Anyone please tell me why we are not using capacitive load while doing line accumulation calibration of ADE7758 IC.
Thanks in advance,
You can absolutely use capacitive loads for calibration of ADE7758. The datasheet refers to inductive load as an example, due to the fact that purely capacitive loads are not that common. Regardless of the type of load you use while calibration, the calibrated meter will be able to operate accurately with both capacitive and inductive loads.
Will you please tell me ,why we are not doing gain calibration of voltage & current of ADE7758.
The current channel gain registers (xIGAIN) in ADE7758 are not recommended to be used. There are voltage gain registers (xVRMSGAIN) though, which you can use to adjust the voltage RMS scaling. In terms of interpretation of RMS register results, you should compute the V/LSB and A/LSB constants as part of the gain calibration for RMS.
The datasheet doesn't go into this because the focus is on Energy calibration. But if you are using RMS registers, you will have to do gain calibration for that.
Please have a look at the first question and answer of this discussion: RMS gain calibration
Thank you very much for reply. I have read that conversation regarding RMS gain calibration.
I am confused about VRMS gain calibration, please tell what I should perform first VRMS gain calibration or offset calibration. And suppose if I calculate V/LSB & I/LSB constant from known source & load in that case I need to do VRMS gain calibration?
The term 'Gain Calibration' involves two steps: the first step is to set the scaling the way you want it (For VRMS gain calibration, it will be done by adjusting the xVRMSGAIN registers). The next step is to find out the conversion constant (getting the V/LSB constant). To get the conversion constant, you will need to apply a known voltage to the meter and read the measurement value (xVRMS in this case. You can do an average of 100 xVRMS register values collected over 1 second to get a stable result). Now, dividing the input voltage by the register result in decimal will give you the conversion constant. You can save this constant in MCU for interpreting future register results.
Also, RMS gain calibration is usually done before doing RMS offset calibration. This order is important particularly when you are setting xVRMSGAIN registers to values other than the default. Offset calibration corrects for errors with very low input signals and the offset correction injected doesn't affect the conversion constants that you obtained at high input signal conditions (usually gain calibration is done with nominal input- so, voltage will be 220Vrms/110Vrms during gain calibration).
Hope this is clear! Please get back to me if you still have concerns!
Thank you very much for reply. I have understood the steps you have mentioned for RMS gain calibration.
Now I am doing burden resistor calculation as per your application note AN639. My CT's Imax as per its datasheet(http://www.electrohms.com/data_pdf/CT1273-A1-RC_01.pdf) is 100A & its turns ratio is 1:2500(i.e CTRN is 2500).CT is in center tap configuration. After calculation Rb is 2.21 Ohm. Is this calculation is correct or I have done some wrong calculation please let me know.
Thanks & Regards,
The burden resistor calculation makes sense but I have just one question. The max current of 100Arms- where is it coming from? This should be the max current in your application and not the CT's maximum allowable current level. So, say if your application nominal current is 10Arms and maximum current is 50Arms, then should use the 50Arms value in the formula to calculate the burden resistor value.
Thank you very much for reply.I had wrong assumption regarding Maximum current in burden register calculation.
The max current in my application is 18Amp & as per calculation burden resistor value is near about 12Ohm.
Thanks & Rgds,
The Rb value of 12ohm is correct for 18Arms max current if you are using 2 of these 'Rb's (burden resistors) in a differential manner. This will help keep the max current level to half of the IC's full scale.
Let me know if you run into any other issues.
I am currently doing offset calibration of voltage & current. In my application I can not give Vfullscale/20(i.e 341/20=17V) to ADE7758 because my controller & input of ADE7758 are on same supply.
So is it correct to give the Vmin=minimum voltage that will drive controller & ADE7758?
Also I am having one more query regarding WATTHr,VAHr energy gain calibration, in this I am not using calibration frequencies, so will you please tell me how to calculate WATTHR_expected & VAHR_expected for gain calculation?
1. It makes sense to do offset calibration with Vmin. Vfs/20 is just an example.
2. Page 50 of the datasheet shows the formulae that you are looking for. You can calibrate either using CF pulses or using line cycle energy accumulation mode.
Thank you very much for reply. I have gone through all formulas given in line accumulation method.
I am not using CF frequencies. If I calculate Wh/LSB constant by giving known WATTHR first and using that constant I will calculate WATTHRmeasred to find out xWG . This approach is correct or I need to use MC as 3200 & then calculate WATTHRexpected ?
Your approach would work well. However, you need to watch out for how you compute the expected xWATTHR value. Start with getting the Wh/LSB value by reading the energy register in line cycle accumulation mode (use Equation 66 in Rev.E datasheet). Then, if you want to set a expected Wh/LSB value, use that expected value in the same formula to get expected xWATTHR. Now, with measured and expected xWATTHR, you can get the gain scaling register value (Equation 63).
Thank you very much.I have understood the Energy gain calibration,
Hope you are doing well. I am facing problem while reading watthr in line accumulation mode with 5 second accumualtion time but I am getting nonlinear values for range of current. I have also read one of your post in which you had mentioned to perform 2's complement if MSB of fetched watthr is negative, but after doing so still I am getting nonliner watthr. Please help me.
What is your LCYCMODE register set to?
What is your LINECYC register set to?
What is the range of currents you are looking at? And what values are you returning in the WATTHR register over that range?
This extra information will help me understand what could be the problem.
Thank you for reply.
Its my bad, I have solved that problem , as my WDIV register was set to 0.
But I am facing another problem , I have done watthr calibration as follows:
I am reading WATTHR in energy accumulation mode for 5 seconds with WDIV=0x20 . I am having programmable load-source so that I have the correct WATTHR for 5 seconds. So I just calculates the Wh/LSB constant first & stores into my controller's non-volatile memory. After that I calculates % error using known WATHHR against , WATTHR fetched from ADE using eq. 63 .
So this gives me accurancy below 1% for power factor 1. But for power factor 0.5 it gives almost 5% error.
ADE7758 register configuration is as follows:
Please suggest if I am doing some wrong.
From the way you describe it, it sounds like you need to perform a phase calibration on your energy channels. I believe you are using CT's in which case a phase calibration will most likely be required, have you performed phase compensation yet?
Page 51 of the data sheet talks about phase calibration using line accumulation for reference.
Thank you very much for reply.
I have understood the scenario from figure 6 from following application note.
So as you said it doesn't matter which kind of load we are using for calibration.
Retrieving data ...