Low Voltage Dump Circuit

[dat209]

I am working with a single board computer that has a strange power supply issue. After trying some work-arounds, I've found a way to fix it, which requires me to dump all the capacitors if the voltage goes beneath a certain threshold in order for the computer to turn on again. I'm trying to design a circuit that can do this automatically.

I need a circuit that will normally operate by forwarding a voltage (22V) to the computer. If this voltage decreases below a threshold (still TBD, but around 7V), I need this circuit to dump everything to a low value load resistor to drain the capacitors fully.

I've been trying to mess around with FETs for this, but all my simulations and math can't give me a solution to this.

I basically just need a circuit that acts as a normally closed switch, but when higher voltages are applied, opens. I'd like to avoid using a relay if at all possible.

Thanks!

 

[betwixt]

Sounds like a very strange power supply!
I think your problem is that while 'dumping' the voltage you are monitoring are held low and it therefore can't restart properly. You need to add a timing element so when triggered it does the 'dump' for a short time then automatically stops it again so the power supply can recover. It would probably be easier to control the reset line of the SBC than overload the power supply though.

Brian.

 

[dat209]

Ok, I think I mis-spoke. So the SBC takes a variable supply of anywhere from 7V to 28V and then does its own conditioning on it. I am supplying 22V (nominal) from a vehicle's power for the computer. When I turn off power to the SBC though, from a scope I can see that the voltage is decaying due to an RC drain. Unfortunately, this process is slow, and if I try to turn the SBC on again before the computer's power conditioning has drained below about 1V, the SBC is unresponsive.

I'm looking for a way so that when I turn off my power switch, that power will not only stop being applied, but all residual charge will be dissipated more quickly.

I know that another option is to build some sort of power-up delay circuit to prevent power from being applied instantly when the switch is turned back on, but right now I would prefer a way to drain the charge instead.

 

[betwixt]

That makes more sense!

the simplest way I can think of is to replace the power switch with a changeover type. Wire it so the common pin is the power TO the SBC and the normally open is power FROM the PSU. Connect the normally closed contact to ground (or through a low value resistor to ground). When switched on, the power is applied normally, when switched off, the PSU is isolated and the SBC power line is grounded to kill the charge.

Does this sound reasonable - it's a one component solution!

Brian.

 

[dat209]

You know, sometimes its the simplest solution that is the hardest to realize lol... Yes, this was a complete hand to forehead moment haha.