OpenEnergyMonitor Community

Calculating the actual use in W out of pulse count


First of all, thank you for this wonderful device and excuse me. I’m new in the emon topic.
I’d like to know if it is possible, to calculate the actual use in W out of the time difference of two pulse counts and how it is programmed as a emocms process. I’m grateful for any advice!

Thank you very much! Markus

Welcome, Markus, to the OEM forum.

In general terms - it is not possible to do it accurately over a short time period - because a pulse is history, it only happens when a specified quantum of energy has passed through the meter. Power comes from the pulse rate, so at best, you only know the average rate (the power) between the intervals when the pulse count is reported. For that reason, we always prefer to measure current and calculate power directly, then you can integrate power over time and compare that with the pulse count and calibrate the power accurately. Over a long period, an hour or a day, then the problem is much easier - but you still have only the average power over the period.

Yes, of course! Your’re absolutely right. So I need 3 current sensors for my 3 conductors. My device only has 2 inputs for 2 sensors. How do I extend them, that I can attach three of them?

What is “My device”?

Oh yes! Sorry and thank you for your patience!
I have the emonPi energy monitor: emonPi Energy Monitor - Shop | OpenEnergyMonitor

Unfortunately, you cannot add an extra channel to the emonPi, nor can you measure power on two phases using the emonPi (it cannot run the 3-phase sketch - it is designed specifically for a single-phase system).

You can add an emonTx running the 3-phase sketch - that will give you a measure of the power and energy on each of your three phases, which the emonPi will record.

If you are measuring “whole house” energy, and if the pulse rate is reasonably fast, then the errors you get from differentiating the pulse count will be less. But when the pulse rate is low (say overnight), then you will see the power recorded in bursts and steps, because the pulses are discrete quantities of energy, not a continuously varying quantity.

If you look at the processes available on the Inputs page of emonCMS (“spanner” icon), there is a “kWh to Power” process, but because it runs whenever a pulse count comes in - normally every 5 s - it will not give reliable power values unless a significant number of pulses arrive in that 5 s interval.

Say you have 1 pulse ≡ 1 Wh, and 1 pulse every 5 s - that’s a load of 720 W. If you have a load of 700 W, you will have a pulse every 5.14 s: in each 5 s interval, you will get either 1 pulse, which will show as 720 W; or sometimes no pulse will arrive, which will show as zero. But if your load is 750 W, you will have a pulse every 4.8 s. You will see 1 pulse most of the time, showing 720 W, and then when two pulses fall inside the 5 s window, it will show as 1440 W. But if the pulse rate varies each side of 1 pulse per second, your power readings might step between 2.88 kW, 3.6 kW & 4.32 kW - still a big variation but less in percentages.

Thank you so, so much for your help. It is even for a newbie like me understandable. I think I have to go for a emonTx for measuring “wohle house”.

A load of about 88 W (yellow trace), 1 pulse per kWh. It shows the “power” from the pulse (blue trace, right axis) as either zero, 720 W or 900 W.

Data recorded every 10 s.

I get much the same with a 2.2 kW heating load - there are just more pulses in the 10 s window: