. Most modern one-use cells are D-type, mostly smaller. YouTube "AA battery factory" turns up stuff like this:
https://www.youtube.com/watch?v=zPINunca1lY
https://www.youtube.com/watch?v=MNT8uUQbitk
https://www.youtube.com/watch?v=JoKasqxD-NI
If you watch several of these you'll see they copy each other's machines pretty closely. The cell-flipper may be yellow steel or blue plastic or shiny rods, but same dimensions; to move down they all use an SSS track (to settle the filler?).
About half of 9V batts are six extra-small AAA cells, others are a stack of six little slabs. Here's a long talky video: https://www.youtube.com/watch?v=c9JGfoLxVCo

[merlinb]

Is there any reason one could not take two 9-volt batteries, and use them for a +/-9v bipolar supply?

You can do that. Lots of hobby books used to do it that way. But batteries are expensive, and you need a double-pole switch to turn them both off.

The most prominent appeal of a bipolar supply is the idea that a true 0v ground would allow one to remove or at least minimize coupling caps between stages;

Coupling caps between stages are the same either way, if you know what you're doing. The opamps don't 'know' the difference. But you're right that a unipolar circuit often needs an extra cap for non-inverting gain stages, between the feedback resistors and ground, and you have to think harder about how pots to 'ground' are implemented.

[merlinb]

Here's a phaser using two batteries, for example. From a hobby magazine; there were loads of circuits like this back in the 80s.

[ElectricDruid]

Here's a phaser using two batteries, for example. From a hobby magazine; there were loads of circuits like this back in the 80s.

+1 agree, there were. This was pretty common.

Wow, using the super-rare and highly-desirable SSM2040 filter chip! Combined with the 741, lol.
Fancy trying it with the re-released SSI2140 version?!

[amptramp]

One other consideration with dual supplies is: what happens in the circuit when one supply comes on before the other or one supply does not come on but the other does?

Some op amps may get upset if there is a ground connection or ground path at an input and either a positive or negative supply but not the other supply.  In some ways, this is similar to operating the device with the input beyond the common-mode input voltage limits.  Some op amps take it in stride, some do not.  The ordinary dual regulated supply comes up slowly enough that it usually avoids any ill effect but switched batteries may cause trouble.

[antonis]

The ordinary dual regulated supply comes up slowly enough that it usually avoids any ill effect but switched batteries may cause trouble.

I've implemented a couple of slow turn-on (delayed) simple configurations (BJTs, electros and a couple of resistors) for bateries dual supply with great success but, to be honest  , I'm not sure if the */- supplies ever came up out of time..

[bluebunny]

Fancy trying it with the re-released SSI2140 version?!

I was thinking the exact same thing. 

[mzy12]

Bipolar supply opamps require less setup to bring them to optimal conditions, like typically better PSR (power supply rejection). Check out analog devices fantastic article on biasing single supply opamps: https://www.analog.com/en/resources/app-notes/an-581.html

That being said, almost every guitar effect under the sun, especially commercial ones, use single supply rails. Guitars don't exactly produce the 'quietest' electrical signals by themselves, and most amps aren't really designed with the most noise free operation in mind (sometimes to their audible detriment). So do incremental noise performance increases really matter when the first thing your guitar sees is, for example, a germanium fuzz? Who's to say