Odd power sampling on power-on?

[CL23]

Hi:

I noticed the behavior shown in the figure below during several power-off-on cycles. We were making changes to the installation and had to remove power from the Iotawatt a few times. Although the monitored power remains mostly constant, something seems odd on he first sample after power-on. As can be seen, the power is sampled too low in relation to the actual power. Could it be that sampling begins before signals are steady?

Thanks a lot,

 Carlos

[overeasy]

@CL23

Could it be that sampling begins before signals are steady?

I think it has to do with steady state, but not the electrical signals, they should be OK instantaneously. My guess is that it has to do with averaging over the interval you are plotting. I’ll look into it, but could you zoom in to 5 second interval on one of those droops, and also show the cvs data for that segment?

[CL23]

Thanks. Here’s it is. If you need anything else, just let me know.

Carlos

[overeasy]

I can’t reproduce this with inputs. Are those plots of calculated outputs, and if so, can you post the outputs configuration menu that shows the functions?

[CL23]

Totally forgot about this one; probably because it’s not a major issue.

These are the outputs. I’m also adding the inputs, just in case. Version was 02_03_10 at the time.

Thanks.

[overeasy]

02_03_17 ALPHA auto-update

New release to ALPHA users: Add LLMNR responder - Greatly improves accessibility from most browsers. Most will now just accept the device name “iotawatt”. Some will still need “iotawatt.local”, but in general works much better. Use latest core 1.8.0 (2.4.2) - As implied by the version number, this is just a maintenance update of 2.4.1 and is lower risk than the previous jump. Save new values before aging - this should fix a problem where initial measurements after startup are low for one meas…

Note the last item.

[Jam]

Firmware 02_03_18. Looks like boot time voltage is reported 0 and/or low first value to influx. Of course power calculated from it will also be 0 or low…
I have 2 boxes, first uses derived 3-phase (A Volt) and other uses direct (L1-3 Volt). A and L1 is same phase. Voltage differences between phases are also calculated (last 3 measurements).
Here is data from one example (dec 2) when other box boots (from grafana query inspector):
response:Object
results:Array[7]
0:Object
series:Array[1]
0:Object
name:“AV”
columns:Array[2]
values:Array[8]
0:Array[1543781550000,233.09]
1:Array[1543781555000,233.42]
2:Array[1543781560000,233.41]
3:Array[1543781565000,233.26]
4:Array[1543781570000,233.32]
5:Array[1543781575000,233.27]
6:Array[1543781580000,233.33]
7:Array[1543781585000,233.48]
1:Object
series:Array[1]
0:Object
name:“L1V”
columns:Array[2]
values:Array[8]
0:Array[1543781550000,233.18]
1:Array[1543781555000,233.22]
2:Array[1543781560000,233.23]
3:Array[1543781565000,0]
4:Array[1543781570000,203]
5:Array[1543781575000,233.22]
6:Array[1543781580000,233.25]
7:Array[1543781585000,233.43]
2:Object
series:Array[1]
0:Object
name:“L2V”
columns:Array[2]
values:Array[8]
0:Array[1543781550000,232.48]
1:Array[1543781555000,232.66]
2:Array[1543781560000,232.69]
3:Array[1543781565000,0]
4:Array[1543781570000,202.01]
5:Array[1543781575000,232.31]
6:Array[1543781580000,232.45]
7:Array[1543781585000,232.06]
3:Object
series:Array[1]
0:Object
name:“L3V”
columns:Array[2]
values:Array[8]
0:Array[1543781550000,231.13]
1:Array[1543781555000,231.16]
2:Array[1543781560000,231.11]
3:Array[1543781565000,0]
4:Array[1543781570000,200.94]
5:Array[1543781575000,230.98]
6:Array[1543781580000,230.86]
7:Array[1543781585000,230.71]
4:Object
series:Array[1]
0:Object
name:“L12Vd”
columns:Array[2]
values:Array[8]
0:Array[1543781550000,0.7]
1:Array[1543781555000,0.56]
2:Array[1543781560000,0.54]
3:Array[1543781565000,0]
4:Array[1543781570000,1]
5:Array[1543781575000,0.91]
6:Array[1543781580000,0.8]
7:Array[1543781585000,1.37]
5:Object
series:Array[1]
0:Object
name:“L13Vd”
columns:Array[2]
values:Array[8]
0:Array[1543781550000,2.06]
1:Array[1543781555000,2.06]
2:Array[1543781560000,2.12]
3:Array[1543781565000,0]
4:Array[1543781570000,2.07]
5:Array[1543781575000,2.24]
6:Array[1543781580000,2.38]
7:Array[1543781585000,2.72]
6:Object
series:Array[1]
0:Object
name:“L23Vd”
columns:Array[2]
values:Array[8]
0:Array[1543781550000,1.35]
1:Array[1543781555000,1.49]
2:Array[1543781560000,1.59]
3:Array[1543781565000,0]
4:Array[1543781570000,1.07]
5:Array[1543781575000,1.34]
6:Array[1543781580000,1.59]
7:Array[1543781585000,1.34]

[overeasy]

Thanks for that data. Very timely, when a competitor has been making assertions that Derived Reference produces meaningless results, this data shows a difference in voltage of less than 1% between L1 and both L2 and L3.

I will look into the startup issue. Could you post the message log for that period?

[Jam]

Here is that part: log.txt (2.0 KB)
That 22:12:40 is good value before 0.

Long term averate voltage for phases is really close each other. Average difference is about 0.5%. Error comes meaningful when there is big load difference between phases. Like when I have (over) 15A more load in 1 phase, that makes over 5V difference (2-3%) to other 2 phases. Here is long term pic about it:

Phase volts are mean values, but diff is maximum values so you see worst case.