I have a situation where I can measure 2.2V where I would expect to see 0V and it took me some tracking down where it comes from. In retrospect, it's pretty obvious and I suspect many people more knowledgeable than me would say straight way...
So, given this schematic (note VDD pin 1 on the MCP2221A is not the same net as +Vdd):
and this section of the PCB:
Then, given JP1 2 & 3 are jumpered together, JP2 1 & 2 are jumpered together and +Vdd has no external power applied to it, why can I measure +2.2V across +Vdd and -Vdd (measured with the right polarity, i.e. -VDD connected to COM on the DMM)?
It disappears if I remove the jumper from JP2 which might imply some leakage from U4 but it has no connection to +Vdd; nor does J5. I've checked JP1 and there is no continuity between any of the pins. The only power being supplied is +5V from the USB lead plugged into J5.
It's not actually causing a problem when +Vdd is provided with its own +5V supply, and I only noticed this because there is enough juice in 2.2V to light the power LED on the Arduino that is connected across +Vdd and -Vdd and connected through fly leads to the sockets on J13.
Turns out that the TX pin (RX on the Arduino) provides 2.2V of power which and it's finding it's way out of the Arduino's +5V port to +Vdd! Pulling TX, the voltage disappears. I would never have thought that power on the TX pin would route through the ATMega to AVCC/VCC and I'm not even sure it's obvious in retrospect but there you go. I thought it useful enough to post up for others new-ish to Arduinos.
Still a lot to learn!!