Line powered audio switch, is this feasible ?

Yeah, not worth the effort.

Running on a microwatt is out of my field of expertise. If I go down to

10mA, I'm really happy ;-)

Yep, I figured that one out. Ah well, sometimes electronics is NOT the answer ;-)

Well, now even I'm confused.

Where is the signal coming from, and where is it going to? You sound like you have inputs acting like outputs - PLEASE draw a picture telling what's coming out of which box, which direction it's going, and what you expect to see, where. ?:-/

Thanks, Rich

I have the pc output I have the TV input I have the speaker input

However, those two inputs are definitely not clean, and the tv puts some sort of feedback noise signal into the line. This happens even if the TV is off (but still plugged) and "goes back" into the speakers if there is no physical separation of the signals.

Read the chapter on low-power design in Art of Electronics, if you're interested in learning more on the topic.

Thanks for the tip, I'll have a look at it.

Let's see if a latching relay might be fired by saving energy. I looked up one that needs 5V at 139ohm for 6mS according to specs. That is 0.18W during 6mS => 1mWS of energy

Then we route audio power (via an uptransformer) to a capacitor for storage. Assuming we might be able to tap 1mW aside without putting too much of a load on the line output, then 1 second of charging would be required. With some lower level signals and less then optimal transformer for the full audio range, quit a bit more of time would be needed, but it does seem feasable. How often do you need to switch signals?

Joop

Once every week, or something like that. Yeah, I know, why bother with electronics for such a low usage? Well, as I said, I'm just toying with ideas, that's my geeky side ;-)

Thanks for the input on the latching relay, that's a very interesting idea. As to the uptransformer with a capacitor, I like that, but I'm not sure how you'd do that. Is there a schematic I could have a look at ?

I read in sci.electronics.design that OBones wrote (in ) about 'Line powered audio switch, is this feasible ?', on Mon, 25 Apr 2005:

Change the bridge rectifier to an op-amp full-wave rectifier to improve efficiency. You want to draw as little power as possible from the audio. For 'logic' you need a D-type bistable.

--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
'What is a Moebius strip?'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

Okay, after giving a few thoughts to it, here is what I came up with:

1:6 Ina ---+---. ,---. .-----. ___ --- .-----. | )|( '--|~ +|--|_R_|---+----o o--|Logic|-------. | )|( .--|~ -|-. |+ '-----' | Inb -+-)---' '---' '-----' | --- .-------. | | | --- .--------| Latch | | | | | | | Relay | | | '--------+-------' | | | | | |-- O1a | '----------------------------------------------| |-- O1b | | |-- O2a '------------------------------------------------| |-- O2b '-------'

(created by AACircuit v1.28.5 beta 02/06/05

Calculating R is not a problem, but I have a hard time figuring out what to put in the "Logic" box. I press the button, it puts the relay in Set position, I press again, the relay goes back to Reset.

Thanks for any input.

I read in sci.electronics.design that OBones wrote (in ) about 'Line powered audio switch, is this feasible ?', on Mon, 25 Apr 2005:

No. The diodes in the op-amp rectifier circuit do that.

Very little.

OK, you could use a monostable, but since you have a push-button, why bother?

Yes.

--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
'What is a Moebius strip?'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

Thanks for the tip on the op-amp, I'll have a look. But does this mean I need to add a diode just before the resistor so that the capacitor does not unload itself into the op-amp ?

I also initially thought of D-type bistable, but this will draw current permanently from the capacitor. Plus the relay does not need permanent injection, just a pulse. Basically, and maybe I'm wrong, what was intended by Joop was to load a capacitor as slowly as possible from the audio line and then use that stored charge to trigger the latching relay when the push button is pressed.

I read in sci.electronics.design that OBones wrote (in ) about 'Line powered audio switch, is this feasible ?', on Mon, 25 Apr 2005:

No.

No, you need another capacitor, with a steering diode, to provide the Vcc for the op-amp But it's still more efficient. This supply also runs your flip-flop.

Correct.

Correct.

Correct.

Ah, well, if you have a 2-coil relay, have two push-buttons and eliminate the flip-flop. Saves energy (well, it uses your instead, but you have plenty to spare).

See above; you have a second capacitor that isn't discharged.

--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
'What is a Moebius strip?'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

Well, now I'm confused. Let me explain what I understood and ask a few questions along the way.

1. The step up transformer is used to get about 6Volts from the 1V audio signal. That's fine. Out of curiosity, could this be replaced by a diode voltage multiplier?

1. The diode bridge is used to make this into a positive voltage. This could be replaced by op amps as you suggested. But where do these get the VCC-VSS from? Directly from their own regulation of the output of the transformer?

2. The positive voltage is fed into the capacitor through a resistor so that the current drawn from the transformer is very small. Charge time is thus longer (one second or more), but it's not a problem for that applicaton.

1. The push button is depressed, the "logic" box connects the capacitor to the first coil for a short time. During this time, the capacitor discharges itself almost completely into the coil. This, to me, is required because of the relatively high energy required to have the coil operate.

2. The capacitor loads itself again.

1. The button is depressed a second time, the logic connects the capacitor to the second coil of the relay, thus changing its position.

What I have a hard time understanding is how a D-Type flip flop would remember its previous state when the charge in the capacitor is gone, thus reducing the voltage to almost nil. Note that I'm indicating a two coil latching as it seems easier to connect one coil or the other rather than invert the polarity of the voltage applied to the singe coil.

Thanks a lot for your advices, they are very well appreciated.

Olivier

I read in sci.electronics.design that OBones wrote (in ) about 'Line powered audio switch, is this feasible ?', on Mon, 25 Apr 2005:

Yes, and the op-amp gain helps too; it overcomes the 'bottom-bend' of the diodes at low input voltages.

[snip]

So don't do that!

--
Regards, John Woodgate, OOO - Own Opinions Only.
There are two sides to every question, except
\'What is a Moebius strip?\'
http://www.jmwa.demon.co.uk Also see http://www.isce.org.uk

Fair enough, that's what I thought.

Ah, that's the trick. I should have thought about that. However, could you explain why 4 diodes in a bridge configuration are less efficient than the opamp solution? Is it because of the cumulated dropout voltage? (about 1.4v, if I'm correct)

That's right, but after having thought about the flip flop idea, it seems to me that I can simply plug the coil of the latching relay accross Q and /Q as those outputs are rated for +-50mA as indicated in the datasheet (example) here:

and the relay is around 25mA. But doesn't this permanently draw 25mA from the supply, and hence put a big stress on the audio line?

Cheers Olivier

Understood.

Fair enough, I've put a schematic together, which I think might work. However, I would appreciate anyone's input wether it is ok or not. Here is the URL:

Thanks a lot for your effort and support.