Low side P-FET usage is opposite of expected, what did I do wrong?

Question

I'm working on a circuit that should close a reed relay (normally open) when TI/COUT is low and open said relay when TI/COUT is high (6V).

I'm using a P-FET low side transistor to ground the relay & status LED, but it's working the opposite way of what I expected: LED is lit and relay closed when TI/COUT is high.

What did I do wrong, and what would be the necessary component or circuit changes needed?

Original design:

Revised design:

Answer 1

If you want the relay to be 'off' when T1 is high, you have to wire it like this:

^{simulate this circuit – Schematic created using CircuitLab}

This will only work in if the control voltage for 'off' is similar to (or a bit higher than) the relay supply voltage.

Answer 2

Your PFET will conduct all the time because it has a body diode that is forward biased in that orientation.

If you use an NFET in place of the PFET, as suggested in another answer, the circuit will "work", but the logic will be reversed. That is, the relay coil will be active when the gate is high, and the relay coil will be de-energized when they gate is low. However, you stated:

I'm working on a circuit that should close a reed relay (normally open) when TI/COUT is low and open said relay when TI/COUT is high (6V).

Spehro Pefhany beat me to the punch of getting a working CircuitLab simulation. But he has the right idea. The source pin of the PFET needs to be connected to the positive supply, making the switch a high-side switch.

One problem you may face with the particular PFET you are using is that its threshold voltage is only 0.4-1.0 V (per the CJ2301 datasheet). Therefore, to turn the relay off reliably, you need to bring the gate voltage to no less than the supply voltage minus 0.4 V. You might have better luck with a FET with a slightly larger threshold voltage.

Answer 3

To switch a load on the low side, you need to use an n-FET instead of a p-type. Then a normal logic high on the gate will turn the FET on.

Fortunately, for SOT-23 FET the p- and n- type pinouts are the same. So just change the FET to a suitable n-type. Then it should work.

What went wrong with the p-FET? Two things:

• Body diode is forward biased (it will be reverse bias with n-FET)
• Gate-source bias will never turn on the FET (will be ok with n-FET.)

Result? The p-FET conducts all the time, with a forward bias of one diode drop, and will not shut off.

One more thing. Add a flyback diode across the relay coil to catch the spike when the FET turns off. Otherwise your FET can be damaged.