Remotely switching between inverted and non-inverted audio?

[ElectricDruid]

Does anyone know a nice simple circuit that would allow me to remotely "flip" a signal from non-inverting to inverting?

I need something like this for an idea I've got, needs to flip the audio when 5V logic pulse arrives.

I thought of the Boss OC-2 as one example, and then I thought of a phase-shift stage as another. I wondered if other people have other ideas?

Thanks!
Tom

[Ben N]

Would a DPDT relay or solid state equivalent bypassing an inverting opamp buffer be too crude?

[ElectricDruid]

I need audio-rate switching, so a relay is out. Sorry, I didn't mention that till now. CMOS switch would do it, for sure. Is there a simpler way?

I think something like the modular synth "CV Attenuverter" circuit, shown at the bottom of this schematic, might work:



I don't need a pot, but I can replace it with a couple of resistors, and a tranny to short one out. It's worth an experiment on the breadboard at least.

[Vivek]

How about an Attenuverter with FET as switching element ?

[merlinb]

I thought of the Boss OC-2 as one example,

You mean the switch hitter? That's certainly an obvious choice.

As is this, if you have a faily low source impedance:

[amz-fx]

How about something like this:



As shown with switch open, it is inverting. Close the switch and it is non-inverting. The switch could be a relay, activated remotely, or it could be a fet switch if you need to flip polarity rapidly. All resistors have the same value; 10k for example. You will have to bias the jfet and op amp properly if using a single supply voltage.

regards, Jack

[antonis]

Can a SPST switch with a looooooong cable be considered "remote"..??

If yes, Jack's proposal is the most plain one..

[R.G.]

Try "Standard - and Strange... - uses of the Polarity Reversing Circuit", geofex.com, 2001.

http://www.geofex.com/article_folders/polarity_reverser/polarity_reverser.htm

Automate the signal grounding with a JFET or analog switch.

[ElectricDruid]

Thanks all. Lots of variations there.

Is the FET crucial? In the cases where the connection being made is just between the +ve input and ground, could I use an ordinary NPN instead? It doesn't seem to me like any signal is going through it, so I don't see why it needs to be a FET. IT's just acting as a switch, right?

Thanks,
Tom

PS: I should have known you'd have an article about it, R.G.!! Whatever I come up with, you've already been there!

[R.G.]

The signal switch doesn't have much in the way of requirements, but you do have to worry about the math. If you think about it, this circuit computes a partial cancellation of half of the signal. The accuracy of the resistors determines how accurately the cancellation happens. Reversing phase in an amplifier on a static basis is easy, as there's no change in noticeable amplitude from one polarity to another. Reversing phase accurately on the fly will mean you have to worry about the switch some.
The JFET has to make the signal at the + input be as nearly zero as possible, so the JFET on resistance may be an issue. With JFETs, you have to have a big enough "off" signal so that any signal swing doesn't pull it into the triode region from cutoff. With NPNs, you have DC offset to be held off by a cap in series with the collector, and control feedthrough can be tougher to quell. This thing goes right at the + input of an opamp, so any control feedthrough is going to get into the signal path.

I first used this circuit back in the late 1970s when I built a magnetometer as an electronic compass for some sailboat racing instrumentation.

[ElectricDruid]

Thanks R.G.

This afternoon I got it working on the breadboard with an NPN, which is easier than a JFET in my book. The resistor value isn't fussy. I tried 100Ks initially. It worked. I swapped it for 10Ks just to see. It worked. One op-amp, three resistors, and a 2N3904 - basic stuff (Perfect!).

I'm hoping to use it to turn a 10-bit unipolar 0-5V DAC into a -5V to +5V 11-bit bipolar DAC by turning the output upside down half the time. As such, I'm working on getting the "Polarity" output which drives the NPN to switch as close to the point at which the DAC is updated as I possibly can to minimise glitches. Below 0.5usec is pretty doable.

I was pretty pleased when it started working. Getting a bipolar DAC out of a unipolar one is pretty nice (although there are other ways, of course) but gaining a bit of resolution in the process is the icing on the cake!

I'll be back again if I manage to do anything useful with this new toy, I'm sure.