Inputs with periodics zero values

Hello, After many days with correct curves with emonpi i see now strange results: AC probes send ‘’zero’’ values or no physical value: goes to zero between two measurements: see curves.

Before the issue, Emonpi works well in House “1 ,” with AC probes, and DIY serial USB arduino and 5 AC probes.

In house ‘’2’’ I reset the data feeds, I change the name of some inputs. And I Have theses results……
The EmonPi TX , and USB arduino probes works well with correct data. The AC CT2 monitor one part of CT1 current : serial circuit. Any way, the zero value is not possible. The issu come from inputs, not from feed.

The delta between two setups are : house , last software update today. And maybe AC polarity.

Any idea?

Many thanks

I am not an expert on emonCMS, I am asking these questions so that someone who knows more about it will have more information and therefore be able to give you a solution.

  1. You installed your emonPi in House No.1. You had there two current transformers and a.c. adapter. I do not understand what a “USB Arduino” is, and what are the 5 a.c. probes?
    But everything there worked as you expected.

  2. You moved everything to House No.2.
    You do not see the same as at House No.1.
    How did you reset the data feeds? Did you empty the feed, or did you delete the feed?
    How did you change the name of the inputs?

  3. At House No.2, exactly what are you measuring? Do you expect a steady value, and not a value that changes rapidly between (say) 200 W and 1000 W; or is your problem the steady value of ~200 W from 18:37 to 18:45?

  4. Do you have duplicate inputs logging to the same feed?

Hello Robert,
1: USB arduino with ac probes is a kind off directserial emontX : always ‘‘good’’ data, no missing values
2: First I empty the feed : Feeds < select fields < garbage< clear data
3 : at house No 2 i’m measuring total or part of total power cunsumption : I know that I can not have this periodic cumsumption with wide drop to zeros watts
4 : yes it is more complicated : i’m using some post processing to add some feeds thus :
5 : today I remove totaly the emonpi : power2 : I start with a ‘‘new’’ input power2 : I see the same result. Emonpi ''miss ‘’ periodialy a value. I’m not using the AC sources saled with emonPI; phase is so critical ? It could not explained a periodic error.

Sorry !

Emonpi is working perfectly. I discover the guilty : Heating radiator with electronic switch. When temperature is OK , it use a short duty cyle : about 15s on, 40s off. It explained my curves…Emonpi help me to discover the real world !

I have an answer for that problem: We are testing it at the moment, it is called emonPiCM, and it will accurately average the power over the 10 s reporting period, just like the emonTx.
Trystan Lea and I are doing the tests, but we I have not finished. I cannot tell you when it will be available.