search

nipunbatra / Home_Deployment

http://iawe.github.io
7 stars 5 forks source link

CT Measurement Fluctuating #53

Open nipunbatra opened 11 years ago

nipunbatra commented 11 years ago

Log for Circuit 4 (only refrigerator on at this time, RO water purifier also connected, but i think it will consume significant current only when it needs to fill its tank)

This is the CT log when compressor was Off. During this time jPlug was showing current ~0

4, 0.1037
4, 0.3110
4, 0.4147
4, 0.2074
4, 0.2074
4, 0.2074
4, 0.3110
4, 0.2074
4, 0.2074
4, 0.5184
4, 0.4147 
4, 0.2074
4, 0.3110
4, 0.3110
4, 0.3110
4, 0.1037
4, 0.2074
4, 0.4147
4, 0.4147

This is the CT log when compressor was On. During this time jPlug was showing a current of 0.42 A

4, 0.6221
4, 0.6221
4, 0.6221
4, 0.7258
4, 0.6221
4, 1.0368
4, 0.8295
4, 0.6221
4, 0.6221
4, 0.7258
4, 0.6221
4, 0.6221
4, 0.7258
4, 0.6221
4, 0.7258
4, 1.0368
4, 0.6221
4, 0.6221
4, 0.7258
4, 0.8295
4, 0.6221
4, 0.6221

cc @amarjeet-iiitd

manojgulati commented 11 years ago

image After plotting this we can identify that some appliance is turned on but I feel some noise is coming. This CT is having only refrigerator or something else as well ?.

nipunbatra commented 11 years ago

https://github.com/nipunreddevil/Home_Deployment/blob/master/configuration/circuits.md

shows the mapping.

Barring RO rest all appliances are Off.

manojgulati commented 11 years ago

Can you do a small test as of now:

  1. Turn off refrigerator and take reading from CT for 10 minutes with only RO turned on.
  2. Similar case with only refrigerator turned on w/o RO.

Than it will be easy to understand from where this noise is originating and how we can reduce that.

nipunbatra commented 11 years ago

Ok. Will do the tests and report the results.

manojgulati commented 11 years ago

I feel this board is more noisy than previous one. I had tried to log data from Winet for 2 hours and data was clean enough. I can add some filters to reduce these low freq spikes.

nipunbatra commented 11 years ago
Case I

All appliances on CT number 4 off (turned OFF Ref. and RO). With no load on this CT, it is showing a lot of fluctuation. Readings from other CT's also mentioned here for reference.

Time, CT#, Current

1372936421.51,4, 0.4147
1372936421.59,0, 0.3110
1372936421.68,1, 0.3110
1372936421.76,2, 0.2074
1372936421.85,3, 0.3110
1372936421.93,4, 0.1037
1372936422.02,0, 0.1037
1372936422.1,1, 0.3110
1372936422.19,2, 0.1037
1372936422.27,3, 0.2074
1372936422.36,4, 0.1037
1372936422.44,0, 0.3110
1372936422.53,1, 0.4147
1372936422.61,2, 0.2074
1372936422.69,3, 0.3110
1372936422.78,4, 0.4147
1372936422.86,0, 0.4147
1372936422.95,1, 0.6221
1372936423.03,2, 0.4147
1372936423.12,3, 0.3110
1372936423.2,4, 0.3110
1372936423.29,0, 0.1037
1372936423.37,1, 0.3110
1372936423.46,2, 0.2074
1372936423.54,3, 0.1037
1372936423.63,4, 0.1037
1372936423.71,0, 0.2074
1372936423.79,1, 0.3110
1372936423.88,2, 0.2074
1372936423.96,3, 0.2074  
1372936424.05,4, 0.4147
1372936424.13,0, 0.4147
1372936424.22,1, 0.3110
manojgulati commented 11 years ago

I feel 2 reasons in this case

  1. Circuit is introducing too much noise.
  2. These CT's are placed too much close to each other and there might be EMI in between them. All 5 cores are placed very close to each other. (And this is unavoidable due to placement of MCB's).

If you feel data is not useful than please remove this circuit i will modify this to ensure minimal noise interference.

nipunbatra commented 11 years ago

Ok. Currently testing the WiNet based CT circuit to see if that has noise in it or not.

nipunbatra commented 11 years ago

Following are 7 test cases i performed

@manojgulati : If you would like to pan and zoom into these plots, the script used is here https://dl.dropboxusercontent.com/u/75845627/Deployment/ct_data/ct_data_plot.py

  1. Nothing barring 2 Routers and Plug Computer on first floor was ON. Second floor PC+Router+RPi Data file: https://dl.dropboxusercontent.com/u/75845627/Deployment/ct_data/ct_1.csv 1
  2. In addition to 1, tubelight in small room on first floor was ON Data file: https://dl.dropboxusercontent.com/u/75845627/Deployment/ct_data/ct_2.csv 2
  3. In addition to 2, fan in small room on first floor was ON Data file: https://dl.dropboxusercontent.com/u/75845627/Deployment/ct_data/ct_3.csv 3
  4. In addition to 3, Plug point powering phone and multisensor was ON in small room first floor Data file: https://dl.dropboxusercontent.com/u/75845627/Deployment/ct_data/ct_4.csv 4
  5. In addition to 4, plug point powering phone and multisensor was ON in big room floor Data file: https://dl.dropboxusercontent.com/u/75845627/Deployment/ct_data/ct_5.csv 5
  6. In addition to 5, tube in big room first floor was ON Data file: https://dl.dropboxusercontent.com/u/75845627/Deployment/ct_data/ct_6.csv 6
  7. In addition to 6, fan in big room first floor was ON https://dl.dropboxusercontent.com/u/75845627/Deployment/ct_data/ct_7.csv 7
manojgulati commented 11 years ago

These graphs seems to be much cleaner than previous one. If you do remember ADC was giving +/- one point fluctuation so similar effect is reflected in final current calculation as well. But if you notice mean is still constant we can find average of this to find out actual current.

Do you feel value of current that you are getting is also erroneous ? Because i am not having idea like how much current should be there in these specific load conditions.