Equalize data or Inverter problem ?

aurora protocol http://www.curtronics.com/Solar/
stetor
Posts: 12
Joined: Fri May 13, 2016 6:42 pm

Re: Equalize data or Inverter problem ?

Post by stetor » Mon May 23, 2016 9:36 am

OK ! the patch work !
two days without the spike in the main graph
Image

but the work isn't finished ... the same "wrong" value, sometimes happen also in the string1 or string2 of the detail inverter data ... see image and data below. I rework the code also for skipping these "bad" data.
Image

Code: Select all

09:50:05,228.625259,3.758124,859.2021,228.448654,3.667091,837.742088,,,,,,,240.428482,6.824213,1628.911255,,,,,,,49.988003,95.990852,42.90321,40.861954,,10036.438
09:55:05,228.213181,3.895882,889.091543,228.016953,3.804849,867.570053,,,,,,,238.714508,7.133376,1687.046387,,,,,,,49.988003,96.037071,43.47784,41.325367,,10036.576
10:00:01,228.389786,4.029611,920.322018,228.173935,3.936967,898.313337,,,,,,,237.411575,7.384788,1745.181641,,,,,,,50.004002,95.961052,44.033092,41.689487,,10036.719
10:05:04,225.779953,4.160924,939.453216,225.583725,4.073919,919.00989,,,,,,,239.431183,7.510992,1776.671631,,,,,,,50.002998,95.598969,44.584415,42.182751,,10036.865
10:15:47,223.817673,4.417911,988.806463,4.419522,223.896164,989.513979,,,,,,,236.875931,7.961128,1882.445435,,,,,,,49.985004,95.153717,4.330906,43.203041,,10037.191
10:20:08,223.817673,4.52425,1012.607006,223.621445,4.439662,992.803519,,,,,,,241.556229,8.030239,1926.046875,,,,,,,49.987003,96.042526,45.879517,43.610008,,10037.333
10:25:02,224.112015,4.641867,1040.298202,223.915787,4.559696,1020.987961,,,,,,,243.711227,8.173458,1976.915161,,,,,,,50.007999,95.906876,46.34848,44.131001,,10037.49
10:30:01,224.170883,4.746595,1064.048479,223.817673,4.662007,1043.439631,,,,,,,241.849548,8.452565,2028.590942,,,,,,,49.996002,96.25634,46.824615,44.519558,,10037.657
10:35:07,223.856918,4.871463,1090.510762,223.680313,4.787681,1070.910008,,,,,,,244.147217,8.582261,2081.881592,,,,,,,50.000999,96.320053,47.337418,44.925915,,10037.833
10:40:10,223.032761,4.956857,1105.54144,222.836533,4.869852,1085.180949,,,,,,,239.394165,8.885756,2112.563965,,,,,,,49.997002,96.432297,47.8284,45.312298,,10038.007

appera to me that at 10:15 two field are swapped
4.419522,223.896164
must be
223.896164,4.419522
really really strange ... this is not a transmission error ... i'm sure this is the firmware that has some problem or the aurora binary (i try to contact Curt as you suggested).

in effect i've tried swapping the data and ... tadah !!
here the correct graph ! :)
Image

and now ? Or i skip all the raw with a control similar to the main graph or ... i haven't other idea ...


:)

jeanmarc
Posts: 1832
Joined: Thu Aug 29, 2013 7:16 am

Re: Equalize data or Inverter problem ?

Post by jeanmarc » Wed May 25, 2016 8:53 pm

hmm. one other trick would be to check if all values are present :
if (isset($$I1V)&&isset($I1A)&& so on )

stetor
Posts: 12
Joined: Fri May 13, 2016 6:42 pm

Re: Equalize data or Inverter problem ?

Post by stetor » Fri May 27, 2016 8:15 am

Hi JeanMarc,

the fields are present but have a bad value or are swapped so the trick to check all the field presence can't work ...
In the mean time i've contacted Curtis but he exclude a problem of the aurora program ... he think that is the firmware of the inverter.
I've also looked at the aurora source and i've had two idea :-)
The first for solving this issue. I thinked to get for two times the same parameter and if the data are different, discard all and retry other two times. The packet for a single read is really little (10 bytes) so i think this change don't impact on the performance.
The second change is for implementing a direct read through the net socket and not only through the serial device. Something like tcp@my.server.com:8080 or best udp@my.server.com:8080 instead of /dev/ttyS0
In this manner i don't have need anymore for the socat wrapper
The code of the aurora is well written and clear. The license is GNU so there is no problem.

What do you think ?

P.S.: I did not dare to ask to Curtis because he don't have a great patience ;)

jeanmarc
Posts: 1832
Joined: Thu Aug 29, 2013 7:16 am

Re: Equalize data or Inverter problem ?

Post by jeanmarc » Fri May 27, 2016 10:55 am

Hi,
I am not sure it is wise to compare twice :roll: The data can vary vary much between 2 samples (Clouds,..)
Maybe, the solution is to contact ABB for a firmware update or even downgrade.

ps: Curt is a busy guy but he takes the time to reply to good questions :)

stetor
Posts: 12
Joined: Fri May 13, 2016 6:42 pm

Re: Equalize data or Inverter problem ?

Post by stetor » Fri May 27, 2016 11:39 am

Hi Jeanmarc,

the time between 2 sample are minimal but ... you are right ... not for the cloud (there is a sufficient time for change due to a cloud) but the value are a float with many decimal number so, also in a short time the value may change ... anyway changing a bit the algorithm that accept minimal difference bewteen the two read and discard when there are major difference, can work ... i hope :)

Post Reply

Who is online

Users browsing this forum: No registered users and 2 guests