2 software features for temperature compensation, Software features for temperature compensation – Maxim Integrated 71M6541 Demo Board User Manual
Page 47

71M6541 Demo Board REV 3.0 User’s Manual
47
Rev 4.0
When analyzing the contribution of thermal errors for power equation 1 for single-phase 3-wire systems, we can
write the equation as follows:
2
)
2
2
)
(
6
2
4
4
1
4
X
S
X
VD
X
S
X
VD
C
C
IB
C
C
VA
C
C
IA
C
C
VA
IB
IA
VA
P
⋅
⋅
⋅
⋅
⋅
−
⋅
⋅
⋅
⋅
⋅
=
−
=
The terms used in the above equation are defined as follows:
•
VA = voltage applied to the meter
•
IA = current applied to the shunt S1 that is connected to the IAP/IAN pins of the 71M6541
•
IB = current applied to the shunt S2 that is connected via the Remote Interface IC
•
C
VD
= error contribution from the voltage-divider
•
C
4X
= error contribution from the voltage reference of the 71M6541
•
C
S1
= error from the shunt resistor that is connected to the IAP/IAN pins of the 71M6541
•
C
S2
= error from the shunt resistor that is connected via the Remote Interface IC
•
C
6X
= error contribution from the voltage reference of the Remote Interface IC
The equation can be simplified as follows:
⋅
⋅
⋅
⋅
⋅
−
⋅
⋅
⋅
⋅
=
2
)
2
6
2
4
2
4
1
X
S
X
VD
X
S
VD
C
C
IB
C
C
VA
C
C
IA
C
VA
P
Or:
}
{
X
S
X
S
X
VD
C
C
IB
C
C
IA
C
C
VA
P
6
2
4
1
4
2
⋅
⋅
−
⋅
⋅
⋅
=
2.4.2 SOFTWARE FEATURES FOR TEMPERATURE COMPENSATION
In the default settings for the Demo Code, the
CECONFIG
register has its
EXT_TEMP
bit (bit 22) set, which
means that temperature compensation is performed by the MPU by controlling the
GAIN_ADJA
and
GAIN_ADJB
registers. In this context,
GAIN_ADJA
controls both current and voltage readings for phase A
(i.e., the VA and IAP/IAN pins) whereas
GAIN_ADJB
controls both current and voltage readings for phase B
(i.e., the VA and the 71M6X0X Remote Sensor Interface IC).
In general, the
GAIN_ADJA
and
GAIN_ADJB
registers offer a way of controlling the magnitude of the voltage
and current signals in the data flow of the CE code. A value of 16385 means that no adjustment is performed
(unity gain), which means that the output of the gain adjust function is the same as the input. A value of 99% of
16385, or 16222, means that the signal is attenuated by 1%.
The Demo Code bases its adjustment on the deviation from calibration (room) temperature
DELTA_T
and the
coefficients
PPMC
and
PPMC2
to implement the equation below:
23
2
14
2
2
_
2
_
16385
_
PPMC
T
DELTA
PPMC
T
DELTA
ADJ
GAIN
⋅
+
⋅
+
=
It can be seen easily that the gain will remain at 16385 (0x4001), or unity gain, when
DELTA_T
is zero.
For complete compensation, the error sources for each channel have to be combined and curve fit to generate
the
PPMC
and
PPMC2
coefficients, as we will see in the following section.
The
PPMC
and
PPMC2
coefficients are in the following MPU RAM locations:
•
Phase A (IAP/IAN pins):
PPMCA
-- 0x0B,
PPMC2A
– 0x0F
•
Phase B (IBP/IBN pins):
PPMCB
-- 0x0C,
PPMC2B
– 0x10