TERIDIAN 71M6515H

71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
GENERAL DESCRIPTION
FEATURES
VOLTAGE
CURRENT
The 71M6515H is a high-accuracy analog front end
(AFE) IC that provides measurements for fourquadrant 3-phase metering. The combination of a
21-bit sigma-delta A/D converter with a six-input
analog front end, a thermally compensated high precision reference and a compute engine results in
high accuracy and wide dynamic range. Patented
Single-Converter Technology™ reduces cross talk
and cost. This IC also provides RTC and battery
backup for time-of-use (TOU) metering.
Battery
RTC
COMPUTE
ENGINE
UART, IRQZ
HOST
PROCESSOR
RAM
SSI
VOLTAGE
REFERENCE
CONTROL
High Accuracy
• < 0.1% Wh accuracy over temperature and
2000:1 range
• Exceeds IEC62053/ANSIC12.20 specifications.
• Precision ultra-stable voltage reference
(10PPM/°C)
• Single Converter Technology™ reduces cross
talk and power consumption
• Six sensor inputs—referenced to V3P3
• Compatible with CTs, resistive shunts and
Rogowski Coil sensors
• -90dB THD max.
• Digital temperature compensation
• Sag detection
• Measures Wh, VARh, VAh, Vrms, Irms, V-to-V
phase and load angle on each phase
• Four-quadrant metering.
• Four low-jitter pulse outputs from selectable
measurements
• Four pulse count registers
• Selectable default status for pulse pins
• Same calibration data for 46Hz to 64Hz line
frequency
• Broad CT phase compensation (±7deg)
DIO
TERIDIAN 71M6515H
Figure 1: Meter block diagram
As shown in the block diagram (Figure 1), the host
processor communicates with the 71M6515H
through a UART interface using the programmable
IRQZ interrupt. The 71M6515H calculates and accumulates meter measurements for each accumulation
interval. A high-speed synchronous serial port (SSI)
is provided to facilitate high-end metering. Integrated
rectifying functions on the battery backup circuit
enable minimal external component usage and
minimum back-up current. Also, eight multi-purpose
pins are provided for control of peripherals.
Page: 1 of 57
Battery Backup
• Protects accumulated data and powers real time
clock during power supply outage
• Compatible with Li-ion, NiCd, or super-capacitor
• Battery backup current 2µA typical at 25°C
External Data Interface
• UART Control Interface, two selectable data
rates
• 8 general-purpose I/O pins with alarm capability
• 5 or 10MHz selectable high speed synchronous
serial output for DSP interface
• IRQ output signal for alarms and end of measurement intervals.
• Alarms on voltage sag, over-voltage, over-current
Low System Cost
• Power Consumption 30mW @ 3.3V typical
• Real-time clock with temperature compensation
• Built-in power fault detection
• Single 32kHz crystal time base
• Single-supply operation (3.3V)
• 64-lead LQFP package
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
VREF
IA
VA
IB
VB
IC
VC
PULSEW
PULSER
V3P3A
GNDA
GNDA
ΔΣ ADC
CONVERTER
GNDD
MUX
VOLT
REG
V3P3A
+
V3P3D
21
GNDD
VREF
VX
TEMP
VREF
GNDD
V2P5
MUXSYNC
VBIAS
(1.5V)
2.5V to logic
CALCULATIONS
SSI INTERFACE
OUTPUT VALUES:
Whr (A, B, C)
VARhr (A, B, C)
VAhr (A, B, C)
Vrms (A, B, C)
Irms (A, B, C)
Iphase (A, B, C)
Frequency (Selected Phase)
Temperature
SSCLK
SSDATA
SFR
SRDY
SYSTEM CLOCKS
CK_GEN
CKTEST
ALARMS:
Voltage Sag (A, B, C)
Zero Cross (Selected Phase)
Over-Voltage (All)
Over-Current (All)
D0-D7 State Change
BATTERY BACKUP
OSC
(32KHz)
XIN
XOUT
RTC
VBAT
PULSE4
PULSE3
PULSE_INIT
BAUDRATE
DATA
RAM
CONTROL
MISC
TX
IRQZ
UART
RX
I/O CONTROL
D0
D4
D1
D5
D2
D6
Change of State
(D0...D7)
D3
D7
TEST
VFLT
FAULT DETECT
VBIAS
(1.5V)
GNDD
V3P3
TMUXOUT
VX
RESERVED
64 PINS -- 64 TQFP
RESETZ
March 3, 2008
Figure 2: IC Functional Block Diagram
Page: 2 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
ELECTRICAL SPECIFICATIONS
ABSOLUTE MAXIMUM RATINGS
Supplies and Ground Pins:
V3P3D, V3P3A
|V3P3D – V3P3A|
VBAT
GNDD
Analog Output Pins:
−0.5V to 4.6V
0V to 0.5V
-0.5V to 4.6V
-0.5V to +0.5V
-1mA to 1mA,
-0.5V to V3P3A+0.5V
-1mA to 1mA,
-0.5 to 3.0V
VREF
V2P5
Analog Input Pins:
IA, VA, IB, VB, IC, VC
VFLT, VX
XIN, XOUT
Digital Input Pins:
RX
D0…D7
All other pins
-0.5V to V3P3A+1.0V
-0.5V to V3P3A+0.5V
-0.5V to 3.0V
-0.5V to 3.6V
-0.5V to 6V
-0.5V to V3P3D+0.5V
Operating junction temperature (peak, 100ms)
Operating junction temperature (continuous)
Storage temperature
Solder temperature – 10 second duration
ESD Stress
Pins IA, VA, IB, VB, IC, VC, RX, TX
All other pins
140 °C
125 °C
−45 °C to 165 °C
250 °C
6kV
2kV
Stresses beyond Absolute Maximum Ratings may cause permanent damage to the device. These are stress ratings only and functional
operation at these or any other conditions beyond those indicated under “recommended operating conditions” is not implied. Exposure to
absolute-maximum-rated conditions for extended periods may affect device reliability. All voltages are with respect to GNDA.
RECOMMENDED OPERATING CONDITIONS
PARAMETER
CONDITION
MIN
3.3V Supply Voltage (V3P3A, V3P3D)+
Normal Operation
Battery Backup
No Battery
Battery Backup
3.0
3.3
3.6
V
0
3.8
V
Externally Connect to V3P3D
2.0
3.8
V
-40
85
ºC
VBAT
+
Operating Temperature
TYP
MAX
UNIT
V3P3A and V3P3D should be shorted together on the circuit board. GNDD and GNDA should also be shorted on the circuit board.
Page: 3 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
LOGIC LEVELS
PARAMETER
CONDITION
Digital high-level input voltage, VIH
Digital low-level input voltage, VIL
ILOAD = 1mA
ILOAD = 15mA
Input pull-up current, IIL
RESETZ
E_RXTX, E_ISYNC/BRKRQ
E_RST
Other digital inputs
Input pull down current, IIH
TEST
Other digital inputs
TYP
2
−0.3
V3P3D
–0.4
V3P3D0.6
0
Digital high-level output voltage VOH
Digital low-level output voltage VOL
MIN
ILOAD = 1mA
ILOAD = 15mA
VIN=0V
MAX
UNIT
V3P3D
0.8
V
V
V3P3D
V
V
0.4
0.8
V
V
10
10
10
-1
100
100
100
+1
µA
µA
µA
µA
10
-1
100
+1
µA
µA
VIN=V3P3D
SUPPLY CURRENT
PARAMETER
V3P3A + V3P3D
V3P3A current
V3P3D current
VBAT current
VBAT current,
VBAT=3.6V
CONDITION
MIN
Normal Operation,
V3P3A=V3P3D=3.3V
VBAT=3.6V
TYP
MAX
UNIT
8.8
3.7
5.1
2
11.5
4.7
6.8
300
4
mA
mA
mA
nA
µA
4
12
µA
TYP
MAX
UNIT
-300
Battery backup,
≤25°C
V3P3A=V3P3D=0V
fOSC = 32kHz
85°C
VREF
PARAMETER
VREF output voltage, VNOM(25)
VREF output impedance
VNOM definition A
CONDITION
MIN
Ta = 25ºC
1.193
1.195
1.197
2.5
ILOAD = 10µA, -10µA
2
VNOM(T) = VREF(22) + (T-22)TC1 + (T-22) TC2
V
kΩ
V
VREF(T) deviation from VNOM(T)
VREF(T ) − VNOM(T )
10 6
Ta = -40ºC to +85ºC
VNOM
max(|T − 22 |,40)
-10
±25
VREF aging
Page: 4 of 57
+10
© 2005-2008 TERIDIAN Semiconductor Corporation
PPM/ºC
PPM/year
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
2.5V VOLTAGE REGULATOR
PARAMETER
CONDITION
Voltage Overhead V3P3D-V2P5
PSRR
ΔV2P5/ΔV3P3D
MIN
Reduce V3P3 until V2P5
drops 200mV
RESETZ=1, ILOAD=0
TYP
-3
MAX
UNIT
440
mV
+3
mV/V
RTC
PARAMETER
CONDITION
Range for date
MIN
TYP
MAX
UNIT
2000
--
2255
year
MIN
TYP
MAX
UNIT
1
µs
µs
MAX
1
3
UNIT
µW
pF
5
5
25
pF
pF
kHz
MAX
UNIT
RESETZ
PARAMETER
CONDITION
Reset pulse width
Reset pulse fall time
5
CRYSTAL OSCILLATOR
PARAMETER
Maximum Output Power to Crystal4
Xin to Xout Capacitance
Capacitance to DGND
Xin
Xout
Watchdog RTC_OK threshold
CONDITION
MIN
TYP
TEMPERATURE SENSOR
PARAMETER
CONDITION
Nominal Sensitivity (Sn)
4
Nominal Offset (Nn) 4
MIN
TYP
-900
LSB/ºC
40000
0
LSB
TA=25ºC, TA=75ºC
Nominal relationship:
N(T)= Sn*T+Nn
Temperature Error
ERR = (T − 25) −
( N (T ) − N (25))
Sn
TA = -40ºC to +85ºC
-3
+3
ºC
MAX
UNIT
7.56
kHz
0.15
kHz
0.15
kHz
PULSE GENERATOR TIMING SPECIFICATIONS
PARAMETER
PULSEW, PULSER maximum rate
PULSE3, PULSE4 maximum rate
Pulse count frequency
Page: 5 of 57
CONDITION
MIN
31
APULSE=2 -1,
15
WRATE=2 -1
31
PULSE3=2 -1,
WRATE=215-1
all pulse outputs
© 2005-2008 TERIDIAN Semiconductor Corporation
TYP
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
THERMAL CHARACTERISTICS
PARAMETER
Thermal resistance, junction to ambient (RθJA )
CONDITION
VALUE
UNIT
Air velocity 0 m/s. Part soldered to PCB.
63.7
°C/W
UART HOST INTERFACE
PARAMETER
CONDITION
Baud Rate
Character set
Data Format
Byte-to-byte delay (6515H times out after
maximum delay)
Byte-to-byte delay
Response time to read command
Response time to read command when
71M6515H is post-processing data
Host sending data to
6515H
6515H sending data to
host
6515H has data ready
Data not ready
CE_ONLY = 1
CE_ONLY = 0 and
VAH_SELECT = 0
CE_ONLY = 0 and
VAH_SELECT = 1
MIN
TYP
MAX
UNIT
19.2
binary
8N1
38.4
kBaud
10
20
ms
0
0.1
ms
0.5
2
ms
40
ms
80
ms
350
ms
MAX
UNIT
-250
250
mV
peak
-10
+10
µV/V
40
-75
-90
90
dB
dB
kΩ
ADC CONVERTER, V3P3 REFERENCED
PARAMETER
CONDITION
Usable Input Range (Vin-V3P3A)
Voltage to Current cross talk:
10 6 ΔNout PK 357 nV / VIN
100 ΔV 3P3 A / 3.3
Input Offset (Vin-V3P3A)
Page: 6 of 57
TYP
Vin = 200mV peak, 65Hz,
on VA, VB, or VC
10 6 *Vcrosstalk
cos(∠Vin − ∠Vcrosstalk ) Vcrosstalk = largest
Vin
measurement on IA, IB, or
THD (First 10 harmonics)
250mV-pk
20mV-pk
Input Impedance
Temperature Coefficient of Input
Impedance
LSB size
Digital Full Scale
ADC Gain Error vs.
%Power Supply Variation
MIN
IC
Vin=65Hz,
64kpts FFT, BlackmanHarris window
Vin=65Hz
Vin=65Hz
Vin=200mV pk, 65Hz
V3P3A=3.0V, 3.6V
-10
© 2005-2008 TERIDIAN Semiconductor Corporation
1.7
Ω/°C
355
+884736
nV/LSB
LSB
50
PPM/
%
+10
mV
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
RECOMMENDED EXTERNAL COMPONENTS
NAME
C1
C2
FROM
V3P3A
V3P3D
TO
AGND
DGND
XTAL
XIN
XOUT
CXS
CXL
C2P5
XIN
XOUT
V2P5
AGND
AGND
DGND
FUNCTION
Bypass capacitor for 3.3V supply
Bypass capacitor for 3.3V supply
32.768kHz crystal – electrically similar to ECS
.327-12.5-17X or Vishay XT26T, load capacitance 12.5pF
Load capacitor for crystal (depends on crystal
specs and board parasitics).
Bypass capacitor for V2P5
VALUE
≥0.1±20%
≥0.1±20%
UNIT
μF
μF
32.768
kHz
27±10%
27±10%
≥0.1±20%
pF
pF
µF
FOOTNOTES:
1 This spec is guaranteed, has been verified in production samples, but is not measured in production.
2 This spec is guaranteed, has been verified in production samples, but is measured in production only at DC.
3 This spec is measured in production at the limits of the specified operating temperature.
4 This spec defines a nominal relationship rather than a measured parameter. Correct circuit operation is verified with other specs that use
this nominal relationship as a reference
VC
V3P3A
GNDA
50
49
VB
52
51
IC
VA
IB
55
53
IA
56
54
VX
VREF
57
VFLT
59
58
XIN
GNDA
61
60
XOUT
GNDD
62
RESERVED
1
63
GNDD
64
PIN CONFIGURATION AND PIN FUNCTION
48
GNDD
RESERVED
TMUXOUT
/RTM
TX
2
47
RESETZ
3
46
V2P5
4
45
VBAT
SSCLK
5
44
RX
CKTEST
6
V3P3D
7
SSDATA
SFR
TERIDIAN
8
43
71M6515H-IGT
9
RESERVED
42
D0
41
IRQZ
40
PULSE_INIT
10
39
D6
RESERVED
11
38
D5
RESERVED
12
37
D4
RESERVED
13
36
PULSE3
14
35
PULSEW
PULSE4
15
34
UARTCSZ
33
D7
18
19
20
21
22
23
24
25
26
27
28
29
30
31
RESERVED
RESERVED
RESERVED
D1
D2
D3
SRDY
MUX_SYNC
RESERVED
GNDD
RESERVED
RESERVED
RESERVED
RESERVED
PULSER
RESERVED
17
16
RESERVED
BAUD_RATE
32
RESERVED
Pins marked RESERVED should be left unconnected during normal use.
Page: 7 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Analog Pin Description
IA,
IB,
IC
VA,
VB,
VC
VFLT
VX
VREF
Pin
No.
56
55
54
53
52
51
59
58
57
XIN,
XOUT
61
63
Name
Type
Circuit
I
6
I
6
I
I
I/O
7
6
9
I
8
Description
Line Current Sense Inputs: Voltage inputs to the internal A/D converter. Typically,
they are connected to the output of a current transformer. The input is referenced
to V3P3A. Unused pins must be tied to V3P3A.
Line Voltage Sense Inputs: Voltage inputs to the internal A/D converter. Typically,
they are connected to the output of a resistor divider. The input is referenced to
V3P3A. Unused pins must be tied to V3P3A.
Power Fault Input. This pin must be tied to V3P3A.
Auxiliary input (not used). This pin should be tied to VREF.
Voltage Reference for the ADC.
Crystal Inputs: A 32768Hz crystal should be connected across these pins.
Typically, a 15pF capacitor is also connected from each pin to GNDA. See the
datasheet of the crystal manufacturer for details.
Pin types: P = Power, O = Output, I = Input, I/O = Input/Output
The circuit number denotes the equivalent circuit, as specified under “I/O Equivalent Circuits”.
Digital Pin Description
Unless otherwise indicated, all inputs and outputs are standard CMOS. Inputs do NOT have internal pull-ups or pull-downs.
Pin
No.
Type
Circuit
CKTEST
6
I/O
4
Clock PLL output. Can be enabled and disabled by CKOUT_DSB (see
Status Mask).
D0
D1
D2
D3
D4
D5
D6
D7
PULSE4
PULSE3
42
21
22
23
37
38
39
33
15
14
I/O
3, 4
Input/output pins 0 through 7. These pins must be terminated to
V3P3D or ground if configured as input pins.
D0 through D7 float after reset or power-up and are configured as
outputs and driven low 140ms after RESETZ goes high.
O
O
4
4
PULSE_INIT
40
I
3
BAUD_RATE
16
I
3
IRQZ
41
O
4
MUXSYNC
25
O
4
RESETZ
47
I
1
Name
Page: 8 of 57
Description
The fourth pulse generator output
The third pulse generator output
The pulse output initial power-up voltage (0: 0V, 1: 3.3V), default is 1.
This pin must be terminated to V3P3D or ground.
The UART baud rate (1: 38.4kbd, 0: 19.2kbd). This pin must be
terminated to V3P3D or ground.
Interrupt output, low active. A falling edge indicates the end of a
measurement frame, as well as alarms. Rises when status word is read.
Internal signal. MUXSYNC falls at the beginning of each conversion
cycle (multiplexer frame).
Chip reset: Input pin with internal pull-up resistor, used to reset the chip
into a known state. For normal operation, this pin is set to 1. To reset
the chip, this pin is driven to 0 for 5 microseconds. No external reset
circuitry is necessary for power-up reset.
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Name
Pin
No.
Type
Circuit
UARTCSZ
34
I
3
SRDY
SFR
SSCLK
SSDATA
24
9
5
8
I
O
O
O
3
4
4
4
RX
44
I
3
TX
TMUXOUT
PULSER
PULSEW
4
3
36
35
O
O
O
O
4
4
4
4
Description
Enables the UART when 0. The UART is disabled when this pin is set
to 1. A positive pulse on this pin will reset the UART. This pin must be
terminated to ground.
High-Speed Synchronous Interface (SSI).
The SRDY input should be tied to ground.
SSI frame pulse output, one SSCLK wide.
SSI clock output (5MHz or 10MHz selectable).
SSI data output, changes on the rising edge of SSCLK.
UART serial Interface receiver input. The voltage at this pin must not
exceed 3.6V. This pin must be terminated to V3P3D or ground.
UART serial Interface transmitter output.
Digital output test multiplexer. Controlled by TMUX[2:0].
Selectable pulse output (default: VARh pulse).
Selectable pulse output (default: Wh pulse).
Power/Ground Pin Description
Name
Pin No.
Type
GNDA
P
Analog ground: This pin should be connected directly to the ground plane.
P
Digital ground: These pins must be connected directly to the ground plane.
V3P3A
49,60
1,27,
48,62
50
P
Analog power: A 3.3V analog power supply should be connected to this pin.
V3P3D
7
P
Digital power supply: A 3.3V digital power supply should be connected to this pin.
VBAT
45
P
V2P5
46
O
GNDD
Description
Battery backup power supply. A battery or super-capacitor is to be connected between
VBAT and GNDD. If no battery is used, connect VBAT to V3P3D.
Output of the 2.5V regulator. A 0.1µF capacitor should be connected from this pin to GND.
Pin types: P = Power, O = Output, I = Input, I/O = Input/Output
The circuit number denotes the equivalent circuit, as specified under “I/O Equivalent Circuits”.
Reserved Pins
Pins labeled RESERVED are not to be connected.
Name
RESERVED
Page: 9 of 57
Pin No.
2,10,11,12,
13,17,18,19,
20,26,28,29,
30,31,32,43,
64
Description
DO NOT CONNECT THESE PINS!
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
I/O Equivalent Circuits
V3P3D
V3P3A
V3P3D
V3P3D
110K
Digital
Input
Pin
CMOS
Input
Analog
Input
Pin
To
MUX
GNDD
from
internal
reference
V2P5
Pin
GNDA
Digital Input Equivalent Circuit
Type 1:
Standard Digital Input or
pin configured as DIO Input
with Internal Pull-Up
GNDD
Analog Input Equivalent Circuit
Type 6:
ADC Input
V2P5 Equivalent Circuit
Type 10:
V2P5
V3P3D
V3P3A
Digital
Input
Pin
CMOS
Input
110K
GNDD
Comparator
Input
Pin
GNDD
Comparator Input Equivalent
Circuit Type 7:
Comparator Input
V3P3D
Power
Down
Circuits
VBAT
Pin
GNDD
GNDA
Digital Input
Type 2:
Pin configured as DIO Input
with Internal Pull-Down
Digital
Input
Pin
To
Comparator
VBAT Equivalent Circuit
Type 12:
VBAT Power
V3P3D
CMOS
Input
Oscillator
Pin
GNDD
To
Oscillator
GNDD
Digital Input Type 3:
Standard Digital Input or
pin configured as DIO Input
Oscillator Equivalent Circuit
Type 8:
Oscillator I/O
V3P3D
V3P3D
V3P3A
Digital
Output
Pin
CMOS
Output
GNDD
GNDD
Digital Output Equivalent Circuit
Type 4:
Standard Digital Output or
pin configured as DIO Output
Page: 10 of 57
from
internal
reference
VREF
Pin
GNDA
VREF Equivalent Circuit
Type 9:
VREF
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
%Error
TYPICAL PERFORMANCE CHARACTERISTICS
0.2
0.15
0.1
0.05
0
-0.05
-0.1
-0.15
-0.2
0 Deg
60 Deg
-60 Deg
180 Deg
1
3
0.3
0.1
200
10
1
25 30
10
100
100
A
1000
Figure 3: Wh Accuracy, 0.3A - 200A/240V
0.2
90 Deg
0.15
150 Deg
% Error
0.1
0.05
3
270 Deg
10
25
1
0
30
0.3
100
-0.05
200
-0.1
-0.15
-0.2
0.1
1
10
100
A
1000
Figure 4: VARh Accuracy for 0.3A to 200A/240V Performance.
2
1
0
Error [%]
-1
-2
-3
50Hz Harmonic Data
60Hz Harmonic Data
-4
-5
-6
-7
-8
1
3
5
7
9
11
13
15
17
19
21
23
25
Harmonic
Measured at current distortion amplitude of 40% and voltage distortion amplitude of 10%.
Figure 5: Meter Accuracy over Harmonics at 240V, 30A
Page: 11 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
y
p
(
Limit
15
Accuracy [PPM/°C]
MARCH 2008
)
Limit
10
typical chip
5
0
-5
-10
-15
-60
-40
-20
0
20
40
60
80
100
Temperature [°C]
Figure 6: Typical Meter Accuracy over Temperature Relative to 25°C (w/ Temperature Compensation)
Performance for Apparent Energy (VAh)
Error [%]
0.2
0.1
0
-0.1
-0.2
0.1
1
10
100
1000
Current [A]
Figure 7: Typical VAh Accuracy for VAh Using Vector Method
Page: 12 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
FUNCTIONAL DESCRIPTION
THEORY OF OPERATION
The 71M6515H integrates the primary functional blocks required to implement a solid-state electricity meter front end. Included
on-chip are an analog front end (AFE), a digital computation engine (CE), a voltage reference, a real time clock, and I/O pins.
Various current sensor technologies are supported including Current Transformers (CT), Resistive Shunts, and Rogowski
(di/dt) Coils.
In a typical application, the 71M6515H sequentially digitizes the voltage inputs on pins IA, VA, IB, VB, IC, VC and performs
calculations to measure active energy (Wh), reactive energy (VARh), and apparent energy (VAh). In addition to these
measurement functions, the real time clock function allows the device to record time of use (TOU) metering information for
multi-rate applications.
The 71M6515H contains a temperature-trimmed ultra-precise voltage reference, and the on-chip digital temperature compensation mechanism includes a temperature sensor and associated controls for correction of unwanted temperature effects on
measurement. RTC accuracy can be greatly improved by supplying correction coefficients derived from crystal
characterization. The combination of both features enables designers to produce electricity meters with exceptional accuracy
over the industrial temperature range.
Meter Equations
The 71M6515H implements the equations in Table 1. Register EQU specifies the equation to be used. In one sample time,
each of the six inputs is converted and the selected equation updated. In a typical application, IA, IB, IC are connected to
current transformers that sense the current on each phase of the line voltage. VA, VB, and VC are typically connected to
voltage sensors (resistor dividers) with respect to NEUTRAL. NEUTRAL is to be connected to V3P3A, the analog supply
voltage. NEUTRAL is the zero reference for all analog measurements.
EQU
Watt & VAR Formula
Application
Channels used from MUX
sequence
Mux State:
0
1
2
3
4
5
0
VA IA
1 element, 2W 1ø
IA
VA
-
-
-
-
1*
VA(IA-IB)/2
1 element, 3W 1ø
IA
VA
IB
-
-
-
2
VA IA + VB IB
2 element, 3W 3 øDelta
IA
VA
IB
VB
-
-
3*
VA (IA - IB)/2 + VC IC
2 element, 4W 3ø Delta
IA
VA
IB
-
IC
VC
4*
VA(IA-IB)/2 + VB(IC-IB)/2)
2 element, 4W 3ø Wye
IA
VA
IB
VB
IC
-
5
VA IA + VB IB + VC IC
3 element, 4W 3ø Wye
IA
VA
IB
VB
IC
VC
Note: Equations 1*, 3*, 4* available only when IMAGE = 00 (CT mode).
Table 1: Meter Equations
Table 2 shows how the elements of the meter are mapped for the six possible equations.
Page: 13 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
EQU
Element Output Mapping
Watt & VAR Formula
(WSUM/VARSUM)
W0SUM/
VAR0SUM
W1SUM/
VAR1SUM
W2SUM/
VAR2SUM
I0SQ
SUM
I1SQ
SUM
I2SQ
SUM
VA*IA
-
-
IA
-
-
VA*(IA-IB)/2
VA*IB
-
IA-IB
IB
-
0
VA IA (1 element, 2W 1φ)
1
VA*(IA-IB)/2
(1 element, 3W 1φ)
2
VA*IA + VB*IB
(2 element, 3W 3φ Delta)
VA*IA
VB*IB
-
IA
IB
-
3
VA*(IA-IB)/2 + VC*IC
(2 element, 4W 3φ Delta)
VA*(IA-IB)/2
-
VC*IC
IA-IB
IB
IC
4
VA*(IA-IB)/2 + VB*(IC-IB)/2
(2 element, 4W 3φ Wye)
VA*(IA-IB)/2
VB*(IC-IB)/2
IA-IB
IC-IB
IC
5
VA*IA + VB*IB + VC*IC
(3 element, 4W 3φ Wye)
VA*IA
VB*IB
IA
IB
IC
VC*IC
Table 2: Meter Element Output Mapping
ANALOG FRONT END
A/D Converter (ADC)
A single delta-sigma A/D converter (ADC) digitizes the inputs to the device. The resolution of the ADC is 21 bits. The ADC
operates at 5MHz oversampling rate and places the digital results in CE memory. Each analog input is sampled at 2520Hz.
Once each accumulation interval, it refreshes the temperature value that is placed in the TEMP_RAW register. The analog reference for all inputs is V3P3A, i.e. the ADC processes voltages between the input pins and V3P3A.
Voltage Reference
The device includes an on-chip precision bandgap voltage reference that incorporates auto-zero techniques as well as
production trims to minimize errors caused by component mismatch and drift. The result is a voltage output with a predictable
temperature coefficient.
The CE compensates for temperature characteristics of the voltage reference by modifying the gain applied to the V and I
channels based on the coefficients PPMC and PPMC2. See the section “TEMPERATURE COMPENSATION” for details.
DIGITAL COMPUTATION
The six ADC outputs are processed and accumulated digitally. The default product summation is based on 42*60 (if the
SUM_CYCLES register is set to 60) samples per accumulation interval. At the end of each accumulation interval, a ready
interrupt (IRQZ) is signaled (if enabled with the READY bit in STMASK), indicating that fresh data is available to the host. For
instance, if SUM_CYCLES =30, the IRQZ rate will be 2Hz (500ms).
A dedicated 32-bit Computation Engine (CE) performs the precision computations necessary to accurately measure energy.
Internal CE calculations include frequency-insensitive offset cancellation on all six channels and a frequency insensitive 90°
phase shifter for VAR calculations. The CE also includes LPF smoothing filters after each product and squaring circuit to
attenuate ripple and eliminate beat frequencies between the power line fundamental and the accumulation time. The CE
directly calculates Watts, VARs, V2, and I2 and accumulates them for one interval.
At the end of each CE computation cycle, the accumulated data are post-processed to calculate RMS amplitudes, phase
angles, and VAh. When post-processing is complete, the IRQZ signal is activated.
The minimum combined cycle time for CE and post-processor is 400ms, which makes the maximum frequency for the IRQZ
signal 2.5Hz.
Page: 14 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
If the 71M6515H is interfacing to an external DSP (typically, but not necessarily through the SSI interface), the host may turn
off post-processing by setting the CE_ONLY bit in the CONFIG word. This will permit setting SUM_CYCLES below its
recommended lower limit of 24. SUM_CYCLES may then be reduced to 1, creating an accumulation interval of only 42 samples.
The outputs available in CE only mode are limited to temperature, frequency, voltage phases, input signal zero crossings, plus
WSUM and VARSUM for each phase and VSQSUM, ISQSUM, and ISQFRACT for each phase.
Pulse Generators
The chip contains four pulse generators connected to the pins PULSEW, PULSER, PULSE3, and PULSE4 that create low jitter
pulses from 32-bit data. The peak time jitter for PULSEW and PULSER is the 397µs MUX frame period, and is independent of
the rate of the generator or the length of time the generator is monitored. Thus, if the pulse generator is monitored for 1
second, the peak jitter is 400PPM. After 10 seconds, the peak jitter is 40PPM.
PULSE3 and PULSE4 are updated at a slower rate and have four times higher jitter, i.e. 160PPM after 10 seconds.
The average jitter is always zero. If it is attempted to drive either pulse generator faster than its maximum rate, it will simply
output at its maximum rate without exhibiting any roll-over characteristics.
Pulse generator inputs may be from three sources:
•
Internal (directly from the CE), PULSEW and PULSER only
•
External (controlled by the host writing to registers APULSEW, APULSER, APULSE3, APULSE4)
•
Post-processed values
The source is selected individually for each pulse output with the PULSEW_SRC, PULSER_SRC, PULSE3_SRC, and PULSE4_SRC
registers. Figure 8 shows internal pulse generation for the PULSEW output selected by writing the value 35 into the
PULSEW_SRC register.
35: WSUM
CE
2: WBSUM
3: WCSUM
35
PULSEW_SRC
4: VARSUM
POST PROCESSOR
0: WSUM
1: WASUM
PULSEW
OUTPUT
34: VAR2SUM_E
36: APULSEW
HOST
Figure 8: Internal Pulse Generation Selected in the PULSEW_SRC Register
Internal data is pulsed out during the accumulation interval immediately following its accumulation interval. Post-processed
values are pulsed out one accumulation interval after that.
Page: 15 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
The pulse generator output rate depends on its input value, WRATE, PULSE_SLOW, and PULSE_FAST. Additionally, its
maximum pulse width (negative going pulse) is controlled by PULSEWIDTH. High frequency pulses will have 50% duty cycle
until their rate slows enough that their pulse width is limited by PULSEWIDTH.
In internal and post-processed modes, the pulse rate, expressed as Kh (Wh per pulse) is given by the formula:
Kh =
VMAX IMAX
1.5757 Wh / Pulse
In _ 8 ⋅ SUM _ CYCLES ⋅ WRATE ⋅ X
where
VMAX is the meter voltage corresponding to an input voltage of 176mV (rms) at the VA, VB, and VC input pins ,
IMAX is the meter current corresponding to an input voltage of 176mV (rms) at the IA, IB, and IC input pins,
In_8 is the additional ADC gain (1 or 8), as controlled by the IA_X, IB_X and IC_X bits in the CONFIG register.
X is the pulse speed factor determined from Table 3.
PULSE_SLOW
0
0
1
1 (default)
PULSE_FAST
0
1
0
1 (default)
X
1.5*22=6
1.5*26=96
1.5*2-4=0.09375
1.5
Table 3: Pulse Speed Factor X
In external pulse mode, the pulse rate is given by the formula:
Rate(Hz) = WRATE * X * input * 35.82*10-12,
where input is the value in registers APULSER, APULSEW. APULSE3 or APULSE4,
X is the pulse speed factor determined from Table 3.
External pulse generation can be seen as providing the raw voltage and current readings equivalent to Vin*Iin / LSB directly to
the pulse generator.
The maximum pulse rate is 7.56kHz for PULSEW and PULSER, and 150Hz for PULSE3 and PULSE4.
In external pulse mode, the pulse generators load their data at the beginning of each CE accumulation interval, preserving any
partially implemented pulses from the previous interval. The source of data is controlled by the entries in the PULSE_SRCS
register. PULSER_SRCS contains 8-bit entries for each pulse source, PULSEW, PULSER, PULSE3, and PULSE4. See the
register description for details.
The procedure for accurate external pulse generation controlled by the host is:
1)
Respond to a READY interrupt by reading the accumulated values.
2)
Process the accumulated values.
3)
Write the processed value(s) to APULSER, APULSEW, APULSE3, or APULSE4. The host must write to APULSER,
APULSEW, APULSE3, and APULSE4 before the next READY interrupt for the pulse generation to be beginning in the
following accumulation interval.
Figure 9 illustrates pulse generator timing.
Regardless of the source, the pulse generators should receive new data during each accumulation interval. If this does not
occur and if the corresponding bit in the STMASK register is set, an APULSE_ERR interrupt will be issued.
The PULSEW, PULSER, PULSE3 and PULSE4 pins are suitable for driving LEDs through a current limiting resistor. The LED
should be connected so it is on when the pulse pin is low.
The pin PULSE_INIT determines the logic level applied to the pulse pins on power-up, i.e. with PULSE_INIT low, the pulse
pins will be initialized to low (default = 1).
Page: 16 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
The pulse width PW is controlled with the PULSEWIDTH register for the PULSER and PULSEW output pins per the following
formula:
PW =
2 ⋅ PULSEWIDTH + 1
2520.6
The PULSE3 and PULSE4 output pins will always generate pulses with 50% duty cycle.
Internal Data (Directly by CE)
Accumulation Interval
CE Operations
Accumulation 1
XFER
Post Processing
Post 0
Pulse Generator
Accumulation 2
XFER
Accumulation 3
XFER
Post 1
Post 2
READY
READY
READY
Pulse 0
Pulse 1
Pulse 2
Post-Processed Data
Accumulation Interval
CE Operations
Accumulation 1
XFER
Post Processing
Accumulation 2
XFER
Post 0
Post 1
READY
Pulse Generator
Accumulation 3
XFER
APULSE write
Pulse -1
Post 2
READY
READY
APULSE write
Pulse 0
APULSE write
Pulse 1
External (Host data is transferred to the pulse generator in the first accumulation interval after the
next READY)
Accumulation Interval
CE Operations
Accumulation 1
XFER
Post Processing
Post 0
Pulse Generator
Accumulation 3
XFER
Post 1
READY
Host Processing
Accumulation 2
XFER
Host 0
APULSE write
Pulse -2
Post 2
READY
Host 1
APULSE write
Pulse -1
READY
Host 2
APULSE write
Pulse 0
Figure 9: Pulse Generator Timing
Page: 17 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Internal Resources
Oscillator
The oscillator drives a standard 32.768kHz watch crystal. Crystals of this type are accurate and do not require a high current
oscillator circuit. The 71M6515H oscillator has been designed specifically to handle watch crystals and is compatible with their
high impedance and limited power handling capability. The oscillator power dissipation is very low to maximize the lifetime of
any battery backup device attached to VBAT. Using PLL techniques, all internal clocks, such as the 4.915MHz clock for the
ADC and the post-processor, are derived from the watch crystal frequency.
Real-Time Clock (RTC)
The RTC is driven directly by the crystal oscillator. In the absence of V3P3, it is powered by the battery-backed up supply. The
RTC consists of a counter chain and output registers. The counter chain consists of registers for seconds, minutes, hours, day
of week, day of month, month, and year. The nominal quadratic temperature coefficient of the crystal is automatically
compensated in the RTC. The RTC is capable of processing leap years.
I/O Peripherals
The 71M6515H includes several I/O peripheral functions that improve the functionality of the device and reduce the component
count for most meter applications. The I/O peripherals include a UART and digital I/O.
Digital I/O
The device includes eight pins of general purpose digital I/O (D0…D7). Each pin can be configured independently as an input
or output with the D_DIR bits. Inputs are standard CMOS with no pull-ups or pull-downs. Outputs are standard CMOS. The
DIO pins are controlled by the D_CONFIG register.
Immediately after reset or power-up, D0 through D7 are in tri-state mode (floating). 140 ms after reset, D0 through D7
are configured as outputs and driven low.
UART Host Interface
The UART is a dedicated 2-wire serial interface, which can communicate with the host processor. The operation of each pin is
as follows:
RX: Is the pin accepting the serial input data. It inputs data to internal registers. The bytes are input LSB first. The voltage
applied to this pin must be restricted to 0 to 3.6V.
TX: Is the pin used for serial output data. It outputs the contents of a block of internal registers. The bytes are output LSB
first.
BAUD_RATE: The baud rate can be selected with the BAUD_RATE pin (38.4bps when high, 19.2bps when low).
UARTCSZ: This pin enables the UART when low. The UART can be reset by taking UARTCSZ briefly to the high state and
then low again.
The 71M6515H has several on-chip registers, which can be read and written. All transfers start with a stream of 8-bit bytes
(LSB first) from the host on the RX input, followed by a (possibly null) stream of 8-bit bytes (LSB first) to the host on the TX
output (see Figure 10 and Figure 11). The UART is configured as 8N1 (8 bits, no parity, 1 stop bit).
If the READY bit in STMASK is enabled, the IRQZ pin can be used to signal data availability to the host. If data read cycles
exceeding 1 second are used, care should be taken to prevent data overflow.
UART Write Register Operation
The registers are written by sending a byte, consisting of a starting register address in the seven MSBs and ‘0’ in the LSB
indicating this is a write operation. It is followed by a one byte length of bytes to write. If more bytes arrive than fit in the
addressed register, subsequent registers will be written. The bytes are processed in “big-endian” order (i.e. most significant
byte first). See Figure 10 (read bits and bytes from left to right).
Page: 18 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
L
W S
B
M
S
B
L
S
B
M
S
B
L
S
B
M
S
B
L
S
B
M
S
B
RX
register address
most significant
data byte
length
least significant
data byte
time
TX
Figure 10: UART Write Operation
UART Read Register Operation
The registers are read by sending a byte, consisting of a start register address in the seven MSBs and ‘1’ in the LSB indicating
this is a read operation. It is followed by a one byte length of bytes to read. If more bytes are asked for than the size of the
addressed register, subsequent registers will be read. The bytes are in “big-endian” order (i.e. most significant byte first). See
Figure 11.
L
R S
B
M
S
B
L
S
B
M
S
B
RX
register address
length
time
L
S
B
TX
M
S
B
most significant
data byte
L
S
B
M
S
B
least significant
data byte
Figure 11: UART Read Operation
Note: In both register read and write operations, the register address can be 0 through 127 (0x7F). The register address byte is
obtained by left-shifting the register address by one bit and setting bit 0 to 1 for read or setting bit 0 to 0 for write.
Synchronous Serial Interface (SSI)
A high speed, handshake, serial interface is available to send a contiguous block of CE data to an external data logger or DSP.
The block of data, configurable as to location and size, is sent at the beginning of each ADC multiplex cycle. The SSI interface
is enabled by the SSI_EN bit and consists of the outputs SSCLK, SSDATA, and SFR and of the SRDY input pin. The interface
is compatible with 16-bit and 32-bit processors. The operation of each pin is as follows:
SSCLK: This pin provides the serial clock. The clock can be 5MHz or 10MHz, as specified by the SSI_10M bit. The
SSI_CKGATE bit controls whether SSCLK runs continuously or is gated off when no SSI activity is occurring. If SSCLK is
gated, it will begin three cycles before SFR rises and will persist three cycles after the last data bit is output.
SSDATA: This pin provides the serial output data. SSDATA changes on the rising edge of SSCLK and outputs the contents
of a block of CE words starting with address SSI_STRT and ending with SSI_END. The words are output MSB first. SSDATA is
stable with the falling edge of SSCLK.
SFR: This pin provides the framing pulse. Although CE words are always 32 bits, the SSI interface will frame the entire data
block as a single field, as multiple 16 bit fields, or as multiple 32 bit fields. The SFR pulse is one clock cycle wide, changes
state on the rising edge of SSCLK and precedes the first bit of each field. The field size is set with SSI_FSIZE: 0-entire data
block, 1-8 bit fields, 2-16 bit fields, 3-32 bit fields. The polarity of the SFR pulse can be inverted with SSI_FPOL. The first SFR
pulse in a frame will rise on the third SSCLK clock period after MUX_SYNC (fourth SSCLK period, if SSCLK is 10MHz).
MUX_SYNC can be used to synchronize the fields arriving at the data logger or DSP.
SRDY: The SRDY input should always be tied to GND.
Page: 19 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
The SSI timing is shown in Figure 12.
If 16bit fields
If SSI_CKGATE =1
If 32bit fields
If SSI_CKGATE =1
SFR (Output)
SCLK (Output)
SSDATA (Output)
31
30
16
15
1
0
31
30
SSI_BEG
16
15
1
SSI_BEG+1
0
31
1
0
SSI_END
MUX_STATE
Figure 12: SSI Timing (SSI_FPOL = SSI_RDYPOL = 0)
Fault and Reset Behavior
Reset Mode
When RESETZ is pulled low or when VFLT < V3P3/2, all activity (i.e. sampling of analog signals, CE, generation of digital
outputs) in the chip stops while the analog circuits are active. The exceptions are the oscillator and RTC module, which
continue to run. Additionally, all I/O Register bits are cleared. As long as VFLT > V3P3/2, the internal 2.5V regulator will
continue to provide power to the digital section.
Once initiated, the reset mode will persist until the reset timer times out. This will occur in 4100 cycles of the real time clock
after RESETZ goes high, at which time the 71M6515H will begin executing its preboot and boot sequences.
Power Fault Circuit
The power fault comparator compares the voltage at the VFLT pin to V3P3/2. The comparator output internally enables the
battery backup protection for oscillator, RTC and RAM during the power fail mode.
Temperature Compensation
Voltage Reference
The internal voltage reference of the 71M6515H is calibrated at 25°C during device manufacture. The 71M6515H is given
additional temperature-related calibrations which further compensate its ADC gain and allow it to achieve 10PPM/°C over
±60°C temperature range.
Temperature Sensor
The device includes an on-chip temperature sensor for determining the temperature of the bandgap reference. The primary
use of the temperature data is to determine the magnitude of compensation required to offset thermal drift in the system. The
temperature sensor is read once per accumulation interval.
Temperature measurement can be implemented with the following steps:
1)
At a known temperature TN, read the TEMP_RAW register and write the value into TEMP_NOM register.
2)
Read the DELTA_T register at the known temperature. The obtained value should be <±0.1°C.
3)
The temperature T (in °C) at any environment can be obtained by reading the DELTA_T register and applying the
following formula:
T = TN +
DELTA _ T
10
Temperature Compensation for Energy Measurements
TEMP_NOM is one of the calibration parameters that must be loaded by the host in order to enable temperature measurement
and thereby temperature compensation.
Page: 20 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
PPMC and PPMC2, the linear and quadratic compensation coefficients, compensate for temperature drift in the 71M6515H
reference that affects the meter performance. PPMC and PPMC2 describe how the 71M6515H calculations are to respond to
temperature. This means they should be the negative of the meter behavior before compensation. PPMC and PPMC2 are
scaled from PPM/°C and PPM/°C2 values. See the register description for details. Temperature compensation can be
selected to operate in one of two modes shown in the table below:
DEFAULT_PPM Bit in
CONFIG Register
PPMC, PPMC2 Calculation
Internal (CE)
1
By post-processor, based on
stored VREF characteristics
External (host)
0
By host
Temperature Compensation Mode
When the part is first powered up, TEMP_NOM, PPMC, and PPMC2 are zero. When the host writes its calibration value into
TEMP_NOM (after setting the DEFAULT_PPM bit on the CONFIG register to 1), PPMC and PPMC2 will automatically be
initialized to the values that best compensate for the temperature drift of the internal reference. These parameters will be
individually customized for 71M6515H parts. If, for some reason, the host writes to TEMP_NOM again, PPMC and PPMC2 will
not be changed since they will no longer be zero.
If TEMP_NOM is not loaded by the host, PPMC and PPMC2 are ignored, and their values are permanently held at zero.
If TEMP_NOM is zero, no temperature compensation occurs, even if PPMC and PPMC2 are loaded.
If the host wishes to provide its own compensation, it should read PPMC and PPMC2 and modify them by merging the
additional compensation into to them. In that case, the DEFAULT_PPM bit in the CONFIG register must be zero.
Temperature Compensation for the Crystal and RTC
The crystal oscillator contributes negligible error to energy calculations. However, sometimes specifications for the real time
clock (RTC) require better accuracy than that provided by the untrimmed watch crystal. The 71M6515H therefore allows
calibration of the RTC clock. Calibration requires that frequency tolerance and frequency stability either be obtained from the
manufacturer or be independently measured (the RTC clock is available on the TMUX pin). Calibration does not change the
frequency of the RTC clock, but rather increments or decrements the clock by one second when sufficient error has
accumulated. Positive correction makes the clock run faster.
The formula for the RTC correction factor is as follows:
CORRECTION [PPM] =
Where
Y _ CALC 0 Y _ CALC1 ΔT
Y _ CALC 2 ΔT 2
+
+
10
100
1000
Y_CALC0 = 10 * crystal frequency deviation from ideal (measured)
Y_CALC1 = 100 * crystal skew (nominally zero)
Y_CALC2 = 1000 * crystal frequency stability (specified)
ΔT = T - Calibration Temperature in °C
Calibration
Calibration Factors for CT and Resistive Shunt
Once installed in a meter, the TERIDIAN 71M6515H IC has to be calibrated for the tolerances of current sensors, voltage
dividers and signal conditioning components. The room temperature reading of its temperature sensor must also be entered.
These calibration factors must be stored by the host and, upon power up, loaded into the TERIDIAN 71M6515H. Typical
calibration constants are listed in Table 4.
Page: 21 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Name
CAL_IA
CAL_VA
CAL_IB
CAL_VB
CAL_IC
CAL_VC
TEMP_NOM
PHADJ_A
PHADJ_B
PHADJ_C
Description
Gain factors for current and voltage of each phase.
The value of TEMP_RAW at nominal temperature.
Phase compensation for each current. If phase compensation is 0 or if
current sensors have predictable phase, PHADJ may not need to be
measured on every meter.
Table 4: Typical Calibration Parameters (CT)
Gain adjustment (CAL_Xn parameters) is used to compensate for tolerances of components used for signal conditioning,
especially the resistive components. A 1% increase in CAL_Xn will cause a 1% increase in the channel gain.
The phase compensation circuit in the TERIDIAN 71M6515H is optimized for operation with current transformers (CT’s). These
devices have a low frequency pole and therefore have a slight amount of phase lead at 50 or 60Hz—more at 50Hz than at
60Hz. The phase lead diminishes at higher harmonics. When PHADJ_n is calibrated as shown below at either 50Hz or 60Hz,
the CT will be correctly compensated from below 25Hz to beyond 1100Hz.
This phase compensator is markedly superior to the more common technique of programming a time delay to compensate for
CT phase. The time delay technique results in phase compensation that is correct at only one frequency, and actually amplifies
the phase error at harmonics of the frequency.
Calibration Factors for Rogowski Coil Sensors
If IMAGE is set to 01, i.e. the 71M6515H can be operated with Rogowski Coil sensors. In this case, one more calibration factor
per phase is needed. The PHADJ parameters have non-zero defaults and do not obey the same formula used for CT
calibration. The feedthrough parameter has to be determined by a separate crosstalk measurement. Table 5 shows the
parameters involved in the calibration procedure for the Rogowski sensor.
Name
CAL_IA
CAL_VA
CAL_IB
CAL_VB
CAL_IC
CAL_VC
TEMP_NOM
PHADJ_A
PHADJ_B
PHADJ_C
VFEED_A
VFEED_B
VFEED_C
Description
Gain constants for current and voltage of each phase.
The value of TEMP_RAW at nominal temperature.
Phase compensation for each current. If phase compensation is 0 or if
current sensors have predictable phase, PHADJ may not need to be
measured on every meter.
Feedthrough compensation for each current.
Table 5: Typical Calibration Parameters (Rogowski)
Page: 22 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
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
(inductive)
Positive
direction
+60°
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 13).
Current
-60°
Current leads
voltage
(capacitive)
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 observable 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
Voltage
Generating Energy
Using Energy
= 1.875Hz will be established.
Figure 13: 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 14. 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 14 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
I
φL
φ L is phase lag
ERRORS
METER
−φS
IRMS
A XI
Π
V
IDEAL = I ,
φS is phase lead
W
V RMS
AXV
ERROR ≡
ACTUAL = I AXI
IDEAL = IV cos(φ L )
ACTUAL = IV AXI AXV cos(φ L − φ S )
IDEAL = V ,
ACTUAL = V AXV
ACTUAL − IDEAL ACTUAL
−1
=
IDEAL
IDEAL
Figure 14: 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: 23 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Calibration with Three Measurements
The simplest calibration method is to make three measurements. Typically, a voltage measurement and two Watt-hour (Wh)
measurements are made.
If the voltage measurement has the error EV and the two Wh measurements have errors E0 and E60, where E0 is measured
with φL = 0 and E60 is measured with φL = 60. These values should be simple ratios—not percentage values. They should be
zero when the meter is accurate and negative when the meter runs slow. The fundamental frequency is f0. T is equal to 1/fS,
where fS is the sample frequency (2520.62Hz). Set all calibration factors to nominal: CAL_IA = 16384, CAL_VA = 16384,
PHADJ_A = 0.
From the voltage measurement, we determine that
1.Î
AXV = EV + 1
We use the other two measurements to determine φS and AXI.
IV AXV AXI cos(0 − φ S )
− 1 = AXV AXI cos(φ S ) − 1
IV cos(0)
2.
E0 =
2a.
AXV AXI =
3.
E 60 =
IV AXV AXI cos(60 − φ S )
cos(60 − φ S )
− 1 = AXV AXI
−1
IV cos(60)
cos(60)
3a.
E 60 =
AXV AXI [cos(60) cos(φ S ) + sin(60) sin(φ S )]
−1
cos(60)
E0 + 1
cos(φ S )
= AXV AXI cos(φ S ) + AXV AXI tan(60) sin(φ S ) − 1
Combining 2a and 3a:
4.
E 60 = E 0 + ( E 0 + 1) tan(60) tan(φ S )
5.
tan(φ S ) =
6.Î
φ S = tan −1 ⎜⎜
E 60 − E 0
( E 0 + 1) tan(60)
⎛
⎞
E 60 − E 0
⎟⎟
⎝ ( E 0 + 1) tan(60) ⎠
and from 2a:
7.Î
AXI =
E0 + 1
AXV cos(φ S )
Now that we know the AXV, AXI, and φS errors, we calculate the new calibration voltage gain coefficient from the previous ones:
CAL _ V NEW =
Page: 24 of 57
CAL _ V
AXV
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
We calculate PHADJ from φS, the desired phase lag:
[
]
⎤
⎡
tan(φ S ) 1 + (1 − 2 −9 ) 2 − 2(1 − 2 −9 ) cos(2πf 0T )
PHADJ = 2 ⎢
⎥
−9
−9
⎣ (1 − 2 ) sin(2πf 0T ) − tan(φ S ) 1 − (1 − 2 ) cos(2πf 0T ) ⎦
20
[
]
Finally, we calculate the new calibration current gain coefficient, including compensation for a slight gain increase in the phase
calibration circuit.
CAL _ I NEW =
CAL _ I
AXI
1
1+
2
− 20
PHADJ (2 + 2 PHADJ − 2(1 − 2 −9 ) cos(2πf 0T ))
1 − 2(1 − 2 −9 ) cos(2πf 0T ) + (1 − 2 −9 ) 2
− 20
Calibration with Five Measurements
The five measurement method provides more orthogonality between the gain and phase error derivations. This method
involves measuring EV, E0, E180, E60, and E300. Again, set all calibration factors to nominal, i.e. CAL_IA = 16384, CAL_VA =
16384, PHADJ_A = 0..
First, calculate AXV from EV:
1.Î
AXV = EV + 1
Calculate AXI from E0 and E180:
IV AXV AXI cos(0 − φ S )
− 1 = AXV AXI cos(φ S ) − 1
IV cos(0)
2.
E0 =
3.
E180 =
4.
E 0 + E180 = 2 AXV AXI cos(φ S ) − 2
5.
AXV AXI =
6.Î
AXI =
IV AXV AXI cos(180 − φ S )
− 1 = AXV AXI cos(φ S ) − 1
IV cos(180)
E 0 + E180 + 2
2 cos(φ S )
( E 0 + E180 ) 2 + 1
AXV cos(φ S )
Use above results along with E60 and E300 to calculate φS.
7.
E 60 =
IV AXV AXI cos(60 − φ S )
−1
IV cos(60)
= AXV AXI cos(φ S ) + AXV AXI tan(60) sin(φ S ) − 1
8.
E 300 =
IV AXV AXI cos( −60 − φ S )
−1
IV cos( −60)
= AXV AXI cos(φ S ) − AXV AXI tan(60) sin(φ S ) − 1
Page: 25 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Subtract 8 from 7:
9.
E 60 − E300 = 2 AXV AXI tan(60) sin(φ S )
use equation 5:
E 0 + E180 + 2
tan(60) sin(φ S )
cos(φ S )
10.
E 60 − E300 =
11.
E 60 − E300 = ( E 0 + E180 + 2) tan(60) tan(φ S )
12.Î
φ S = tan −1 ⎜⎜
⎛
⎞
( E 60 − E300 )
⎟⎟
⎝ tan(60)( E 0 + E180 + 2) ⎠
Now that we know the AXV, AXI, and φS errors, we calculate the new calibration voltage gain coefficient from the previous ones:
CAL _ V NEW =
CAL _ V
AXV
We calculate PHADJ from φS, the desired phase lag:
[
]
⎡
⎤
tan(φ S ) 1 + (1 − 2 −9 ) 2 − 2(1 − 2 −9 ) cos(2πf 0T )
PHADJ = 2 20 ⎢
⎥
−9
−9
⎣ (1 − 2 ) sin(2πf 0T ) − tan(φ S ) 1 − (1 − 2 ) cos(2πf 0T ) ⎦
[
]
Finally, we calculate the new calibration current gain coefficient, including compensation for a slight gain increase in the phase
calibration circuit.
CAL _ I NEW =
CAL _ I
AXI
1
1+
2
− 20
PHADJ (2 + 2 PHADJ − 2(1 − 2 −9 ) cos(2πf 0T ))
1 − 2(1 − 2 −9 ) cos(2πf 0T ) + (1 − 2 −9 ) 2
− 20
Alternative Calibration Procedures
It is possible to implement a fast calibration based on only one measurement with a zero-degree load angle. Details can be
found in the TERIDIAN Application Note AN_651X_022 (Calibration Procedures).
Calibration Procedure for Rogowski Sensor
Rogowski coils generate an output signal that is the derivative of the input current. The 6515H Rogowski module implemented
in the Rogowski CE image digitally compensates for this effect and has the usual gain and phase calibration adjustments.
Additionally, calibration adjustments are provided to eliminate voltage coupling from the sensor input.
Current sensors built from Rogowski coils have relatively high output impedances that are susceptible to capacitive coupling
from the large voltages present in the meter. The most dominant coupling is usually capacitance between the primary of the
coil and the coil’s output. This coupling adds a component proportional to the derivative of voltage to the sensor output. This
effect is compensated by the voltage coupling calibration coefficients.
As with the CT procedure, the calibration procedure for Rogowski sensors uses the meter’s display to calibrate the voltage
path and the pulse outputs to perform the remaining energy calibrations. The calibration procedure must be performed to each
phase separately, making sure that the pulse generator is driven by the accumulated real energy for just that phase. In other
words, the pulse generator input should be set to WhA, WhB, or WhC, depending on the phase being calibrated. The IC has to
be configured for Rogowski mode (IMAGE=01). In preparation of the calibration, all calibration parameters are set to their
default values. VMAX and IMAX are set to reflect the system design parameters. WRATE and PULSE_SLOW, PULSE_FAST are
adjusted to obtain the desired Kh.
For details on calibrating a meter for Rogowski coil sensors, see the TERIDIAN Application Note AN_6515_036.
Page: 26 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Meter Design - Scaling of Measured Values
An actual meter will always use sensors that scale the voltages and currents managed by the meter to small voltages that can
be processed by the 71M6515H. This scaling is reflected in the system parameters VMAX and IMAX. Scaling is physically
implemented with resistor dividers for the voltage signals and current transformers, shunt resistors or Rogowski coils for the
current signals.
IMAX is the RMS meter current that results in 250mV peak signal (or 177mV RMS) at the ADC input (IA, IB, IC pins). VMAX is
the RMS meter voltage that results in 250mV peak signal at the ADC input (VA, VB, VC pins). In_8 is either 1 or 8, depending
on In_8x, the ADC gain configuration bit for element n (see IA_8, IB_8, IC_8) in the CONFIG register.
Only the host is aware of the system parameters VMAX and IMAX, while the CE and the post-processor know signal
amplitudes only as values relative to their maximum peak levels (250mV). This makes the host itself responsible for translating
the measured values from the 71M6515H registers into real-world values by applying the parameters VMAX, IMAX and In_8.
Equally, the host is responsible for non-volatile storage of accumulated energy values, calibration factors, default settings et
cetera.
Measured values and values determining the function of the 71M6515H, as controlled by the registers described in the
following section, are often stated as fractions or multiples of the system parameters VMAX, IMAX and In_8.
Host Interface - REGISTER DESCRIPTION
Communication between the host and the 71M6515H is established by writing to and reading from the registers described in
this section. The registers are accessible via the UART (see UART Write and Read Operation).
The tables below contain the registers that can be accessed by the host to obtain data from the 71M6515H or to control and
configure the IC.
Bits with a W (write) direction are written by the host. Bits with R (read) direction can only be read by the host. Write operations
attempted to read-only registers will result in the CMD_IGNORED bit set in the STATUS register. Unless stated otherwise, all
registers are four bytes (32 bits), 2’s complement, and have a range of (231-1) to -(231-1).
Register Groups
Each register belongs to one of the following functional groups:
•
Pulse Generation
•
Calibration
•
Control of Basic Functions
•
Temperature
•
Temperature Compensation
•
Output Signals
•
Accumulated Energy and V/I Values
•
Alarms and Thresholds
•
Time (RTC)
•
Test
•
Digital I/O Control (pins D0…D7)
Page: 27 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Registers in Alphabetical Order
Address
(hex)
R/W
Default
APULSEW
APULSER
APULSE3
APULSE4
0x30
0x31
0x32
0x33
R/W
R/W
R/W
R/W
0
0
0
0
CAL_IA
CAL_VA
0x24
0x25
R/W
R/W
214
214
Calibration
CAL_IB
CAL_VB
0x26
0x27
R/W
R/W
214
214
Calibration
CAL_IC
CAL_VC
0x28
0x29
R/W
R/W
214
214
Calibration
CE_DATA
0x63
R/W
N/A
Control of Basic Functions
CE_DATA_ADDR
0x61
R/W
N/A
Control of Basic Functions
CE_DATA_INC
0x65
R/W
N/A
Control of Basic Functions
CE_PROG
0x62
R/W
N/A
Control of Basic Functions
CE_PROG_ADDR
0x60
R/W
N/A
Control of Basic Functions
CE_PROG_INC
0x64
R/W
N/A
Control of Basic Functions
Control of Basic Functions
Name
Group
Comment
Pulse Generation Control
CONFIG
0x16
R/W
0
CREEP_THRSLD
0x1D
R/W
6000
Alarms and Thresholds
DEG_SCALE
0x1C
R/W
22721
Temperature
D_CONFIG
0x1A
R/W
15
FREQ_DELTA_T
0x11
R
N/A
I/O Control
Outputs, Temperature
Temperature
Compensation
GAIN_ADJ
0x4E
R
16384
IASQFRACT
IBSQFRACT
ICSQFRACT
0x4A
0x4B
0x4C
R
R
R
N/A
N/A
N/A
Outputs
IASQSUM
IBSQSUM
ICSQSUM
0x39
0x3A
0x3B
R
R
R
N/A
N/A
N/A
Outputs
INSQFRACT
0x4D
R
N/A
Outputs
INSQSUM
0x3C
R
N/A
Outputs
IPHASE_ABC
0x0F
R
N/A
Outputs
Uses the postprocessor
IRMS_A
IRMS_B
IRMS_C
0x0C
0x0D
0x0E
R
R
R
N/A
N/A
N/A
Outputs
Uses the postprocessor
KVAR
0x2F
R
6444
Calibration
Do not change
MAIN_EDGE_ COUNT
0x35
R
N/A
OP_TIME
0x1E
R/W
0
Page: 28 of 57
Outputs
Time
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Address
(hex)
R/W
Default
PHADJ_A
PHADJ_B
PHADJ_C
0x2A
0x2B
0x2C
R/W
R/W
R/W
CT: 0
CT: 0
CT: 0
PPMC1_2
0x1B
R/W
0
PULSE3_4_ CNTS
0x42
R
N/A
PULSE_SRCS
0x43
R/W
PULSEW_R_ CNTS
0x41
R
N/A
Outputs
PULSE_WIDTH
0x34
R/W
50
Pulse Generation Control
QUANT_W
QUANT_VAR
QUANT_I
0x36
0x37
0x38
R/W
R/W
R/W
0
0
0
Calibration
Name
Group
Comment
Calibration
Defaults are –3973
for Rogowski
operation
Temperature
Compensation
Outputs
Pulse Generation Control
RTC_DATE
0x20
R/W
N/A
Time
RTC_TIME_DAY
0x1F
R/W
N/A
Time
RTM
0x21
R/W
0
Test
SAG
0x2E
R/W
80,
26000
SSI
0x22
R/W
0
STATUS
0x14
R
N/A
Control of Basic Functions
STMASK
0x15
R/W
0
Control of Basic Functions
TEMP_NOM
0x13
R/W
0
Outputs
TEMP_RAW
0x12
R
N/A
Outputs
VASQSUM
VBSQSUM
VCSQSUM
0x3D
0x3E
0x3F
R
R
R
N/A
N/A
N/A
Outputs
VAH_A
VAH_B
VAH_C
0x06
0x07
0x08
R
R
R
N/A
N/A
N/A
Outputs
VARH_A
VARH_B
VARH_C
0x03
0x04
0x05
R
R
R
N/A
N/A
N/A
Outputs
VFEED_A
VFEED_B
VFEED_C
0x44
0x45
0x46
R/W
R/W
R/W
0
0
0
VI_PTHRESH
0x17
W
21000
Alarms and Thresholds
VI_THRESH
0x40
W
21000
Alarms and Thresholds
Page: 29 of 57
Alarms and Thresholds
Test
Uses the postprocessor
Calibration
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Address
(hex)
R/W
Default
Group
VPHASE_ABC
0x10
R
N/A
Outputs
VRMS_A
VRMS_B
VRMS_C
0x09
0x0A
0x0B
R
R
R
N/A
N/A
N/A
Outputs
WH_A
WH_B
WH_C
0x00
0x01
0x02
R
R
R
N/A
N/A
N/A
Outputs
WRATE
0x2D
R/W
683
Name
Uses the postprocessor
Pulse Generation Control
Y_DEG0
0x18
R/W
0
Temperature
Compensation
Y_DEG1_2
0x19
R/W
0
Temperature
Compensation
Page: 30 of 57
Comment
© 2005-2008 TERIDIAN Semiconductor Corporation
Holds Y_CALC0
Holds Y_CALC1 and
Y_CALC2
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Individual Register Descriptions
Registers for Pulse Generation Control
APULSEW (0x30), APULSER (0x31), APULSE3 (0x32), APULSE4 (0x33),
Figure 15 shows the registers that control the PULSEW, PULSER, PULSE3 and PULSE4 output pins if the corresponding
PULSE_SRCS register contains the decimal value 36. This figure uses the PULSER_SRC 8-bit portion of the PULSE_SRCS
register as an example: The internal pulse generation (CE) and the post-processor pulse generation are deselected, and
the APULSER register acts as the pulse generation source. In this setting (external pulse generation), the host is
responsible for updating the data in the APULSER register.
35: WSUM
CE
0: WSUM
POST PROCESSOR
1: WASUM
2: WBSUM
3: WCSUM
36
PULSER_SRC
4: VARSUM
PULSER
OUTPUT
34:
34: VAR2SUM_E
VAR2SUM_E
36: APULSER
HOST
Figure 15: Pulse Generation via APULSER Selected in the PULSER_SRC Register
PULSE_SRCS (0x43)
This register contains the pulse source selectors for the pulses generated by the PULSEW, PULSER, PULSE3 and
PULSE4 output pins. Pulse sources can be selected individually for internal (CE), external (post-processor) or external
(supplied by the host). The internal selection is valid for the PULSER and PULSEW generators only. The allocation of the
bytes in PULSE_SRCS is as follows:
31
24
23
16
15
8
7
0
PULSEW_SRC
PULSER_SRC
PULSE3_SRC
PULSE4_SRC
Source selector register for
the PULSEW generator
Source selector register for
the PULSER generator
Source selector register for
the PULSE3 generator
Source selector register for
the PULSE4 generator
Table 6 shows the codes used to select pulse sources.
PULSE_WIDTH (0x34)
This register contains the numerical value controlling the pulse width for the PULSEW and PULSER output pins. The
default value is 50, which amounts to 40.07ms. The pulse width PW follows the formula:
PW <= (2 * PULSE_WIDTH + 1)/2520.6
Page: 31 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
At high pulse rates, duty cycle is 50%. At rates less than 1/(2*PWmax), the negative going pulse width is PWmax. The
31
allowed range is 0 to 2 -1.
The pulse width for the PULSE3 and PULSE4 outputs is always at a 50% duty cycle.
The initial voltage level of the pulse pins is defined with the PULSE_INIT pin.
Value in Pulse
Source Register
(hex) (dec)
0x00
0
0x01
1
0x02
2
0x03
3
0x04
4
0x05
5
0x06
6
0x07
7
0x08
8
0x09
9
0x0A
10
0x0B
11
0x0C
12
Name
WSUM
WASUM
WBSUM
WCSUM
VARSUM
VARASUM
VARBSUM
VARCSUM
VASUM
VAASUM
VABSUM
VACSUM
INSQSUM
Description
The signed sum: W0SUM + W1SUM + W2SUM
The sum of Wh samples from individual wattmeter elements.
LSB = 9.4045*10-13 VMAX IMAX / In_8 Wh
The signed sum: VAR0SUM + VAR1SUM + VAR2SUM
The sum of VARh samples from individual wattmeter elements.
LSB = 9.4045*10-13 VMAX IMAX / In_8 VARh
The sum of VAh samples from individual samples.
The sum of VAh samples from individual wattmeter elements.
LSB = 9.4045*10-13 VMAX IMAX / In_8 VAh
The sum of the square of the calculated neutral current.
∑ ( I0 + I1 + I 2 )2.
LSB = 9.4045*10-13 IMAX2 / In_82 A2h
0x0D
0x0E
0x0F
13
14
15
IASQSUM
IBSQSUM
ICSQSUM
The sum of squared current samples from each element.
LSB = 9.4045*10-13 IMAX2 / In_82 A2h
Table 6: Pulse Sources Defined by the PULSE_SRCS Register (1/2)
Page: 32 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Value in Pulse
Source Register
(hex)
(dec)
0x10
16
0x11
17
0x12
18
0x13
19
0x14
20
0x15
21
0x16
22
Name
VASQSUM
VBSQSUM
VCSQSUM
WSUM_I
WASUM_I
WBSUM_I
WCSUM_I
Description
The sum of squared voltage samples from each element.
LSB = 9.4045*10-13 VMAX2 V2h
Imported Energy: W0SUM_I + W1SUM_I + W2SUM_I
Imported energy from individual wattmeter elements. Never negative.
LSB = 9.4045*10-13 VMAX IMAX / In_8 Wh
Imported VARh:
VAR0SUM_I + VAR1SUM_I + VAR2SUM_I
0x17
23
VARSUM_I
0x18
0x19
0x1A
0x1B
0x1C
0x1D
0x1E
24
25
26
27
28
29
30
VARASUM_I
VARBSUM_I
VARCSUM_I
WSUM_E
WASUM_E
WBSUM_E
WCSUM_E
Exported Energy: W0SUM_E + W1SUM_E + W2SUM_E
0x1F
31
VARSUM_E
Exported VARh:
VAR0SUM_E + VAR1SUM_E + VAR2SUM_E
0x20
0x21
0x22
32
33
34
VARASUM_E
VARBSUM_E
VARCSUM_E
0x23
35
Internal (CE)
0x24
36
External
(host)
Exported reactive energy from individual wattmeter elements. Never negative.
LSB = 9.4045*10-13 VMAX IMAX / In_8 VARh
Exported energy from individual wattmeter elements. Never negative.
LSB = 9.4045*10-13 VMAX IMAX / In_8 Wh
Exported reactive energy from individual wattmeter elements. Never negative.
LSB = 9.4045*10-13 VMAX IMAX / In_8 VARh
Connects the pulse generator to the WSUM or VARSUM values internally
calculated by the CE. Not selectable for PULSE3 and PULSE4 pulse generators.
Indicates that the host will provide values in the APULSER, APULSEW,
APULSE3, and APULSE4 registers and that the pulse generator should be updated on the next READY interrupt..
Table 6: Pulse Sources Defined by the PULSE_SRCS Register (2/2)
WRATE (0x2D)
This register controls the rate of the pulse generation for the PULSEW, PULSER, PUSE3 and PULSE4 output pins. The
inverse pulse rate, expressed as Wh per pulse is:
Kh =
VMAX IMAX
1.5757 Wh / Pulse (X = value formed by PULSE_SLOW and
In _ 8 ⋅ SUM _ CYCLES ⋅WRATE ⋅ X
PULSE_FAST bits in the CONFIG register)
Page: 33 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Registers Used for Calibration
CAL_IA (0x24), CAL_VA (0x25), CAL_IB (0x26), CAL_VB (0x27), CAL_IC (0x28), CAL_VC (0x29):
These registers adjust the gain for the current and voltage measurements of each phase for the purpose of calibration. The
calibration factors have to be stored by the host and written to the registers of the 71M6515H after power-up. The allowed
range is (215 – 1) to –(215 – 1). The default value of 16384 equals unity gain.
If a voltage measurement of phase C is higher than expected, CAL_VC has to be adjusted to:
CAL_VC = 16384 / (1 + error)
Error must be expressed as a fraction, not a percentage value.
If the percent error is +3.5%, the relative error is 0.035, and the calibration factor becomes:
CAL_VC = 16384 / (1 + 0.035) = 15829.952,
which is rounded up to15830.
KVAR (0x2F)
This register holds the relative gain of the VAR calculation with respect to the Watt calculation. The value should always be
6444.
PHADJ_A (0x2A), PHADJ_B (0x2B), PHAD_C (0x2C)
These registers hold the phase correction factors for channels A, B, and C. The values are used by the CE to compensate
15
15
for phase errors induced by current transformers. The allowed range is (2 -1) to -(2 -1). See the Calibration Procedure
section for applicable values.
If the CE is operated in Rogowski Coil mode, no phase compensation should be required. The default value is not zero and
should need to be changed only slightly, if at all. See the Calibration section for details.
QUANT_W (0x36), QUANT_VAR (0x37), QUANT_I (0x38)
These registers hold DC values that are added to each calculated product in order to compensate for internal quantization
(a very small amount) and external noise. These values are normally set to zero. The LSB values for these variable are
listed in Table 7.
Variable
LSB Value
QUANT_W
(VMAX IMAX/ In_8)* 1.04173*10-9
QUANT_VAR
QUANT_I
Unit
(VMAX IMAX / In_8)* 1.04173*10
2
2
-9
W
VAR
-13
(IMAX / In_8 )* 5.08656*10
A (rms)
Table 7: LSB Values for QUANT Variables
Nonlinearity is most noticeable at low currents, and can result from input noise and truncation. Nonlinearities can be
eliminated using the QUANT_W register. The error can be seen as the presence of a virtual constant noise current that
becomes dominant at small load currents.
The value to be used for QUANT_W can be determined by the following formula:
error
V ⋅ I ⋅ In _ 8
QUANT _ W = − 100
VMAX ⋅ IMAX ⋅ LSB
Where error = observed error at a given voltage (V) and current (I),
VMAX = voltage scaling factor, as described in section Scaling of Measured Values
IMAX = current scaling factor, as described in section Scaling of Measured Values
LSB = QUANT LSB value = 1.04173*10-9W
Page: 34 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Example: For a Wh measurement, an error of +1% was observed at 1A. If VMAX is 600V and IMAX = 208A,
and if the measurement was taken at 240V, we determine QUANT_W as follows:
1
240 ⋅1
100
QUANT _ W = −
= −18460
600 ⋅ 208 ⋅1.04173 ⋅10 −9
The negative value obtained by the calculation will compensate for the positive error. It does not matter which current value
is chosen as long as the corresponding error value is significant (5% error at 0.2A used in the above equation will produce
the same result for QUANT_W).
Input noise and truncation can cause similar errors in the VAR calculation that can be eliminated using the QUANT_VAR
register.
VFEED_A (0x44), VFEED_B (0x45), VFEED_C (0x46)
These registers hold the compensation factors for feedthrough used for calibrating Rogowski coils. The allowed range is
(215-1) to -(215-1). See the section on Rogowski Coil Calibration for details.
Registers Controlling Basic Function and Settings of the 71M6515H
CONFIG (0x16)
The four bytes written to this register determine the basic operation of the 71M6515H. The function of the 28 bits used for
this register are explained below:
Bit 0: This bit (VAH_SELECT) determines the method used by the post-processor for determining the apparent energy
measured in VAh. If the bit is 0 (default), the calculation is based on VRMS, IRMS and the time (t) per the formula:
VAh = VRMS * IRMS * t
The accuracy of the result can be improved for low currents by setting Bit 0 to 1. The calculation is then based on the Wh
and VARh values per the formula:
VAh = Wh 2 + VARh 2
Using the calculation method above, high accuracy can be achieved for low currents.
Bit 3: This bit (RTM_EN) enables the Real-Time Monitor function when set to 1.When used, the RTM signal is available at
the TMUX pin.
Bit 4: This bit (CE_EN) enables the CE when set to 1. The CE has to be enabled for most metering functions.
Bits 7-5: These three bits (EQU) define the equation (see table below) to be implemented by the CE.
Watt & VAR Formula
Application
0
VA IA
1 element, 2W 1ø
1*
VA(IA-IB)/2
1 element, 3W 1ø
2
VA IA + VB IB
2 element, 3W 3 øDelta
3*
VA (IA - IB)/2 + VC IC
2 element, 4W 3ø Delta
4*
VA(IA-IB)/2 + VB(IC-IB)/2)
2 element, 4W 3ø Wye
5
VA IA + VB IB + VC IC
3 element, 4W 3ø Wye
EQU
Page: 35 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Bits 13-8: These six bits (SUM_CYCLES) define the length of the accumulation interval τ per the formula:
τ=
SUM _ CYCLES ⋅ 42
2520.6
Allowed values are 24 (400ms) through 60 (1000ms), unless the post-processor is disabled. Bit 8 is the LSB.
It is important to note that the length of the accumulation interval, as determined by SUM_CYCLES, is not an exact multiple
of 1000ms. For example, if SUM_CYCLES = 60, the resulting accumulation interval is:
τ=
60 ⋅ 42
2520
=
= 999.75ms
32768 Hz 2520.62 Hz
13
This means that accurate time measurements should be based on the RTC, not the accumulation interval.
Bit 14: This bit (CKOUT_DISB) disables the CKOUT pin when set. The CKOUT pin can be used for diagnostics. For EMC
compliance and power saving reasons, CKOUT_DISB should always be set.
Bit 15: This bit (ADC_DIS) disables the ADC when set, e.g. to save power. Of course, no metering or measuring can be
performed with the ADC disabled.
Bits 18-16: These three bits (TMUX) select the source for the TMUX diagnostic output pin. For EMC compliance and
power saving reasons, TMUX should be zero (default) if unused.
Bit 18
TMUX2
0
0
0
0
1
1
1
1
Bit 17
TMUX1
0
0
1
1
0
0
1
1
Bit 16
TMUX0
0
1
0
1
0
1
0
1
TMUX
0
1
2
3
4
5
6
7
Signal Selected for the TMUX Pin
GND
MUX_SYNC
RTM
RTC Output
CE_BUSY
XFER_BUSY
VX_OK (Comparator Output)
V3P3/2 =1.5V internal analog voltage
Bits 20-19: These two bits (F_SELECT) select the phase that is to be used for frequency measurement. The frequency will
be shown in bits 31-16 of the FREQ_DELTA_T register (and as bit 4 of the STATUS word – in this form as a digitized zero
crossing signal).
Bit 20
F_SELECT1
0
0
1
1
Bit 19
F_SELECT0
0
1
0
1
F_SELECT
0
1
2
3
Phase
Selected
Phase A
Phase B
Phase C
Not allowed
Since the signal at the input selected with F_SELECT is used to synchronize filters and other processing stages in the CE,
accuracy for most measurements will be reduced if no voltage is present at the selected phase input. Accuracy can be
established by selecting the phase that carries a stable signal (A, B, or C).
Bit 21: This bit (CE_ONLY) disables the post-processor when set to 1. When the post-processor is disabled, the timeintensive computations of IPHASE, IRMS, VAh and VRMS are not performed, and therefore smaller accumulation times
(SUM_CYCLES < 24) are permitted. In this case, the host is responsible for calculating IPHASE, IRMS, VAh and VRMS.
Page: 36 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Bits 23-22: These two bits (IMAGE) select the code to be used by the CE. The CE can be operated in standard mode
when using CTs and/or shunt resistor sensors or in Rogowski mode when using Rogowski coil sensors. In order to switch
the operation mode, the CE has to be disabled first by clearing the CE_EN bit.
Bit 23
IMAGE 1
0
0
1
1
Bit 22
IMAGE 0
0
1
0
1
IMAGE
CE Code Selected
0
1
2
3
Standard (CT/shunt)
Rogowski coil
Standard (CT/shunt)
Standard (CT/shunt)
Bit 24: This bit (RESET), when reset, forces all internal states of the 71M6515H to their power-up default.
Bits 26-25: These two bits (PULSE_SLOW, PULSE_FAST) modify the speed of the pulse generator. PULSE_SLOW and
PULSE_FAST determine the factor X in the equation used for Kh as shown in the table below.
PULSE_SLOW
0
0
1
1 (default)
PULSE_FAST
0
1
0
1 (default)
X
1.5*22 = 6
1.5*26 = 96
1.5*2-4 = 0.09375
1.5
PULSE_SLOW and PULSE_FAST will affect the operation of all four pulse outputs. See the Pulse Generation section for
details.
Bits 29-27: These three bits (IA_8X, IB_8X, IC_8X) apply an additional gain of 8 to the IA, IB, and IC channels when set to
1. This is a useful tool when very small signals are encountered, as is the case when using current shunt resistors with very
low resistance while operating at low currents. Care must be taken to avoid clipping. If the input to the meter exceeds
IMAX/8, clipping will occur. These bits should normally be zero, unless additional gain following the ADC stage is needed.
Bit 30: This bit (DEFAULT_PPM) defines the source of temperature compensation. When DEFAULT_PPM is 1, the
71M6515H will automatically apply compensation coefficients derived from the stored VREF temperature characteristics to
the PPMC and PPMC2 registers. When DEFAULT_PPM is zero, the host is allowed to write its own values to the PPMC and
PPMC2 registers.
STATUS (0x14)
The four bytes in this register reflect the status of the various measurement functions of the 71M6515H. This register is
read only. When a bit in the STMASK register is set, an interrupt (IRQZ) is generated as soon as the corresponding bit in
the STATUS register is set.
Bit 0: This bit (BOOTUP) signals a request from the 71M6515H to the host to be initialized.
Bit 1: This bit (SAGA ), when set, indicates that the voltage applied to phase A has sagged below SAGTHR. See the for SAG
register for a detailed description.
Bit 2: This bit (SAGB), when set, indicates that the voltage applied to phase B has sagged below SAGTHR.
Bit 3: This bit (SAGC), when set, indicates that the voltage applied to phase C has sagged below SAGTHR.
Bit 4: This bit (F0) follows the polarity of the input voltage selected with the F_SELECT bits in the CONFIG register. It
represents a smoothed, filtered and squared copy of the fundamental waveform.
Bit 5: This bit (MAXV), when set, indicates that a voltage greater than the voltage limit defined in the VI_PTHRESHOLD
register had been detected in the previous accumulation interval.
Bit 6: This bit (MAXI), when set, indicates that a current greater than the current limit defined in the VI_PTHRESHOLD
register had been detected in the previous accumulation interval.
Page: 37 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Bit 7: This bit (1SECI, toggles every second. It is controlled by the RTC.
Bit 8: This bit (VXEDGE), when set, indicates a change in state of VX comparator. This bit is updated every accumulation
interval.
Bit 9: This bit (DEDGE), when set, indicates a change in state of any selected DIO pin. This bit is updated every
accumulation interval. Pins have to be configured to generate the DEDGE flag using the DIO_INT_CTRL bits in the
D_CONFIG register.
Bit 10: This bit (XOVF), when set, indicates that the host failed to read at least one of the Wh values. Between interrupts
(indicated by the READY bit in the STATUS word), the 71M6515H expects the host to read at least one of the WATTHR_A,
WATTHR_B, or WATTHR_C values.
Bit 11: This bit (READY), when set, indicates that the 71M6515H has fresh output values ready for the host. Setting this bit
in STMASK will enable the hardware interrupt output pin IRQZ.
Bit 14-12: These bits (bit 12 for phase A, bit 13 for phase B, bit 14 for phase C), when set, indicate that the energy received
from element A, B, or C is below the creep threshold defined in the CREEP_THRSHLD register or that the current in
elements A, B, or C is below the threshold defined in bits 15-0 of the START_THRESHLD register. The creep condition
flagged by bits 14-12 of the STATUS register indicates that Wh, VARh, and IRMS measurements of element A, B, or C have
been zeroed out. Consequently, accumulation did not occur.
Bit 15: This bit (CMD_IGNORED), when set, indicates that the 71M6515H ignored the last command received from the
host. The reason can be any type of command incompatibility, e.g. attempts to write to a read-only register.
Bit 16: This bit (PULSEW_ERR), when set, indicates that the pulse generator PULSEW is configured for external (host)
input, but did not receive an update during the previous accumulation interval.
Bit 17: This bit (PULSER_ERR), when set, indicates that the pulse generator PULSER is configured for external (host) input,
but did not receive an update during the previous accumulation interval.
Bit 18: This bit (PULSE3_ERR), when set, indicates that the pulse generator PULSE3 is configured for external (host) input,
but did not receive an update during the previous accumulation interval.
Bit 19: This bit (PULSE4_ERR), when set, indicates that the pulse generator PULSE4 is configured for external (host) input,
but did not receive an update during the previous accumulation interval.
STMASK (0x15)
The four bytes in this register enable interrupts when the corresponding bit in the STATUS register is set. The default value
for STMASK is zero.
When a bit in the STMASK register is set, an interrupt (IRQZ) is generated as soon as the corresponding bit in the STATUS
register is set. Interrupts indicated by IRQZ do not necessarily have to be synchronized with accumulation intervals. For
example, the toggling of a signal applied to the DIO pins (D0…D7), when the interrupt is enabled with the DIO_INT_CTRL
register, can cause an interrupt at any time.
Page: 38 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Registers Controlling Temperature Measurement and Compensation
DEG_SCALE (0x1C)
This register holds the scale factor used to calculate the internal temperature value provided by the temperature sensor to
temperature in degrees Celsius (°C). The default value is 22721 and should not be changed.
FREQ_DELTA_T (0x11)
This register holds frequency and temperature information in two bytes each.
31
16
FREQUENCY (see Output Registers)
15
0
DELTA_T
Bits 15-0: These bits (DELTA_T) represent the temperature information relative to the value stored in the TEMP_NOM
register. One LSB is equivalent to 0.1°C. The formula used for DELTA_T is:
DELTA_T = -DEGSCALE*2-22*(TEMP_RAW-TEMP_NOM)
TEMP_NOM (0x13)
This register holds the nominal (reference) temperature. During calibration, the host must write the value read from
TEMP_RAW to TEMP_NOM in order to enable temperature compensation. See the Meter Calibration section for details. The
temperature available in register DELTA_T is based on the difference between the current temperature, as provided in
TEMP_RAW, and the reference temperature provided by TEMP_NOM.
TEMP_RAW (0x12)
This read-only register holds the raw temperature provided by the temperature sensor on the 71M6515H chip. During
calibration, the host must write the value read from TEMP_RAW to TEMP_NOM in order to enable temperature
compensation.
Example: At calibration time, the raw temperature value of 853030 was read from the TEMP_RAW register and
written to the TEMP_NOM register. At a later time, the raw temperature register reads 844866. The 71M6515H
calculates the temperature difference to:
DELTA_T = -DEGSCALE*2-22*(TEMP_RAW-TEMP_NOM) = 44
This value is interpreted as +4.4°C.
PPMC1_2 (0x1B)
This register holds the linear and squared compensation factors for the ADC temperature compensation. The allowed
range is (215 – 1) to – (215 – 1). Both words are reset to zero if TEMP_NOM equals zero.
31
16
PPMC = ADC linear factor (PPMC = 26.84 * PPM/°C )
15
0
PPMC2 = ADC quadratic factor (PPMC2 = 1374 * PPM/°C2)
Changes to PPMC1_2 by the host are only allowed if the DEFAULT_PPM bit in the CONFIG register is zero. If additional
temperature compensation by the host, e.g. for external components, is required, the procedure is as follows:
1)
The host sets the DEFAULT_PPM bit in the CONFIG register to 1 and then reads PPMC and PPMC2.
2)
The host then adds the compensation factors to PPMC and PPMC2, resets the DEFAULT_PPM bit in the CONFIG
register to 0 and then writes the modified values to PPMC and PPMC2.
Page: 39 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Y_DEG0 (0x18)
This register holds the constant compensation factor for the RTC temperature compensation. One LSB is equivalent to
0.1PPM.
Bits 31-16: These bits (Y_CALC0) represent the constant compensation factor.
Y_DEG1_2 (0x19)
This register holds the linear and quadratic compensation factors for the RTC temperature compensation.
31
16
Y_CALC1 = linear compensation factor. One LSB is equivalent to 0.01PPM/°
15
0
Y_CALC2 = quadratic compensation factor. One LSB is
equivalent to 0.001PPM/°C
Both Y_DEG0 and Y_DEG1_2 can be used to compensate the RTC to be accurate over the whole temperature range by
characterizing the crystal.
Registers for Output Signals
PULSEW_R_CNTS (0x41)
This register contains the pulse count for the PULSEW and PULSER output pins for the past accumulation interval. The
counters will be cleared at the beginning of each accumulation interval and then start counting up with each generated pulse.
Bit 15-0: The counter for the PULSER (VARh) generator.
Bit 31-16: The counter for the PULSEW (Wh) generator.
31
16
Counter for the PULSEW generator (Wh)
15
0
Counter for the PULSER generator (VARh)
At pulse rates that do not result in generation of whole counts per accumulation interval, e.g. 3 1/3 pulses, the count equivalent
to the next lower natural number will be generated until the residue accumulates to a full count, i.e. the pulse sequence
generated will be 3, 3, 3, 4, 3, 3, 3, 4…
PULSE3_4_CNTS (0x42)
This register contains the pulse count for the PULSE3 and PULSE4 output pins for the accumulation interval. The counters
will be cleared at the beginning of each accumulation interval and then start counting up with each generated pulse.
Bit 15-0: The counter for the PULSE4 generator.
Bit 31-16: The counter for the PULSE3 generator.
31
16
Counter for the PULSE3 generator
15
0
Counter for the PULSE4 generator
IASQSUM (0x39), IBSQSUM (0x3A), ICSQSUM (0x3B)
These registers hold the sum of the squared current samples collected during the previous accumulation interval. The
values for IASQSUM, IBSQSUM, and ICSQSUM are provided directly by the CE and are not post-processed. The magnitude
of the accumulated samples is determined by:
LSB =
IMAX 2
9.4045 ⋅ 10 −13 A 2 h
2
In _ 8
Page: 40 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
IASQFRACT (0x4A), IBSQFRACT (0x4B), ICSQFRACT (0x4C)
These read-only registers hold the difference between the 10-bit residual squared current sum of the current accumulation
interval and the 10-bit residual squared current sum of the previous interval. If IASQSUM, IBSQSUM, or ICSQSUM are
used to calculate current, the value obtained over several accumulation intervals will be accurate due to averaging, but the
individual values will have a higher uncertainty than when using IASQFRACT, IBSQFRACT, and ICSQFRACT.
The most accurate calculation of the squared current for a phase X (IXSQ) uses the formula:
IXSQ = IXSQSUM + 2-10 IXSQFRACT
INSQFRACT (0x4D)
This register holds the difference between the 10-bit residual squared neutral currents of the current accumulation interval
and the 10-bit residual squared neutral currents of the previous interval. The value can be used to improve the accuracy of
the squared neutral current reading by applying the formula:
INSQ = INSQSUM + 2-10 INSQFRACT
INSQSUM (0x3C)
This register holds the sum of the square of the calculated neutral current collected during the previous accumulation
interval. The calculation is implementing the following equation:
INSQSUM = ∑ (I 0 + I1 + I 2)
2
The magnitude of the accumulated samples is determined by:
LSB =
IMAX 2
9.4045 ⋅ 10 −13 A 2 h
2
In _ 8
IPHASE_ABC (0x0F)
This register holds voltage-to-current phase information for all three phases. Positive phase means lagging current
(inductive load). One LSB is equivalent to 1 degree. The range is from (28 – 1) to –(28 – 1). Since the phase calculation
involves the post-processor, these registers will not be functional when the CE_ONLY bit in the CONFIG register is set.
Bits 8-0: These bits (IPHASE_C) represent the voltage-to-current phase angle in phase C.
Bits 17-9: These bits (IPHASE_B) represent the voltage-to-current phase angle in phase B.
Bits 26-18: These bits (IPHASE_A) represent the voltage-to-current phase angle in phase A.
IRMS_A (0x0C), IRMS_B (0x0D), IRMS_C (0x0E)
These registers hold the post-processed RMS current for each phase. Only the 16 most significant bits are used. The
magnitude of the values is determined by:
LSB =
6.8781 ⋅ 10 −9 IMAX
In _ 8 SUM _ CYCLES
A rms
Since the RMS calculation involves the post-processor, these registers will not be functional when the CE_ONLY bit in the
CONFIG register is set. If higher precision is required, the host must calculate the RMS currents from the values in the
IASQSUM, IBSQSUM, and ICSQSUM registers. For even higher precision, the IASQFRACT, IBSQFRACT, ICSQFRACT
registers should be used.
Page: 41 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Example: The register IRMS_C reads the value 2,079,670. Assuming IMAX to be 208A, and using the formula
above, we determine the RMS current of phase C to:
I RMS = 2079670 ⋅
6.8781 ⋅10 −9 208
= 0.0385 A
60
VASQSUM (0x3D), VBSQSUM (0x3E), VCSQSUM (0x3F)
These registers hold the sum of the squared voltage samples collected during the previous accumulation interval. The
values for VASQSUM, VBSQSUM, and VCSQSUM are provided directly by the CE and are not post-processed. The
magnitude of the accumulated samples is determined by:
LSB = VMAX 2 9.4045 ⋅ 10 −13 V 2 h
VAH_A (0x06), VAH_B (0x07), VAH_C (0x08)
These registers hold the apparent energy collected during the previous accumulation interval. The magnitude of the
accumulated samples is determined by:
LSB = 9.4045 ⋅10 −13
VMAX IMAX
In _ 8
VAh
Since the VAh calculation involves the post-processor, these registers will not be functional when the CE_ONLY bit in the
CONFIG register is set.
VARH_A (0x03), VARH_B (0x04), VARH_C (0x05)
These registers hold the reactive energy collected during the previous accumulation interval. The magnitude of the
accumulated samples is determined by:
LSB = 9.4045 ⋅10−13
VMAX IMAX
In _ 8
VARh
VPHASE_ABC (0x10)
This register holds the phase angle between the voltages of phases A/C and A/B. The LSB is one degree.
Bits 15-0: These bits (VPHASE_AC) hold the phase angle between VA and VC.
Bits 31-16: These bits (VPHASE_AB) hold the phase angle between VA and VB.
VRMS_A (0x09), VRMS_B (0x0A), VRMS_C (0x0B)
These registers hold the post-processed RMS voltage for each phase. Only the 16 most significant bits are used. The
magnitude of the values is determined by:
LSB =
6.8781 ⋅10 −9 VMAX
SUM _ CYCLES
V rms
Since the RMS calculation involves the post-processor, these registers will not be functional when the CE_ONLY bit in the
CONFIG register is set.
If higher precision is required, the host must calculate the RMS voltages from the values in the VASQSUM, VBSQSUM, and
VCSQSUM registers.
Example: The register VRMS_B reads the value 425,778,000. Assuming VMAX to be 600V, and using the
formula above, we determine the RMS voltage of phase B to:
V RMS = 425778000 ⋅
Page: 42 of 57
6.8781 ⋅10 −9 600
= 226.85V
60
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
WH_A (0x00), WH_B (0x01), WH_C (0x02)
These registers hold the real energy collected during the previous accumulation interval. The magnitude of the values is
determined by:
LSB = 9.4045 ⋅ 10 −13
VMAX IMAX
In _ 8
Wh
Example: The register WH_A reads the value 236,675 for one accumulation interval of one second. Assuming
600V for VMAX, 208A for IMAX, and unity gain, we determine the real energy to be:
E = 236,675*600*208*9.4045*10-13 Wh = 0.0277781Wh.
By multiplying with 3,600, we get 100Wh/h, which means the applied power is 100W.
FREQ_DELTA_T (0x11)
This register holds frequency and temperature information in two bytes each.
Bits 31-16: These bits (FREQUENCY) represent the frequency of the input signal selected with the F_SELECT bit of the
CONFIG register. One LSB is equivalent to 0.1Hz.
MAIN_EDGE_CNT (0x35)
This register holds the number of zero crossings of the input phase selected by the F_SELECT bits in the CONFIG register
detected in the previous accumulation interval. The value in MAIN_EDGE_CNT can be used by the host to correct its own
RTC or to synchronize events to the line voltage.
Registers Controlling Alarms and Thresholds
CREEP_THRSLD (0x1D)
The four bytes written to this register determine the creep threshold. Setting a creep threshold helps suppressing I2H, Wh
and VARh readings when the values of WSUM and VARSUM are determined to be below the creep threshold.
Example: The creep threshold of a meter operating with an accumulation interval of 1000ms is to be configured
to be 15mA at 240V. The meter is using a VMAX of 600V, an IMAX of 208A, and is not using the additional gain
of 8. The numerical value for the CREEP_THRSLD register is to be determined.
With 15mA, the power per phase will be 3.6W, or 0.001Wh per second. With the LSB of WSUM readings given
as 9.4045*10-13*VMAX*IMAX [Wh], we determine the value to:
-13
n = 0.001Wh / (9.4045*10 *VMAX*IMAX /In_8 Wh) = 8520.2
The rounded down value of 8520 is written to the CREEP_THRSLD register.
SAG (0x2E)
This register holds the voltage and timing threshold for sag detection.
Bits 15-0: These bits (SAG_CNT) hold the sag count. A sag condition must persist for at least SAG_CNT samples before a
15
sag alarm is generated. The allowed range is 1 to (2 -1), and the default is 80 (31.7ms). The time period defined by
SAG_CNT is:
T = SAG_CNT*397μs
Bits 31-16: These bits (SAGTHR) hold the voltage threshold that is to be applied for sag detection. The peak voltage must
exceed SAGTHR once each SAG_CNT samples in order to prevent a SAG warning. One LSB is defined as:
LSB = 7.8798 ⋅ 2 16 ⋅10 −9 VMAX
Page: 43 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Example: A meter operating at 50Hz and 240V (RMS) is supposed to apply a sag threshold of 180V (RMS) for
at least four periods before a sag warning is issued. VMAX is 600V. Which values are to be selected for SAG?
Four periods translate to T = 4 * 20ms = 80ms. This means that
SAG_CNT = T/397μs = 202.
180V (RMS) translate to 255V (peak), so SAGTHR is determined by
SAGTHR =
255
255
=
= 823
LSB 7.8798 ⋅ 2 16 ⋅10 − 9 600
START_THRESHLD (0x40)
This register holds the voltage and current thresholds that apply to the calculation of frequency, zero crossings, voltage
phase and energy values. If the current is below the value stored in I_START, calculation of RMS current, and energy (Wh,
VARh and VAh) is suppressed.
Bits 15-0: These bits (I_START) hold the threshold to be applied for under-current. If ISQSUM< I_START, all postprocessed values are set to zero for that phase. This includes reported values for Wh, VARh, VAh, IRMS, VPHASE,
IPHASE, PULSER, PULSEW, PULSE3 and PULSE4. This applies only if CREEP_THRSLD is not set to zero. One LSB is
defined as:
LSB =
IMAX 2
⋅ 9.4045 ⋅ 10 −13 A 2 h
In _ 8
Elements with ISQSUM< I_START will set the creep bits in the STATUS register.
Bits 31-16: These bits (V_START). hold the threshold to be applied for under-voltage. If VRMS<V_START, the values for
frequency (register FREQ_DELTA_T), zero crossings (register MAIN_EDGE_CNT), and voltage phase (register
V_PHASE_ABC) are set to zero. Additionally, if VRMS< V_START, the reported value of VRMS is set to zero. One LSB is
defined as:
LSB = VMAX 2 ⋅ 2 16 ⋅ 9.4045 ⋅10 −13 V 2 h
Elements with VRMS< V_START will not set the creep bits in the STATUS register.
VI_PTHRESHOLD (0x17)
This register holds the threshold for over-voltage and over-current alarms.
Bits 15-0: These bits (I_PTHRESHOLD) hold the threshold to applied for over-current alarm. The threshold comparison is
applied to the upper 16 bits of the values in IRMS: One LSB is defined as:
LSB =
450 .76 ⋅ 10 −6 IMAX
In _ 8 SUM _ CYCLES
A rms
Bits 31-16: These bits (V_PTHRESHOLD). hold the threshold to applied for over-voltage alarm. The threshold comparison
is applied to the upper 16 bits of the values in VRMS One LSB is defined as:
LSB =
450 .76 ⋅ 10 −6 VMAX
SUM _ CYCLES
V rms
Example: A meter designed with VMAX=600V and IMAX=208A is supposed to apply an overvoltage threshold
of 520V (RMS) and an over-current threshold of 400A (RMS). Which values are to be selected for
V_PTHRESHOLD and I_PTHRESHOLD?
Page: 44 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
V _ PTHRESHOLD =
520
520
=
= 14,893
LSB
450.76 ⋅ 10 − 6 600
SUM _ CYCLES
I _ PTHRESHOLD =
400
400
=
= 33,046
LSB
450.76 ⋅ 10 −6 208
SUM _ CYCLES
Registers Controlling Time and RTC Functions
OP_TIME (0x1E)
This register holds the operating time expressed in 1/100 hours. The register will be reset to zero whenever the host writes
time or date to the RTC.
RTC_DATE (0x20)
This register holds the date information provided by the RTC. The register can be written to in order to set the date.
Bits 15-8: These bits contain the year information (0 to 255). The value 0 represents the year 2000.
Bits 23-16: These bits contain the month information (01 to 12). The value 01 represents January.
Bits 31-24: These bits contain the day of month information (01 to 31).
RTC_TIME_DAY (0x1F)
This register holds the time and day information provided by the RTC. The register can be written to in order to set the
time.
Bits 7-0: These bits contain the day of the week (01 to 07). The value 01 represents Sunday.
Bits 15-8: These bits contain the hour information (00 to 23). The value 00 represents midnight.
Bits 23-16: These bits contain the minutes information (00 to 59).
Bits 31-24: These bits contain the seconds information (00 to 59).
Example: The value 200835 is read from the OP_TIME register. This means that the meter has been running
since the last reset or power-up for
t = 800850/100h = 8008.50h or 8,008h and 30 minutes (333.6875 days or 333 days and 16.5h).
Registers Used for Test Functions
RTM (0x21)
This register controls the Real-Time Monitor. When the RTM_EN bit in the CONFIG register is set, the values of CE RAM
locations specified with the RTM register can be routed to the TMUX pin as a serial data stream. The TMUX bits in the
CONFIG register must be set to 3 in order to select the RTM output for the TMUX pin.
Bits 7-0: These bits (RTM3) select the CE address for RTM3.
Bits 15-8:. These bits (RTM2) select the CE address for RTM2.
Bits 23-16: These bits (RTM1) select the CE address for RTM1.
Bits 31-24: These bits (RTM0) select the CE address for RTM0.
Page: 45 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
SSI (0x22)
This register controls the function of the Serial Synchronous Interface (SSI). The function of the SSI is described in the
Internal Resources section of this data sheet. If the SSI is enabled (bit 23, SSI_EN, in the SSI register), a block of CNT words
starting at the address BEG will be transmitted each 397µs.
Bits 7-0: These bits (CNT) select the number of CE RAM address locations to be transmitted. The value in CNT must be >=
0.
Bits 15-8: These bits (BEG) define the start address of the transfer region of the CE data RAM address.
Bit 16: This bit must be set to zero.
Bit 17: This bit must be set to zero.
Bit 18: This bit (SSI_FPOL) defines the polarity of the SFR pulse signal (0: positive, 1: negative).
Bits 20-19: These bits (SSI_FSIZE) define the frame pulse format as follows:
Bit 20
Bit 19
SSI_FSIZE
SSI Frame Pulse Format
0
0
0
Once at beginning of SSI sequence
0
1
1
Every 8 bits
1
0
2
Every 16 bits
1
1
3
Every 32 bits
Bit 21: This bit (SSI_CGATE) enables the clock to be gated. When low, the SSCLK signal is active continuously, when high,
the SSCLK signal is held low when no data is being transferred.
Bit 22: This bit (SSI_10M) defines the speed of the SSCLK signal: 0: 5MHz, 1: 10MHz.
Bit 23: This bit (SSI_EN), when set to 1, enables the SSI interface.
Registers Used for I/O Control
D_CONFIG (0x1A)
This register holds three bytes that are used to manipulate the DIO pins of the 71M6515H.
Bits 7-0: These bits (DIO_VALUE) form the data register for the DIO pins D0 through D7. When a byte is written to
DIO_VALUE, the pins configured as outputs (using the D_DIR register) will change their state accordingly. Pins configured
as inputs will ignore the byte written to DIO_VALUE.
Reading DIO_VALUE will return a byte that reflects the state of all pins regardless whether they are configured as inputs or
outputs.
Bits 15-8: These bits are not used
Bits 23-16: These bits (D_DIR) define the data direction. Bit 0 controls the D0 pin, bit 7 controls the D7 pin. Setting the bit
corresponding to a pin to 1 makes the pin an output, clearing it to 0 makes it an input.
Bits 31-24: These bits (DIO_INT_CTRL) form a mask enabling the DEDGE interrupt (bit 9 of the STATUS word register). Bit
8 controls the D0 pin, bit 15 controls the D7 pin. Setting the bit corresponding to a pin to 1 enables the DEDGE interrupt,
clearing the bit disables the interrupt. If the bit in DIO_INT_CTRL is set and a transition from high to low or from low to high
occurs, the DEDGE interrupt bit in the STATUS word register will be set in the following accumulation interval.
Page: 46 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Example: DIO pins D0 through D3 are to be configured as outputs, while D4 through D7 are to be inputs. DIO7
must generate a DEDGE interrupt when the input value changes, and D0 through D3 must apply the hexadecimal
pattern 0x05. This makes the selection for the D_CONFIG registers as follows:
DIO_INTCTRL = 0x80, D_DIR = 0x0F, DIO_VALUE = 0x05
The values are combined into the 32-bit pattern 0x800F0005.
Example: D0 through D5 are to be configured as outputs, while D6 and D7 are to be inputs. D6 and D7 must
generate a DEDGE interrupt when their input value changes. This makes the selection for the D_CONFIG registers
as follows:
DIO_INTCTRL = 0xC0, D_DIR = 0x3F
The values are combined into the 32-bit pattern 0xC03F0000.
Page: 47 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Registers in Numerical Order
Address
Name
Address
Name
Address
Name
0x00
WH_A
0x20
RTC_DATE
0x40
START_THRESHLD
0x01
WH_B
0x21
RTM
0x41
PULSEW_R_CNTS
0x02
WH_C
0x22
SSI
0x42
PULSE3_4_CNTS
0x03
VARH_A
0x23
RESERVED
0x43
PULSE_SRCS
0x04
VARH_B
0x24
CAL_IA
0x44
VFEED_A
0x05
VARH_C
0x25
CAL_VA
0x45
VFEED_B
0x06
VAH_A
0x26
CAL_IB
0x46
VFEED_C
0x07
VAH_B
0x27
CAL_VB
0x47
0x08
VAH_C
0x28
CAL_IC
0x48
0x09
VRMS_A
0x29
CAL_VC
0x49
0x0A
VRMS_B
0x2A
PHADJ_A
0x4A
IASQFRACT
0x0B
VRMS_C
0x2B
PHADJ_B
0x4B
IBSQFTACT
0x0C
IRMS_A
0x2C
PHADJ_C
0x4C
ICSQFRACT
0x0D
IRMS_B
0x2D
WRATE
0x4D
INSQFRACT
0x0E
IRMS_C
0x2E
SAG
0x4E
GAIN_ADJ
0x0F
IPHASE_ABC
0x2F
KVAR
0x10
VPHASE_ABC
0x30
APULSEW
0x11
FREQ_DELTA_T
0x31
APULSER
0x12
TEMP_RAW
0x32
APULSE3
0x13
TEMP_NOM
0x33
APULSE4
0x14
STATUS
0x34
PULSE_WIDTH
0x15
STMASK
0x35
MAIN_EDGE_CNT
0x60
CE_PROG_ADDR
0x16
CONFIG
0x36
QUANT_W
0x61
CE_DATA_ADDR
0x17
VI_PTHRESH
0x37
QUANT_VAR
0x62
CE_PROG
0x18
Y_DEG0
0x38
QUANT_I
0x63
CE_DATA
0x19
Y_DEG1_2
0x39
IASQSUM
0x64
CE_PROG_INC
0x1A
D_CONFIG
0x3A
IBSQSUM
0x65
CE_DATA_INC
0x1B
PPMC1_2
0x3B
ICSQSUM
0x1C
DEG_SCALE
0x3C
INSQSUM
0x1D
CREEP_THRSLD
0x3D
VASQSUM
0x1E
OP_TIME
0x3E
VBSQSUM
0x1F
RTC_TIME_DAY
0x3F
VCSQSUM
Page: 48 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
APPLICATION INFORMATION
Meter Circuits
Bits 7 through 5 (EQU) of the CONFIG register allow the selection of the metering equation that is to be implemented by the
71M6515H. The equation to be used depends on the meter configuration.
Figure 16 shows how the 71M6515H is connected for the most common configuration, the three-phase, four-wire WYE. The
neutral wire connects to V3P3A. For this configuration, equation 5 must be selected.
Distribution
transformers
CT
A
LOAD A
CT
B
LOAD B
C
CT
LOAD C
N
71M6515H
IA
IB
IC
V3P3A
VA
VB
VC
ANSI: Form 16S
EQU = 5
P=VA*IA+VB*IB+VC*IC
Figure 16: 4-Wire 3-Phase WYE Connection
In many three-phase three-wire configurations, one phase can be grounded, as shown in Figure 17. Again, the grounded wire
is connected to V3P3A. For this configuration, equation 2 must be selected.
The four-wire three-phase delta configuration is shown in Figure 20. In this case, the center tap of the transformer that
provides the A-C voltage is grounded. the grounded wire is connected to V3P3A. For this configuration, equation 2 must be
selected. For this configuration, equation 3 must be selected, i.e. P = VA*(IA-IB)/2 + VC*IC.
Page: 49 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
B
LOAD B
C
A
LOAD A
IA
IB
IC
V3P3A
VC
VA
VB
EQU = 2
P=VA*IA+VB*IB
71M6515H
Figure 17: 3-Wire 3-Phase Delta Connection
C
LOAD C
LOAD B
B
NEUTRAL
A
LOAD A
IA
IB
IC
V3P3A
VC
VA
VB
71M6515H
Figure 18: 4-Wire 3-Phase Delta Connection
Page: 50 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Communication between the 71M6515H and the Host Processor
General
To ensure proper transfer of energy and other values from the 71M6515H to the host, the output data of the 71M6515H must
be read by the host each time they are ready, and no energy-related datum can be missed. This requires close synchronization
between the 71M6515H and the host.
Control Signals
Figure 19 shows the control signals between the 71M6515H and the host processor. These signals are:
1)
2)
3)
4)
5)
6)
7)
TX: Serial transmit output pin of the 71M6515H
RX: Serial receive input pin of the 71M6515H
IRQZ: Interrupt output, used as “Data Ready” hardware signal of the 71M6515H (low-active)
RESETZ: Reset input pin of the 71M6515H (low-active, should be pulled up to V3P3)
UARTCSZ: UART reset input pin of the 71M6515H (low-active)
BAUD_RATE: Baud rate select input pin of the 71M6515H (optional)
PULSE_INIT: Pulse polarity select input pin of the 71M6515H (optional)
TX and RX are the most essential signals for the communication between the 71M6515H and the host processor. The IRQZ
pin provides a useful output signal that can be used by the host to determine whether the 71M6515H has fresh data ready.
IRQZ can be connected to either an interrupt input or general I/O input of the host processor.
RESETZ can be used to force a hardware reset of the 71M6515H, and UARTCSZ can be used to reset (purge) the UART of
the 71M6515H communication buffers for reconfiguration. The additional pins BAUD_RATE and PULSE_INIT can be hardwired for configuring the communication baud rate and the pulse status, or controlled on power up by the host processor.
71M6515H
Host
TX
RX
RX
TX
IRQ or DIO_1
IRQZ
3.3V
DIO
DIO0...DIO7
RESETZ
DIO_2
UARTCSZ
DIO_3
BAUD_RATE
DIO_4
PULSE_INIT
DIO_5
Figure 19: Connections between 71M6515H and Host
Note that the DIO pins of the host processor used to control the 71M6515H are not lost, since the 71M6515H can provide eight
DIO pins ((DIO0…DIO7) to act as general-purpose DIO pins.
Since the communication between the 71M6515H and its host is based on a binary protocol, it is imperative for the host to
issue a clean character stream without added bytes (UART drivers of high-level operating systems often add extra bytes to the
character stream, relying on error-detecting protocols). In case the 71M6515H loses synchronization due to unexpected bytes
sent by the host, it times out after 10ms (maximum 20ms). The 71M6515H flags a time-out condition by setting the BOOTUP
flag in the STATUS register. This happens because the incomplete or garbled data could have included a write command to the
CONFIG register or other important registers.
Page: 51 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Methods of Control
Two different methods of control can be used by the host processor:
1)
Synchronization using the IRQZ pin of the 71M6515H (interrupt or DIO pin polling method, see Figure 20):
a. Interrupt Method: The IRQZ pin of the 71M6515H is connected to a pin of the host processor that can generate an
interrupt for the host processor. This is the easiest method for synchronization between the 71M6515H and the host.
The CONFIG register of the 71M6515H is set up to generate an interrupt on the IRQZ pin whenever fresh data are
ready, and the interrupt service routine of the host processor reads the fresh data out of the 71M6515H.
b. DIO pin polling method: The IRQZ pin of the 71M6515H is connected to a DIO pin of the host processor. The host
processor polls the status of the DIO pin as frequently as possible or through a timer-based polling method. The
CONFIG register of the 71M6515H is set up to have the IRQZ pin to go low on every fresh data ready status, and the
timer-serviced polling of the host processor will monitor the status of the DIO pin and initiate the serial communication
when IRQZ is detected. For this method to be effective, the firmware of the host processor must maintain the timer
interrupt to be the highest priority, followed by the serial communications priority.
2)
Polling the READY bit in the STATUS word of the 71M6515H (status polling method, see Figure 21).
This method requires that the host processor utilizes a timer with 1ms to 5ms resolution tied into the highest-priority
interrupt. The interrupt service routine must initiate reading the STATUS register, preferably at least every 10ms, in
order to monitor the READY bit, but the host processor must wait for the response of each status request. Otherwise,
the STATUS register read operations will be stacked in the 71M6515H resulting in multiple responses.
If a delayed response is received upon a STATUS register read, the host processor will know that the 71M6515H is
within its post-processing period, which makes it necessary hat the host waits for the response. Every time the READY
bit in the STATUS register is not set, indicating that data is not available, the host should poll again
read
command
from host
read
command
from host
6515H data
6515H data
IRQZ
post-processor
active
post-processor
active
80ms
200ms
time
80ms
300ms
400ms
500ms
Figure 20: Timing Diagram (Using IRQZ, SUM_CYCLES = 12)
The communication between the 71M6515H and the host processor can always be reset without disturbing the metering
function by utilizing the UARTCSZ and BAUDRATE pins. Configuring the BAUDRATE pin without resetting the UART buffers
is not recommended. The UART of the 71M6515H can be “reset” by pulling the UARTCSZ pin low. This will force the UART
back into the default configuration while clearing all buffers (UART buffers, UART-related buffers in the firmware).
Page: 52 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
read
commands
from host
read
commands
from host
6515H not
ready
6515H data
6515H not
ready
6515H data
post-processor
active
post-processor
active
330ms
330ms
1000ms
time
2000ms
Figure 21: Timing Diagram (Host Polling)
Communication Steps for Interrupt Method:
The following is a list of commands required from the host processor to establish communication with the 71M6515H.
1)
Establish host baud rate and data format as required by the 71M6515H.
2)
Configure the DIO pins (D0…D7), if required, using the D_CONFIG register
3)
Configure the 71M6515H by selecting the CE image (bits 23-22), equation (bits 7-5), pulse rate (bits 26-25), followed
by enabling the CE (bit 4 in the CONFIG register). The bit pattern sent to CONFIG should have all bits set to their
default state, as given in the Register Table.
4)
Write the TEMP_RAW value obtained at calibration into the TEM_NOM register to enable temperature compensation.
5)
Write the calibration coefficients into registers CAL_IA, CAL_VA, CAL_IB, CAL_VB, CAL_C, CAL_VC, PHADJ_A,
PHADJ_B, PHADJ_C. Write calibration values to the VFEED_A/B/C registers if the Rogowski image is used.
6)
Configure WRATE with the value required to generate the desired pulse rate.
7)
Establish creep, sag and over/under voltage/current thresholds, if necessary, by writing values to the
CREEP_THRSLD, SAG, VI_PTHRESH, and VI_THRSHLD registers.
8)
Set the READY bit in the status mask STMASK in order to enable the generation of interrupts on the IRQZ pin.
9)
Read at least one accumulated value register (WATTH_X, or VAH_X, or VARH_X).
10) Wait for the IRQZ pin to go low.
11) After receiving the interrupt, read at least one accumulated value register (WATTH_X, or VAH_X, or VARH_X) in order
to maintain interrupt generation. Read the bits in the STATUS word to detect potential faults (overflow signaled by the
XOVF flag or uninitialized condition signaled by the BOOTUP flag) or warnings and events (sag, creep, excessive
voltage or current, DIO signal changes). If necessary, take corrective action.
12) After reading the required data from the 71M6515H, configuration changes should be made, if necessary. These configuration changes should be completed before the pre-processing period begins again.
Page: 53 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Timing
The fundamental factor for all timing considerations is SUM_CYCLES, which determines the length of the accumulation interval
for the 71M6515H per the equation:
τ=
SUM _ CYCLES ⋅ 42
2520.6 Hz
The default setting for SUM_CYCLES is 60, which yields an accumulation interval close to 1,000ms. A conservative minimum
number for SUM_CYCLES is 24, which yields an accumulation interval close to 400ms. Both calculations by the post-processor
in the 71M6515H and the communication between 71M6515H and the host have to be completed within the accumulation
interval. If an accumulation interval has passed, and the energy values have not been read by the host, they are lost forever.
In order to analyze the timing of the communication between the 71M6515H and the host, it is useful to know the basic timing
requirements of the post-processor of the 71M6515H. Some timing parameters are listed in Table 8.
CE_ONLY
Disabled
VAh Calculation
Resulting Calculation Time
Vector method:
350ms
VAh = Wh 2 + VARh 2
Disabled
Vrms*Irms method
80ms
Enabled
X
40ms
Table 8: Post-processor Timing
As can be seen in Table 8, the calculation time can be greatly reduced if the VAh values are calculated using the method of
multiplying Vrms by Irms (by resetting bit 0 in the CONFIG register), which is less accurate at low currents.
Further improvement can be achieved by disabling the post-processor using the CE_ONLY bit in the CONFIG register. This is
possible for applications where the registers IPHASE, IRMS, VAh and VRMS are not required.
It becomes clear now that the minimum value of 24 for SUM_CYCLES, equivalent to a 400ms accumulation interval, accommodates the worst-case scenario, using the vector method, which requires 350ms post-processing time. This setting leaves
around 50ms for the communication to take place between the 71M6515 and the host. If the simpler Irms*Vrms method is
chosen for VAh, a lower value can be selected for SUM_CYCLES.
Lower values for SUM_CYCLES, for example 12, yielding a 200ms accumulation interval, are possible, leaving still 120ms for
host communications, as long as the Vrms*Irms method for VAh is used.
Some other timing parameters are listed in Table 9.
Parameter
Value
Comment
Time from power-up to UART being functional
370ms ±10%
Time from HW reset to UART being functional
370ms ±10%
Time from soft reset to UART being functional
245ms ±10%
Soft reset = RESET bit in CONFIG register
is set high.
20ms
UARTCSZ is polled just before the
71M6515H checks its data buffer for a
command. This means that the command
latency specified in the Electrical Specifications section also applies to the
UARTCSZ pin.
Time from UARTCSZ pin low to UART being functional
Table 9: UART Timing Parameters
Page: 54 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
PACKAGE OUTLINE
11.7
12.3
11.7
+
12.3
PIN No. 1 Indicator
9.8
10.2
0.60 Typ.
0.50 Typ.
0.00
0.20
0.14
0.28
1.40
1.60
NOTE: Controlling dimensions are in mm
Ordering Information
PART DESCRIPTION
ORDER NO.
PACKAGING MARK
71M6515H 64-Pin LQFP Lead-Free, 0.1% accuracy
71M6515H-IGT/F
71M6515H-IGT
71M6515H 64-Pin LQFP Lead-Free, 0.1% accuracy, T&R
71M6515H-IGTR/F
71M6515H-IGT
Page: 55 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
REVISION HISTORY
Revision
Rev. 1.0
Rev. 1.1
Date
October 26, 2005
December 11, 2006
Rev. 1.2
March 15, 2007
Rev. 1.3
August 17, 2007
Rev. 1.4
March 5, 2008
Page: 56 of 57
Description
First publication.
Changed capacitor values for XIN/XOUT
Corrected addresses for CE_DATA and made addresses CE_DATA etc. visible in
table “Registers in Numerical Order”.
Changed frequency range to 46-64Hz and “Real-time clock for TOU” to “Real-time
clock with temperature compensation” on title page.
Added complete chapter on communication between host and 6515H.
Changed pin name V1 to VFLT in Electrical Specification.
Consolidated spelling for CREEP_THRSLD register. Added explanation for X in
formula for WRATE.
Added information on processing time for registers involving post-processing.
Changed default value for WRATE to 683.
Deleted note on low-pass filter in the CE in CONFIG register description.
Added note at CONFIG register description stating that phase with stable voltage
can be selected with F_SELECT to avoid inaccurate measurements and note stating
that PULSE_SLOW and PULSE_FAST affect all four pulse sources.
Added diagram “Connections between 71M6515H and host”.
Added I/O Equivalent Circuit diagrams and circuit type numbers in Pin Descriptions.
Added note in Pin Descriptions stating that the voltage at RX must not exceed 3.6V.
Added VREF aging data.
Clarified polarity of SSDAT and SSCLK pins.
Changed recommended value for capacitors at XIN/XOUT to 22pF.
Changed recommended crystal to ECS ECX-3TA series.
Added note on exact length of default accumulation interval.
Corrected bit locations for D_CONFIG register.
Added note in pin description for D0-D7 and Digital I/O” section: D0 through D7 float
after reset or power-up and are configured as outputs and driven low 140ms after
RESETZ goes high.
Updated default values for VIPTHRESH and VI_THRESH.
Changed pin and register names to PULSEW and PULSER and updated block
diagram (Figure 2), updated pin-out diagram with corrected pin names.
Removed COMPARATOR table in ELECTRICAL SPECIFICATIONS.
Changed note for SRDY in pin description table to “SRDY should be tied to ground”,
deleted Figure 13 (SRDY function) and removed all text describing the function of
SRDY (except that SRDY should be grounded), deleted Figure 13 (SSI Timing w/
SRDY) and removed references to SRDY in Figure 12.
Consolidated spelling of Y_CALC etc. constants in section “Temperature Compensation for the Crystal and RTC” and in the register tables.
Added reference to fast calibration procedure (AN_651X_022).
Added diagrams for metering configurations.
Improved Table 6 and explanation of PULSE_SRC register.
Changed capacitor values for XIN/XOUT to 27pF and added recommended load
capacitance value (12.5pF).
Changed register name at location 0x38 from QUANT_V to QUANT_I.
Removed non lead-free packages from Ordering Information.
Updated Teridian street address.
Added text at explanation of bits 14-12 of the STATUS register stating that these bits
are also set when the current is below the threshold defined by bits 15-0 of the
START_THRESHLD register. Added at explanation of START_THRESHLD : “Elements
with VRMS< V_START will not set the creep bits in the STATUS register” and
“Elements with ISQSUM< I_START will set the creep bits in the STATUS register”.
Added value for RθJA in Electrical Specification.
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4
71M6515H
Energy Meter IC
DATA SHEET
MARCH 2008
Final Data Sheet: This Final Data Sheet is proprietary to TERIDIAN Semiconductor Corporation (TSC) and sets forth design goals for the
described product. The data sheet is subject to change. TSC assumes no obligation regarding future manufacture, unless agreed to in
writing. If and when manufactured and sold, this product is sold subject to the terms and conditions of sale supplied at the time of order
acknowledgment, including those pertaining to warranty, patent infringement and limitation of liability. TERIDIAN Semiconductor Corporation
(TSC) reserves the right to make changes in specifications at any time without notice. Accordingly, the reader is cautioned to verify that a
data sheet is current before placing orders. TSC assumes no liability for applications assistance.
TERIDIAN Semiconductor Corp., 6440 Oak Canyon, Suite 100, Irvine, CA 92618
TEL (714) 508-8800, FAX (714) 508-8877, http://www.teridian.com
3/13/2008
Page: 57 of 57
© 2005-2008 TERIDIAN Semiconductor Corporation
V1.4