Phase correction and different inverters

Hello, I created my own hardware / software with the use of emonlib. The results i’m getting are very satisfying. I measure grid, solar panels, battery charging / decharging, waterheater current and redirect the ‘too much’ current with a triac to my 200l waterheater. Finding the correct phase error was an important part to get the correct values.

However I still have a problem and as a sofware developer the electricity side is less clear for me. The problem i’m facing is that with 2000watt solar power (aps microinverters), 1500watt water heater, 10A and 20a battery chargers and grid use the data is corresponding to the data I read on the display of the electricity company but: if I use the battery decharging inverter (only at night when no charging / solar is going on) ( 500W Micro Grid tie inverter for 12V 24V 36V 48V battery discharge adjustable output power solar panel grid tie inverter|Inverters & Converters| - AliExpress ) things are very ‘off’ ) The central grid clamp doesn’t measure correct and the decharging current is measured to low.

If I change the phase correction factor, i’m able to get values more in line with the values seen on the enerymonitor of my energy supplier. However, in that case, if the inverter is off during daytime, the other items (solar, charging, water heater) do not measure correctly.

Does anybody have an idea about what could be going on here?


Welcome, Henk, to the OEM forum.

Do you mean that the 230V output active power from the inverter is measured low?

It is quite possible that you have discovered that the phase error of a current transformer changes according to the value of current that you are measuring. This has no significant effect on the accuracy of the current measurement, but it will have a measurable and possibly significant effect on the value of real (active) power that you calculate, especially if the power factor is low. And the smaller the current, the greater the error.

The error is inherent in the way the strength of the magnetic field affects the losses in the core of the transformer. I posted the theoretical explanation here: Why does current lead voltage when measured by a CT? - #2 by Robert.Wall

There are two things you can do: the easy way is buy a very expensive current transformer - but this will only have a phase error that changes less. The hard way is to adjust the value of phase error correction that you apply automatically, according to the value of current that it is measuring.

You can read the results of tests on some current transformers in the ‘Learn’ section, and see how the phase error changes not only as current changes, but also between different samples of the same current transformer…

Yes, exactely.

Thanks for your answer, I will first chew a bit on it. It is all quite new for me.

The problem i’m facing is not that the active power is a bit more or less than what the electricity company is billing me. The problem is dat I would like to have the exact zero point with the grid. With zero point I mean that the active power goes to the grid (more solar or battery production than I use in the house) or that I use (buy) the power from the grid. It is working quite allright but I would like it 100%. I do not even know if that is possible. Varying the phase has a huge impact I saw, but the problem was, that the battery inverter needs(I think for the moment), a different phase setting.

That’s OK. We were all new once. We understand.

You didn’t mention which emonLib you are using. If it’s “emonLib”, then you must remember that it only measures for 100 - 200 ms (I can’t remember which now) every 10 s, so unless everything is absolutely steady, you must rely on averages for the correct reading. If you’re using emonLibCM, then this does read continuously (“Continuous Monitoring”) and it’s much more accurate when something is changing rapidly. But the OEM Shop sells the same transformers for voltage and current for use with both, and it’s those which show a varying phase error over their operating ranges.

I think your problem isn’t that the inverter needs a different phase correction setting, but that the current when the inverter is working (compared to the current supplied normally by the grid) means the current transformer has a different phase error which in turn means your monitor needs a different phase error correction to be set - ideally one whose value depends on the c.t’s phase error at the current flowing at the time. And that’s not easy.

Lots of time have passed and I finally understood the problem.

I used the zmpt101b - module. This thing adds tons of noise in the circuit and the correct phasecal was a pain in the ***. I ended with -1.8 which gave reasonable results but not good enough.

I started experimenting with the zmpt101b only (not the module) and things started to go better. Still had to learn à lot en trial and error takes time. However, noise is gone, and the measurements are finally spot on with the data from the energy company. In or out (grid solar battery), problems are gone :slight_smile: And now the incoming data makes sense, things are lots clearer.

Thanks for the reply.

You should have mentioned that you used the ZMPT101B MODULE. Unfortunately, and I’ve written this several times previously:

The ZMPT101B MODULE is totally and completely useless for measuring real power.

It’s OK to use the transformer, on its own.

It includes an op.amp and a filter circuit that introduces a large and variable phase shift, which makes any attempt to synchronise the voltage and current measurements - and you must do that to read real power correctly - almost impossible.

Had you written that you were using this module, I could have saved you two months of problems and investigation…

:smiley: advise taken (too late)! Sooo much info on the net but it’s hard to choose between good and bad info when you start as a novice. Luckily I ended up with lots of things learned and confidence to continue.

1 Like

thumbs_up thumbsup