Connecting power with incorrect polarity is an easy mistake to make. Fortunately, protecting your device from reverse polarity is also quite easy.

Bad things can happen when you reverse the polarity of your device’s power supply. Swapping the positive and negative power leads is probably the primary method of “letting the smoke out” of a shiny new PCB, and that is actually a better scenario than causing some sort of subtle damage that leads to perplexing or intermittent malfunctions. Reverse polarity can also occur after the testing and development phase. A device will generally be designed so as to prevent the end user from incorrectly plugging in a power cable, but even the best of us might occasionally insert a battery without looking at the polarity diagram....

I prefer to use whatever means are available to make reverse polarity physically impossible, but the bottom line is that the device is never truly safe unless the circuit itself is able to survive a reversed supply voltage. In this article, we’ll examine two simple yet highly effective ways to make your circuitry robust against power-supply-polarity mishaps.

 

What Is a Reverse Polarity Protection Diode?

You can, in fact, get reverse polarity protection with a diode. Yes, all you need is one diode. This really does work, but of course a more sophisticated solution could provide superior performance.

The idea here is to put the diode in series with the supply line.

 

 

If you’re not familiar with this technique, it might initially seem a bit strange: Can a diode change the polarity of an applied voltage? Can it really “isolate” the downstream circuitry from the applied voltage?

It certainly cannot “undo” the reverse polarity, but it can isolate the rest of the circuit from this condition simply because it will not conduct current when the cathode voltage is higher than the anode voltage. Thus, in a reverse-polarity situation, damaging reverse currents cannot flow and the voltage across the load is not the same as the reversed power-supply voltage because the diode functions like an open circuit.

The LTspice schematic shown above allows us to investigate the transient and steady-state behavior of the diode-based protection circuit. The power-supply voltage is initially at 0 V, then it abruptly changes to –3 V. My idea here is to simulate the effect of incorrectly inserting two 1.5 V batteries (or one 3 V battery). The simulation includes load resistance (corresponding to a circuit that consumes about 3 mA) and load capacitance (corresponding to decoupling caps for a few ICs).

 

 

You can see that some reverse (i.e., cathode-to-anode) current does flow through the diode. The transient current is very small and the longer-term current is miniscule. However, current is flowing and consequently the cathode side is not completely floating; instead, there is a very small reverse voltage across the load circuitry. This is not the steady-state condition, though. If we extend the simulation out to 300 ms, we see the following:

 

 

So as the load capacitance charges up and becomes an open circuit, the current falls to zero (or more precisely, 0.001 femtoamps, according to LTspice), and consequently there is no reverse voltage across the load. The conclusion here is that the diode isn’t perfect, but as far as I’m concerned it’s close enough, because I can’t imagine that any realistic circuit would be negatively affected by ~100 ms of a few microvolts of reverse polarity.

 

Pros and Cons

By now the advantages of this circuit should be clear: it’s cheap, exceedingly simple, and highly effective. There are definitely disadvantages that need to be considered, though:

  • During normal operation, the diode drops its typical ~0.6 V. That could be a significant portion of the supply voltage, and as the battery voltage decreases the device may stop working prematurely.
  • Any component that has a voltage drop across it and current flowing through it is consuming power. If that dissipated energy comes from a battery, the diode is reducing battery life. This may not be an acceptable trade-off in devices that have very low risk of experiencing reverse polarity.

 

Reverse Polarity Protection with a Schottky Diode

An easy way to mitigate both of the above disadvantages is to use a Schottky diode instead of a normal diode. This approach reduces voltage loss and power dissipation. I’m not sure how low Schottky diodes can go, but in some cases the forward voltage can be below 300 mV.

Here is the new simulation circuit:

 

 

The following specs give you an example of the forward-voltage characteristics of a BAT54 diode:

 

Table taken from this Vishay datasheet.

 

Here is a plot of the transient and steady-state response of the Schottky-based reverse-polarity-protection circuit.

 

 

You can see that the reverse current and the reverse voltage across the load are much larger than what we observed with the non-Schottky diode. This higher reverse leakage current is a known disadvantage of Schottky diodes, though in this particular application the reverse current is still far lower than anything that would cause serious concern. So when it comes to reverse-polarity protection, Schottky diodes are definitely preferred.

 

Conclusion

We’ve seen that a single diode is a surprisingly effective way to incorporate reverse-polarity protection into a device’s power-supply circuitry. Schottky diodes have lower forward voltage and consequently are generally a better choice than normal diodes. An AAC contributor who has experience with these circuits recommends p/n 1N4001 (if for some reason you want to use a normal diode) or p/n MBRA130 (this one is a Schottky).

 

Comments

4 Comments


  • vanderghast 2018-06-29

    Someone can also use a PMOS in reverse (its drain, not its source pin, toward the source of alimentation), load on the low side and then, just a low 0.15V is lost (depends on the exact PMOS that you use) which is even less than a Schottky diode, with a better blocking.  Got that idea at https://www.youtube.com/watch?v=IrB-FPcv1Dc&t=184s. It also explain WHY it works (not necessary evident at first glance that just a small 0.15V is lost under “usual” operations). The author speaks of a P-FET, but the term PMOS is more usual (to me, at least).

  • recklessrog 2018-06-29

    My preference would be to put a fuse inline and then a diode normally reverse biased in parallel with the supply after the fuse.
    Reverse polarity would forward bias the diode and blow the fuse. No forward voltage drop to worry about.

  • Bismarck49 2018-06-29

    I’m not sure if I understand why the current goes back to zero after the transient response when the capacitor charges up. The small reverse current would charge up the capacitor and cause it to be an open circuit, I agree with you there. But your diagram has the load resistance in parallel with the load capacitance. So even though the capacitor is open circuit, reverse current could still flow through the resistor could it not? Or am I to assume the resistance is large enough that virtually no current makes it through? Thanks!

    • RK37 2018-07-02

      Yes, current can still flow through the resistor, but the steady-state current that can flow through the resistor is small compared to the transient current that flows through the capacitor.