Archived Forum

What is your mains voltage?

Are your current transformers on the correct cables (there must be two if you have "Use" and "Gen")?

Which voltage adapter, current transformer and burden are you using?

Which sketch?

What did you do to change the calibration? Changing a number shouldn't cause a reset.

How are you reading the 'wrong' value? What are the numbers?

AC voltage 162VRMS

Current sensors SCT-013-000 with standard 18ohm burden. Standard emontx kit built. (i have two). 

CT1 connected to Use

CT2 connected to Gen 

Type 2 setup

AC monitor 'Mascot'

sketch emontx_CT123_Voltage with following code setup

ct1.voltageTX(234.26, 1.7); 

ct1.currentTX(1, 111.1);  

loop{
ct1.calcVI(20,2000);

With a 1.6kw kettle the use displayed was 790W but with another current monitor it was 1.6kW. 

For completeness is CT3 set to 0 ?

Yep.

It was all running OK. But then i decided to calibrate it against another current measuring device.

I was a bit suspicious that the house was only using 200W at night when most things are off, and the kettle only used 800W extra.

Looking at the the calculations for the calibration values the figures all make sense in theory.

Also checked the CT ferrites were clean. Ans swapped CT sensors over.

For info. I am a Software engineer with a  electronic hardware background.

I think I see at least part of your problem. Is your voltage monitor is a standard one sold for the UK, or one made for your local voltage?  Look in 'Building Blocks' at the article about calibration: http://openenergymonitor.org/emon/buildingblocks/ct-and-ac-power-adaptor... This explains how the calibration factor 234.26 is arrived at. It is likely that this is wrong for your set-up.

If you struggle, you can let me know the voltage you get out of the adapter and at what mains voltage, and tomorrow I can work through the numbers for you.

Can you put some Serial.print statements in to see whether the voltage or current or neither are being measured correctly?

Just above the LED flash is as good a place as any.

  Serial.print("CT1 V= "); Serial.println(ct1.Vrms);
  Serial.print("CT1 I= "); Serial.print(ct1.Irms);
  Serial.print(" P= "); Serial.print(ct1.realPower);
  Serial.print(" VA= "); Serial.print(ct1.apparentPower);
  Serial.print(" pf= "); Serial.println(ct1.powerFactor);
  Serial.print("CT2 V= "); Serial.println(ct2.Vrms);
  Serial.print("CT2 I= "); Serial.print(ct2.Irms);
  Serial.print(" P= "); Serial.print(ct2.realPower);
  Serial.print(" VA= "); Serial.print(ct2.apparentPower);
  Serial.print(" pf= "); Serial.println(ct2.powerFactor);

Note also the true voltage and the load, please.

Sorry not had a lot of free time to setup equipment. 

But what i did notice was that the voltage measured ct1.vrms = 165V which i thought was right but it should be 230VRMS not 165 x 1.414 = 230v pk

The voltage out of the AC adaptor is 11V unloaded measured with a DVM which is about right i think. So don't know where the 165VRMS comes from. Both Emontx PCB give the same results.

I will try and add the debugging later this week.

When I ask you for your mains voltage, I need you to measure the mains voltage directly using a relatively accurate meter. If that is wildly different to the declared voltage provided by your electricity company, I need to know that too, because it may mean that you meter is faulty and cannot be trusted.

If you give the voltage that the emonTx reads, that is just adding total confusion and misdirection and I won't be able to solve your problem.

The following is is a capture of the debug o/p, taken on a sunny day when panels generating about 800W, house load approx 500W

The numbers before CT1 are power1,power2,volts but power1 and power2 have x2 fix to get it to read right.

Mains voltage measured as 235VAC

CT part no. SCT-013-000

9V adaptor Mascot

emonTX CT123 Voltage example
OpenEnergyMonitor.org
Node: 10
-351 1044 167.58
CT1 V= 167.58
CT1 I= 7.59 P= -175.51 VA= 1271.27 pf= -0.14
CT2 V= 163.49
CT2 I= 7.48 P= 522.10 VA= 1222.55 pf= 0.43
-518 791 159.72
CT1 V= 159.72
CT1 I= 2.00 P= -259.34 VA= 318.82 pf= -0.81
CT2 V= 158.99
CT2 I= 2.63 P= 395.91 VA= 418.46 pf= 0.95
-545 826 164.89
CT1 V= 164.89
CT1 I= 1.80 P= -272.93 VA= 297.19 pf= -0.92
CT2 V= 162.70
CT2 I= 2.53 P= 413.14 VA= 412.32 pf= 1.00
-535 795 162.30
CT1 V= 162.30
CT1 I= 1.82 P= -267.76 VA= 295.29 pf= -0.91
CT2 V= 159.69
CT2 I= 2.50 P= 397.60 VA= 399.07 pf= 1.00
-547 823 164.93
CT1 V= 164.93
CT1 I= 1.80 P= -273.69 VA= 296.79 pf= -0.92
CT2 V= 162.79
CT2 I= 2.52 P= 411.79 VA= 410.98 pf= 1.00
-533 791 160.45
CT1 V= 160.45
CT1 I= 1.80 P= -266.69 VA= 288.69 pf= -0.92
CT2 V= 158.73
CT2 I= 2.49 P= 395.62 VA= 394.57 pf= 1.00
-572 834 166.91
CT1 V= 166.91
CT1 I= 1.87 P= -286.23 VA= 312.05 pf= -0.92
CT2 V= 162.45
CT2 I= 2.56 P= 417.37 VA= 416.32 pf= 1.00
-516 784 158.12
CT1 V= 158.12
CT1 I= 1.78 P= -258.32 VA= 281.58 pf= -0.92
CT2 V= 157.89
CT2 I= 2.48 P= 392.28 VA= 391.41 pf= 1.00
-586 827 169.33
CT1 V= 169.33
CT1 I= 1.89 P= -293.25 VA= 320.24 pf= -0.92
CT2 V= 162.88
CT2 I= 2.54 P= 413.83 VA= 413.03 pf= 1.00
-528 787 160.31
CT1 V= 160.31
CT1 I= 1.81 P= -264.11 VA= 289.44 pf= -0.91
CT2 V= 160.01
CT2 I= 2.46 P= 394.00 VA= 393.54 pf= 1.00

Obviously for some reason the Voltage is being read wrong.

OK, You need to check the voltage input circuitry. This picture should help.

1. With the ac input unplugged, do you measure 1.65 V d.c. (or close) on the a.c input socket (either the centre or the barrel contact)?

2. Do you have the correct value components in R11, R12, R13, R14? (These are the 4 resistors, two each side of the ac-in socket).

3. Is C10  the correct way round? (Broad white band (-) to the edge of the board).

4. With the ac input plugged in, do you measure about 1.1 V rms across R14 (10 K - brown-black-orange)?

If all these are correct, I'm stumped.

1.   1.62Vdc

2.   yes checked OK

3.   C10 correct

4.   0.98V rms

i have 2 emontx boards and both built the same and behave the same. 

Check to see if you have the correct version of EmonLib.cpp.  I had a similar problem and found that the function readVcc at the end of EmonLib.cpp was incorrect.  The source on github was updated with this fix on 26 Oct.

Good point - I'd forgotten that!

That fixed it. My Emonlib was dated Oct16th.

Remembering to restart Arduino IDE 

Voltage now 245 Vrms (still a bit high) but power monitoring roughly right.

Thanks Rob and Stephen. Sort of knew it had to be software

And I spotted the fault in the library. 

You just need to tweak the voltage calibration now to get it correct.

I had to adjust the constant VoltageCalibration in emonTx_CT123_Voltage.ino to get an accurate voltage measurement.  I used a 9V ac adaptor I had spare so wasn't surprised that i needed to change it from the default value.  I see above that you have the means to measure the mains voltage which you'll need to adjust the calibration.  I also set the PhaseShift contstant to suit my adaptor.

When comparing your power measurement with another power monitor remember that the other device, assuming it's only sensor is a CT, will use a nominal mains voltage and power factor in its power calculation.