I want to link a Pro Mini Arduino to a computer using an XBee transmission and a XBee to USB adapter on the computer side.
I have checked the XBee transmitters using a XBee shield and a Uno board. They function correctly together. I have also checked that the code loaded on the Pro Mini transmits correctly using an FTDI adapter and Arduino IDE's serial monitor.
However when connected to the Pro Mini, the XBee emitter doesn't work. The connection is made this way :
Pro Mini XBee
TX <-> DIN
RX <-> DOUT
(there is a switch between the XBee and the Pro Mini to disconnect the XBee transmitter when using the FTDI).
The circuit is fed by a 3.7 battery over a Pololu 5V step-up regulator as I need 5V to feed a series of LEDs. The XBee gets its 3.3V from a LF33 regulator which can provide much more than the current demanded by the XBee transmitter (I have checked that the XBee gets its 3.3V).
Could it be a power cleanliness issue ? Something else ?
PaulS:
Does the actual baud rate that the underclocked Mini is running at match what the XBee expects?
I've never thought a 3.3 V/8 MHz Pro Mini could be considered as an undeclocked version of the 5V/16 MHz version. Does it imply that a direct match between the clock rate on the Mini and the rate on the XBee is the issue and that the rate in my code should be adapted (doubled ?) from XBee's rate ? I thought Arduino's IDE would deal with such issues according to the board chosen before uploading the program.
PaulS:
Does the actual baud rate that the underclocked Mini is running at match what the XBee expects?
I've never thought a 3.3 V/8 MHz Pro Mini could be considered as an undeclocked version of the 5V/16 MHz version. Does it imply that a direct match between the clock rate on the Mini and the rate on the XBee is the issue and that the rate in my code should be adapted (doubled ?) from XBee's rate ? I thought Arduino's IDE would deal with such issues according to the board chosen before uploading the program.
I have taken into account this possibility of a mismatch between clock rates and modified the code according to it but it hasn't changed anything, so the issue remains. Any idea ?
The Arduino is a 3.3V version so no need for a level shifter I suppose.
The baud rate is 19200. There's obviously no connection. The RISS LED on the USB board is not lit.
The code is fine, as it works flawlessly if I switch off the connection between the Pro Mini and the XBee and instead plug it to the computer over a FTDI adapter. So yes, I suspect an electrical problem.
Here's the complete schematics. Please note the XBee is not powered by the Arduino. Please note also that the IR LEDs are not fully powered to 1W, as I have noticed a current of 100 mA makes them bright enough for my project.
I have also tested the Pro Mini connected to the XBee only (and powering it) on a breadboard and the communication with the other Xbee module was established directly.
I have also tried to bypass the switch on the lines between the Pro Mini and the XBee module but it solved nothing.
I really have no clue where to search from here. =(
[quote author=Jack Christensen link=topic=270731.msg1915984#msg1915984 date=1412968918]
LF33 datasheet recommends 0.1µF on the input and 2.2µF on the output.
Ah! That's an information! Thanks a lot, Jack!
[/quote]
Don't know if it's likely to be the problem, but it is definitely something I'd do.
Actually only the 2.2 µF capacitor seems to be required. And indeed by replacing the ceramic 100 nF by this one everything is running correctly. It was a bit hard to find the room for the capacitor however.
