HP HCPL-0870 Isolated 15-bit a/d converter Datasheet

H
Isolated 15-bit A/D Converter
Technical Data
HCPL-7860
HCPL-0870, -7870
Features
• 12-bit Linearity
• 700 ns Conversion Time
(Pre-Trigger Mode 2)
• 5 Conversion Modes for
Resolution/Speed Trade-Off;
12-bit Effective Resolution
with 18 µs Signal Delay
(14-bit with 94 µs)
• Fast 3 µs Over-Range
Detection
• Serial I/O (SPI®, QSPI® and
Microwire® Compatible)
• ± 200 mV Input Range with
Single 5 V Supply
• 1% Internal Reference
Voltage Matching
• Offset Calibration
• -40 °C to +85 °C Operating
Temperature Range
• 15 kV/µs Isolation Transient
Immunity
• Regulatory Approvals; UL,
CSA, VDE
DIGITAL CURRENT SENSOR
ISOLATION
BOUNDARY
+
HPx870
YYWW
INPUT
CURRENT
HP7860
YYWW
OUTPUT
DATA
ISOLATED
MODULATOR
DIGITAL
INTERFACE IC
MICRO-CONTROLLER
+
Hewlett-Packard’s Isolated A/D Converter delivers the reliability, small size, superior
isolation and over-temperature performance motor drive designers need to accurately
measure current at half the price of traditional solutions.
CAUTION: It is advised that normal static precautions be taken in handling and assembly of this component to
prevent damage and/or degradation which may be induced by ESD.
SPI and QSPI are trademarks of Motorola Corp.
Microwire is a trademark of National Semiconductor Inc.
1-260
5965-5255E
Digital Current Sensing
Circuit
than 1 µs, the fast over-range
detector for quickly detecting
short circuits, different conversion
modes giving various resolution/
speed trade-offs, offset calibration mode to eliminate initial
offset from measurements, and
an adjustable threshold detector
for detecting non-short circuit
overload conditions.
shunt resistance, any range of
current can be monitored, from
less than 1 A to more than 100 A.
As shown in Figure 1, using the
Isolated 2-chip A/D converter to
sense current can be as simple as
connecting a current-sensing
resistor, or shunt, to the input
and reading output data through
the 3-wire serial output interface.
By choosing the appropriate
Even better performance can be
achieved by fully utilizing the
more advanced features of the
Isolated A/D converter, such as
the pre-trigger circuit which can
reduce conversion time to less
NON-ISOLATED
+5V
ISOLATED
+5V
+
INPUT
CURRENT R
SHUNT
0.02
C1
0.1 µF
VDD1
VDD2
VIN+
MCLK
VIN-
MDAT
GND1
GND2
HCPL-7860
C2
0.1 µF
CCLK
VDD
CLAT
CHAN
CDAT
SCLK
MCLK1
SDAT
MDAT1
CS
MCLK2
THR1
MDAT2
OVR1
GND
3-WIRE
SERIAL
INTERFACE
+
C3
10 µF
RESET
HCPL-x870
Figure 1: Typical Application Circuit.
Product Overview
Description
The HCPL-7860 Isolated Modulator and the HCPL-x870 Digital
Interface IC together form an
isolated programmable two-chip
analog-to-digital converter. The
isolated modulator allows direct
measurement of motor phase
currents in power inverters while
the digital interface IC can be
programmed to optimize the
conversion speed and resolution
trade-off.
In operation, the HCPL-7860
Isolated Modulator (optocoupler
with 3750 VRMS dielectric withstand voltage rating) converts a
low-bandwidth analog input into
a high-speed one-bit data stream
by means of a sigma-delta (∑∆)
oversampling modulator. This
modulation provides for high
noise margins and excellent
immunity against isolation-mode
transients. The modulator data
and on-chip sampling clock are
encoded and transmitted across
the isolation boundary where they
are recovered and decoded into
separate high-speed clock and
data channels.
The Digital Interface IC converts
the single-bit data stream from
the Isolated Modulator into
fifteen-bit output words and
provides a serial output interface
that is compatible with SPI®,
QSPI®, and Microwire® protocols, allowing direct connection
to a microcontroller. The Digital
Interface IC is available in two
package styles: the HCPL-7870 is
in a 16-pin DIP package and the
HCPL-0870 is in a 300-mil wide
SO-16 surface-mount package.
Features of the Digital Interface
IC include five different conversion modes, three different pretrigger modes, offset calibration,
fast over-range detection, and
adjustable threshold detection.
Programmable features are configured via the Serial Configuration port. A second multiplexed
input is available to allow
measurements with a second
1-261
isolated modulator without
additional hardware. Because the
two inputs are multiplexed, only
one conversion at a time can be
made and not all features are
available for the second channel.
The available features for both
channels are shown in the table
at right.
HCPL-x870 Digital Interface IC
Feature
Conversion Mode
Offset Calibration
Pre-Trigger Mode
Over-Range Detection
Adjustable Threshold Detection
Channel #1
✓
✓
✓
✓
✓
Channel #2
✓
✓
Functional Diagrams
ISOLATION
BOUNDARY
VDD1
VIN+
VIN–
GND1
1
2
3
8
SIGMADELTA
MOD./
ENCODE
4
7
VDD2
MCLK
DECODE
SHIELD
6
5
MDAT
CCLK
1
CLAT
2
CDAT
3
MCLK1
4
MDAT1
5
MCLK2
6
CONVERSION
INTERFACE
15 CHAN
14 SCLK
13 SDAT
CH1
MDAT2
7
GND
8
GND2
HCPL-7860 Isolated Modulator
16 VDD
CONFIG.
INTERFACE
12 CS
CH2
THRESHOLD
DETECT
&
RESET
11 THR1
10 OVR1
9
RESET
HCPL-x870 Digital Interface IC
Pin Description, Isolated Modulator
Symbol
VDD1
VIN+
VIN–
GND1
1-262
Description
Supply voltage input (4.5 V to 5.5 V)
Positive input (± 200 mV
recommended)
Negative input
(normally connected to GND1)
Input ground
Symbol
VDD2
MCLK
Description
Supply voltage input (4.5 V to 5.5 V)
Clock output (10 MHz typical)
MDAT
Serial data output
GND2
Output ground
Pin Description, Digital Interface IC
Symbol
Description
CCLK Clock input for the Serial Configuration
Interface (SCI). Serial Configuration
data is clocked in on the rising edge
of CCLK.
CLAT Latch input for the Serial Configuration
Interface (SCI). The last 8 data bits
clocked in on CDAT by CCLK are
latched into the appropriate
configuration register on the rising
edge of CLAT.
CDAT Data input for the Serial Configuration
Interface (SCI). Serial configuration
data is clocked in MSB first.
MCLK1 Channel 1 Isolated Modulator clock
input. Input Data on MDAT1 is clocked
in on the rising edge of MCLK1.
Symbol
Description
VDD
Supply voltage (4.5 V to 5.5 V).
CHAN
Channel select input. The input level on
CHAN determines which channel of
data is used during the next conversion
cycle. An input low selects channel 1,
a high selects channel 2.
SCLK
Serial clock input. Serial data is clocked
out of SDAT on the falling edge of SCLK.
SDAT
Serial data output. SDAT changes from
high impedance to a logic low output
at the start of a conversion cycle.
SDAT then goes high to indicate that
data is ready to be clocked out. SDAT
returns to a high-impedance state after
all data has been clocked out and CS
has been brought high.
Conversion start input. Conversion
begins on the falling edge of CS. CS
should remain low during the entire
conversion cycle and then be brought
high to conclude the cycle.
Continuous, programmable-threshold
detection for channel 1 input data. A
high level output on THR1 indicates
that the magnitude of the channel 1
input signal is beyond a user
programmable threshold level between
160 mV and 310 mV. This signal
continuously monitors channel 1
independent of the channel select
(CHAN) signal.
High speed continuous over-range
detection for channel 1 input data. A
high level output on OVR1 indicates
that the magnitude of the channel 1
input is beyond full-scale. This signal
continuously monitors channel 1
independent of the CHAN signal.
Master reset input. A logic high input
for at least 100 ns asynchronously
resets all configuration registers to
their default values and zeroes the
Offset Calibration registers.
CS
MDAT1
Channel 1 Isolated Modulator data
input.
MCLK2
Channel 2 Isolated Modulator clock
input. Input Data on MDAT2 is clocked
in on the rising edge of MCLK2.
THR1
MDAT2
Channel 2 Isolated Modulator data
input.
OVR1
Digital ground.
RESET
GND
1-263
Isolated A/D Converter Performance
Electrical Specifications
Unless otherwise noted, all specifications are at VIN+ = -200 mV to +200 mV and VIN- = 0 V; all Typical
specifications are at TA = 25°C and VDD1 = VDD2 = VDD = 5 V; all Minimum/Maximum specifications are at
TA = -40°C to +85°C, VDD1 = VDD2 = VDD = 4.5 to 5.5 V.
Parameter
Symbol
Min.
Typ.
Max.
Units
Test Conditions
Fig. Note
STATIC CONVERTER CHARACTERISTICS
Resolution
Integral Nonlinearity
15
INL
6
0.025
Differential Nonlinearity
DNL
Uncalibrated Input Offset
VOS
-1
Offset Drift vs. Temperature
dVOS/dTA
Offset drift vs. VDD1
dVOS/dVDD1
Internal Reference Voltage
VREF
Absolute Reference Voltage
-4
Tolerance
Reference Voltage
-1
Matching
VREF Drift vs. Temperature dVREF /dTA
VREF Drift vs. VDD1
dVREF /dVDD1
Full Scale Input Range
-VREF
Recommended Input
-200
Voltage Range
1
4
0.7
326
4
bits
LSB
%
LSB
mV
µV/ °C
mV/V
mV
%
1
%
30
0.14
1
2.5
190
0.9
+VREF
+200
3
4
1
2
3
VIN+ = 0 V
5
4
6
5
TA = 25°C.
See Note 5
ppm/°C
%
mV
6
DYNAMIC CONVERTER CHARACTERISTICS
(Digital Interface IC is set to Conversion Mode 3.)
Signal-to-Noise Ratio
Total Harmonic Distortion
Signal-to-(Noise
+ Distortion)
Effective Number of Bits
Conversion Time
Signal Delay
Over-Range Detect Time
Threshold Detect Time
Signal Bandwidth
Isolation Transient
Immunity
1-264
SNR
THD
SND
62
73
-67
66
ENOB
tC2
tC1
tC0
tDSIG
tOVR1
tTHR1
BW
CMR
10
12
0.7
18
37
18
2.7
10
22
20
2.0
18
15
1.0
22
44
22
4.2
dB
VIN+ = 35 Hz,
400 mVpk-pk
(141 mVrms) sine
wave.
bits
µs
8
Pre-Trigger Mode 2 7,
Pre-Trigger Mode 1 14
Pre-Trigger Mode 0
10
VIN+ = 0 to 400 mV 12
step waveform
kHz
kV/µs
2,9
11
VISO = 1 kV
7
8
9
10
11
12
13
Notes:
1. Resolution is defined as the total
number of output bits. The useable
accuracy of any A/D converter is a
function of its linearity and signal-tonoise ratio, rather than how many
total bits it has.
2. Integral nonlinearity is defined as
one-half the peak-to-peak deviation
of the best-fit line through the
transfer curve for VIN+ = -200 mV to
+200 mV, expressed either as the
number of LSBs or as a percent of
measured input range (400 mV).
3. Differential nonlinearity is defined as
the deviation of the actual difference
from the ideal difference between
midpoints of successive output
codes, expressed in LSBs.
4. Data sheet value is the average
magnitude of the difference in offset
voltage from TA = 25°C to
TA = -40°C, expressed in microvolts
per °C.
5. All units within each HCPL-7860
standard packaging increment (either
50 per tube or 1000 per reel) have an
Absolute Reference Voltage tolerance
of ± 1%. An Absolute Reference
Voltage tolerance of ± 4% is
guaranteed between standard
packaging increments.
6. Beyond the full-scale input range the
output is either all zeroes or all ones.
7. The effective number of bits (or
effective resolution) is defined by the
equation ENOB = (SNR-1.76)/6.02
and represents the resolution of an
ideal, quantization-noise limited A/D
converter with the same SNR.
8. Conversion time is defined as the
time from when the convert start
signal CS is brought low to when
SDAT goes high, indicating that
output data is ready to be clocked
out. This can be as small as a few
cycles of the isolated modulator clock
and is determined by the frequency of
the isolated modulator clock and the
selected Conversion and Pre-Trigger
modes. For determining the true
signal delay characteristics of the A/D
converter for closed-loop phase
margin calculations, the signal delay
specification should be used.
9. Signal delay is defined as the effective delay of the input signal through
the Isolated A/D converter. It can be
measured by applying a -200 mV to
± 200 mV step at the input of modulator and adjusting the relative delay
of the convert start signal CS so that
the output of the converter is at midscale. The signal delay is the elapsed
time from when the step signal is
applied at the input to when output
data is ready at the end of the conversion cycle. The signal delay is the
most important specification for
determining the true signal delay
characteristics of the A/D converter
11.
12.
13.
16
75.0
0.08
14
VDD1 = 4.5 V
VDD1 = 5.0 V
VDD1 = 5.5 V
VDD1 = 4.5 V
VDD1 = 5.0 V
VDD1 = 5.5 V
12
SNR
74.0
73.5
0.07
0.05
8
6
0
20
40
60
TEMPERATURE – °C
Figure 2. SNR vs. Temperature.
85
0.04
0.03
4
0.02
2
0.01
73.0
-20
0
-40
VDD1 = 4.5 V
VDD1 = 5.0 V
VDD1 = 5.5 V
0.06
10
INL – %
INL – LSB
74.5
72.5
-40
10.
and should be used for determining
phase margins in closed-loop applications. The signal delay is determined
by the frequency of the modulator
clock and which Conversion Mode is
selected, and is independent of the
selected Pre-Trigger Mode and,
therefore, conversion time.
The minimum and maximum overrange detection time is determined by
the frequency of the channel 1 isolated modulator clock.
The minimum and maximum threshold detection time is determined by
the user-defined configuration of the
adjustable threshold detection circuit
and the frequency of the channel 1
isolated modulator clock. See the
Applications Information section for
further detail. The specified times
apply for the default configuration.
The signal bandwidth is the frequency
at which the magnitude of the output
signal has decreased 3 dB below its
low-frequency value. The signal
bandwidth is determined by the frequency of the modulator clock and
the selected Conversion Mode.
The isolation transient immunity (also
known as Common-Mode Rejection)
specifies the minimum rate-of-rise of
an isolation-mode signal applied
across the isolation boundary beyond
which the modulator clock or data
signals are corrupted.
-20
0
20
40
60
85
TEMPERATURE – °C
Figure 3. INL (Bits) vs. Temperature.
0
-40
-20
0
20
40
85
60
TEMPERATURE – °C
Figure 4. INL (%) vs. Temperature.
1-265
400
100
0
-100
-200
-300
-400
VDD1 = 4.5 V
VDD1 = 5.0 V
VDD1 = 5.5 V
1.5
1.0
0.5
0
-0.5
-500
-1.0
-600
-40
-1.5
-40
0
-20
20
40
60
85
PRE-TRIGGER
MODE 2
100
80
60
40
0
0
-20
40
20
60
85
1
14
85
13
80
2
3
4
5
CONVERSION MODE #
Figure 6. VREF Change vs.
Temperature.
Figure 7. Conversion Time vs.
Conversion Mode.
100
90
SIGNAL DELAY – µs
80
75
SNR
12
11
10
70
65
60
9
55
8
50
1
2
3
4
5
Figure 8. Effective Resolution vs.
Conversion Mode.
1
2
3
4
5
100
90
VIN+ (200 mV/DIV.)
70
OVR1 (200 mV/DIV.)
60
THR1
(2 V/DIV.)
50
40
30
20
10
0
1
2
3
4
5
2 µs/DIV.
CONVERSION MODE #
Figure 11. Signal Bandwidth vs.
Conversion Mode.
60
50
40
30
10
Figure 9. SNR vs. Conversion Mode.
80
70
20
CONVERSION MODE #
CONVERSION MODE #
SIGNAL BANDWIDTH – kHz
120
160
TEMPERATURE – °C
Figure 5. Offset Change vs.
Temperature.
1-266
140
20
TEMPERATURE – °C
EFFECTIVE RESOLUTION (# BITS)
CONVERSION TIME – µs
VDD1 = 4.5 V
VDD1 = 5.0 V
VDD1 = 5.5 V
200
PRE-TRIGGER
MODE 0
PRE-TRIGGER
MODE 1
180
2.0
VREF CHANGE – %
OFFSET CHANGE – µV
200
2.5
300
Figure 12. Over-Range and Threshold
Detect Times.
0
1
2
3
4
CONVERSION MODE #
Figure 10. Signal Delay vs.
Conversion Mode.
5
Isolated Modulator
Ordering Information
Specify Part Number followed by Option Number (if desired).
Example:
HCPL-7860#XXX
No Option = Standard DIP package, 50 per tube.
300 = Gull Wing Surface Mount Option, 50 per tube.
500 = Tape and Reel Packaging Option, 1000 per reel.
Option data sheets available. Contact Hewlett-Packard sales representative or authorized distributor.
Package Outline Drawings
8-pin DIP Package
9.40 (0.370)
9.90 (0.390)
TYPE NUMBER
8
7
6
5
REFERENCE VOLTAGE
MATCHING SUFFIX*
DATE CODE
HP 7860X
YYWW
PIN ONE
1.19 (0.047) MAX.
1
2
3
5° TYP.
4
1.78 (0.070) MAX.
4.70 (0.185) MAX.
PIN ONE
0.51 (0.020) MIN.
2.92 (0.115) MIN.
0.76 (0.030)
1.24 (0.049)
0.18 (0.007)
0.33 (0.013)
6.10 (0.240)
6.60 (0.260)
7.36 (0.290)
7.88 (0.310)
0.65 (0.025) MAX.
PIN DIAGRAM
1
VDD1
VDD2 8
2
VIN+
MCLK 7
3
VIN–
MDAT 6
4
GND1 GND2 5
2.28 (0.090)
2.80 (0.110)
DIMENSIONS IN MILLIMETERS AND (INCHES).
*ALL UNITS WITHIN EACH HCPL-7860 STANDARD PACKAGING INCREMENT (EITHER 50 PER TUBE OR 1000 PER REEL)
HAVE A COMMON MARKING SUFFIX TO REPRESENT AN ABSOLUTE REFERENCE VOLTAGE TOLERANCE OF ± 1%.
AN ABSOLUTE REFERENCE VOLTAGE TOLERANCE OF ± 4% IS GUARANTEED BETWEEN STANDARD PACKAGING
INCREMENTS.
1-267
8-pin DIP Gull Wing Surface Mount Option 300
PIN LOCATION (FOR REFERENCE ONLY)
9.65 ± 0.25
(0.380 ± 0.010)
8
7
6
1.02 (0.040)
1.19 (0.047)
5
4.83 TYP.
(0.190)
6.350 ± 0.25
(0.250 ± 0.010)
MOLDED
1
2
3
9.65 ± 0.25
(0.380 ± 0.010)
4
0.380 (0.015)
0.635 (0.025)
1.19 (0.047)
1.78 (0.070)
9.65 ± 0.25
(0.380 ± 0.010)
1.780
(0.070)
MAX.
1.19
(0.047)
MAX.
7.62 ± 0.25
(0.300 ± 0.010)
0.255 (0.075)
0.010 (0.003)
4.19 MAX.
(0.165)
1.080 ± 0.320
(0.043 ± 0.013)
0.635 ± 0.25
(0.025 ± 0.010)
0.51 ± 0.130
(0.020 ± 0.005)
2.540
(0.100)
BSC
12° NOM.
DIMENSIONS IN MILLIMETERS (INCHES).
TOLERANCES (UNLESS OTHERWISE SPECIFIED): xx.xx = 0.01
xx.xxx = 0.005
LEAD COPLANARITY
MAXIMUM: 0.102 (0.004)
Package Characteristics
Unless otherwise noted, all specifications are at TA = +25°C.
Parameter
Input-Output Momentary
Withstand Voltage
(See note ** below)
Resistance (Input - Output)
Capacitance
(Input - Output)
Input IC Junction-to-Case
Thermal Resistance
Output IC Junction-to-Case
Thermal Resistance
Symbol Min.
VISO
3750
Typ.
1012
1011
1013
Ω
CI-O
0.7
pF
θjci
96
°C/W
θjco
114
°C/W
RI-O
Max. Units
Vrms
Test Conditions
RH ≤ 50%, t = 1 min.
VI-O = 500 Vdc
TA = 100°C
f = 1 MHz
Note
14,15
15
Thermocouple located at
center underside of
package
** The Input-Output Momentary Withstand Voltage is a dielectric voltage rating that should not be interpreted as an input-output
continuous voltage rating. For the continuous voltage rating refer to your equipment level safety specification or HP Application Note
1074, Optocoupler Input-Output Endurance Voltage.
1-268
TEMPERATURE – °C
Maximum Solder Reflow Thermal Profile
260
240
220
200
180
160
∆T = 145°C, 1°C/SEC
∆T = 115°C, 0.3°C/SEC
140
120
100
80
60
40
20
0
∆T = 100°C, 1.5°C/SEC
0
1
2
3
4
5
6
7
8
9
10
11
12
TIME – MINUTES
(NOTE: USE OF NON-CHLORINE ACTIVATED FLUXES IS RECOMMENDED.)
Regulatory Information
The HCPL-7860 (isolated modulator) has been approved by the following organizations:
UL
Recognized under UL 1577,
Component Recognition
Program, File E55361.
VDE (Pending)
Approved under VDE 0884/06.92
with VIORM = 848 VPEAK.
CSA
Approved under CSA Component
Acceptance Notice #5, File CA
88324.
VDE 0884 Insulation Characteristics
Description
Installation classification per DIN VDE 0110/1.89, Table 1
for rated mains voltage ≤ 300 Vrms
for rated mains voltage ≤ 600 Vrms
Climatic Classification
Pollution Degree (DIN VDE 0110/1.89)
Maximum Working Insulation Voltage
Input to Output Test Voltage, Method b*
VIORM x 1.875 = VPR, 100% Production Test with tm = 1
sec, Partial Discharge < 5 pC
Input to Output Test Voltage, Method a*
VIORM x 1.5 = VPR, Type and Sample Test, tm = 60 sec,
Partial Discharge < 5 pC
Highest Allowable Overvoltage
(Transient Overvoltage tini = 10 sec)
Safety-Limiting Values–Maximum Values Allowed in the
Event of a Failure, also see Figure 13.
Case Temperature
Input Power
Output Power
Insulation Resistance at TSI, VIO = 500 V
Symbol
Characteristic
Unit
VIORM
I - IV
I - III
40/85/21
2
848
V PEAK
VPR
1590
V PEAK
VPR
1273
V PEAK
VIOTM
6000
V PEAK
TS
IS, INPUT
PS, OUTPUT
RS
175
80
250
≥ 109
°C
mW
mW
Ω
*Refer to the optocoupler section of the Optoelectronics Designer's Catalog, under Product Safety Regulations section, (VDE 0884)
for a detailed description of Method a and Method b partial discharge test profiles.
Note: Isolation characteristics are guaranteed only within the safety maximum ratings which must be ensured by protective circuits in
application.
1-269
300
PSi – POWER – mW
250
PSi, OUTPUT
200
PSi, INPUT
150
MAX. OPERATING
TEMP. IS 100 °C
100
50
0
0
50
100
150
200
TA – TEMPERATURE – °C
Figure 13. Dependence of SafetyLimiting Values on Temperature.
Insulation and Safety Related Specifications
Parameter
Minimum External Air Gap
(Clearance)
Minimum External Tracking
(Creepage)
Minimum Internal Plastic Gap
(Internal Clearance)
Symbol Value Units
L(I01)
7.4
mm
L(I02)
Tracking Resistance
(Comparative Tracking Index)
Isolation Group
CTI
8.0
mm
0.5
mm
175
Volts
IIIa
Conditions
Measured from input terminals to output
terminals, shortest distance through air.
Measured from input terminals to output
terminals, shortest distance path along body
Insulation thickness between emitter and
detector; also known as distance through
insulation.
DIN IEC 112/VDE 0303 Part 1
Material Group (DIN VDE 0110, 1/89, Table 1)
Option 300 - surface mount classification is Class A in accordance with CECC 00802.
Absolute Maximum Ratings
Parameter
Storage Temperature
Ambient Operating Temperature
Supply Voltages
Steady-State Input Voltage
Two Second Transient Input Voltage
Output Voltages
Lead Solder Temperature
Solder Reflow Temperature Profile
Symbol
TS
TA
VDD1, VDD2
VIN+, VIN-
Min.
Max.
Units
Note
-55
125
°C
-40
+85
°C
0
5.5
Volts
-2.0
VDD1 + 0.5
Volts
16
-6.0
MCLK, MDAT
-0.5
VDD2 +0.5
Volts
260°C for 10 sec., 1.6 mm below seating plane
17
See Maximum Solder Reflow Thermal Profile section
Recommended Operating Conditions
Parameter
Ambient Operating Temperature
Supply Voltages
Input Voltage
1-270
Symbol
TA
VDD1, VDD2
VIN+, VIN-
Min.
-40
4.5
-200
Max.
+85
5.5
+200
Units
°C
V
mV
Note
16
Electrical Specifications, Isolated Modulator
Unless otherwise noted, all specifications are at VIN+ = 0 V and VIN- = 0 V, all Typical specifications are at
TA = 25°C and VDD1 = VDD2 = 5 V, and all Minimum and Maximum specifications apply over the following
ranges: TA = -40°C to +85°C, VDD1 = 4.5 to 5.5 V and VDD2 = 4.5 to 5.5 V.
Parameter
Symbol Min. Typ. Max. Units
Test Conditions
Fig. Note
Average Input Bias Current
IIN
-1.0
µA
14
18
Average Input Resistance
RIN
270
kΩ
Input DC Common-Mode
CMRRIN
55
dB
19
Rejection Ratio
Output Logic High Voltage
VOH
3.9
4.9
V
IOUT = -100 µA
Output Logic Low Voltage
VOL
0.3
0.6
V
IOUT = 1.6 mA
Output Short Circuit Current
|IOSC|
10
mA VOUT = VDD2 or GND2
20
Input Supply Current
IDD1
9.5
15
mA VIN+ = -350 mV
15
Output Supply Current
IDD2
8.8
15
mA to +350 mV
16
Output Clock Frequency
fCLK
9
11
14
MHz
17
Data Hold Time
tHDDAT
15
ns
21
Notes:
14. In accordance with UL1577, for devices with minimum VISO specified at 3750 Vrms, each isolated modulator (optocoupler) is
proof-tested by applying an insulation test voltage greater than 4500 Vrms for one second (leakage current detection limit
II-O < 5 µa). This test is performed before the Method b, 100% production test for partial discharge shown in VDE 0884
Insulation Characteristics Table.
15. This is a two-terminal measurement: pins 1-4 are shorted together and pins 5-8 are shorted together.
16. If VIN- (pin 3) is brought above VDD1 - 2 V with respect to GND1 an internal optical-coupling test mode may be activated. This test
mode is not intended for customer use.
17. HP recommends the use of non-chlorinated solder fluxes.
18. Because of the switched-capacitor nature of the isolated modulator, time averaged values are shown.
19. CMRRIN is defined as the ratio of the gain for differential inputs applied between VIN+ and VIN- to the gain for common-mode
inputs applied to both VIN+ and VIN- with respect to input ground GND1.
20. Short-circuit current is the amount of output current generated when either output is shorted to VDD2 or GND2. Use under these
conditions is not recommended.
21. Data hold time is amount of time that the data output MDAT will stay stable following the rising edge of output clock MCLK.
1-271
1
10.5
0
-1
10.0
-40 °C
25 °C
85 °C
-3
IDD1 – mA
IIN – mA
-2
-4
-5
9.5
9.0
-6
-7
8.5
-8
-9
-6
-4
-2
0
2
4
8.0
-400
6
Figure 15. IDD1 vs. VIN.
11.05
9.4
CLOCK FREQUENCY – MHz
9.2
IDD2 – mA
9.0
8.8
8.6
8.4
-40 °C
25 °C
85 °C
8.2
-200
0
VIN – mV
Figure 16. IDD2 vs. VIN.
1-272
400
200
VIN – mV
Figure 14. IIN vs. VIN.
8.0
-400
0
-200
VIN – V
200
400
11.00
10.95
10.90
VDD1 = 4.5 V
VDD1 = 5.0 V
VDD1 = 5.5 V
10.85
10.80
-40
-20
0
20
40
60
85
TEMPERATURE – °C
Figure 17. Clock Frequency vs. Temperature.
Digital Interface IC
Ordering Information
Specify Part Number followed by Option Number (if desired).
Example
HCPL-7870
Standard 16-pin DIP package, 25 per tube.
HCPL-0870#XXX
No Option = Standard 16-pin SO package, 47 per tube.
500 = Tape and Reel Packaging Option, 1000 per reel.
Option data sheets available. Contact Hewlett-Packard sales representative or authorized distributor.
Package Outline Drawings
Standard 16-pin DIP Package
15
16
14
13
12
11
10
9
TYPE NUMBER
R 0.030 x 0.030 DP
DATE CODE
HP 7870
YYWW
1
2
3
4
5
6
7
8
0.310 ± 0.010
(OUTER TO OUTER)
0.754
0.258
7°
7°
0.130
0.150 ± 0.010
0.130 ± 0.010
0.060
0.060
0.260
0.010 ± 0.002
0.018 ± 0.003
0.060
0.100 ± 0.010
0.310/0.380
(CENTER TO CENTER)
DIMENSIONS IN INCHES.
TOLERANCES (UNLESS OTHERWISE SPECIFIED): xx.xx = ± 0.01
xx.xxx = ± 0.002
1-273
Standard 16-pin SO Package
TOP VIEW
0.33 x 45°
(0.013 x 45°)
BOTTOM VIEW
PIN NO. 1 IDENTIFIER
∅ 1.27 (0.050) x 0.075 (0.003) DEPTH
SHINY SURFACE
16 15 14 13 12 11 10 9
7.544 ± 0.05
(0.297 ± 0.002)
10.00–10.65
(0.394–0.419)
(TIP TO TIP)
HP 0870
YYWW
1.27
(0.050)
1
2
3
4
5
6
7
1.90
(0.075)
1.90
(0.075)
∅ 1.27 (0.050)
x 0.075 (0.003)
DEPTH
(2x) EJECTOR PIN
SHINY SURFACE
XX
TH
8
1.27 (0.050)
SIDE VIEW
R 0.18 (R 0.007)
ALL CORNERS
AND EDGES
1.27 BSC
0.33–0.51
(0.050 BSC) (0.013–0.020)
1.016 ± 0.025
(0.040 ± 0.001)
END VIEW
0.10–0.30
(0.004–0.0118)
7°
PARTING
LINE
2.286
(0.090)
0.01 (0.004)
SEATING PLANE
10.21 ± 0.10
(0.402 ± 0.002)
A
0.23–0.32
(0.0091–0.0125)
2.386–2.586
(0.094–0.1018)
1.016 REF.
(0.040)
DIMENSIONS IN MILLIMETERS (INCHES).
TOLERANCES
(UNLESS OTHERWISE SPECIFIED): xx.xx = ± 0.010
xx.xxx = ± 0.002
0° – 8°
0.40 – 1.27
(0.016 – 0.050)
1-274
DETAIL A
TEMPERATURE – °C
Maximum Solder Reflow Thermal Profile
260
240
220
200
180
160
140
120
100
80
60
40
20
0
∆T = 145°C, 1°C/SEC
∆T = 115°C, 0.3°C/SEC
∆T = 100°C, 1.5°C/SEC
0
1
2
3
4
5
6
7
8
9
10
11
12
TIME – MINUTES
(NOTE: USE OF NON-CHLORINE ACTIVATED FLUXES IS RECOMMENDED.)
Absolute Maximum Ratings
Parameter
Storage Temperature
Ambient Operating Temperature
Supply Voltage
Input Voltage
Output Voltage
Lead Solder Temperature
Solder Reflow Temperature Profile
Symbol
Min.
Max.
Units
TS
-55
+125
°C
TA
-40
+85
°C
VDD
0
5.5
V
All Inputs
-0.5
VDD + 0.5
V
All Outputs
-0.5
VDD + 0.5
V
260°C for 10 seconds, 1.6 mm below seating plane
See Reflow Thermal Profile
Note
17
Note:
17. HP recommends the use of non-chlorinated solder fluxes.
Recommended Operating Conditions
Parameter
Ambient Operating Temperature
Supply Voltage
Input Voltage
Symbol
TA
VDD
All Inputs
Min.
-40
4.5
0
Max.
+85
5.5
VDD
Units
°C
V
V
Note
1-275
Electrical Specifications, Digital Interface IC
Unless otherwise noted, all Typical specifications are at TA = 25°C and VDD = 5 V, and all Minimum and
Maximum specifications apply over the following ranges: TA = -40°C to +85°C and VDD = 4.5 to 5.5 V.
Parameter
Symbol
Min. Typ. Max. Units Test Conditions Fig. Note
Supply Current
IDD
20
35
mA
fCLK = 10 MHz
DC Input Current
IIN
0.001
10
µA
Input Logic Low Voltage
VIL
0.8
V
Input Logic High Voltage
VIH
2.0
V
Output Logic Low Voltage
VOL
0.15
0.4
V
IOUT = 4 mA
Output Logic High Voltage
VOH
4.3
5.0
V
IOUT = -400 µA
Clock Frequency (CCLK,
fCLK
20
MHz
MCLK and SCLK)
Clock Period (CCLK,
tPER
50
ns
18,
MCLK and SCLK)
19
Clock High Level Pulse
tPWH
20
ns
Width (CCLK, MCLK
and SCLK)
Clock Low Level Pulse
tPWL
20
Width (CCLK, MCLK
and SCLK)
Setup Time from DAT to
tSUCLK
10
18
Rising Edge of CLK
(CDAT, CCLK, MDAT
and MCLK)
DAT Hold Time after
tHDCLK
10
Rising Edge of CLK
(CDAT, CCLK, MDAT
and MCLK)
Setup Time from Falling
tSUCL1
20
Edge of CLAT to First
Rising Edge of CCLK
Setup Time from Last
tSUCL2
20
Rising Edge of CCLK
to Rising Edge of CLAT
Delay Time from Falling
tDSDAT
15
19
Edge of SCLK to SDAT
Setup Time from Data
tSUS
200
Ready to First Falling
Edge of SCLK
Setup Time from CHAN
tSUCHS
20
to falling edge of CS
Reset High Level Pulse
tPWR
100
Width
1-276
tSUCL1
tSUCL2
CLAT
CDAT
B7
tSUCLK
B6
B5
B4
B3
B2
B1
B0
tHDCLK
tPWH
CCLK
tPWL
tPER
Figure 18. Serial Configuration Interface Timing.
CHAN
tSUCHS
CS
SDAT
B14
B13
B12
tDSDAT
1
SCLK
tC
2
tSUS
3
tPWL
B11
B10
B1
6
15
B0
tPWH
4
5
16
tPER
Figure 19. Conversion Timing.
1-277
Applications
Information
Product Description
The HCPL-7860 Isolated Modulator (optocoupler) uses sigmadelta modulation to convert an
analog input signal into a highspeed (10 MHz) single-bit digital
data stream; the time average of
the modulator’s single-bit data is
directly proportional to the input
signal. The isolated modulator’s
other main function is to provide
galvanic isolation between the
analog input and the digital
output. An internal voltage
reference determines the fullscale analog input range of the
modulator (approximately
± 320 mV); an input range of
± 200 mV is recommended to
achieve optimal performance.
The primary functions of the
HCPL-x870 Digital Interface IC
are to derive a multi-bit output
signal by averaging the single-bit
modulator data, as well as to
provide a direct microcontroller
interface. The effective resolution
of the multi-bit output signal is a
function of the length of time
(measured in modulator clock
cycles) over which the average is
taken; averaging over longer
periods of time results in higher
resolution. The Digital Interface
IC can be configured for five
conversion modes which have
different combinations of speed
and resolution to achieve the
desired level of performance.
Other functions of the HCPLx870 Digital Interface IC include
a Phase Locked Loop based pretrigger circuit that can either give
more precise control of the
effective sampling time or reduce
conversion time to less than 1 µs,
a fast over-range detection circuit
that rapidly indicates when the
magnitude of the input signal is
beyond full-scale, an adjustable
threshold detection circuit that
indicates when the magnitude of
the input signal is above a useradjustable threshold level, an
offset calibration circuit, and a
second multiplexed input that
allows a second Isolated
Modulator to be used with a
single Digital Interface IC.
The digital output format of the
Isolated A/D Converter is 15 bits
of unsigned binary data. The
input full-scale range and code
assignment is shown in Table 1
below. Although the output contains 15 bits of data, the effective
resolution is lower and is determined by selected conversion
mode as shown in Table 2 below.
Table 1. Input Full-Scale Range and Code Assignment.
Analog Input
Full Scale Range
Minimum Step Size
+Full Scale
Zero
-Full Scale
Voltage Input
640 mV
20 µV
+320 mV
0 mV
-320 mV
Digital Output
32768 LSBs
1 LSB
111111111111111
100000000000000
000000000000000
Table 2. Isolated A/D Converter Typical Performance Characteristics.
Conversion
Mode
1
2
3
4
5
Signal-toNoise Ratio
(dB)
83
79
73
66
53
Effective
Resolution
(bits)
13.5
12.8
11.9
10.7
8.5
Note: Bold italic type indicates Default values.
1-278
Conversion Time (µs)
Pre-Trigger Mode
0
1
2
188
94
95
47
37
18
0.7
19
10
10
5
Signal
Delay
(µs)
94
47
18
10
5
Signal
Bandwidth
(kHz)
3.4
6.9
22
45
90
Digital Interface
Timing
Power Up/Reset
At power up, the digital interface
IC should be reset either
manually, by bringing the RESET
pin (pin 9) high for at least
100 ns, or automatically by
connecting a 10 µF capacitor
between the RESET pin and VDD
(pin 16). The RESET pin operates
asynchronously and places the IC
in its default configuration, as
specified in the Digital Interface
Configuration section.
Conversion Timing
Figure 19 illustrates the timing
for one complete conversion
cycle. A conversion cycle is
initiated on the falling edge of the
convert start signal (CS); CS
should be held low during the
entire conversion cycle. When CS
is brought low, the serial output
data line (SDAT) changes from a
high-impedance to the low state,
indicating that the converter is
busy. A rising edge on SDAT
indicates that data is ready to be
clocked out. The output data is
clocked out on the negative edges
of the serial clock pulses (SCLK),
MSB first. A total of 16 pulses is
needed to clock out all of the data.
After the last clock pulse, CS
should be brought high again,
causing SDAT to return to a highimpedance state, completing the
conversion cycle. If the external
circuit uses the positive edges of
SCLK to clock in the data, then a
total of sixteen bits is clocked in,
the first bit is always high
(indicating that data is ready)
followed by 15 data bits. If fewer
than 16 cycles of SCLK are input
before CS is brought high, the
conversion cycle will terminate
and SDAT will go to the high-
impedance state after a few
cycles of the Isolated Modulator’s
clock.
The amount of time between the
falling edge of CS and the rising
edge of SDAT depends on which
conversion and pre-trigger modes
are selected; it can be as low as
0.7 µs when using pre-trigger
mode 2, as explained in the
Digital Interface Configuration
section.
Serial Configuration
Timing
The HCPL-x870 Digital Interface
IC is programmed using the
Serial Configuration Interface
(SCI) which consists of the clock
(CCLK), data (CDAT), and
enable/latch (CLAT) signals.
Figure 18 illustrates the timing
for the serial configuration interface. To send a byte of configuration data to the HCPL-x870, first
bring CLAT low. Then clock in
the eight bits of the configuration
byte (MSB first) using CDAT and
the rising edge of CCLK. After the
last bit has been clocked in,
bringing CLAT high again will
latch the data into the appropriate configuration register inside
the interface IC. If more than
eight bits are clocked in before
CLAT is brought high, only the
last eight bits will be used. Refer
to the Digital Interface Configuration section to determine appropriate configuration data. If the
default configuration of the
digital interface IC is acceptable,
then CCLK, CDIN and CLAT may
be connected to either VDD or
GND.
Channel Select Timing
The channel select signal (CHAN)
determines which input channel
will be used for the next conver-
sion cycle. A logic low level
selects channel one, a high level
selects channel 2. CHAN should
not be changed during a conversion cycle. The state of the CHAN
signal has no effect on the
behavior of either the over-range
detection circuit (OVR1) or the
adjustable threshold detection
circuit (THR1). Both OVR1 and
THR1 continuously monitor
channel 1 independent of the
CHAN signal. CHAN also does not
affect the behavior of the pretrigger circuit, which is tied to
the conversion timing of channel
1, as explained in the Digital
Interface Configuration section.
Digital Interface
Configuration
Configuration Registers
The Digital Interface IC contains
four 6-bit configuration registers
that control its behavior. The two
LSBs of any byte clocked into the
serial configuration port (CDAT,
CCLK, CLAT) are used as address
bits to determine which register
the data will be loaded into.
Registers 0 and 1 (with address
bits 00 and 01) specify the
conversion and offset calibration
modes of channels 1 and 2,
register 2 (address bits 10)
specifies the behavior of the
adjustable threshold circuit, and
register 3 (address bits 11)
specifies which pre-trigger mode
to use for channel 1. These
registers are illustrated in Table 3
below, with default values
indicated in bold italic type. Note
that there are several reserved
bits which should always be set
low and that the configuration
registers should not be changed
during a conversion cycle.
1-279
Table 3. Register Configuration.
Register
0
Configuration Data Bits
Bit 7
Bit 6
Bit 5
Bit 4
Bit 3
Channel 1 Conversion Mode
Channel 1
Offset Cal
High
High
Low
Low
Channel 2 Conversion Mode
1
2
Low
Channel 2
Offset Cal
Bit 2
Reserved
Address Bits
Bit 1
Bit 0
Low
Reserved
Low
Low
High
High
Threshold
Detection Time
Low
Low
Low
Threshold Level
Low
Low
High
High
Low
Pre-Trigger Mode
Low
Low
Low
Reserved
Low
High
Low
Low
Low
Low
High
High
3
Low
Low
Low
Note: Bold italic type indicates default values. Reserved bits should be set low.
Conversion Mode
The conversion mode determines
the speed/resolution trade-off for
the Isolated A/D converter. The
four MSBs of registers 0 and 1
determine the conversion mode
for the appropriate channel. The
bit settings for choosing a particular conversion mode are shown
in Table 4 below. See Table 2 for
Table 4. Conversion Mode Configuration.
Conversion
Mode
1
2
3
4
5
Bit 7
Low
Low
High
High
High
Configuration
Bit 6
High
Low
High
High
Low
Note: Bold italic type indicates default values.
1-280
Data Bits
Bit 5
Low
High
High
Low
High
Bit 4
High
High
Low
Low
Low
a summary of how performance
changes as a function of conversion mode setting. Combinations
of data bits not specified in Table
4 below are not recommended.
Pre-Trigger Mode
The pre-trigger mode refers to
the operation of a PLL-based
circuit that affects the sampling
behavior and conversion time of
the A/D converter when channel 1
is selected. The PLL pre-trigger
circuit has two modes of operation; the first mode allows more
precise control of the time at
which the analog input voltage is
effectively sampled, while the
second mode essentially
eliminates the time between when
the external convert start
command is given and when output data is available (reducing it
to less than 1 µs). A brief
description of how the A/D converter works with the pre-trigger
circuit disabled will help explain
how the pre-trigger circuit affects
operation when it is enabled.
ately following the convert start
command. The weighting function increases for half of the conversion cycle and then decreases
back to zero, at which time the
data ready signal is given,
completing the conversion cycle.
The analog signal is effectively
sampled at the peak of the
weighting function, half-way
through the conversion cycle.
This is the default mode.
If the convert start signal is
periodic (i.e., at a fixed frequency) and the PLL pre-trigger
circuit is enabled (pre-trigger
modes 1 or 2), either the peak of
the weighting function or the end
of the conversion cycle can be
aligned to the external convert
start command, as shown in
Figure 20. The Digital Interface
IC can therefore synchronize the
conversion cycle so that either
the beginning, the middle, or the
end of the conversion is aligned
with the external convert start
command, depending on whether
pre-trigger mode 0, 1, or 2 is
selected, respectively. The only
requirement is that the convert
start signal for channel 1 be
With the pre-trigger circuit is
disabled (pre-trigger mode 0),
Figure 20 illustrates the relationship between the convert start
command, the weighting function
used to average the modulator
data, and the data ready signal.
The weighted averaging of the
modulator data begins immedi-
periodic. If the signal is not
periodic and pre-trigger mode 1
or 2 is selected, then the pretrigger circuit will not function
properly.
An important distinction should
be made concerning the difference between conversion time
and signal delay. As can be seen
in Figure 20, the amount of time
from the peak of the weighting
function (when the input signal is
being sampled) to when output
data is ready is the same for all
three modes. This is the actual
delay of the analog signal through
the A/D converter and is independent of the “conversion time,”
which is simply the time between
the convert start signal and the
data ready signal. Because signal
delay is the true measure of how
much phase shift the A/D
converter adds to the signal, it
should be used when making
calculations of phase margin and
loop stability in feedback
systems.
There are different reasons for
using each of the pre-trigger
modes. If the signal is not
WEIGHTING
FUNCTION
CONVERT START – CS
DATA READY – SDAT
A) PRE-TRIGGER MODE 0
B) PRE-TRIGGER MODE 1
C) PRE-TRIGGER MODE 2
Figure 20. Pre-Trigger Modes 0, 1, and 2.
1-281
periodic, then the pre-trigger
circuit should be disabled by
selecting pre-trigger mode 0. If
the most time-accurate sampling
of the input signal is desired,
then mode 1 should be selected.
If the shortest possible conversion time is desired, then mode 2
should be selected.
The pre-trigger circuit functions
only with channel 1; the circuit
ignores any convert start signals
while channel 2 is selected with
the CHAN input. This allows
conversions on channel 2 to be
performed between conversions
on channel 1 without affecting
the operation of the pre-trigger
circuit. As long as the convert
start signals are periodic while
channel 1 is selected, then the
pre-trigger circuit will function
properly.
The three different pre-trigger
modes are selected using bits 6
and 7 of register 3, as shown in
Table 5 below.
Table 5. Pre-Trigger Mode Configuration.
Pre-Trigger Mode
0
1
2
Configuration Data Bits
Bit 7
Bit 6
Low
Low
Low
High
High
Don’t Care
Note: Bold italic type indicates default values.
Offset Calibration
The offset calibration circuit can
be used to separately calibrate
the offsets of both channels 1 and
2. The offset calibration circuit
contains a separate offset register
for each channel. After an offset
calibration sequence, the offset
registers will contain a value
equal to the measured offset,
which will then be subtracted
from all subsequent conversions.
A hardware reset (bringing the
RESET pin high for at least
100 ns) is required to reset the
offset calibration registers to
zero.
The following sequence is
recommended for performing an
offset calibration:
1. Select the appropriate channel
using the CHAN pin (low =
channel 1, high = channel 2).
2. Force zero volts at the input of
the selected isolated
modulator.
3. Send a configuration data byte
to the appropriate register for
1-282
the selected channel (register
0 for channel 1, register 1 for
channel 2). Bit 3 of the
configuration byte should be
set high to enable offset
calibration mode and bits 4
through 7 should be set to
select conversion mode 1 to
achieve the highest resolution
measurement of the offset.
4. Perform one complete conversion cycle by bringing CS low
until SDAT goes high, indicating completion of the conversion cycle. Because bit 3 of the
configuration has been set
high, the uncalibrated output
data from the conversion will
be stored in the appropriate
offset calibration register and
will be subtracted from all
subsequent conversions on
that channel. If multiple
conversion cycles are
performed while the offset
calibration mode is enabled,
the uncalibrated data from the
last conversion cycle will be
stored in the offset calibration
register.
5. Send another configuration
byte to the appropriate register for the selected channel,
setting bit 3 low to disable
calibration mode and setting
bits 4 through 7 to select the
desired conversion mode for
subsequent conversions on
that channel.
To calibrate both channels,
perform the above sequence for
each channel. The offset
calibration sequence can be
performed as often as needed.
The table below summarizes how
to turn the offset calibration
mode on or off using bit 3 of
configuration registers 0 and 1.
Table 6. Offset Calibration
Configuration.
Offset
Calibration
Mode
Off
On
Configuration
Data Bits
Bit 3
Low
High
Note: Bold italic type indicates default
values.
Over-Range Detection
The over-range detection circuit
allows fast detection of when the
magnitude of the input signal on
channel 1 is near or beyond full
scale, causing the OVR1 output to
go high. This circuit can be very
useful in current-sensing applications for quickly detecting when a
short-circuit occurs. The overrange detection circuit works by
detecting when the modulator
output data has not changed state
for at least 25 clock cycles in a
row, indicating that the input
signal is near or beyond fullscale, positive or negative.
Typical response time to overrange signals is less than 3 µs.
The over-range circuit actually
begins to indicate an over-range
condition when the magnitude of
the input signal exceeds approximately 250 mV; it starts to
generate periodic short pulses on
OVR1 which get longer and more
frequent as the input signal
approaches full scale. The OVR1
output stays high continuously
when the input is beyond full
scale.
The over-range detection circuit
continuously monitors channel 1
independent of which channel is
selected with the CHAN signal.
This allows continuous monitoring of channel 1 for faults while
converting an input signal on
channel 2.
Adjustable Threshold
Detection
The adjustable threshold detector
causes the THR1 output to go
high when the magnitude of the
input signal on channel 1 exceeds
a user-defined threshold level.
The threshold level can be set to
one of 16 different values
between approximately 160 mV
and 310 mV. The adjustable
threshold detector uses a smaller
version of the main conversion
circuit in combination with a
digital comparator to detect when
the magnitude of the input signal
on channel 1 is beyond the
defined threshold level. As with
the main conversion circuit, there
is a trade-off between speed and
resolution with the threshold
detector; selecting faster detection times exhibit more noise as
the signal passes through the
threshold, while slower detection
times offer lower noise. Both the
detection time and threshold level
are programmable using bits 2
through 7 of configuration
register 2, as shown in Tables 7
and 8 below.
As with the over-range detector,
the adjustable threshold detector
continuously monitors channel 1
independent of which channel is
selected with the CHAN signal.
This allows continuous monitoring of channel 1 for faults while
converting Channel 2.
Table 7. Threshold
Detection Configuration.
Threshold
Detection
Time
2 - 6 µs
3 - 10 µs
5 - 20 µs
10 - 35 µs
Configuration
Data Bits
Bit 7
Bit 6
Low
Low
Low
High
High
Low
High
High
Note: Bold italic type indicates default
values.
Table 8. Threshold Level Configuration.
Threshold Level
± 160 mV
± 170 mV
± 180 mV
± 190 mV
± 200 mV
± 210 mV
± 220 mV
± 230 mV
± 240 mV
± 250 mV
± 260 mV
± 270 mV
± 280 mV
± 290 mV
± 300 mV
± 310 mV
Configuration Data Bits
Bit 5
Bit 4
Bit 3
Bit 2
Low
Low
Low
Low
Low
Low
Low
High
High
Low
High
High
Low
Low
High
High
Low
High
High
Low
Low
Low
High
High
Low
High
High
Low
Low
High
High
Low
High
Note: Bold italic type indicates default values.
1-283
Analog Interfacing
Power Supplies and
Bypassing
The recommended application
circuit is shown in Figure 21. A
floating power supply (which in
many applications could be the
same supply that is used to drive
the high-side power transistor) is
regulated to 5 V using a simple
zener diode (D1); the value of
resistor R1 should be chosen to
supply sufficient current from the
existing floating supply. The
voltage from the current sensing
resistor or shunt (Rsense) is
applied to the input of the HCPL7860 (U2) through an RC antialiasing filter (R2 and C2). And
finally, the output clock and data
of the isolated modulator are
connected to the digital interface
IC. Although the application
circuit is relatively simple, a few
recommendations should be
followed to ensure optimal
performance.
The power supply for the isolated
modulator is most often obtained
from the same supply used to
power the power transistor gate
drive circuit. If a dedicated
supply is required, in many cases
it is possible to add an additional
winding on an existing transformer. Otherwise, some sort of
simple isolated supply can be
used, such as a line powered
transformer or a high-frequency
DC-DC converter.
An inexpensive 78L05 threeterminal regulator can also be
used to reduce the floating supply
voltage to 5 V. To help attenuate
high-frequency power supply
noise or ripple, a resistor or
inductor can be used in series
with the input of the regulator to
form a low-pass filter with the
regulator’s input bypass
capacitor.
As shown in Figure 21, 0.1 µF
bypass capacitors (C1 and C3)
should be located as close as
possible to the input and output
power-supply pins of the isolated
modulator (U2). The bypass
capacitors are required because
of the high-speed digital nature of
the signals inside the isolated
modulator. A 0.01 µF bypass
FLOATING
POSITIVE
SUPPLY
capacitor (C2) is also recommended at the input due to the
switched-capacitor nature of the
input circuit. The input bypass
capacitor also forms part of the
anti-aliasing filter, which is
recommended to prevent highfrequency noise from aliasing
down to lower frequencies and
interfering with the input signal.
PC Board Layout
The design of the printed circuit
board (PCB) should follow good
layout practices, such as keeping
bypass capacitors close to the
supply pins, keeping output
signals away from input signals,
the use of ground and power
planes, etc. In addition, the layout
of the PCB can also affect the
isolation transient immunity
(CMR) of the isolated modulator,
due primarily to stray capacitive
coupling between the input and
the output circuits. To obtain
optimal CMR performance, the
layout of the PC board should
minimize any stray coupling by
maintaining the maximum
possible distance between the
input and output sides of the
circuit and ensuring
+5V
HV+
GATE DRIVE
CIRCUIT
R1
D1
5.1 V
C1
0.1 µF
R2 39 Ω
MOTOR
+
C2
0.01 µF
RSENSE
VDD1
VDD2
VIN+
MCLK
VIN-
MDAT
GND1
GND2
HCPL-7860
C3
0.1 µF
CCLK
VDD
CLAT
CHAN
CDAT
SCLK
MCLK1
SDAT
MDAT1
CS
MCLK2
THR1
MDAT2
GND
OVR1
RESET
HCPL-X870
HV-
Figure 21. Recommended Application Circuit.
1-284
TO
CONTROL
CIRCUIT
that any ground or power plane
on the PC board does not pass
directly below or extend much
wider than the body of the
isolated modulator.
Shunt Resistors
The current-sensing shunt
resistor should have low
resistance (to minimize power
dissipation), low inductance (to
minimize di/dt induced voltage
spikes which could adversely
affect operation), and reasonable
tolerance (to maintain overall
circuit accuracy). Choosing a
particular value for the shunt is
usually a compromise between
minimizing power dissipation and
maximizing accuracy. Smaller
shunt resistances decrease power
dissipation, while larger shunt
resistances can improve circuit
accuracy by utilizing the full
input range of the isolated
modulator.
MOTOR OUTPUT POWER – HORSEPOWER
The first step in selecting a shunt
is determining how much current
the shunt will be sensing. The
graph in Figure 22 shows the
RMS current in each phase of a
three-phase induction motor as a
function of average motor output
power (in horsepower, hp) and
motor drive supply voltage. The
40
440
380
220
120
35
30
25
20
15
10
5
0
0
5
10
15
20
25
30
35
maximum value of the shunt is
determined by the current being
measured and the maximum
recommended input voltage of
the isolated modulator. The
maximum shunt resistance can be
calculated by taking the maximum recommended input voltage
and dividing by the peak current
that the shunt should see during
normal operation. For example, if
a motor will have a maximum
RMS current of 10 A and can
experience up to 50% overloads
during normal operation, then the
peak current is 21.1 A
(=10x1.414x1.5). Assuming a
maximum input voltage of
200 mV, the maximum value of
shunt resistance in this case
would be about 10 mΩ.
The maximum average power
dissipation in the shunt can also
be easily calculated by multiplying the shunt resistance times the
square of the maximum RMS
current, which is about 1 W in
the previous example.
If the power dissipation in the
shunt is too high, the resistance
of the shunt can be decreased
below the maximum value to
decrease power dissipation. The
minimum value of the shunt is
limited by precision and accuracy
requirements of the design. As
the shunt value is reduced, the
output voltage across the shunt is
also reduced, which means that
the offset and noise, which are
fixed, become a larger percentage
of the signal amplitude. The
selected value of the shunt will
fall somewhere between the
minimum and maximum values,
depending on the particular
requirements of a specific design.
MOTOR PHASE CURRENT – A (rms)
Figure 22. Motor Output Horsepower
vs. Motor Phase Current and Supply
Voltage.
When sensing currents large
enough to cause significant
heating of the shunt, the
temperature coefficient (tempco)
of the shunt can introduce
nonlinearity due to the signal
dependent temperature rise of the
shunt. The effect increases as the
shunt-to-ambient thermal
resistance increases. This effect
can be minimized either by
reducing the thermal resistance
of the shunt or by using a shunt
with a lower tempco. Lowering
the thermal resistance can be
accomplished by repositioning
the shunt on the PC board, by
using larger PC board traces to
carry away more heat, or by
using a heat sink.
For a two-terminal shunt, as the
value of shunt resistance
decreases, the resistance of the
leads becomes a significant
percentage of the total shunt
resistance. This has two primary
effects on shunt accuracy. First,
the effective resistance of the
shunt can become dependent on
factors such as how long the
leads are, how they are bent, how
far they are inserted into the
board, and how far solder wicks
up the lead during assembly
(these issues will be discussed in
more detail shortly). Second, the
leads are typically made from a
material such as copper, which
has a much higher tempco than
the material from which the
resistive element itself is made,
resulting in a higher tempco for
the shunt overall.
Both of these effects are eliminated when a four-terminal shunt
is used. A four-terminal shunt has
two additional terminals that are
Kelvin-connected directly across
the resistive element itself; these
two terminals are used to monitor
the voltage across the resistive
element while the other two
terminals are used to carry the
load current. Because of the
Kelvin connection, any voltage
1-285
drops across the leads carrying
the load current should have no
impact on the measured voltage.
Several four-terminal shunts from
Isotek (Isabellenhütte) suitable
for sensing currents in motor
drives up to 71 Arms (71 hp or
53 kW) are shown in Table 9; the
maximum current and motor
power range for each of the PBVseries shunts are indicated. For
shunt resistances from 50 mΩ
down to 10 mΩ, the maximum
current is limited by the input
voltage range of the isolated
modulator. For the 5 mΩ and
2 mΩ shunts, a heat sink may be
required due to the increased
power dissipation at higher
currents.
When laying out a PC board for
the shunts, a couple of points
should be kept in mind. The
Kelvin connections to the shunt
should be brought together under
the body of the shunt and then
run very close to each other to
the input of the isolated modulator; this minimizes the loop area
of the connection and reduces the
possibility of stray magnetic
fields from interfering with the
measured signal. If the shunt is
not located on the same PC board
as the isolated modulator circuit,
a tightly twisted pair of wires can
accomplish the same thing.
Also, multiple layers of the PC
board can be used to increase
current carrying capacity.
Numerous plated-through vias
should surround each non-Kelvin
terminal of the shunt to help
distribute the current between the
layers of the PC board. The PC
board should use 2 or 4 oz.
copper for the layers, resulting in
a current carrying capacity in
excess of 20 A. Making the
current carrying traces on the PC
board fairly large can also
improve the shunt’s power
dissipation capability by acting as
a heat sink. Liberal use of vias
where the load current enters and
exits the PC board is also
recommended.
Shunt Connections
The recommended method for
connecting the isolated modulator to the shunt resistor is shown
in Figure 21. VIN+ (pin 2 of the
HPCL-7860) is connected to the
positive terminal of the shunt
resistor, while VIN- (pin 3) is
shorted to GND1 (pin 4), with the
power-supply return path functioning as the sense line to the
negative terminal of the current
shunt. This allows a single pair of
wires or PC board traces to
connect the isolated modulator
circuit to the shunt resistor. By
referencing the input circuit to
the negative side of the sense
resistor, any load current induced
noise transients on the shunt are
seen as a common-mode signal
and will not interfere with the
current-sense signal. This is
important because the large load
currents flowing through the
motor drive, along with the
parasitic inductances inherent in
the wiring of the circuit, can
generate both noise spikes and
offsets that are relatively large
compared to the small voltages
that are being measured across
the current shunt.
If the same power supply is used
both for the gate drive circuit and
for the current sensing circuit, it
is very important that the connection from GND1 of the isolated
modulator to the sense resistor
be the only return path for
supply current to the gate drive
power supply in order to
eliminate potential ground loop
problems. The only direct connection between the isolated
modulator circuit and the gate
drive circuit should be the
positive power supply line.
Table 9. Isotek (Isabellenhütte) Four-Terminal Shunt Summary.
Shunt Resistor
Part Number
PBV-R050-0.5
PBV-R020-0.5
PBV-R010-0.5
PBV-R005-0.5
PBV-R002-0.5
Shunt
Resistance
mΩ
50
20
10
5
2
Tol.
%
0.5
0.5
0.5
0.5
0.5
Note: Values in brackets are with a heatsink for the shunt.
1-286
Maximum
RMS Current
A
3
7
14
25 [28]
39 [71]
Motor Power Range
120 Vac-440 Vac
hp
kW
0.8-3
0.6-2
2-7
1.4-5
4-14
3-10
7-25 [8-28]
5-19 [6-21]
11-39 [19-71]
8-29 [14-53]
In some applications, however,
supply currents flowing through
the power-supply return path may
cause offset or noise problems. In
this case, better performance
may be obtained by connecting
VIN+ and VIN- directly across the
shunt resistor with two conductors, and connecting GND1 to the
shunt resistor with a third
conductor for the power-supply
return path, as shown in Figure
23. When connected this way,
both input pins should be
bypassed. To minimize electromagnetic interference of the
sense signal, all of the conductors
(whether two or three are used)
connecting the isolated modulator to the sense resistor should be
either twisted pair wire or closely
spaced traces on a PC board.
The 39 Ω resistor in series with
the input lead (R2) forms a lowpass anti-aliasing filter with the
0.01 µF input bypass capacitor
(C2) with a 400 kHz bandwidth.
The resistor performs another
important function as well; it
dampens any ringing which might
be present in the circuit formed
by the shunt, the input bypass
capacitor, and the inductance of
wires or traces connecting the
two. Undamped ringing of the
input circuit near the input
sampling frequency can alias into
the baseband producing what
might appear to be noise at the
output of the device.
FLOATING
POSITIVE
SUPPLY
HV+
Voltage Sensing
The HCPL-7860 Isolated Modulator can also be used to isolate
signals with amplitudes larger
than its recommended input
range with the use of a resistive
voltage divider at its input. The
only restrictions are that the
impedance of the divider be
relatively small (less than 1 kΩ)
so that the input resistance
(280 kΩ) and input bias current
(1 µA) do not affect the accuracy
of the measurement. An input
bypass capacitor is still required,
although the 39 Ω series damping
resistor is not (the resistance of
the voltage divider provides the
same function). The low-pass
filter formed by the divider
resistance and the input bypass
capacitor may limit the achievable
bandwidth. To obtain higher
bandwidth, the input bypass
capacitor (C2) can be reduced,
but it should not be reduced
much below 1000 pF to maintain
adequate input bypassing of the
isolated modulator.
GATE DRIVE
CIRCUIT
R1
D1
5.1 V
C1
0.1 µF
R2a 39 Ω
R2b 39 Ω
MOTOR
+
C2b
C2a
0.01 µF 0.01 µF
RSENSE
VDD1
VDD2
VIN+
MCLK
VIN-
MDAT
GND1
GND2
HCPL-7860
HV-
Figure 23. Schematic for Three Conductor Shunt Connection.
1-287
Similar pages