Driving a 24volt mosfet circuit requires power, so parts will get hot whichever way you choose.
In this case the transistor and R2 (0.17W).
This is a clever circuit, and a small heatsink or doubling the resistor values should get you out of trouble.
Leo..
While Nick Gammon's circuit is nice it won't work here due to the large input voltage variation (it "eats" 5 V from the supply so only 5 V is left for Vgs in the worst case).
The proposed fix from the linked stack exchange discussion is even worse. The solution from OP is the only that will work.
I see that I can never convince you.
I know that switch circuits (including those with MOSFET) are chosen in order to decrease power, not to increase.
You can use this circuit, and I will use a different one.
I think it will be right.
I also said that the circuit will work in post # 6.
A similar circuit (but with MOSFET as Q1) is recommended by Vishay Siliconix application notes P-Channel MOSFETs, the Best Choice for High-Side Switching on Figure 3.
He is right! Assuming the same drive current (meaning the same switching speed) both circuits consume about the same power. Nick's circuit dissipate some of it in Q1 while your circuit dissipate it in the resistors. If Q1 is able to handle the power it is OK. If Q1 is hot in Nick's circuit a resistor will be hot in the other.
If the absolute maximum gate-source voltage is +/-20V, make that zener much lower voltage such as 6V8 - this must be a logic-level pFET?
Typically MOSFETs that are logic level have absolute maxes of +/-20V or less, and the rest are +/-30V. You never want to go anywhere near the gate oxide breakdown voltage - 18V is not effective protection for a 20V "absolute maximum never exceed" spec'd device.
I exchanged the zener diode to one with 12V. That should be enough to ensure a low R_DS_on and it's still far away from 20V. The MOSFETs are switching at around 4V, but the resistance is too high for what I intend to do. With 12V everything works as expected.
Of course. That is why you must ignore the VGS(th) specification and look only at the VGS specified for RDS(on).
But you never said what you expect from the circuit.
What is the MOSFET, what is the load?
MOSFET could be a IRF5305, the load will be a solenoid with a low DC resistance.
I think the circuit works well and could be used for other purposes as well.
The zener diode does not provide low R_DS_on, it protects the gate from overvoltage.
Resistance is too high, how much is it?
Why IRF5305? This is usual MOSFET, but you need p MOSFET with low gate threshold voltage like NDP6020P, AO3401A, NTD25P03L etc.
Low resistance, how much?
You do not say any real value and say that the circuit does not work well.
The IRF5305 is a p-channel MOSFET.
The zener will limit the gate voltage to 12V, and this 12V will be enough to ensure a low R_DS_on (60 mOhm). The solenoid will have <1 Ohm DC resistance.
Everything is fine, my initial question is answered.
Load resistance <1 ohm? This was not in your question.
You really want to feed the 30V to solenoid with <1 ohm?
This is 1000 watts of power!
I think you're just fooling us and wasting our time.
No, I‘m not fooling anybody. I know it‘s a huge power, but as I only discharge a small capacitor, dissipated heat will still be bery low.
Hi,
Okay I'm game, what is the REAL WORLD application?
Why do you need to pulse a solenoid?
For how long does the pulse need to be?
What does the solenoid activate?
What will be your power supply?
Thanks.. Tom.. 
Many solenoids are "pulsed" -- they're actuated for a short period of time.
This topic was automatically closed 120 days after the last reply. New replies are no longer allowed.