2
\$\begingroup\$

Designing a circuit to measure short circuit current of a power supply every second. Measurement will be on for about 50ms every second. Heat design has been made for this about of time not constant current flow. Current can be up to 15A and the voltage up to 60V when in voltage source.

Want to provide reverse current protection in case the user connects polarity wrong so the components don't overheat. If this occurs I have tested the board and it goes over 180º. In normal mode no component goes over 45º.

Have designed following circuit but it does not work. Works fine in forward mode. In reverse polarity mode once I switch ON both MOSFETs and its fine but I cannot switch them back off even if the gate is set to GND and heat rises drastically.

enter image description here

Any thoughts what could be wrong?


Built new circuit but it's failing when V1 is higher than 50V.

enter image description here

\$\endgroup\$
15
  • \$\begingroup\$ Well, if your voltage source is reversed, then (+) is at the bottom, and the gate is not set to GND, it's set to (+), so your bottom NMOS is switched on and the top NMOS is conducting through the body diode. \$\endgroup\$
    – brhans
    Jul 27, 2020 at 17:02
  • \$\begingroup\$ Why not just use a single PMOS for reverse polarity protection? \$\endgroup\$
    – brhans
    Jul 27, 2020 at 17:03
  • \$\begingroup\$ I dont think the single PMOS will work in my application as once the PMOS goes ON voltage drops to almost 0 at the source of the PMOS, so I would not be able to switch it back OFF. I haven't tried it but that's my thought \$\endgroup\$
    – SolarTec
    Jul 27, 2020 at 17:25
  • 1
    \$\begingroup\$ Oh, I was confused. But the pulldown is to the drain of the lower fet. If the pulldown is to the mutual source, then that will guarantee that Vgs is zero unless you drive it some other way. Under reverse polarity, the drain of the lower fet is a pullup not a pulldown. \$\endgroup\$
    – user57037
    Jul 27, 2020 at 17:55
  • 1
    \$\begingroup\$ No problem with the resitor. It is actually 16mOhm. Three 50mOhm in parallel. 1W each. About 10º temperature rise \$\endgroup\$
    – SolarTec
    Jul 27, 2020 at 18:21

2 Answers 2

1
\$\begingroup\$

This is more or less what I have in mind. When I use back-to-back FET's I almost always start by pulling the gates to the sources with a resistor. This is a foolproof way to make sure they will be off in normal operation. Then I go from there.

In this case, D1 is needed to make sure that during reverse polarity connection no current flows through the body diode of M2 into the control circuitry I assume is connected to G. If that were to happen, the MOSFET's could be damaged as well as anything connected to G.

In addition to what I have shown, you may want Zener protection in parallel with R1 to make sure Vgs never gets high enough to damage M1 and M2.

Note that if and when M1 and M2 fail, you will have a short circuit across the power rail. So you should give some thought to the consequences of that to make sure they are acceptable. It is a foreseeable failure, not a one in a million freak occurrence. So the consequences need to be acceptable, or if not, you may need to add a fuse or some form of detection, etc.

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
9
  • 1
    \$\begingroup\$ Circuit is correct and working. It is however failing at switching voltages over 50V. I am going to add a zener diode and test it and will post back. \$\endgroup\$
    – SolarTec
    Jul 27, 2020 at 18:42
  • \$\begingroup\$ 0 So I added two zener diodes bak to back parallel to R1 but circuit is failing when V1 is higher than 50V, even when polarity is OK. Diode from M1 is going to short-circuit. I am using IRFB4510 NMOS specified at 100V. Following circuit is what I built. Any ideas why M1 is failing? \$\endgroup\$
    – SolarTec
    Jul 28, 2020 at 16:53
  • \$\begingroup\$ I can see your deleted post. Not really sure why its failing. What EXACTLY are the steps to make it fail? Does it fail when you turn on the FET, or when you turn it off, or the moment you connect the power supply? Sometimes depending on what is going on with wires or inductive loads, etc, there can be large voltage transients when you switch a FET. Even 2x over-voltage is possible for example if you try to turn off an inductive load. \$\endgroup\$
    – user57037
    Jul 28, 2020 at 19:01
  • \$\begingroup\$ Another possibility is that the FET is turning on too slowly. During the slow turn-on time it may be dissipating a lot of power. Especially if there is capacitance on the 50V rail. \$\endgroup\$
    – user57037
    Jul 28, 2020 at 20:26
  • \$\begingroup\$ Its failing when I turn the FET ON. Inmediately when it goes on. I though about it turning on slowly and the power thing so I turn the current of the power source to only 400mA to avoid that and it's still failing so it must be a voltage thing \$\endgroup\$
    – SolarTec
    Jul 29, 2020 at 8:27
1
\$\begingroup\$

Built new circuit but its failing when V1 is higher than 50V

It should be be failing well below 50 V. Your circuit does not provide reverse polarity protection.

Ground should be applied to the MOSFET Source terminals, so that V2 is directly connected between the Gate and Source of each FET. The circuit looks like this:-

schematic

simulate this circuit – Schematic created using CircuitLab

(R1 is provided to ensure that the FETs stay turned off if V2 is disconnected)

The FETs should be rated for at least the maximum expected voltage (including spikes caused by wiring inductance etc.) preferably 20% higher to give a safety margin. Most FETs rated higher than 60 V are 'standard' Gate drive, which requires 10 V to turn on fully.

\$\endgroup\$
7
  • \$\begingroup\$ Are you commenting on the 2nd circuit (which I suggested to the OP) or the original circuit the OP drew on paper? I think the 2nd circuit does in fact provide reverse polarity protection. It is actually not all that different from yours. It SOUNDS like the OP is having a problem when switching with 50V or above. But I am not 100% sure. \$\endgroup\$
    – user57037
    Jul 29, 2020 at 0:17
  • \$\begingroup\$ @mkeith I simulated his new circuit in LTspice and it did not work properly with polarity reversed (drew high current continuously, not controlled by V2). The problem above 50V may be due to Drain-Source breakdown (few FETs that turn on fully with 5V can withstand >50V). \$\endgroup\$ Jul 29, 2020 at 6:17
  • \$\begingroup\$ The new circuit I built and suggested by mkeith is failing in normal polarity, not in reverse. The FET I'm using is IRFB4510 and has a max gate threshold voltage of 4V. In reverse polarity below 50V it works fine. \$\endgroup\$
    – SolarTec
    Jul 29, 2020 at 8:31
  • \$\begingroup\$ Did some more testing and found the problem to be the G-S voltage. With voltages above 6V I have tested the input at 63V and 5A for over 50 switches and works fine. When I bring G-S voltage down to 5V it fails inmediately. According to the datasheet it seems like it should work at 5V but I guess its not all true \$\endgroup\$
    – SolarTec
    Jul 29, 2020 at 8:55
  • \$\begingroup\$ With respect to whether the ground is at the sources or at the Drain of M2 I guess being at the sources is better but it complicates my circuit as I have ADC for measuring voltage at the source and the shunt resistor and are referenced to the drain of M2 \$\endgroup\$
    – SolarTec
    Jul 29, 2020 at 9:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.