Best method for reverse polarity protection and without power loss?

MikeML

Joined Oct 2, 2009
5,444
Hey Ernie,
then how do you protect a mobile transceiver that costs ~$800, operates on a 12Vdc automotive supply, where it draws 25A, and will not tolerate more than a few 10s of mV of series drop in the power lines, and must prevent the input from ever going more than ~1V reverse?

I'd say that having to replace a $0.10 fuse and installing a $1 rectifier is a pretty good tradeoff...
 

ian field

Joined Oct 27, 2012
6,536
My power source is 9v for the entire circuit. As I undertand, the common way to prevent damages to a circuit would be to use a diode. Another method is a bridge rectifier However, I get voltage drops and lose power with a diode, and twice the loss of power with a bridge rectifier.

Taking to consideration for voltage drops with circuit protection, my question is if there is a possible way to make a reverse polarity protection without any loss of power at all? Or will there always be losses of power no matter what for this type of feature in any circuit?
You can use a low RDSon MOSFET wired so the body diode isn't reverse biased (it would defeat the object if it was) as long as the gate polarity is correct in relation to the channel, the channel will conduct with a much lower volt drop than the Vf of the body diode.

This was explained in some detail in a recent issue of Everyday Practical Electronics.

If volt drop is the critical issue but you have abundant current supply - you can energise a relay coil with a single diode in series, then the contacts will only close with the applied power the right way round.
 

Thread Starter

Fuji

Joined Nov 8, 2014
100
As Figi is asking for a zero power loss protection circuit then any and all schemes employing a diode are missing the point. A MOSFET will only loose power due to the drain to source resistance, which can be very very small for modern devices.

A MOSFET is a most useful way of dealing with things AS LONG AS THE INPUT VOLTAGE IS LIMITED. If the input can exceed the gate to source voltage then either the method is not valid, or some voltage limiting is necessary. A resistor and zener can work well there in many cases, though it may loose some power to run the zener for circuits constantly running over the Vgs limit.
Is a schottky diode enough? I heard it has a voltage drop of about 0.2
 

tracecom

Joined Apr 16, 2010
3,944
Yes, a Schottky diode has a very low voltage drop, but that isn't what you asked for.

Taking to consideration for voltage drops with circuit protection, my question is if there is a possible way to make a reverse polarity protection without any loss of power at all?
Had you stated that you wanted a method with a very low voltage drop, a Schottky diode would have been suggested several posts ago.

ETA: My mistake! Rifaa made that suggestion in post 11.
 

ronv

Joined Nov 12, 2008
3,770
OK, Here you go. Lets use 1 amp for your current.
Fuse .09 ohms = .09 watts
PFET .005 ohms .005 watts
NFET .002 ohms .002 watts
Diode .5 volts .5 watts.
1.2 inches of 24 awg wire .005 ohms .005 watts
 

cmartinez

Joined Jan 17, 2007
8,220
OK, Here you go. Lets use 1 amp for your current.
Fuse .09 ohms = .09 watts
PFET .005 ohms .005 watts
NFET .002 ohms .002 watts
Diode .5 volts .5 watts.
1.2 inches of 24 awg wire .005 ohms .005 watts
Interesting list, but pardon my lack of experience with NFETs configured as diodes. I know that NFETs are more efficient devices than PFETs. But isn't it easier, and perhaps more convenient, to use a PFET instead? I've never seen an NFET used as rectifier before, but I'm guessing that it has to be placed at the circuit's return to ground (instead of right after the circuit's power supply, as the PFET), right?
 

ErnieM

Joined Apr 24, 2011
8,377
Is a schottky diode enough? I heard it has a voltage drop of about 0.2
Do you mean a schottky in series with a lead? That is normally good protection too, though it also has limits such as schottky diodes are not available with a very high lower breakdown voltage such that you can find in junction diodes.

It would help to flesh out a complete answer by flipping over all the cards and stating where this 9V comes from, how much current your load will be taking, and perhaps hinting at what this load is.

Hey Ernie,
then how do you protect a mobile transceiver that costs ~$800, operates on a 12Vdc automotive supply, where it draws 25A, and will not tolerate more than a few 10s of mV of series drop in the power lines, and must prevent the input from ever going more than ~1V reverse?

I'd say that having to replace a $0.10 fuse and installing a $1 rectifier is a pretty good tradeoff...
Hey Mike, if I had an $800 mobile anything that could not handle a drop of more that a few millivolts off an automotive supply that in normal operation can vary by VOLTS then I would simply put it back in the box it came in and return it, instead of hijacking somebody else's thread for suggestions.

Or I may just get a device such as a PSMN0R9-25YLC MOSFET which has a specfied Rds on of only .99 milliohms at 25A for a drop of less than 25 millivolts no load to full load.
 

ronv

Joined Nov 12, 2008
3,770
Interesting list, but pardon my lack of experience with NFETs configured as diodes. I know that NFETs are more efficient devices than PFETs. But isn't it easier, and perhaps more convenient, to use a PFET instead? I've never seen an NFET used as rectifier before, but I'm guessing that it has to be placed at the circuit's return to ground (instead of right after the circuit's power supply, as the PFET), right?
Yes, It goes in the ground side, but it really doesn't matter. See post 5.
While it's true it acts as a diode, it does that only until it is turned on - then it's a low loss switch.
 

MikeML

Joined Oct 2, 2009
5,444
...
Hey Mike, if I had an $800 mobile anything that could not handle a drop of more that a few millivolts off an automotive supply that in normal operation can vary by VOLTS then I would simply put it back in the box it came in and return it, instead of hijacking somebody else's thread for suggestions.
Since when is citing common practice, and giving an example of a real-world application to the thread subject "hijacking a thread"?

Getting the maximum available power out of a 120Wrf SSB transciever, which draws ~25A on voice peaks, requires minimizing the series voltage drop in the power feed, especially when operating with the engine off, where the battery voltage might be only 11.9V to begin with. I, for one, am looking for suggestions that minimize the voltage drop.

Or I may just get a device such as a PSMN0R9-25YLC MOSFET which has a specfied Rds on of only .99 milliohms at 25A for a drop of less than 25 millivolts no load to full load.
I just looked, and that is an NFET, which would have to be installed in the negative lead of the power feed, which cannot be done in the typical mobile installation I have described. This can only be done with a PFET.
 
Last edited:

MrAl

Joined Jun 17, 2014
11,393
You misunderstand how the circuit is put together. The diode is parallel to the load and (usually) reversed bias. When the power is connected correctly, there is no current through the diode. If the power is connected incorrectly, current flows through the diode and blows the fuse.
View attachment 77572
Hi,

That is exactly how i added reverse connection protection to my more expensive power supply.

The power supply already had a reverse diode, but no fuse in series. By placing a fuse in series if the load is connected backwards (the load can also be a battery itself in this case) the fuse blows because the diode conducts. In some cases the diode might blow short also so it would have to be replaced as well.
I would suggest using a high current diode that can take a large current. They uses a 1N5400 series diode in a 10 amp supply which i didnt like too much, but it did protect it one time already. The diode had to be replaced too after that.

The main idea is to not bother the output voltage too much during normal operation yet still get the needed fast acting protection. Diodes turn on very fast (maybe not off as well but who cares). so they are good at protecting things.
 

ian field

Joined Oct 27, 2012
6,536
OK, Here you go. Lets use 1 amp for your current.
Fuse .09 ohms = .09 watts
PFET .005 ohms .005 watts
NFET .002 ohms .002 watts
Diode .5 volts .5 watts.
1.2 inches of 24 awg wire .005 ohms .005 watts
A scheme that satisfies the OP's request to the letter, is a fuse and a shunt protection diode.

It has the downside that reverse polarity blows the fuse - and if the shunt diode isn't generously rated, it can fail short circuit and blow the next new fuse regardless of polarity.
 

cmartinez

Joined Jan 17, 2007
8,220
A scheme that satisfies the OP's request to the letter, is a fuse and a shunt protection diode.

It has the downside that reverse polarity blows the fuse - and if the shunt diode isn't generously rated, it can fail short circuit and blow the next new fuse regardless of polarity.
I'm guessing that one also has to be practical... maybe the reverse diode technique is also the simplest... and besides, I don't think that the OP means to connect and disconnect his device so frequently so as to make a blown fuse an annoying event, right?
 

crutschow

Joined Mar 14, 2008
34,285
So for Mike's SSB receiver requirements the parallel diode and fuse seems best

For the op's requirements I would still go with the MOSFET and resistor. (I hate fuses).
 
Top