Auto-bypass when loss of power- stompboxes and on board

[Eddododo]

So I'm curious as to some of the approaches used to automatically bypass when power is lost in stompboxes- I'm curious for its own sake, but also I'd like to think of some ways to apply this type of solution to an on board eq- in the latter case some goals would be a) automatically going to passive when he battery goes BELOW A CERTAIN POINT, rather than an absolute loss of power, b) for this solution the current drawn by the solution should be very minimal c) no popping and weird stuff.

Relays, transistor switching, optical switching, I'd love to hear any ideas and I'd love any links to material

[Unlikekurt]

Sure, have a switching arrangement that utilizes relay(s).
Have the effect be "ON" when the relay coil is energized and bypassed when the relay coil is not energized.
Now, should you lose power the effect will default to the bypassed position.

[Eddododo]

I guess i should clarify- I am curious of the general approaches used in stompboxes, but I am looking specifically to optimize it for a battery powered onboard (bass) preamp. I'm aware of the general possibility in stompboxes already, all i really desire in that application are specific schematics that I can try to learn from and/or adapt

[Eddododo]

So my thinking out loud- relay is tricky (as far as I understand them) based on my assumption that anything that would be 'held' until voltage drops would be drawing current the whole time (right?).
This is a little circuit that I'm trying to remember the origin of... it was some kind of preamp and basically when you plug it in the LED flashes once, as an indicator that you have 'at least one gig' left... I think it stops flashing around 8.5 V (from a ~9.5 fresh 9 volt). I like the idea of something that has a moment of action when you plug in your 1/4', that would turn off via cap discharge or something when voltage drops too low...




What I picture, out of my pure and innocent ignorance, is something like this: latching relay (those don't draw power while latched, right?). charge stored on a capacitor that is in *some kind* of transistor arrangement such that when the available voltage becomes too low the transistor would close or open causing the capacitor to discharge one last fatal gasp of voltage to switch the relay back.

I realize I'm grasping at straws and I need to learn a lot more about  some pretty basic stuff, but if I have the slightest grasp of an idea and someone can point me in the right direction I'd greatly appreciate it. In the meantime I'll be researching all of these basic blocks 

[Eddododo]

update- read R.G's relay stuff.

[antonis]

Not to be rude but why you don't use some kind of visual battery state indicator..??

(I can't imagine any person who might like a sudden & unannounched interruption of his wah/fuzz/tremolo/delay/flanger/chorus/compress riff..) 

[GibsonGM]

Not to be rude but why you don't use some kind of visual battery state indicator..??

(I can't imagine any person who might like a sudden & unannounched interruption of his wah/fuzz/tremolo/delay/flanger/chorus/compress riff..) 

It is not a bad idea to want to have a safe fall-back like that...but really, what Antonis says - perhaps a monitor to illuminate an amber LED to let you know the battery is getting low would be a more 'professional' approach?  Especially if it comes on with plenty of time for you to react, such as 8.2V or some level like that...   

If one is worried enough about losing effects on stage, this would give you MORE security than auto-bypass! 

[Eddododo]

Well firstly, if I wasn't clear enough the purpose is for an ONBOARD preamp, not so much for pedals; the pedal request was more in terms of getting examples for context that would likely be more available here. Second, it's more of an academic/ problem solving exercise- I'm wondering  how I could do this for the purpose of building knowledge through targeted problem solving, so that I could use this solution to adapt other more nuanced issues- I am NOT looking for a different approach because frankly they're easy to think of

[antonis]

If so, use a single Voltage Comparator with BJT (optionally) to switch the LED..
BJT could be N-P-N or P-N-P depending on the OUT state you like to activate the switch..

You can even use a general purpose cheap Op-Amp (like 741 or so) wired as Schmitt-triger (perhaps with some hysteresis, to avoid threashold "fluctuations")..
(IMHO, that's closer to your wish of building knowledge..)

edit: Forget it, cause it requires a "steady" (constant) voltage for one of comparator inputs.. 
       (the need for coffee becomes imperative..)

[GibsonGM]

Latching relay with a second battery.  Voltage sense circuit...trips relay to 'bypass' when voltage drops below threshold.  Relay can be a lower voltage model...5, 6 volts....if you were crafty, you could set this up to work from a single battery as it will still have switch juice when below your threshold voltage.

[R.G.]

Actions on loss of power have something of an ongoing-puzzle fascination to me. A lot depends on exactly how power is lost, and how certain you have to be that you've taken the desired action when power is on the way down.

Starting with taxonomy, the three classes I can think of are when power (1) drifts down gradually enough that you can pick a threshold and still use the remaining power (2) falls fast enough that you have to have some backup power source and (3) stops what amounts to instantly.

If power drifts down, any number of ways are open. The scheme of using a power detector of some kind works here. There is time to decide what to do, and power left to do it.

If power falls rapidly, but detectably, you can keep a reserve power source, a battery or even a capacitor that stores just enough juice to do the emergency action.

If power fails substantially instantly, you're forced to use "keep-alive" circuits where the desired action is what happens simply because power is gone. There's no time to have a circuit decide what to do. The circuits being run must revert to the desired condition simply because there is no power.

Expanding on that last one (#3), the general idea might be a relay that is held in the on state simply because power exists to hold it there. When power to hold the relay on is gone, it falls back to the no-power condition. This is what you'd want for your onboard preamp, excepting for the issue that holding a relay in constantly eats battery pretty quickly just on its own. You'd be paying for proper dropout in very short battery life.

A latching relay needs zero hold-in power, but also does not fall back to the desired state with power removed. So for a latching relay, you have to reserve a final pulse to put it back in the desired power-fail state, so this automatically puts latching relays into the #2 category. It can be done - you just have to charge a capacitor or mini-battery with enough energy to make the final turn-off pulse. It's do-able, if complex.

If not relays, then what? There's really only one choice for signal switching - JFETs. JFETs are conductive in the no-bias state. So with power removed, they're always "on" - you have to actively turn them off. It is conceivable that a circuit could be set up so the no-power state lets signal through, and application of power routes the signal through the preamp. Some MOSFETs are depletion mode like JFETs, but they are very rare.

Other semiconductor switches are going to be open in the no-power state. There is an asterisk here. CMOS switches use so very little power that nearly-zero may be enough to keep them in the state you want.  CMOS needs a voltage, and remarkably near zero current. I could concoct a situation where a capacitor stored enough voltage to keep a CMOS switch alive for hours. Or for a string of solar cells provided enough voltage (at nearly no current) to keep CMOS switches operating.

Just at this first glance, JFETs and CMOS look like the way to go. A latching relay might work if you assume power is drifting down slowly enough so the turn-off pulse can be applied at just the right moment. That kind of also implies some tricky timing on the switching, and that implies a uC for best control of what happens.