I need to control a DC relay coil that is connected to positive and negative supply rails with a logic signal. I figured a FET would do the job, but I'm not sure what type. Any suggestions would be appreciated

 

With the Source of the FET at some negative voltage it will be on all the time if the gate is at GND or positive. You CANNOT do this with a logic gate as the circuit is drawn.

ETA: what is the magnitude of V+ and V-?

 

With the Source of the FET at some negative voltage it will be on all the time if the gate is at GND or positive. You CANNOT do this with a logic gate as the circuit is drawn.

ETA: what is the magnitude of V+ and V-?

+/-12VDC

I simulated (TI-TINA) this with an N-channel FET and it does in fact work.

 

Any suggestions would be appreciated

I would suggest a Mosfet. What are the specs on the relay?
You mentioned a bi-polar supply so 24 vdc from -V to +V, correct?

 

I simulated (TI-TINA) this with an N-channel FET and it does in fact work.

If you look carefully at the voltages and currents you will see that it does, in fact, not work (for the circuit you posted).

Below is the LTspice sim of a circuit that will work, using an NPN BJT and a P-MOSFET:
Note the 24V voltage across the relay coils (yellow trace) generated by the 3.3V input control voltage (green trace) with the ±12V supplies.
The P-MOSFET can be just about any with a 40V or greater voltage rating and an on-resistance of <10% of the relay coil resistance.

 

+/-12VDC

I simulated (TI-TINA) this with an N-channel FET and it does in fact work.

Then you did something wrong, because it does not work.

 

The very first question would be "what is the reference level of the control logic signal??" To be clearer, what does this undefined control logic signal change relative to??? Every voltage level is relative to some reference point.
Next, what part of the relay is connected to the voltages of the power supply?? It could be coil or contacts based on the description.

 

[IMG align="left" alt="crutschow"]https://forum.allaboutcircuits.com/data/avatars/m/19/19451.jpg?1681483160[/IMG]
crutschow
Thanks for the detailed reply. I simulated the circuit you provided with some of the additional details pertaining to the relay and it works great. The actual rails are +/-16V so I added a couple of series Rs to drop the voltage on either side of the relay. I increased R1 to 300k which balanced the current through those resistors. Thanks again!

 

The actual rails are +/-16V so I added a couple of series Rs to drop the voltage on either side of the relay.

You just need one 1.6kΩ resistor in series with the relay.
No need (or advantage) for two.

I increased R1 to 300k which balanced the current through those resistors

Do you mean R2?
R2 has nothing to do with those resistors' current since they are in series and thus have to always carry the same current.

 

You just need one 1.6kΩ resistor in series with the relay.
No need (or advantage) for two.
Do you mean R2?
R2 has nothing to do with those resistors' current since they are in series and thus have to always carry the same current.

Yes, R2 in mine. I did see slight differences in the currents through each resistors and adjusting R2 balanced them out. Shouldn't there normally be more current through R3 as it flows thru the BJT?

 

[IMG align="left" alt="crutschow"]https://forum.allaboutcircuits.com/data/avatars/m/19/19451.jpg?1681483160[/IMG]
crutschow
Thanks for the detailed reply. I simulated the circuit you provided with some of the additional details pertaining to the relay and it works great. The actual rails are +/-16V so I added a couple of series Rs to drop the voltage on either side of the relay. I increased R1 to 300k which balanced the current through those resistors. Thanks again!

View attachment 330271

I think I'd reduce R6 to 4.3k, so at least 1ma is flowing in the base, and the PMOS gate is pulled as close to ground as possible when turned on.

EDIT: Or change the 2N3904 to 2N7000 NMOS and remove R6

 

Yes, R2 in mine. I did see slight differences in the currents through each resistors and adjusting R2 balanced them out. Shouldn't there normally be more current through R3 as it flows thru the BJT?

Yes, the small current through R2.
But what's the reason for wanting the two currents to balance?
There's no electronic need for that.