71M6515H
A Maxim Integrated Products Brand
Energy Meter IC
DATA SHEET
JULY 2011
General Notes on Calibration
The calibration procedures described below should be followed after
interfacing the voltage and current sensors to the 71M6515H chip. When
properly interfaced, the V3P3 power supply is connected to the meter
neutral and is the DC reference for each input. Each voltage and current
waveform, as seen by the 6515H, is scaled to be less than 250mV
(peak).
Voltage
Current lags
voltage
Each meter phase must be calibrated individually. The procedures below
show how to calibrate a meter phase with either three or five
measurements. Note that there is no need to calibrate for VARh if the
Wh measurement is calibrated correctly. Note that positive load angles
correspond to lagging current (see Figure 12).
(inductive
)
Positive
direction
+60°
Current
-60°
For a typical calibration, a meter calibration system is used to apply a
calibrated load, e.g. 240V at 30A, while interfacing the voltage and
current sensors to the 71M6515H. This load should result in an ob-
servable pulse rate at the PULSEW output depending on the selected
energy per pulse. For example, 7.2kW will result in an energy rate
corresponding to 7200Wh/3600s = 2Wh/s, i.e., when 7.2kW are applied
per phase (resulting in a total power of 21.6kW, equivalent to 6Wh/s) and
a Kh of 3.2 (Wh/pulse) has been configured, a pulse rate of 6Wh/3.2Whs
= 1.875Hz will be established.
Current leads
voltage
(capacitive
)
Voltage
Using Energy
Generating Energy
Figure 12: Definition of Load Angles
It is entirely possible to calibrate piece-wise, i.e. in segments, to compensate for non-linear sensors. For example, one set of
calibration factors can be applied by the host when the current is below 0.5A, while another set is applied when the current is
at or above 0.5A.
Calibration Procedure for CT and Resistive Shunt
A typical meter has phase and gain errors as shown by φS, AXI, and AXV in Figure 13. Following the typical meter convention of
current phase being in the lag direction, the small amount of phase lead in a typical current sensor is represented as -φS. The
errors shown in Figure 13 represent the sum of all gain and phase errors. They include errors in voltage attenuators, current
sensors, signal conditioning circuits, and in ADC gains. In other words, no errors are made in the ‘input’ or ‘meter’ boxes.
INPUT
ERRORS
METER
IRMS
I
φL
−φS
AXI
=
=
IDEAL I, ACTUAL I AXI
φ
L is phase lag
φS is phase lead
=
φ
)
IDEAL IV cos(
L
W
Π
=
φ −φ
ACTUAL IV AXI AXV cos(
)
S
L
VRMS
V
AXV
=
=
IDEAL V , ACTUAL V AXV
−
ACTUAL IDEAL ACTUAL
≡
=
−
1
ERROR
IDEAL
IDEAL
Figure 13: Watt Meter with Gain and Phase Errors.
During the calibration phase, we measure errors and then introduce correction factors to nullify their effect. With three
unknowns to determine, we must make at least three measurements. If we make more measurements, we can average the
results.
Page: 25 of 60
© 2005−2011 Teridian Semiconductor Corporation
1.6
TERIDIAN [ TERIDIAN SEMICONDUCTOR CORPORATION ]
