Powering EmonPi2 with and without EmonVS in North America

Hello,
I’m a complete beginner and excited to dive into the OpenEnergyMonitor ecosystem! I’m currently setting up an emonPi2 (12-channel) for the first time in North America (110V), and I have a few questions to get me started.

  1. Using emonPi2 in Current-Only Mode:

I’m interested in using the emonPi2 in current-only mode. From what I’ve read on the project page, it mentions that:

The emonPi2 can be used without an emonVs in current-only mode. This is only supported by the 6-channel firmware.

My understanding is that I need to update the EmonPi2 to the emonPi2_CM_6CT firmware to enable this. However, I’ve plugged in the EmonPi2 via USB-C, but it doesn’t power on. How is the EmonPi2 powered without the emonVs?

  1. Integrating the EmonVS:

In the future, I plan to integrate the EmonVS. However, the EmonVS unit I received did not include a power cable. Based on my research, I believe to start monitoring power, I need to:

  1. Purchase a NEMA 5-15P plug like this one from Digikey
  2. Connect the L1, E, and N screw terminals of the emonPi2 to the 5-15P plug

Could anyone confirm if this setup is correct, and if there’s anything else I should be aware of before proceeding?

Thanks in advance for any guidance!

Welcome, Joseph, to the OEM forum.

The big question, before I go into detail, is what do you expect and want from your emonPi - in the fullness of time. Do you want the best accuracy you can have, are you prepared to cut corners in the name of cost savings, is yours a fairly typical N.America installation with a mix of 120 V (not 110 V) loads connected to alternate legs of the supply, and 240 V loads connected across the two legs?

Inherently, the emonPi2 and its software is designed to accurately measure real power - the quantity your electricity supplier’s meter measures and what you pay for. And to do this, it needs to know the current and the voltage on each leg (the black and red wires with respect to the white neutral wire in your load centre), and the relationship between the current and voltage waves.

If you do this, you won’t get accurate measurements of real power. You will only have an estimate of apparent power. For something like an electric kettle, this will be inaccurate by whatever fraction the voltage varies from your nominal 120 V. If it’s a pool pump you’re measuring, it is likely to be wrong by a much larger amount, maybe reading twice the real power (because of the poor power factor of low power induction motors).

This is because it uses a different software library that was written for a different processor with fewer analogue inputs. There’s no inherent reason why your emonPi with its expansion board cannot be used in current-only mode - provided it can ‘sniff’ a small sample ( no more than ⅓V ) of the a.c. wave to allow it to count (not measure) mains cycles, it’s only a relatively simple change in the software that’s required. In fact, there’s a demonstration ‘sketch’ published and available which reports all 12 currents, and could easily report 12 apparent powers with 12 small changes.

Via the USB-C connector! I don’t know about the emonPi2 specifically, but the emonTx4 which uses essentially the same design of circuit board did not have a reversible USB-C - it is sensitive to orientation. So I suggest you turn the connector over and try again.

Did you get a ‘single phase (UK)’ emonVs or a 3-phase one? If you bought the 3-phase one, you can connect your A & B legs (black and red wires) to L1 & L2 and the white neutral to N and you’ll be able to get an accurate voltage for the loads on each leg, and the total voltage for the 240 V loads. This is the most accurate configuration. You’ll need to use a NEMA 14-15R 4-pole connector: 2 legs, neutral and protective earth (ground to you).
If you have the single-phase emonVs, then your 5-15P connector will be appropriate, connect green to E, white to N and black to L1. This loses some accuracy because you must assume the voltage on the unmeasured leg of your supply is the same as the one the emonVs is connected to.

Whichever emonVs you have (or none), there is sadly no pre-loaded software suitable for use on the N.American split-phase domestic supply, and you will need to make some fairly simple changes, which will include doubling the power reported to get the correct answer for your 240 V loads. Unfortunately, getting the changed software into the processor is a little more involved.

You might like to check this Topic: Only emonPi2 CT inputs 1 and 4 work - @haversian faced much the same problem as you have.

1 Like

Looking at the emonPi2 schematics, VBUS1 is connected; VBUS2 is not; neither are the CC pins. So yes, it will be rotation-sensitive, and won’t work with a bunch of pickier adapters. Proper USB-PD support is a bit of a pain in the ass, so I understand not bothering…I just wish they’d done it despite the difficulty:)

Also, the schematic says C14 is only a 6V-rated cap which is no good for a 5V rail. I’m sure it works: that LDO probably isn’t very sensitive to the value of the input cap, so the fact that DC bias will reduce the effective capacitance isn’t likely a big problem; and these will generally be used at moderate temperatures so the lifetime hit from using it so close to its rated voltage is probably okay too. But not an ideal component choice: a 10V or 16V cap would have been better.

Griping aside… @jminnich, I’ll probably get around to messing with Arduino development in a couple weeks and will try to take some notes on anything non-obvious, to help out other split-phase users.

Bear in mind @haversian and @jminnich that there’s a page in Docs about using the emon* in N. America. Use in North America — OpenEnergyMonitor 0.0.1 documentation
You might like to read the first two sections that will help to explain why we do things the way we do, then skip to:
Use in North America — OpenEnergyMonitor 0.0.1 documentation
which also (though it needs updating to say so) covers the use of the emonPi2 too.