Monitoring 5 appliances

Hello, first post here. I want to monitor (log, chart etc.) the power consumption of 5 appliances (indoor A/C condensers). There are power sockets and network ports close by to all of them
I am hoping to use an emonPi and an emonTx. I would connect 4x clip-on current sensors to the Tx, and 1x clip-on current sensor to the emonPi, to allow me to monitor all 5.

Two questions:

  1. Is there a way that the emonTx can connect to the emonPi “wired” - since they will be sitting next to each other I figure this would save me messing around with wireless reliability

  2. Is this the right approach for a 5-appliance setup, or should I be looking at 2 emonTx’s?

Thank you!

  1. Yes there is:
    The wired connection with a link to what you need to do in emonHub (in the emonPi) is here:
    EmonTx V3.4 - OpenEnergyMonitor Wiki
    [Edit] The link above is for the emonESP. I think the correct information for a direct wired link is this: emonhub/conf/interfacer_examples/directserial at emon-pi · openenergymonitor/emonhub · GitHub

  2. Probably yes. You’ll have 6 current inputs available in total, plus local logging via emonCMS. If you want more, then another emonTx is probably needed, but presumably you will still want the emonPi for logging? So for now, 2 x emonTx + emonPi seems to be overkill.
    N.B. The 4th input of the emonTx is high sensitivity. You can change that by adding a resistor inside - there are holes ready for a wire-ended resistor - and bring it down to the same as the others if needs be.

Are all the AC units close together? Just checking you are aware that you need a ‘bare wire’ for the CT to work (i.e. the live or the neutral) or are you planning on connecting the CTs back at the distribution board (i.e. each AC is on its own circuit)?

Thanks both! Look forward to tinkering with it
Brian - yes, the units are close together and each one is plugged into a normal 13A socket. I was planning to open up a few inches of mains flex just near the plug and get the clamp around the live, if that’s not wildly bad practice…

Unfortunately yes it is bad practice. There a difference between the insulation on the conductors, which is electrically good but mechanically poor, and the outer sheath which is mechanically tough and is there to protect the insulation from damage. Doing as you suggest is OK for a test under supervision, it’s not OK as a permanent installation. So what I’d do is open up the A/C unit and put the c.t. inside, where the line conductor will be separate anyway, and everything is nicely protected.
If you can’t or aren’t allowed to do that, then one way would be to put the stripped cable with the c.t. (maybe all 5 together) inside a box, to give some protection.

Ah, gotcha. I think I need to do some more thinking before diving in! Thank you for the guidance

I did wonder if that was your thinking, as Robert says, not the best. You also need to consider the distance from the furthest AC unit to the EmonTX.

There are some Sonoff devices available that will measure power, unfortunately none are plug in. I’ve used them by either wiring into the actual cable from the respective unit or by wiring the Sonoff units into a short extension lead. It really depends on what the environment is you are using them in. If a commercial or public place you do need to consider the electrical safety issues (employee or member of public)

If you do feel you can go down the Sonoff route, and do not want to use the built-in mechanisms, there is an alternative firmware available called Tasmota you might want to investigate (search on here for related posts). You then could do it with just an EmonPi, or an EmonBase.or even just an EmonSD & a Pi as all the data would be transmitted via MQTT over WiFi. This method isn’t super accurate, especially at any instant, but it may be good enough depending on what you are trying to achieve.

Using the term bare wire might not be a good idea.

To some, that connotes an exposed conductor.

1 Like

If the condensers (air handlers) are each on a separate circuit (which would generally be required in the USA, but might be different where you are) you could use an Iotawatt to measure the power on each circuit and have 9 inputs left to measure other circuits.

The Sonoff Pow2 could also work. Just put it between the condenser and its plug, will need to cut the wire to do so. I use Tasmota on all my esp8266 based devices and it works well.

In the US, there is the S31 which does not require cutting cords/wires to use. I don’t see one for the sockets common there.

Depending on the equipment, the condensers and air handlers might not be co-located,
or may not even be near each other. In all of the buildings I’ve worked in, the condensers were
outdoors, and the air handlers (essentially large fans and and associated ductwork) were indoors
and therefore on different branch circuits. (both chilled water and non chilled-water type systems)

Of course, like everything else, there are exceptions… :wink:

When you read / if you read the introduction to “Use in North America” (Learn→Electricity Monitoring→AC Power Theory→Use in North America→Use in North America), you’ll see that the power distribution systems in our two countries have significant differences, so what might be normal and acceptable practice for you could well be unacceptable here.

I answered the question based on the facts as they were presented. I cannot speculate on what might or might not be the facts of the installation when no data is available other than that which is given.

Thanks all!
I guess my AC installation is the exception, so I was deliberately a bit vague. The condensers are part of a water-cooled system, so they are indoors (in a private plant room to which only I have access). Each condenser is plugged into a standard 13A socket on a mains ring. They are not particularly power hungry. I think I will end up going with Robert’s suggestion of fitting the stripped wires (for all 5 units) in a box. Besides it’ll look a lot better than a bunch of mains flex with insulation removed…


They are all on the same phase? Because while there’s a sketch for the emonTx to use 3 phases, the emonPi is strictly single-phase.

Yep, all same phase thankfully