This might be a dumb question, but is there a way to hack myself a poor man's depletion mode mosfet? depletion modes are way too expensive, but I need an energy efficient way to enable a disableable circuit even if I don't (temporarily) have access to ground directly (in that case I could simply use a p-channel). Do you guys have any good ideas?
TlDr: I can intercept the power line btween High and Load. I still wanna be able to allow all the current through even if somebody rips out my ground. What do?
What you see here is an attempt, however I lose ~50% of the voltage. can we make it more efficient?
Klicka klacka.
>>988833
the sound of mechanical relays makes me hard
>>988833
a pcb mount bistable relay costs 2 bucks a pop...
>>988820
>depletion modes are way too expensive
What? There like a few dollars on eBay
>I still wanna be able to allow all the current through even if somebody rips out my ground. What do?
Current won't flow without a path to ground. Need more details.
>>988820
>I still wanna be able to allow all the current through even if somebody rips out my ground.
use the Normally Closed section of a relay
>>988892
>Need more details.
In OP post, see the connection between GND and Q2? that could potentially be severed by some idiot, but the load still needs to be supplied (load is most likely actually some cmos, so low current, but I can't actually know that)
>>988892
>What? There like a few dollars on eBay
I need 32 of these circuits, which are part of a bigger circuit. that shit adds up real fast.
>>988898
Doing that will draw a lot of power (relatively) when the contact is severed. monostable relays will drain the battery fast. and bistable relays are expensive too.., see >>988850
:(
>>988833
And now I can't get that sound out of my head.
>>988820
Let me get this straight:
1. You only have access to two points on the circuit: the +ve rail and one side of the load (the other side of which is connected to the -ve rail).
2. You need to implement a "switch", i.e. something which when closed will have little or no voltage across it.
3. The control signal is referenced to some point which isn't either of the points you actually have a connection to.
Right?
If that's the case, such a circuit is physically impossible. If it isn't the case, you need to state whatever it is you omitted.
Note that a depletion-mode FET wouldn't work either, it just means that when the ground gets disconnected the switch fails open rather than closed.
>>988920
>3. The control signal is referenced to some point which isn't either of the points you actually have a connection to.
the control signal is relative to the common ground
it's a box with four wires.
1) +12V,
2) +5V (which powers a circuit that controls Q2),
3) Ground, and
4) Source.
if either +5V or Ground fails, it should default to closed circuit. (+12V - Source)
I shoulda drawn a better diagram.
A relay would do exactly what I want, except that it's too power hungry. I was hoping there was some creative (and cheap) MOS way to do it
>>988933
> the control signal is relative to the common ground
... which may or may not be connected.
How much voltage drop can you tolerate across the switch? Because if +5V and ground are disconnected, that's all you have.
If you can live with a 1V drop, pic related should work.
