Powering from external 5V regulated source

I have seen a number of similar post regarding this that provide conflicting info so I would like to get it answered once and for all.

Using the USB to power a unit also results in the USB 5V to be placed on the 5V rail, and therefore also the the output of the NCP1117 regulator, which many people state as being a bad thing to do to regulators. Since it is designed to work this way by people I assume are much smarter that I am:

  1. How is this NOT a potential problem with back loading the regulator.
  2. How is this any different than applying a 5V regulated source to the 5V pin, NOT Vin, to power the board.

It seams that in half the threads I see where it is asked if you can power the board from an external 5v source via the V5 pin, someone always mentions how this could damage the regulator because it is being back loaded and to be safe you should put a diode across Vout and Vin of the regulator to protect against this.

Thanks for clearing this up for me.

What board?

Before I start, I refer to the UNO R3 board:-
As you mention, when 5V is applied from USB, it connects to the 5V regulator output and does no damage.
I always power my UNOs with a regulated 5V applied to the 5V rail, and have never had a problem.
As I understand it, there will only be a potential problem if there's a load connected between Vin and ground, which could draw excessive current back through the regulator. As long as you avoid this situation, there's nothing to worry about.

  1. How is this any different than applying a 5V regulated source to the 5V pin, NOT Vin, to power the board.

As I say, no difference. (Take a look at the UNO R3 schematic diagram.)

Sorry, I thought it would be obvious which boards this would be relevant to. I was referring to the Uno R3 or the Mega2560. I was hoping the post would serve two purposes. 1) Directly answer the question that is asked quite often with engineering facts related to the circuit as it is designed. 2) To present a good learning experience based on solid electronic design principles with respect to similar PCB system power section design requirements.

I have seen a number of posts that say applying a voltage to the output of a regulator with no input is bad and should be avoided, and equal number saying it is not bad, depending on certain things such as the regulator being used and/or the presence of a diode between Vin and Vout of the regulator. Many boards these days need to implement this capability of providing multiple power options and would therefore I think run into this same design requirement. With respect to this one area, is the circuit used in the UNO and/or Mega boards an example of a great, good, or not so good design, and why?

Thanks again

See page 10-11

See this topic also
http://forum.arduino.cc/index.php/topic,82046.msg758170.html

Thanks for the respose, but:

  1. I had already read the data sheet for the NCP1117 and did see the stuff you are referencing, but I am not sure the case/situation it refers to is equivalent to 5V being present constantly at the 1117’s output when there is no input to the 1117. I have an email into chip manufacture to get clarification.

  2. The thread you list is exactly why I posted this thread. In that thread you have a team member stating that they do not recommend powering the newer mega or unoR3 boards from the 5V connector due to “new way we use to bring power into the boards will not ensure the boards to work properly”. If a external regulated 5V source is applied to the 5V connector what is the risk? A 5Vdc medical grade PS from a reputable manufacture such as TDK Lambda is certainly just as safe as the onboard regulator isn’t it? From what I can tell the only issue would be some sort of issue with the regulator’s output being fed 5V without any input, but if that was the case then powering with the USB port would cause the same issue.

Please note that I am not talking about powering from both the USB and 5V connections at the same time, or trying to apply an unregulated unreliable high voltage to the 5V connection. Very simply, powering the board from a reliable external 5V PS. Am I missing something here?

Thanks

visinet:
Thanks for the respose, but:

  1. I had already read the data sheet for the NCP1117 and did see the stuff you are referencing, but I am not sure the case/situation it refers to is equivalent to 5V being present constantly at the 1117’s output when there is no input to the 1117. I have an email into chip manufacture to get clarification.

  2. The thread you list is exactly why I posted this thread. In that thread you have a team member stating that they do not recommend powering the newer mega or unoR3 boards from the 5V connector due to “new way we use to bring power into the boards will not ensure the boards to work properly”. If a external regulated 5V source is applied to the 5V connector what is the risk? A 5Vdc medical grade PS from a reputable manufacture such as TDK Lambda is certainly just as safe as the onboard regulator isn’t it? From what I can tell the only issue would be some sort of issue with the regulator’s output being fed 5V without any input, but if that was the case then powering with the USB port would cause the same issue.

Please note that I am not talking about powering from both the USB and 5V connections at the same time, or trying to apply an unregulated unreliable high voltage to the 5V connection. Very simply, powering the board from a reliable external 5V PS. Am I missing something here?

Thanks

I should have added yesterday that if you’re really concerned, add a diode from the output to the input of the regulator, and of course, as you mention, don’t connect the USB supply while you have an external 5V connected.

Now, having said that, I have always powered my UNO boards from the 5V rail, as do many other people, and I have never had a problem. Early on, I did check the 5V regulator regularly for temperature, and it didn’t get hot at any time, so I personally will continue to power my UNO boards this way.

The UNO design itself ‘back-powers’ the regulator when the USB 5V supply is connected, so I see absolutely no difference. The only additional problem when you don’t power via Vin is that you have to make sure that you don’t connect your supply in reverse-polarity, since there’s no reverse-polarity protection diode.
I was a bit iffy about it originally, but since, as I mention, the UNO design itself applies 5V to the regulator’s output, a user-supplied 5V is exactly the same.

Referring to the official UNO schematic, you can see that when nothing is connected to Vin, the USB supply connects directly to the 5V rail, and you can also see that the 5V rail connects directly to the 5V regulator’s output:-

UNO 5V power.JPG

So while it is technically incorrect according to the regulator datasheet, I decided that if it’s good enough for the UNO design, it’s good enough for me.
CrossRoads is quite right to point out that it’s not technically correct, though, and that the potential is there to damage/kill the regulator. A diode across the regulator as mentioned in the datasheet is the correct thing to do.

Thanks for the feedback and info. I guess in the end it is all about what the regulator chip manufacture says. Here is the response I got back from one of them.

My Question:

"I have a quick question regarding your Regulator line, and specifically the AMS1117 part. The question is in regards to having a voltage applied to the output of the device if there is not input voltage. Is this a problem? The application is a PCB that provides 2 power options. 1 PS input is say a 12V unregulated input that feeds the input of the AMS117 part to produce the system 5V. The other PS input accepts a regulated 5V supply and connects directly to the system 5V. When the board is powered using the 5V input, the ASM1117 is not used and does not have any input applied to it, but it's output will see the 5V from the boards regulated PS input. Is this an issue? If so, how should it be handled?"

Their response:

"AMS1117 is a series regulator. The input is connected to the collector and the output is the emitter of the power transistor. With respect to ground, there are two resistors that sets-up the output voltage. There will be some current flowing through these resistors to GND, usually 10-12mA. As long the input is floating you should not have any problem. If you ground the input then you have a reverse voltage E-C on the power transistor."

visinet:
Their response:

"AMS1117 is a series regulator. The input is connected to the collector and the output is the emitter of the power transistor. With respect to ground, there are two resistors that sets-up the output voltage. There will be some current flowing through these resistors to GND, usually 10-12mA. As long the input is floating you should not have any problem. If you ground the input then you have a reverse voltage E-C on the power transistor."

Yes, that's exactly what I said in reply #2.

OldSteve:
As I understand it, there will only be a potential problem if there's a load connected between Vin and ground, which could draw excessive current back through the regulator. As long as you avoid this situation, there's nothing to worry about.

But the datasheet does say different to what the manufacturers said in their email. I guess the datasheet is just allowing for the fact that there might be a load connected to the input. :slight_smile: