Limit 12VDC motor to momentary run time

Question

My latest project involves running a 12VDC motor very, very briefly via a momentary DPDT switch that I have wired to reverse polarity depending on which direction the switch is thrown.

In the end I just want the motor to engage very quickly as when the switch is flicked on in either direction because it is attached in a way that will stall if run continuously. Think of electric door locks -- this motor spins to pull a latch, and can spin in the reverse to push a latch, but it should not run beyond the very brief push/pull. The switch itself works for this purpose if it is just flipped on and let go to return to the center/off position. But I want to dummy-proof the setup so someone can't physically hold the switch in the on position and burn the motor out.

I'm talking less than a second ON. Maybe even 250ms. The time it takes to flip the switch on and back, it should cut power if it ever exceeded that.

I've seen some discussions of using a setup with transistors to function as on/off timers, as well as 555s paired with a relay. Others talk about current limiters to cut in the event of a stall. Is one way more suitable than another for my situation?

I'm also a little unsure if the switch reversing polarity would affect possible solutions -- i.e., I believe I read that you should use an NPN transistor after the load, but a PNP before the load... Wouldn't reversing the polarity also effectively reverse the before/after load relationship?

Here is the basic circuit I'm working with now that has no protection against holding the motor in an ON state. I'm not sure how to draw (ON)-OFF-(ON) switches in circuitlab yet, so this drawing shows ON-ON, but my actual switch is (ON)-OFF-(ON):

Answer 1

Here's a solution.

Even with the toggle switch held 'on', the motor would be de-energised by the relay as the capacitor gets charged.

Component values are specific to a working trial circuit, that was built with available components.

Answer 2

^{simulate this circuit – Schematic created using CircuitLab}

The circuit below is one of the many ways of achieving this behavior. R1 adjusts on-time from about 0.1s to 1.4s. SW1 is shown as an SPDT switch, but it should be a reversing switch. R100 is the motor load.

SW2 and SW3 are for simulation purposes only; S2 simulates turning SW1 on at 0.1s, and SW3 simulates turn-off at 1.7s.

Q1+Q2 are a latch, Q3 is the control signal driver for Q4, the latter being the supply switch for the circuit. M1 is the load switch.

• When the supply voltage is turned on, with SW1 open, Q3 conducts, as its base is driven low via R8-R7-R2||C1 in series. This keeps Q4 off for now.

• As soon as SW1 is turned on, D1 conducts, C1 charges up, Q3 gets cut off via R7||R9, and both Q4 and M1 turn on. D1 turns off, bypassed by Q4. Q4 powers the circuit for the duration of the active state, and M1 powers the load.

• After a time determined by R1, C2 charges enough to turn Q2 on, which in turn turns Q1 on, and latches the "finished" state. M1 turns off and Q3 turns on, finally Q4 turns off.

• As long as SW1 is kept on, D1 powers the circuit and maintains the state of the latch Q1+Q2.

The M1 mosfet is not critical - any N-channel power type, rated for the load current, should work in this application. A free-wheeling diode is not necessary. M1's body diode fills that role.

This circuit is set up for simulation. The relevant voltage and current traces, for R1 set to 0%, 50% and 100%, are shown below.