EEVblog Electronics Community Forum

Electronics => Projects, Designs, and Technical Stuff => Topic started by: panoss on April 02, 2019, 01:55:54 pm

Title: 1117 LDO and reverse polarity protection
Post by: panoss on April 02, 2019, 01:55:54 pm
I 've built a device which is powerεd by a 5V PSU.
The voltage firstly goes through an FQP27P06 (P-Channel MOSFET) for reverse polarity protection.
Then it goes to the 1117.
So my question is: do I need the MOSFET for reverse polarity protection?
Or the 1117 has it's own protection against reverse polarity?
Title: Re: 1117 LDO and reverse polarity protection
Post by: balage on April 02, 2019, 02:37:36 pm
Hi,

Here is the simplified internal schematic. I would say it is not reverse polarity protected. However I cannot see any path to the reverse current to flow as all of the transistors are blocking that way.

I use a schottky diode for revesre polarity protection. I don't know which one is the better... I always see the least in TI's reference designs. They must have a reason.
Title: Re: 1117 LDO and reverse polarity protection
Post by: panoss on April 02, 2019, 02:41:59 pm
If a reverse polarity PSU gets connected the current will stop at the transistors at the bottom of the schematic?
Title: Re: 1117 LDO and reverse polarity protection
Post by: balage on April 02, 2019, 02:47:40 pm
I think yes, they block the way. But the question is the voltage that those transistors can carry on their E-C without blowing.

When I get home I will try. :)
Title: Re: 1117 LDO and reverse polarity protection
Post by: schmitt trigger on April 02, 2019, 02:58:25 pm
My take on this is, unless the datasheet specifically indicates that it is reverse-polarity protected, and the reverse voltage withstanding level, I would assume that it is not protected.

For LDOs, I also prefer a PMOS rather than a schotky where extremely low dropouts are required.
Title: Re: 1117 LDO and reverse polarity protection
Post by: Gyro on April 02, 2019, 07:30:44 pm
Agreed, it's an essential rule of product design - If it isn't in the datasheet then it doesn't exist. No amount of peering at simplified schematics changes that. Even if you get away with it on a specific sample, it can't be assumed that it will work for other production batches or second-sources.
Title: Re: 1117 LDO and reverse polarity protection
Post by: David Hess on April 03, 2019, 12:26:32 am
I have heard that 1117s from different manufacturers have significant differences.

As shown the reverse voltage cannot exceed 1 Vbe + 1 reverse Vbe so about 6 volts.  I think all of the junction isolation substrate diodes go to the adjustment pin so the series resistance at that pin will limit current.
Title: Re: 1117 LDO and reverse polarity protection
Post by: panoss on April 03, 2019, 10:56:59 am
Ok, thank you all. I 'll keep the MOSFET.
Title: Re: 1117 LDO and reverse polarity protection
Post by: Peabody on April 03, 2019, 01:35:09 pm
There's a possibility the MOSFET wouldn't be needed at 5V.  When I've tested this with a PNP transistor, I found that applying reverse polarity produced no current - up to 8V.  So it acted like a Zener.  You could test this with your meter.  Apply reverse voltage through an ammeter, and see at what voltage current begins to flow.

Also, is there actually a risk of reverse polarity here?  The risk usually comes with batteries, but you said 5V, which doesn't match most battery types.
Title: Re: 1117 LDO and reverse polarity protection
Post by: Psi on April 03, 2019, 01:48:53 pm
LM2937IMP-5.0  have reverse polarity protection and 60V transient protection.

Downside is they're not cheap
Title: Re: 1117 LDO and reverse polarity protection
Post by: panoss on April 03, 2019, 01:56:32 pm
The risk for reverse polarity is, indeed, very small because the 5V comes (from a PSU) from a standard connector like this one:
(https://images-na.ssl-images-amazon.com/images/I/21XVreSxZpL.jpg)

But for this 1 in a million case...