I hoped to use one EmonPi and one EmonTx so I could read the two metering elements as well as some circuits.
Can I use a single Ac-Ac adapter to provide a Voltage reference to both units?
9V AC 1.33A Compact Power Adapter with 2.1mm plug (auselectronicsdirect.com.au)
PS. Am I right in assuming the Tx will communicate to the Pi and the Pi to my wifi router?
Welcome, Dave, to the OEM forum.
Yes, that’s how it works. The RJ45 connector on the emonTx is NOT Ethernet, they talk via a 433 MHz ISM band radio.
If you look at the circuit diagrams (via the Wiki), you’ll see that the a.c. input for the emonTx is connected with the outer of the connector to GND, as it is on the emonPi. So yes, provided that you connect the outer sleeves together and the centre conductors together, they can share an a.c. adapter and a 5 V d.c. USB supply.
This adapter has been used in the past: https://www.jaycar.com.au/9v-ac-1-amp-plugpack/p/MP3027 , the no-load voltage is 10.4 V. the nominal voltage calibration constant is 300.0
i use two of the jarcar AC/AC power supplies.
My calibration below
float vCal = 298.22; // 268.97 X (240.6 / 217) Calibration for AUS Jaycar AC-AC adapter MP3027
The nominal calibration was 0.6% out - tut, tut. We really must do better. 
I forget what the original sketch values are (or were) when I first setup my emontx ~5 years ago. But the template values meant Voltage was reported at 216V as apposed to the more accurate true voltage closer to 240V. It was a massive difference.
As a newbie it was a significant undertaking for me to understand how todo this back in the day… in Australia… using a non Open Energy Monitor supplied power supply.
Lots of OEM config is bewildering and then it starts to make sense when tinkering with it all.
They jaycar ac/ac power bricks are rock solid & stable. Recommended for anyone in aus.
A note for anyone else confused by this…
The emonTx / emonPi measures the voltage at the a.c. connector input, nominally 9 V, but as we all know, the no-load voltage on a transformer (which is what the a.c. adapter is, along with a plug, a fuse, a case and a lead) is a lot higher. All the a.c. adapters we use and recommend have different output voltages, which is why we need a different calibration constant for each. And that’s before you start taking their manufacturing tolerances into account - and those of all the other components that affect the accuracy of the measurement.
So after some trouble shooting it appears to be something todo with the antenna.
If I remove AC power adaptor from both emontx
Switch antennas (move antennas between emontx’s), then plug in the AC adaptor to each, the problem moves to the other emontx. The problem follows the “bad” antenna.
The emontx with the suspect antenna… If I apply 5V USB power and reset, emontx starts up fine.
The misbehaving antenna causes the emontx light to remain on almost constantly and it occasionally flickers.
Serial output via FTDI is no good to me to troubleshoot because that gives the emontx 5V DC power.
I have quite a bit of test gear and electronics knowledge.
Of course, if you don’t put anything in your profile, and you don’t mention it, we have no means of knowing just how much you or anyone knows, so we have to assume that it’s little or nothing until we know differently.
Now that is weird. I’d never have suspected that. The RFM69CW is not (unless you changed it) working at max power, yet the ‘bad’ antenna (does the good one still transmit to the emonPi?) must be mismatched to the extent that the power drawn by RFM is enough to hit the current limit on the a.c. supply, which causes the voltage to collapse and then the processor restarts as the current falls and the voltage recovers.
What’s happening there is, as it restarts, it goes through the setup procedure - reading setting from EEPROM, setting the I/O ports, etc, then it fires up the radio and sends a test transmission. But by the time it’s got down to 8, the extra current has drained the reservoir capacitor and the voltage collapses. We can’t take more current because it puts too big a dent in the voltage you’re trying to measure.
Definitely time for an email to The Shop: [email protected] and link to this topic.
Thanks Robert,
I greatly appreciated being able to discuss the problem and your experience valued.
If I find the solution or root cause I’ll post back here too.