Powering Uno from regulated +5VDC Safe/Possible?

Have a Uno board right now that I am playing with. I have a nice BK Precision power supply on my bench that outputs a very stable and clean +5VDC to my breadboard setup, to which my Uno is velcro’ed to. I would much prefer to power my Uno from the +5VDC supply instead of separate wall-wart or 9V battery for simplicity of wiring and convenience. The other two output from my power supply are being used to give +15VDC and -15VDC to some op-amps used on the same project. From what I have gleaned from the web, this is very possible. First, am I correct in this regards. Secondly, do I just plunge right into the +5vDC bus on the board, or somewhere else specific. Lastly, what pit-falls do I need to be aware of if I go down this route. One thing that pops to mind is to make sure I never have the USB hooked up and my +5VCD power supply attached at the same time. Anything else?

Yes, I know that the Uno board might not be my best option. I'm still working through the hardware and software requirements for my project, and the Uno offers me all options until I figure out exactly what I don't need. Thanks!

Send anywhere from 7-12V into the socket marked "Vin", and you should be good to go.

You should be good with just connecting (very carefully), 5VDC into the 5V pin. and Ground into the GND pin.

Pitfalls; there aren't too many -don't connect into 3V3, or in reverse or bad things may happen :P -if you you get any kind of voltage spike then it might damage the Arduino

Supplying a regulated +5vdc to the arduino's 5v shield connector pin to power the board is something that many do without reporting problems.

However there remains the issue, if you do it that way, that when you then plug the board into your PC's USB port to upload sketches, monitor serial data, etc, you are effectively hardwiring your external +5vdc source to your PC's +5vdc source. This is not considered a good engineering practice, tying two voltage sources together. However as I said many do it that way without having any problems.

Lefty

Couldn't you hack into the 5v pin of the USB connector on the Uno and send you 5v from the benchtop supply there? Even better would be to cut the end off a USB cable and figure out which wires are the 5v and ground inside the cable and connect them to your power supply.

This is issue has been a real problem for me. Both the UNO and Duemilanove boards were laid out so cutting the one trace at the USB connector in a way that can be reconnected is almost impossible. It seems almost intentional. A cut-the-trace feature right behind the USB connector would have been easy. It is so difficult that I wonder if there is some other issue the designer knows about and I cannot see.

Moderator! why no really simple way to isolate Vcc from the USB? For either board it is simple to isolate the RESET-EN signal, why not the power?.

JC

Moderator! why no really simple way to isolate Vcc from the USB? For both of them it easy for the RESET-EN signal.

I don't think the moderator designed the arduio auto voltage switching circuit or did the board layout. LOL

I don't own any 'official' arduino boards, so I don't have a dog in the fight so to speak. I have a RS-232 arduino compatible board that of course only uses external power via a on-board +5vdc regulator. My other board is a Seeeduino mega board which uses a manual 2 position slide switch to select USB or external power. Interestingly the Seeeduino's FTDI chip is hardwired to the USB +5vdc which saves a little current if you aren't powering the board via USB, cleaver little detail.

I have felt sense the released the auto-voltage selector circuit that it was a overly complex, added too many components, and limited what one could do. They could have added a RTC chip in the same space taken. I much prefer their prior Diecimila ( and older) method of just using a 3 pin male header where you placed a 2 pin female shorting clip to select which voltage source you wanted to use, or no jumper and then power the board from an external regulated +5vdc voltage source via the +5vdc shield connector pin. But I guess that was to much learning curve for the Arduino target market to handle? :D

Lefty

One could always remove Fuse F1 and eliminate the 5v path back into the USB connector, no?

There certainly are work arounds, removing the fuse is the slam dunk answer to the problem. Even if that was a foreseen solution, it would have been nice to have a spot to tag it back down to the board so it was available should someone need to reinstall it. If I take it off it is almost certain to be instantly lost. It seems a bit uneven that the option of disconnecting the reset was important enough to facilitate, but not the power source. The pins and jumper were good. It was even simple to bend them around and absolutely secure the choice with a bit of solder.

You are correct, the moderator did not lay out the board, but occasionally there are forum responses from names that are a lot closer to the source than I am. Maybe someone will notice. Our firm actually consumes a number of the things into data gathering applications. We use them because they are cheap, available, well supported, quick to configure and compared to other platforms, fun to work with. Accommodating our need for the trace cut option would help make them more attractive to other commercial consumers as well. Thus increasing consumption and helping to hold down costs for everyone.

See my reply here.

http://arduino.cc/forum/index.php/topic,54334.0.html

Never had a problem, and I’ve used lots of Arduinos, every version right back to NG.

"You can just plug in you +5V pwer device to Vin" NO: you bring your exernal +5 to the 5V pin or to the VCC pin (on promini for example), bypassing the regulator.

"One thing that pops to mind is to make sure I never have the USB hooked up and my +5VCD power supply attached at the same time. " That may be correct - the external power will pass back into the PC - I have found I need to disconnect while downloading a program for example

You see that pin on the power connector called 5V? That's where you plug in the external 5V & bypass the regulator.

Magician, you are talking about bringing in Vin after the reverse protection diode , but still going thru the regulator. I am saying use the 5V pin instead which is on the other side of the regulator. The regulator will not be bothered by 5V on its output from an external source when no Vin is applied. The promini uses the same regulator, and I have 14 scoring machines at my fencing club powered that way, the first 6 built have been running like that since September 2010, the balance since Dec 2010 & Jan 2011. No problems.

Others have reported power feeding back into their USB port with 5v applied this way, I posted a schematic elsewhere showing a path thru the regulator where that could occur. Hence the observation to disconnect external power when downloading & powering via USB.

So for the non-technical folks reading this, bring external 5V in thru the 5v pin, and do not make it go thru the regulator.

Plug +5 in to the 5V pin and Ground into the GND pin. Not the Vin pin. Ignore the regulator and diode etc - just do as instructed and it will work fine :D

That's the answer you require. It is safe if you do it like that and it will work perfectly.

Whilst we're on the subject of bypassing the regulator: I'm running an Uno board right now with a 3.0V supply connected to the 5V pin. Technically the ATmega328 chip requires at least 3.8V to run at 16MHz, but mine seems happy enough. As others have mentioned above, it is essential to disconnect the external power before connecting to USB.

Technically the ATmega328 chip requires at least 3.8V to run at 16MHz, but mine seems happy enough.

I think it depends on the version of the chip. Also, some functions may not work properly but you probably aren't using those functions ;)

I too have an external BK supply I use as well as the 7805-based 5V supply from Solarbotics that plugs into both sides of the breadboard rails. I've been running my projects from all three sources at various times with the ground planes tied together and run into just one issue: when running on the external power supply plugged into the Vin pin (which seems like the logical place to put that considering it's labelled as such), there's a significant fluctuation in the reference voltage the ADC sees. Code that works when the power supply is on the 5V pin or powered from the USB cable would read erroneous values when running from the Vin pin.

In fact, I'd posted a new topic on this exact thing having pulled most of my hair out over the last couple of days trying to sort this. After reading this read and seeing the comments to use the 5V pin and NOT the Vin pin I tested it out and all the problems disappeared. I promptly deleted the topic and posted here.

My question is 2-fold: Why not use what appears to be labelled and referenced as the external voltage supply input pin; why does the ADC voltage range change when using the Vin pin?

Well, the documentation does say that Vin needs to be a higher voltage:

"Input Voltage (recommended) 7-12V Input Voltage (limits) 6-20V "

So you as a user bear some responsibility to read the documents and supply the correct power voltage.

The ADC needs sufficient voltage to work with. Having the 5V coming in and going thru a 5V regulator doesn't give the regultor enough to work with, so the ADC performance suffers as a result. I'm thinking the default setting is to use the external ARef, which would be connected to 5V. If you had used the internal 1.1V reference, you may have gotten better readings since the chip

My question is 2-fold: Why not use what appears to be labelled and referenced as the external voltage supply input pin; why does the ADC voltage range change when using the Vin pin?

Because the default voltage used by the D/A conversion is Vcc wired to the Avcc pin of the chip. If you carefully measure your +5vdc pin while on just USB power and then on just Vin power you will probably see that there can be several tenths of a volt difference between these two voltage sources. That will change your top of range for analogRead() conversions by the same amount.

Lefty

CrossRoads: So you as a user bear some responsibility to read the documents and supply the correct power voltage.

Indeed I do - lesson learned, hazards of being new to it all.

Lefty - I did measure with a Fluke 87 for just that reason. I measured a difference of 3 millivots but the sensor outputs jumped on the range of 167mv. It would appear there's also some bugs in my code and the apparently correct readings were fluke.