EmonTX-Shield with AC Adaptor - Readings divert to 0

Hey Guys,

I’ve been messing around with my Emon TX Shield and I am out of ideas what’s wrong.
The voltage readings of my AC/AC PSU divert to 0 after a few readings, even though I calibrated everything according to the manual.
My gear:

  • No-Name AC-AC Adapter with 10.5Vrms (Says its 9V) and 600mA. A nice sine wave according to my oszilloscope.
  • Emon TX V2.5, Arduino Uno over Serial for debugging, powered over the AC/AC-Adapter with a diode and a 4700uF Capacitor to smooth out the sinewaves.

What I did:

  • Measured the resistors of the TX Shield for the Probes, they gave readings of 25 Ohms ± 0,1 (But they say 330 on them… strange?!). Calculated calibration factor (100A/0,05A / 25 Ohms); shoud be 78.43.
  • Measured Mains with two Multimeters: 229,5V (totally ok for Germany, we should have 230V)
  • Loaded up the Sketch without the AC/AC Plug use, took a 2000W heater and adjusted the Calibration factor until Serial output said, that the heater had the same reading as the wallplug-powermeter. → Real calibration factor is 69.230 (no wonder, the resistors are 25 Ohms instead of 33)
  • Loaded up the Sketch with the AC/AC Plug-feature and went on with the EU-Plug startpoint of 260.0 for voltage calibration.
    However, the values decrease to 0V after the first reading and stay there. If I leave out the real power calculations, the values given for my 3 CT probes make somewhat sense.

What happened here? and what can I do to get actual readings?
I tried punching in wild numbers for voltage calibration, all divert to 0. I also looked again at the AC/AC adaptor with my oszi. All fine.
Redid all soldering points for the AC/AC adaptor connection.
And everything is working fine with the no-voltage sketch…

Serial output: My calibration:
emonTX Shield CT123 Voltage example
OpenEnergyMonitor.org
Node: 6 Freq: 433Mhz Network: 210
11 5 23 222.96
0 0 0 1.82
0 0 0 0.04

Complete Github Stock Sketch:
emonTX Shield CT123 Voltage example
OpenEnergyMonitor.org
Node: 6 Freq: 433Mhz Network: 210
14 5 21 222.00
0 0 0 1.83
0 0 0 0.04 (and so on for hours)

No real power calculation calibrated sketch:
emonTX Shield CT123 example
OpenEnergyMonitor.org
Node: 6 Freq: 433Mhz Network: 210
43 46 373
36 31 379 (and so on for days)

Was that with the c.t connected, or not?

What d.c. voltage do you measure at the voltage input? It should be ~2.5 V.

If you remove that circuit and use the normal Arduino power supply (7 - 12 V?), does the voltage reading still fall to zero?

ok this is strange.
no matter what I do, as soon as I power the Arduino from the AC/AC the measurements go nuts. Either they go to 60V or to 0V depening which phase I connect (so twist wires).
I also build a new PSU, which uses a L7805CV, 2 electrolyte capacitors and 2 ceramic capacitors as well as 2 shottky diodes to prevent current flowing anywhere where I don’t want it.

I also powered the RaspberryPi which I connect to USB for serial readout through a powerbank, so there is no other way the current can flow. I suspected the RPi making a second unwanted sink over the grounded usb wire. The only 2 physical connections the arduino/RPi setup now has are plugs from the AC/AC adapter. The CT’s are just clamp on and the pi is connected via wifi and powererd via powerbank… STRANGE!

If I measure the voltage on VCC and GND it says a steady 5.04V, I also looked at the voltage with my oscilloscope: steady (without the extra 470 uF capacitor I added on the VIN I saw a small sawtooth, the VCC had steady 5V also without the capacitor)

Since I was tired of messing around with saving another 5$ on a Wallplug I installed one and since than it’s running fine. I am now suppling 5V through the screwterminals on the emontx-shield (GND and VCC).

Anyway, I measured the resistance again without the CT’s plugged in. Now it’s 33,3 Ohms. Fine :smiley:
On which pins should the voltage be ~2.5V?

And just out of curiosity: what could have gone wrong?
The AC AC adapter has 800mA, the power is stable in VIN and VCC, I added diodes and ceramic capacitors to get rid of whatever wavy form my oscilloscope could have missed and I also insulated everything from false current sinks.

Clearly, you are doing something wrong, but I cannot tell what that might be from your description. Can you supply a circuit diagram of the arrangement that caused the incorrect readings?