Enhancing setup

[element2202 (Elmnt22)]

Hi Robert,

Thank you for your responses.

@Robert.Wall

These are really one question, and I might have a better idea. The 3-phase sketch depends on the assumption that the voltages of phases 2 & 3 track the voltage of phase 1 - the one with the a.c. adapter on it. Therein lies a source of inaccuracy. If you want accuracy, then I’d suggest using one emonTx per phase, each with its own a.c. adapter measuring the voltage (and powering it - unless you’re using the ESP8266, when you also need a USB power supply). Each emonTx would then revert to using the standard single-phase sketch, and you’d combine the powers/energies in emonCMS.

Whether you can do this with 3 emonTx’s and measure the grid connection depends on which phases the single-phase boards are on - they need to be on different phases if you want to measure the grid as well, so on 2 emonTx’s you have Grid, 2 × 3-phase & 1 × 1-phase, and on the 3rd you have Grid, 2 × 3-phase.

I really like the sound of this idea. The single phase boards are on different phases, so it would look something like this:

Garage DB - 3-phase
Basement DB - 3-phase
FF DB - single-phase (L1)
Office DB - single-phase (L2)

emonTx 1:
AC adapter on L1 for voltage reference
Garage DB (L1)
Basement DB (L1)
FF DB
optional: desensitise 4th input for grid L1

emonTx 2:
AC adapter on L2 for voltage reference
Garage DB (L2)
Basement DB (L2)
Office DB
optional: desensitise 4th input for grid L2

emonTx 3:
AC adapter on L3 for voltage reference
Garage DB (L3)
Basement DB (L3)
optional: use 3rd input, or desensitise 4th input, for grid L3

Have I understood correctly?

@Robert.Wall

You can “desensitise” the 4th input of the emonTx so that it will go above 16 A, by adding a second resistor in parallel with the existing c.t. burden. 27 Ω will make it the same as the other 3 inputs - 100 A max

Could you expand on this further please as I think I’d like to capture the grid supply too – it seems like a waste not to. I found your response to someone in 2018, below, is this still current/correct?

I can’t seem to find a 27 Ω / 0.5% / 0.1W - do you have any ideas? From RS I can get a 27 Ω ±1% / 0.25W (here) but I’m happy to shop around if it exists.

Would it make sense to modify all three emonTx units and put the grid on input 4 on each, so that each phase of the grid supply is on a “modified” input… or doesn’t it really make any difference?

What is the reference to the calibration constant in the sketch when using this fourth input?

@Robert.Wall

The change to input 4 of the emonTx is the burden resistor - this determines the voltage that the c.t’s secondary current generates. It’s the resistor immediately behind the jack socket, and in the case of input 4, it is 120 Ω (22 Ω for the others). There are two holes either end of it for a wire-ended component, so you need to add a parallel resistor to bring the total resistance down to 22 Ω - 27Ω is very close. A 0.5%, 0.1 W would be my preference, but you might need to settle for 1%, and 0.25 or 0.33 W won’t hurt provided the body is short enough.
You’ll also need to change the calibration constant in the sketch, so if you don’t have one, a FTDI programmer needs to go on your shopping list.

Thank you for your help here and on my previous posts – you are a real asset to this forum!

Ben