3

I'm building a P-MOS based reverse polarity protection circuit.

The input requirements are as follows.

It needs to carry 8–40 V input voltage to the output and it needs reverse polarity protection up to −40 V.

I predict that the system consumes a maximum of 8.8 W of power.

For this application, using the DMP6023LFGQ-7 P-channel MOSFET, and using the BZX84-A10,215 Zener diode rated 10 V 250 mW, and I'm using a 10 kΩ, ⅛ W resistor. I understand how it should work in theory, but I'm a junior engineer. I'm afraid of not seeing overheating and other problems. Can you suggest other configurations that are in active stock on Digikey?

enter image description here

JRE
  • 67,678
  • 8
  • 104
  • 179
Electronx
  • 483
  • 2
  • 9
  • 1
    Why don't you try to simulate it with all the required maximum input voltages to see the effect ? Look at the estimated maximum currents and work out the required power ratings of components... – citizen Mar 13 '23 at 08:22
  • 2
    Welcome! _"I understand how it should work in theory, but I'm a junior engineer"_ Luckily, no simulation software I'm aware of will bother with your title but rather give you same result regardless. – winny Mar 13 '23 at 08:36
  • 5
    `and I'm using a 10 kohm 1/8 W resistor` it's 100R in your schematic. Match the schematic to your text, please. – Rohat Kılıç Mar 13 '23 at 08:43
  • "Luckily, no simulation software I'm aware of will bother with your title ... ". This is a rather cryptic comment ;-) Not sure if I understand it ... – citizen Mar 13 '23 at 12:00
  • 1
    @citizen OP claims he/she's "just" junior engineer. My point is that no SPICE simulation is trying to push OP down because of his/her title. It's a joke. – winny Mar 13 '23 at 14:01
  • @winny ahh ok, get it - the uncertainty of "just" ;-) – citizen Mar 13 '23 at 16:20
  • What do you do if you can't find the specific spice model of a product? – Electronx Mar 16 '23 at 06:18

2 Answers2

6

All I see so far is that the 10 kΩ on the gate is a bit low in value, (your schematic is still showing 100R). With a possible 40 V on the load and 10 V on the Zener that leaves 30 V on the resistor, with 10 kΩ that's gives 3 mA. With 3 mA and 30 V the resistor dissipates 90 mW, that's getting fairly close to 125 mW (1/8 W). The resistor will run a bit warm. You could move up to a 1/4 W resistor but then one of the ideas for using a MOSFET in this way is to minimize power loss. Using a 100 kΩ resistor instead would drop the dissipation down to 9 mW. At such low Zener currents the Zener voltage may not be as accurate but high accuracy is not needed here.

Your MOSFET can take 60 V max Vdss, your max should only be 40 V.
The Vgs max is +/-20 V, the zener keeps that down near 10 V.
Rds-On is very low at 25 mΩ max at -10 Vgs, so if passing 1.1 A (e.g. 8 V at 8.8 W) you would only dissipate 30 mW.

So for the most part you seem good to go, (with or with out the higher gate resistor). A simulation, even on a simple circuit, is still a good thing to do to ensure yourself that everything is as expected.

winny
  • 13,064
  • 6
  • 46
  • 63
Nedd
  • 7,939
  • 15
  • 19
5

Aside from the obvious of analysis and verification by testing at maximum load current, and verifying that the voltage stress and power dissipation of MOSFET, resistor and zener are well within limits, consider what could happen under (possibly pathological) conditions such as low input voltage.

Many devices will draw even more current as the voltage drops (since they incorporate switching supplies with constant output power and more-or-less fixed losses). At some voltage they will stop working, often due to some deliberately incorporated undervoltage lockout. If your MOSFET stops conducting sufficiently before the external UVLO kicks in then it could overheat at low voltage. Even a resistive load could be a problem with a small device with insufficient heat sink. The device you linked might be of concern below perhaps 3-4V worst case, and a volt or so less than that typically.

Spehro Pefhany
  • 376,485
  • 21
  • 320
  • 842