Diode for reverse polarity protection

[daxliniere]

I'm designing a piece of equipment which will draw up to 6 amps and will be powered from a 12VDC/6A switchmode power supply. The PSU itself will have a DC 'barrel' connector.
I'm concerned about someone connecting a non-standard PSU with negative tip plug and blowing up the units.

A diode is obviously the easiest way, but I would like to minimise the voltage drop. Would a 6A Schottky be a good option and, if so, any models you could recommend, please?
It's not going to be used in high temperature environments.

Thank you and all the best,
Dax.

[madires]

You can also use a simple MOSFET based solution which has a lower drop than a Schottky diode.

[Peabody]

A Schottky diode is still going to drop 3V or more, and at 6 amps it will get warm.  A better choice is a mosfet mounted in reverse orientation, which will drop almost no voltage at all.  Here's an excellent video on this option:

Edit: I meant 0.3V of course.



I should add that while you typically see this on the high side with a P-channel mosfet, you can do exactly the same thing on the low side with an N-channel mosfet, and N-channels typically have lower on-resistance than P-channels.  However, the low-side option can be the wrong choice if you have other devices on other supplies sharing a common ground.

Edit:  I should have mentioned that the mosfet is likely to be smaller and cheaper than a 6A Schottky diode.

[TimFox]

The first Schottky diode rated at 6 A average current that I found on Google (SK64 series) quotes maximum forward voltage of 0.65 V at 6 A or 0.85 V for higher voltage versions, at a junction temperature of 25^o C.  At higher junction temperature, the voltage should decrease.

[madires]

And a Schottky diode rated for a much larger current, e.g. an MBR20xx, would give you a slightly lower voltage drop, still much higher than a MOSFET.

[magic]

Schottkys have a bit of reverse leakage, so depending on the tolerance of your circuit for being fed a few mA in reverse, you may want to follow the schottky with another tiny schottky shunting reverse current to ground.

But I recall seeing a superior and still simple MOSFET-based solution somewhere.

[w2aew]

You could always put a reverse-biased diode in parallel with the input power.  If someone connects power in reverse polarity, it'll blow the fuse.  Advantage is that there's no voltage drop under normal conditions.

[CaptDon]

I believe you meant .3v or more? Some mobile equipment uses an inline fuse with a shunt diode after the fuse. Accidental reverse polarity blows the fuse. Was very common on old tape decks and C.B. radios.

[Peabody]

Yes.  0.3V.  Thanks for the correction.

[Terry Bites]

A well chosen P-ch MOSFET will will have lower losses and a 1000x improvement in reverse leakage over a fat ass Schottkey- Here is a good guide:  components101.com/articles/design-guide-pmos-mosfet-for-reverse-voltage-polarity-protection

Of couse this will all be trivial compared with finding and buying MOSFET you love.  I've quickly simmmed this to show you the trade-offs. Based on 1-off prices. If its a consumer product I'd go with the diode. The enviromnental impact (and your pricing strategy) willl be detemined by Mr Putin and the new world order he will be announcing shortly.

Did you watch the prophetic TV series Okkupert (https://en.wikipedia.org/wiki/Occupied)?

That'll make your engineering choices clearer. Net Zero, Net Zero, hahaha hahaha.....hahaha...........

Use the big Fat Diode and a linear PSU to save the earth OK!!!

[TimFox]

The fundamental question is whether or not the application can tolerate roughly 0.5 V of voltage drop.  The series diode is simple and least parts count, the shunt diode and fuse are good for unexpected reversal, series and shunt diodes are belt and suspenders, and the MOSFET has the least forward loss.

[radiolistener]

A diode is obviously the easiest way, but I would like to minimise the voltage drop. Would a 6A Schottky be a good option and, if so, any models you could recommend, please?

if you're designing circuit which doesn't allow voltage drop due high current, I would suggest to not use diodes or transistors, just use relay which enables load only when proper voltage polarity is applied.

Here is example for reverse polarity and overvoltage protection for 20-40 Amps current:


When incorrect polarity is applied, the load is just disconnected and "REVERSE" LED is on to notify operator that polarity is wrong. And when overvoltage is applied the thyristor will be opened to make short circuit in order to protect the load until fuse will be burned out.

Of course, relay consume some current, about 18 mA for a cheap Chinese 12V relay, but that is not an issue when you need to protect high current load which consume 100-1000 times higher current

[exmadscientist]

Be careful with that relay circuit... the contact resistance of a highish-current relay is not all that great. A popular example is the Song Chuan 832A series, which specifies "<50mΩ, @1A/6VDC, initial". This is... not great if you are very concerned about IR drop over the lifetime of the thing. I don't think they're actually that bad in practice, but the spec sure allows them to be. Alas, you won't (okay, I couldn't) do better with a different vendor; many are even worse at 100mΩ.

So if you are very bothered about such things, MOSFETs are often better.

[madires]

You could always put a reverse-biased diode in parallel with the input power.  If someone connects power in reverse polarity, it'll blow the fuse.  Advantage is that there's no voltage drop under normal conditions.

Same story as the relay. Fuses have resistance too. But they are cheaper than a MOSFET.

[SL4P]

6A through a barrel connector ?

[mariush]

I'm designing a piece of equipment which will draw up to 6 amps and will be powered from a 12VDC/6A switchmode power supply. The PSU itself will have a DC 'barrel' connector.
I'm concerned about someone connecting a non-standard PSU with negative tip plug and blowing up the units.

A diode is obviously the easiest way, but I would like to minimise the voltage drop. Would a 6A Schottky be a good option and, if so, any models you could recommend, please?
It's not going to be used in high temperature environments.

Thank you and all the best,
Dax.

If your equipment draws UP TO 6 AMPS , you would NOT use a 6A max. power supply, you would use something with at least 10-20% margin, for safety.  You would also NOT use a diode rated for exactly 6A, you'd also set some safety margins there, like 10-20% extra.

Also account for the cable length ... if you get a laptop adapter that outputs 12v 6A, you may have 0.1-0.3v drop on the cable at 6v so account for that.

It would make more sense to use a higher voltage, like 18-20v and lower current (ex 18.5v 3.5A laptop adapter = 65w or 16.5v 3.65A = 60w  or 19v 4.74A = 90w

[Zipdox]

A dead simple cheap solution is a diode across the input with the anode on ground, combined with a fuse. That way, if an incorrect power supply is used then the diode will short the supply and the fuse will blow.

[Psi]

6A through a barrel connector ?

They make 10A ones, but yes lots of them would melt. Need to get the right ones.

[radiolistener]

A dead simple cheap solution is a diode across the input with the anode on ground, combined with a fuse. That way, if an incorrect power supply is used then the diode will short the supply and the fuse will blow.

the problem with diode+fuse solution is that it still apply some reverse polarity voltage to the load. Because any diode has voltage drop. Since power source can provide very high current, such a small voltage can be enough to burn out electronics, make fire and physically destroy the load.

This is why relay solution is more safe, especially if you deal with very high current (more than 10-20 Amps). With realy solution the load is always physically disconnected from power supply while power supply with proper polarity is not connected.

[magic]

the problem with diode+fuse solution is that it still apply some reverse polarity voltage to the load.

And such a shunt regulator is a very low impedance voltage source.

This is why relay solution is more safe, especially if you deal with very high current (more than 10-20 Amps). With realy solution the load is always physically disconnected from power supply while power supply with proper polarity is not connected.

Except when the supply changes polarity in a few ms
But that's probably not one of the most likely scenarios.

[daxliniere]

I just discovered that my notifications for thread replies were disabled for some reason! Sorry for not responding!
Thank you everyone for taking the time to respond; there's a lot of fantastic information in here. When I get back to this design (other priorities took over), the first thing I'll do is revisit this polarity protection business.

[Veketti]

Use bridge rectifier if you can afford the small voltage drop. Then you don't need to worry which way you connect the power leads and it always works.

[daxliniere]

Yes, bridge rec is a great solution, but I am trying to minimise forward voltage drop in this scenario.

[Psi]

There's potential to use a polyfuse instead of a normal fuse in crowbar circuit, but you need a big enough diode to handle that current for long enough to trip the polyfuse. And the polyfuse will need some current to hold it in a tripped state. So the diode will have to burn that off too.

Also your circuit will still get 0.8-1V or so of reversed voltage, that may or may not be an issue.

Also the powersupply may not have the current needed to trip the polyfuse and may just switch off, or start pulsing trying to start back up.  Clamped to 0.8V obviously.

[MrAl]

I'm designing a piece of equipment which will draw up to 6 amps and will be powered from a 12VDC/6A switchmode power supply. The PSU itself will have a DC 'barrel' connector.
I'm concerned about someone connecting a non-standard PSU with negative tip plug and blowing up the units.

A diode is obviously the easiest way, but I would like to minimise the voltage drop. Would a 6A Schottky be a good option and, if so, any models you could recommend, please?
It's not going to be used in high temperature environments.

Thank you and all the best,
Dax.

Another way to do this is to use a reverse connected diode from the input or output to ground, with a fuse in series with the actual input or output.
This way the power source current does not flow though any diode or transistor only through the fuse, unless it is connected in reverse polarity, and then it will conduct through the fuse, through the diode to ground, and the fuse blows thus disconnecting the reversed power source.  The maximum voltage the original input (or output) will see is less than 1 volt usually.