Archived Forum

I am tempted to suggest that you should use an emonTx V2, or build your own interface for an Arduino board, for this. The difficulty is that the uncommitted digital inputs that you need for the pulse inputs aren't brought out on the emonTx V3.4 PCB. (OK, there are digital pins available, but they have on-board connections intended for other uses.)
If you do build your own front end, note that what we've heard reported indicates that the noise performance, i.e. the ability to record accurately low powers, may not be very good.

I've very recently posted a sketch for 3-phase 3-wire power measurement (using one voltage input and using the 3rd wire to replace the 'neutral' as the voltage reference for measurement purposes). However, measuring the power for 3 phases and 2 distribution boards could pose a problem with the number of voltage and current inputs required, and that too might indicate that a more powerful processor than the the 328P would be needed.

Robert,

Thanks for your feedback, much appriciated, this made me decide that I want to change the scope for now :

• Pulse count of PV power generation -> Already covered by the current arduino.
• Clamp power consumption of my 3 phase feed 
• Pulse count gas consumption (using Flukso FLS02-F1) 
• ​Pulse count water consumption (using Flukso FLS02-T1) 

I read that monitoring the 3 phases is also a bit of a challenge ? Any suggestions about that ?

I can only monitor the net consumption (meaning energy consumption - generated PV power), i presume that there is a way to add those 2 values together in emonCMS, the get what we are consuming, regardless of what the PV's are generating.

Chris

3-phase is not normal in a domestic installation in the UK, which is where the emonTx's home is. So there is no emonTx hardware specifically for a 3-phase set-up. There are three ways to go about this: one is to have 3 emonTx's monitoring one phase each, either independently or synchronised as in MartinR's "Full-fat" design, second is to synthesise the voltage wave for the 2nd & 3rd phases from the first, as the "3-phase" sketches do. That's inherently less accurate, how much depending on how accurately the phase voltages and angles are matched. The final way is to design your own, based on the emonTx (as per Robin's 3-phase diverter recently posted).

"I can only monitor the net consumption"
Does that mean you can only access the cables at the point where you only see nett import/export? If yes, then provided you can get both sets of numbers - Nett & Generation - into emonCMS, then you can do the sum Consumption = PV + Import, bearing in mind Import can go negative if you're exporting. (If 'no', then it's a different sum but the same principle.)
Or you could do the sum in the emonTx.

The 3-phase PCB that I have developed allows the flow of energy on each phase to be accurately monitored at the supply point.  By combining these values, the overall energy state of the premises can be known at all times, and any surplus energy diverted for some useful on-site purpose. 

With all six of the analogue ports in use, no more CTs can be supported.  However, several of the digital IO pins are spare, each of which could be used to monitor an optical indicator (LED) on a utility meter on a polled basis.  This information could then be transmitted by RF for datalogging purposes.

My 3-phase PCB includes all of the necessary interfacing circuits for this kind of application and can be easily connected to one or more output stages if surplus power is to be diverted.  With suitable interfacing circuits and adapters in place, the same approach could be used on any Atmega 386 platform such as an Arduino Uno.