[Anthrox]

Quote:

Originally Posted by tangent Trust me, this really is not a complex circuit I'm proposing. Give me a few more days to get back to it. It shouldn't be that big. About the same size as a CMoy circuit. Suitably packed, you could get an amp, plus this circuit and a single 9V into a regular mint tin.

Sounds good. Looking forward to putting this together.

 

[ChickenScrtchBoy]

http://www6.head-fi.org/forums/showthread.php?t=137429 may be of some interest

 

[Clutz]

Coudln't this be done with the jack, a relay? I imagine it would be a pretty specific relay - one that would have the control ciruit open when it's powered, and closed when unpowered.. then the switching jack would simply turn the relay off (turning the rest of the amp on), or turn the relay on (turning the amp off)?

 

[Anthrox]

Quote:

Originally Posted by ChickenScrtchBoy http://www6.head-fi.org/forums/showthread.php?t=137429 may be of some interest

Thanks for the link. Looks good but from what I gathered there seemed to be some questionable aspects to the circuit? I still don't understand why there's no way to have a circuit that doesn't have to constantly monitor (as it's not truely a power-on as opposed to coming out of a "sleep" mode). But then again I'm not able to build any sort of circuit like this so who am I to complain

 

[tangent]

Quote:

Originally Posted by Clutz Coudln't this be done with the jack, a relay?

Your average relay requires tens of mA of energization current for the coil. That's why you almost never see relays in battery-powered equipment. Also, I don't believe the original poster wants to unplug the source to turn off the amp.

Quote:

Originally Posted by Anthrox there seemed to be some questionable aspects to the circuit?

The main argument in that thread was over the input impedance of the circuit, which Rickcr42 was arguing wasn't high enough for some things. I've you're using all solid-state equipment, it'll be fine.

I have other concerns, but I'll bring them up in that thread. I'm abandoning my plan: this other circuit is the same thing as what I proposed but without the precision rectifier. It uses a non-precision half-wave rectifier instead, but because of where it is in the circuit, this doesn't matter. So, half the op-amps go away, which also eliminates nearly half the operating current, complexity and board space.

I have other concerns about the circuit, but I'll bring it up in that thread.

Quote:

I still don't understand why there's no way to have a circuit that doesn't have to constantly monitor

Ah...so, you want active logic that takes no power, is that what I'm hearing? Show me a product that does this.

 

[Anthrox]

Quote:

Originally Posted by tangent Ah...so, you want active logic that takes no power, is that what I'm hearing? Show me a product that does this.

It all works very nicely in the my head

But then again you have the experience and I'm just theorizing with very little actual knowledge. I guess the theory was that the signal sent by the headphone jack would sort of "jump start" the circuit but that's just me wishing for the impossible I guess. If the circuit's low enough voltage that's fine.

If it's not possible 'sall good. I'll just stay tuned to the other thread.

 

[tangent]

Quote:

Originally Posted by Anthrox the signal sent by the headphone jack would sort of "jump start" the circuit

The most studly portable source I know of is an iPod, and they can only put out about 1.5 V at full loudness. Let's say we're using inefficient headphones like Sennheiser 500 or 600 series ones, which require about that same voltage to sound really loud. If we want our circuit to activate when the music is quiet -- such as with some songs that start low before ramping up -- the voltage would be about 1000x lower.

Problem number 1: There isn't a transistor on the planet that will activate from a 1.5 mV signal. The most sensitive types require several hundred mV. That meanst that even with our inefficient heapdhones, the amp can't possibly turn on until the music is pretty loud.

Let's be super-optimistic and assume that there is some way to boost that 1.5 mV signal to something more like 1.5 V, which is near minimal for practical circuits. We'll say our circuit requires only a microamp to operate. That's a 1000x voltage gain, so assuming our converter is 100% efficient, it has to draw 1000x the current that it puts out, or else we're going to have to find a way to create free energy. This means the whole circuit draws at least 1 mA.

Problem number 2: The whole point of using a headphone amp is to avoid putting a load on the source. 1 mA is on the low side, but remember, we're being very optimistic so far. If we're not using a studly source, and we're using efficient headphones, and our circuit requires a lot more than 1 uA to run, and the converter isn't 100% efficient, we could end up drawing more mA from the source than it can even put out.

That brings us to:

Problem number 3: All of this is predicated on the availability of a DC-DC converter that can up-convert from millivolt-level signals. This doesn't exist, because that, too would have to be built from transistors and such. We have a chicken-and-egg problem: we can't run transistors from our puny voltage, and we can't up-convert that voltage because we need transistors to do it.

"Ya canna change the laws of physics!" -- Scotty