Just finished the install over the weekend! Here is the finished project:
I used the popular Orbit box, need to enlarge the holes at both the box and the panel with a step bit. 1/2" and 1" PVC conduits.
I used Leviton 5835-W ($7.11 on Amazon) and wired the two receptacles with different voltages:
14 VT/CT wires almost fully filled the 1" conduit. The last few were tricky to pull.
I need to reduce the calibration of VT by half to get the correct voltage reference:
I like it. Very resourceful locating that receptacle. I used a standard US 240V plug and changed the plug on a Euro power strip to convert to two-pin 240V. It’s a kludge. This is much cleaner. BTW/ you probably didn’t need to go with dual voltage as I believe the standard CUI USB power supply is good to 270V.
For anyone wondering about the VT calibration comment, this install uses the 230V two-pin Euro VT connected to the phase-to-phase 240V. He then calibrates it down to 120V. With this as a voltage reference, there is no variation due to neutral wire resistance, so always right for the 240V loads that usually account for the bulk of energy consumed. I run my household unit the same way.
Nice job.
Edit: I just realized that the tables don’t have 60Hz phase-shift for that VT. I’ll measure that and update in the next release due out soon.
I wired the 120V receptacle on the off chance I need to backfeed my house with a 120V only generator. Then both legs will be at the same voltage. Will power the IoTaWatt through a UPS, and probably move the emonpi Zero inside the box with a 5V splitter.
May be it’s a good idea to add a checkbox for “half” in the VT configuration page for this use case? So that we don’t need to manually change the calibration value.
Edit: Actually, I might try to open a PR…
Where would you add that? The logical place would be the CTs, and leave the voltage at 240V. Then for 120V CTs, half the voltage the way 120V VT users double for 240V CTs.
But there are very few users that do this, and my approach is to avoid adding a lot of options that complicate the user interface. So lets just let it be. The solution to change the calibration works fine.
That’s the logical approach, but then the majority of circuits need to be selected “half” which is too much of a hassle. I agree to just let it be.
I finally have the solar panels installed and hooked up my reserved CT. Using the last 6 months of data I reduced the calibration factor of the Ideal 77DE-06-09(EU) from 9.56 to 9.52 to better match the power company’s smart meter output. With the line-side taps and solar consumption CTs, the upper space in the panel is almost full:
I found large discrepancies in the consumption between IotaWatt and Enphase CTs. After some investigation I found that a magnet in a solar CT is broken. Without IotaWatt I can never discover this early:
Much better granular data than Enphase offered during this heat wave in the US:
