I have an Uno R2, Ethernet shield, and RF 433 MHz receiver module attached. When I connect via USB, the sketch launches perfectly every time. When I power via a 12v 0.3A wall wart, the Ethernet lights flash at around a 2-3 Hz rate and the sketch does not run. If I press the reset button, the Uno starts working. Does anyone have any idea why this would happen? I have a USB power meter that indicates that the setup draws 0.19A so I am fairly confident that the power supply is sufficient. I understand that using 12V will not be as efficient as using a 5v regulated input so the 0.2mA will certainly rise.
My primary concern is that this will be installed by my structured media/home automation and I expect it to run autonomously. I am not going to accept needing to reset the Uno every time the power fluctuates. I will have the Uno behind a PC UPS, so it should not go down frequently, but reliability is important to me. Should I upgrade the power supply to a 7v 1A wall wart or is that not likely the issue? If there are details I left out that are important, I can supply them.
I went back to check the current draw and found that the scale was actually A not mA on the USB meter (updated the original post). I also found that when I plugged it in to the USB port for this test, the setup flashed lights instead of loading as it did with the wall power. So, this doesn't appear to be a power supply method issue as I originally thought.
For more details on what the lights are doing, I could post a video but I'm not sure where to store it. This is what the lights are doing:
the light for power stays solidly illuminated
the lights for RX, COLL, FULLD flash at around a 1 Hz speed
the light for 100M comes on with the other 3 but flashes 2x before going off with the others
If I press reset, the onboard LED (pin 13) flashes a few times and everything starts working as normal. I don't see any anomalies with power (read by the USB reader) during any of this.
Everything is still pointing towards a power issue.
A power supply of 7V is too low. Do you have a (regulated) power supply of 7.5 1A or 9V 1A for the power jack ? I use a switching wall-wart with has a voltage selector to start with. For a final project I use a DC-DC converter with 5V output that I connect to the 5V pin of the Arduino (that is not recommended, and a protection diode from 5V pin to VIN pin is needed, and current does flow back into the computer sometimes).
The USB current of 0.19A seems low. I think you need at least 300mA for an Ethernet Shield with the W5100 chip. However, a 12V 300mA power supply will cause the 5V voltage regulator (on the Arduino Uno board) to get very hot and shutting down.
I didn't get a chance to measure voltage but that sounds like a good test.
If I get a PSU, I have a bit of an issue that is difficult to resolve. I run X10 automation equipment in my home so I have to be selective about the type of electrical equipment I install. If the component creates noise on the power line, I have to use a filter to keep the system working reliably. But, those filters cost $30 retail. I found many domestic and international sources for Arduino compatible PSU's, but none of them provide a schematic (which doesn't surprise me at all). This means that I could buy one for $2 and spend another $30 to make it work after waiting a month for the boat to arrive. The issue is with the switching type units (110/220v) when they have either a capacitor across the inputs to filter or choke something (not an EE so I don't know) or when the frequency of the components runs at ~120kHz.
Does anyone know of a source for an inductive type PSU or a X10 compatible switching PSU that works with Arduino? The ones I have lying around from old cell phones are the wrong voltage/current/plug type. The only one that does work is already installed on another Arduino controlling my attic fan.
I can't be the only one on the forum who uses X10, so this seems like something that may have been covered but I can't find it with a search.
I know this is an old thread, but I have been looking for a source for a 9V 1A PSU and have not located one. The closest I have found is from AllElectronics:
This is a linear (non-switching), 12VDC 1.25A output, 2.1mm power supply. It appears to be a good choice, but I'm not sure if there's a better fit elsewhere. The price is ok ($7.50ea) but I believe shipping is a fixed $8 or so. I'd probably order at least 2 since I just burned up my second Arduino's PSU (it's making too much noise for X10, so I'd consider it dead).
If there's another option that I should consider, I'd love to hear it. I looked on ebay (only switched mode), Mouser (nothing), DigiKey (nothing), google (hard to search), Amazon (looked like switched mode primarily)...
I swapped PSU's with the one on my attic fan controller.
I found that my other installed Arduino (attic fan controller) had an issue with the power supply making noise on the power line which interfered with my X10 equipment. I don't know if this was due to the environment the PSU was in (humidity, temperature, etc) or if it just decided to cause issues. I swapped that PSU with the one I mentioned in this thread and things seemed to be working better. The attic setup didn't need as much power and was being supplied by a better quality/more powerful PSU. Also, the equipment in my basement where the RF receiving Arduino is located is all behind noise filters.
I added a delay in the setup() loop
I found a number of threads indicating that the early W5100 based Ethernet boards had an issue with starting up prior to being ready. The described behavior appeared identical to what I'm seeing. So, I added a 250ms delay at the head of the setup() loop to give the board a chance to fire up prior to initializing it with the Arduino.
This seemed to work well for the last few weeks. This morning I noticed that the Ethernet board was flashing all it's lights again. So, I think I might have figured out the issue but would like some suggestions. I turned off the power strip with this PSU the other day and then later on turned it back on and apparently the Arduino didn't power up correctly. Is it possible that the PSU could provide bad power when it is initially plugged in? If so, then pulling the plug on the Arduino end would be acceptable but pulling the wall wart from the outlet would be bad. There would be little that could be done during a power failure if that were true.