A5191Hrt Hart Modem: Application Note
A5191Hrt Hart Modem: Application Note
A5191Hrt Hart Modem: Application Note
1.0 Introduction
HART is a registered trademark of the HART Communication Foundation of Austin, Texas, USA. Any time that the term 'HART' is used
hereafter in this document or in any document referenced by this document, that term implies the registered trademark.
This application note describes a demonstration circuit that permits a user to implement a HART slave or master interface between a
microprocessor and a process loop using the AMIS A5191HRT HART modem integrated circuit. The information in this application note
is correct to the best of our knowledge. Development of a circuit suitable to the user's particular system and application environment is
the responsibility of the user.
The HART (highway addressable remote transducer) communication protocol provides digital communication for microprocessor-based
process control instruments. HART uses the Bell-202 forward channel signaling frequencies and bit rate (1200 bits/second) as making it
a subset of the Bell-202 standard. HART-speaking devices can use virtually any Bell-202 standard modem. However, the AMIS
A5191HRT single-chip modem has been designed to meet the low power requirements of 2-wire process instruments.
The AMIS A5191HRT modem is designed to allow the user to easily implement a HART compliant physical layer design (HART FSK
Physical Layer Specification; HCF_SPEC-54, Revision 8.1 November 24, 1999). The A5191HRT is intended to replace the LSI/NCR
20C15 for all existing and future HART applications. The A5191HRT is a near pin-for-pin replacement for the 20C15. The A5191HRT
can be used in any circuit that uses the LSI 20C15 with no circuit topology changes. Only the values of four external resistors in the
receive filter need to be changed.
This application note explains how to interface the A5191HRT modem to the HART network, as well as other general advice on using
the modem and on designing HART devices. A block diagram showing a typical application of the A5191HRT in a HART Slave is
shown in Figure 1.
HART devices are connected in a current loop arrangement shown in Figure 2. The process transmitter (field instrument in HART
specifications) signals by varying the amount of current flowing through itself. The controller (primary master) detects this current
variation by measuring the DC voltage across the current sense resistor. The loop current varies from 4 to 20mA at frequencies usually
under 10Hz.
The HART (digital) signal is superimposed on the 4-20mA (analog) signal as shown in Figure 3. The master transmits HART signals by
applying a voltage signal across the current sense resistor and it receives a voltage signal by detecting the HART current signal across
the sense resistor. Conversely, the slave transmits by modulating the loop current with HART signals and receives HART signals by
demodulating the loop current.
HART signals using phase-continuous frequency-shift-keying (FSK) at 1200 bits/second. Phase-continuous FSK requires the phase
angle of the mark (1200Hz) and the space (2200Hz) to remain continuous at the 1200Hz bit boundaries. A field instrument transmits a
HART signal by modulating a high-frequency carrier current of about 1mA p-p onto its normal output current. This is illustrated in Figure
3 for a 6mA analog signal.
HART slaves transmit by modulating the process 4-20mA DC loop current with a 1mA p-p AC current signal as shown in Figure 3.
Since the average value of the HART signal is zero, the DC value of the process loop remains unchanged. Receive circuits in a HART
slave device amplify, filter and demodulate the current signal.
Some current loops (called networks in HART documents) use only digital signaling. The field instrument current is fixed at 4mA or
some other convenient value, and only digital communication occurs. Up to 15 such field instruments with unique addresses of 1
through 15 may be connected in parallel. A device that is not multi-dropped will usually have its address set to 0.
HART masters transmit by driving the loop with a low impedance voltage source as shown in Figure 2. Regardless of whether a master
or field device is transmitting, a signal voltage of about 500mV p-p is developed across the conductors of the current loop (assuming a
500Ω current sense resistor), and is seen by both devices. Receive circuits in each device filter and demodulate the signal voltage.
In general, a HART primary master is the device that provides the communications between the control system (DCS) and the remote
process instruments with the intent to receive process information and perform maintenance operations. A HART network that has a
HART master interface integrated into the DCS will usually be configured as a primary master.
In general, a second HART interface connected to a network that contains a master will be a secondary master. An example of a
secondary master is a hand-held communicator that would be connected directly across a HART. Such a network may have a primary
master. A HART network can only have one primary and one secondary master connected at a time.
To reduce the design complexity of a multiple loop HART master, the physical layer can be multiplexed to two or more process loops.
This is usually done with analog switches that allow signals as high as 16V p-p to pass and exhibit an extremely low on resistance. The
added impedance of the switch directly affects the output impedance of the master device.
The multiplexer can switch only the HART signal or it can switch both the HART signal and the associated signal return. Switching the
signal return insures the physical layer interface will be non-intrusive to the HART network if a failure were to occur. Typically, the
analog switches connect to each process loop through a coupling capacitor (about 2.2µF).
The greatest disadvantage of multiplexing HART signals is the reduction in communication throughput to each slave device.
Because of the relatively low HART frequencies there is little cable attenuation and delay distortion. This results in very few restrictions
on constructing networks. The complete topology requirements and electrical requirements for HART devices are given in the HART
physical layer specification (2).
In most applications, HART communications can be performed up to a distance of 5000 feet (1500 meters) using existing field wiring for
a 2-wire process instrument.
Normally, one HART device talks while others listen. Talking means applying the modulated carrier to the network cable. A given device
applies carrier in one unbroken segment called a frame. Between frames the network is silent. Field instrument frames are usually
responses to commands by a master. Further information on network protocol is found in the HART data link layer specification (1).
* Note: On the 32 LQFP the analog ground and digital ground pins must be connected together externally and connected to the system power ground. The separate 32 LQFP
grounds help decrease the ground noise coupled into the chip. On the 28 PLCC version there are not separate analog and digital grounds. In the 28 PLCC these two grounds
are connected inside the package to minimize pin count.
4.1.1. Demodulator
The demodulator accepts an FSK signal at its IRXA input and reproduces the original modulating signal at its ORXD output shown in
Figure 5.
The modem uses shift frequencies of nominally 1200Hz (logical one, mark) and 2200Hz (logical zero, space). The bit rate is nominally
1200 bits/second. The output of the modulator ORXD is qualified with the carrier detect signal OCD. Therefore only IRXA signals large
enough to be detected (100mV p-p typical) by the carrier detect function will produce demodulated output at ORXD.
4.1.2. Modulator
The modulator accepts digital data in NRZ form at its ITXD input and generates the FSK modulated signal at its OTXA output. INRTS
must be a logic low for the modulator to be active.
The A5191HRT generates a HART compliant trapezoidal FSK modulated signal at its OTXA output. Shown below are actual transmit
signals from a A5191HRT.
The A5191HRT implements a carrier detect (CD) that is compliant with the HART physical layer specification which requires the
receiver to activate CD between the incoming signal levels of 80 and 120mV p-p . Carrier detect will be compliant when ICDREF -
IAREF = 0.08VDC. The circuit shown in the appendix is designed for a nominal CD level of 100mV p-p .
Internal to the A5191HRT, a comparator asserts a logic low if the IRXAC voltage is below ICDREF. The comparator output is fed into
the carrier detect block, which asserts the output pin OCD to a logic high if INRTS is a logic high, and four consecutive pulses out of the
comparator have arrived. The carrier detect output OCD stays at a logic high as long as INRTS is at a logic high and the next
comparator pulse is received in less than 2.5ms. Once OCD goes inactive to a logic low, it takes four consecutive pulses out of the
comparator to set OCD to a logic high again. Four consecutive pulses amount to 3.33ms when the received signal is 1200Hz and to
1.82ms when the received signal is 2200Hz.
The A5191HRT requires an external 460.8kHz clock, ceramic resonator, or crystal, accurate to at least 1%. The oscillator requires two
external capacitors and one external resistor, which varies depending on the type of the crystal/resonator. The specifications are listed
in the two tables below. The A5191HRT has an internal oscillator cell, only requiring an external ceramic resonator and two capacitors.
The oscillator cell is designed to use either a crystal, ceramic resonator or an external clock. When using a ceramic resonator or crystal,
care should be taken to keep the circuit board traces between the A5191HRT and the external oscillator components as short as
possible.
Ceramic resonators are less expensive than crystals, but are not as accurate. Unfortunately, ceramic resonators at the needed
frequency require special ordering in very large quantities. Ceramic resonators that oscillate at 460kHz are available from:
It may be desirable to use an external clock (460.8kHz) rather than the internal oscillator cell because of the cost and availability of
ceramic resonators. In addition, the A5191HRT consumes less current when an external clock is used as shown below in Figure 8. An
external clock associated with the microprocessor (running at a frequency that is a multiple of 460.8kHz) can be used as an input to the
oscillator cell. The interface between the microprocessor clock and the A5191HRT could be as simple as a direct connection or a single
integrated circuit.
If you use the same time base for both the modem and the UART, a 1% accurate time base may not be good enough. The problem is a
combination of receive data jitter and clock skew between transmitting and receiving HART devices. If the transmit time base is at 99%
of nominal and the receive time base in another device is at 101% of nominal, the receive data (at the receiving UART) will be skewed
by roughly 21% of one bit time at the end of each 11-bit byte. This is shown below in Figure 9. The skew time is measured from the
initial falling edge of the start bit to the center of the 11th bit cell.
This 21% skew by itself isn't bad. However, there is another error source for bit boundary jitter. The phase lock loop demodulator in the
A5191HRT produces jitter in the receive data that can be as large as 12% of a bit time. Therefore, a bit boundary can be shifted by as
much as 24% of a bit time relative to its ideal location based on the start-bit transition. (The start-bit transition and a later transition can
be shifted in opposite directions for a total of 24%.)
The clock skew and jitter added together is 45%, which is the amount that a bit boundary could be shifted from its expected position.
UARTs that sample at mid-bit will not be affected. However, there are UARTs that take multiple samples during each bit to try to
improve on error performance. These UARTs may not be satisfactory, depending on how close the samples are to each other, and how
samples are interpreted. A UART that takes a majority vote of three samples is acceptable.
Even if your own time base is perfect, you still must plan on a possible 35% shift in a bit boundary since you don't have control over
time bases in other HART devices.
To remove the interfering analog signal, a high-pass filter is required in the HART signal receive path. The filter requirements are found
as follows. From section 7.1 of the HART physical layer specification, the interfering signal can be as high as 16V p-p at 25Hz. (NOTE:
This is directly related to limits on analog signaling. The difference in specifications for analog interference as output versus analog
interference as input is the result of the loads being different.) The interfering analog signal should be reduced to at least ten times
smaller than the smallest HART signal, or under about 7.5mV. Therefore, the high-pass filter should have an attenuation of 63dB at
25Hz. The HART signal band covers approximately 950Hz to 2500Hz, which means that the high-pass filter should begin rolling off
somewhere below 950Hz and be 63dB down at 25Hz. This is illustrated in Figure 10.
The A5191HRT has an internal active filter to attenuate the frequencies outside the HART bandpass. In addition to the internal active
filter, an external passive filter is necessary to complete the filtering requirements. The external capacitors and resistor were too large in
size to cost effectively integrate into the A5191HRT silicon.
The external components required for the receive filter is shown in Figure 11. All the external capacitors are ±5% and the resistors are
±1% components (except the 3MΩ is ±5%).
The external components on the receiver create a three pole high pass filter at 624Hz and a one pole low pass filter at 2500Hz.
Internally, the A5191HRT has a high pass pole at 35Hz and a low pass pole at 90kHz, each of which can vary by as much as ±30%.
The input impedance to the entire filter is greater than 150KΩ at frequencies less than 50kHz.
In Figure 11 there are four resistors circled. The value of these resistors is different than those used in the LSI 20C15. The four resistor
values were changed because the A5191HRT is not exactly the same as the 20C15. They are both made using 0.5µm CMOS
processes and the general architectures of the two chips are identical. However, the receive filter amplifier and receive comparator (see
Figure 8) in the A5191HRT are slightly different than in the 20C15. It is impossible to make an analog part identical in two different
processes in two different fabrication facilities.
The biggest difference between the two parts and the major contributor to the need to change the four resistor values when using the
A5191HRT are in the receive demodulator. This block is quite different in the A5191HRT. This section was changed because even
though the 20C15 passed all the conformance tests, it barely passed. The different resistor values change the shape of the lower pass
band for the receive filter. These changes provide more margin in passing the out-of-band noise interference tests. This in turn makes
for a more robust and noise immune receiver.
The values shown in Figure 11 for R9 and R10 may be different than those used in some 20C15 applications. The ratio of these two
resistors sets the receive threshold. In some 20C15 designs the value of R9 is 15k and/or the value of R10 may be 215k. The exact
values may differ from design to design. The values shown for all external components in Figure 11 and all other circuits in this
application note are those used in the circuitry which was used to pass the HART physical layer conformance test for the A5191HRT.
IAREF sets the DC operating point of the internal op-amps and comparators and is usually selected to split the DC potential between
VDD and VSS. The A5191HRT requires a voltage reference (low power if necessary) with an output voltage level between 1.2 and
2.6VDC. In the case of a 5VDC supply, IAREF is typically 2.5VDC. When operating as low as 3.0VDC, a 1.235VDC reference (Analog
Devices AD589) is suitable as IAREF.
The level at which CD (carrier detect) becomes active is determined by the DC voltage difference between ICDREF and IAREF.
Selecting IAREF - ICDREF = 0.08VDC will set the carrier detect to a nominal 100mV p-p.
The A5191HRT requires a bias current set resistor tied between OCBIAS and VSS. The bias current controls the operating parameters
of the internal op-amps and comparators. The value of the bias current set resistor is determined by the reference voltage IAREF.
The recommended value of the bias current set resistor is 499kΩ when IAREF = 1.235VDC.
Rbias = IAREF/2.5mA
Current consumption of internal circuits is important in any two-wire field instrument. It becomes critically important in a microprocessor-
based field instrument featuring both analog and digital signaling. The available current is derived here and some techniques are
examined for reducing current consumption.
The nominal HART signal transmitted by a field instrument is 0.5mA peak. When this is superimposed upon a 4mA analog signal, the
terminal current must vary between 3.5mA and 4.5mA. During the peak of the HART waveform, the instrument has 4.5mA available,
but during the valley it has only 3.5mA. Energy storage techniques can be used to allow the internal circuits to draw a steady 4mA at all
times. However, to be effective at HART frequencies, the storage capacitor is quite large. A large capacitor (or any form of energy
storage device) complicates the circuit design if intrinsic safety is required. Therefore, normally only 3.5mA is used to run everything.
Another 200 to 400µA is often subtracted from this to allow some margin and to satisfy other conditions. Assuming that the value is
200µA, during transmit the internal circuits of the two-wire field instrument have to live on a diet of 3.3mA. During receive there is 3.8mA
available.
A characteristic of the A5191HRT is that its current consumption is approximately 350 to 450µA. This leads to the fortunate
circumstance that the remaining (non-modem) circuits always have at least 3.25mA available. Margins may be needed to cover current
consumption over temperature. One way of reducing the A5191HRT current consumption is to operate it at reduced voltage. Since the
A5191HRT is a CMOS part, current consumption is roughly proportional to supply voltage. This is evident in the graph contained in the
appendix. Operation at 3.5V is common.
In applying the A5191HRT be careful not to let inputs float. The IRTS pin of the A5191HRT will often be driven by an I/O pin of a
microprocessor. During power-up or after reset an I/O pin may be tristated, allowing it to float. This can cause the A5191HRT to draw
excess current. In some cases the current may be large enough to prevent the field instrument from starting up properly. There should
be a 1MΩ pullup resistor on IRTS.
The transmitter lift-off voltage is the minimum terminal voltage at which it is guaranteed to operate. When only analog signaling is
involved, liftoff voltage is an unambiguous quantity. But when HART digital signaling is added, the available voltage can swing by as
much as 0.75V above and below the DC level. The minimum applied voltage is the DC level minus 0.75V. You should either design the
field instrument to accommodate the dips in voltage, or else specify the lift-off voltage to include them.
The slave interface to the network is typically a current regulator, as illustrated in simplified form in Figure 12. Its output current is
controlled by varying a much smaller current into an op-amp summing junction. This junction is a convenient point at which to sum the
analog and digital signals, thereby achieving superposition of the digital signal onto the analog. The digital transmit signal (OTXA) is
capacitively coupled into the summing junction to preserve DC accuracy of the 4-20mA analog signal. Also shown in Figure 12 is the
receive path, consisting of a capacitor from the collector of the current regulator transistor to a high-impedance amplifier. Both the
receive amplifier and the transmit current source should present a high impedance (greater than 100kΩ) to the network.
The amplitude of OTXA is nominally 500mV p-p. The HART physical layer requires the slave to modulate the loop with a 1mA p-p
signal. The amplitude of modulation of the loop current will have to be adjusted before the amplifier in the current loop regulator or at
the regulator itself. It is recommended that the amplitude can be adjusted with voltage gain circuits. Using a small series capacitance to
attenuate the signal may cause distortion of the transmitted signal.
All HART receivers require non-disruptive coupling to the network current loop. This connection can be made with capacitive or
inductive coupling without corruption of the process loop current. Typically a HART slave uses a small value (1000pF) capacitor to
couple an adequate signal level to the receive filter.
The HART master interfaces to the network as a voltage source as illustrated in simplified form in Figure 13. The network is driven by a
low output impedance (600Ω or less) voltage amplifier circuit that can be switched to a high impedance state using INRTS. The output
of the voltage amplifier needs to be a high impedance while the HART master receives, to insure a high input impedance for the
receiver. Typically a HART master uses a 2.2µF coupling capacitor to insure the transmitter circuit meets the output impedance
requirements specified in the HART physical layer specification.
The amplitude of OTXA is nominally 500mV p-p. The 500mV p-p meets the requirements for a HART master. However, the A5191HRT
is unable to source enough current to drive a HART network, therefore a low impedance voltage driver is required between the
A5191HRT and the network.
During a HART message transmission from the master, the output impedance of the voltage source is quite low (0 - 600Ω). However
during reception of a HART message, the input impedance of the receive section must be quite high. Therefore, the transmitter section
must be disabled when not active (INRTS "high").
Several methods can be used to achieve a high output impedance of the transmitter while INRTS is inactive. One method is choosing
an opamp with an active low current programming resistor that will basically turn-off the op-amp during receive operations. A second
method can be implemented with a correctly biased discrete FET in series with the output stage. Lastly, a discrete solid-state relay with
a characteristically low on resistance can be used.
All three techniques will provide an adequate solution for the impedance requirements. However, only the third solution will allow proper
circuit operation per the bit-error-rate requirements and need not be powered by a dual supply. Under a worse case scenario, a
transmitter could dynamically change from 4 - 20mA and a HART master could have as much as 1000Ω of loop resistance. This results
in a 16V p-p low frequency carrier with HART superimposed. The transmit switch must be designed to prevent this 16V p-p signal from
sinking current through the HART transmitter's output stage. Any clipping of the 16V p-p signal through the HART transmitter or the use
of transient suppressors will result in the clipping of the superimposed HART signal. The last solution will block and not clip the 16V p-p
process signal and will not corrupt the input impedance. The third solution can be implemented easily as shown in Figure 21.
HART signal in which the master is always connected across a current sense resistor and share the same ground. Where ground
isolation is required, another coupling capacitor may be necessary in the signal return connection.
To insure complete isolation from the network, galvanic isolation is the preferred method. Typically a transformer and a series capacitor
are used to couple the HART signal. With this type of isolation, the master can be connected across the field instrument or the current
sense resistor regardless of the polarity or grounding configuration.
A HART master must be able to receive voltage signals that are developed across the sense impedance. In addition, it must be able to
apply a transmit signal across the current sense impedance. In some cases the master will signal across the field device as well. In all
cases, the presence of the HART master must not disrupt the analog signaling of the process loop. Therefore, the HART signals must
be non-intrusively coupled from the current sense resistor to the master through one of the various coupling techniques listed below.
Coupling to the network is most commonly done with capacitors and transformers.
Capacitive coupling can work well when masters have their power isolated from ground. When the system power configuration is
unknown, the use of transformer coupling will insure the elimination of ground effects. An example of transformer coupling is shown in
Figure 16.
If a master (that does not have an isolated power source) with single capacitive coupling is connected to a sense resistor or field device
that is not at ground potential, a DC ground loop will occur as shown in Figure 17.
If the master has a ground connection that is not at the same ground potential as the analog ground from the process loop, it is possible
to create AC ground loops. A noisy 110VAC ground signal could be coupled directly into the process loop ground (single capacitive
coupling) or coupled into the process loop (dual capacitive coupling) ground through the connection. The AC noise potentially could
impede HART communications and corrupt the accuracy of the DCS measuring the analog current as shown in Figure 18.
Band limiting the analog signal applies a HART slave and determines how fast it can control its output current. For a HART master the
information in this section is provided to better understand the analog and digital signaling characteristics of a HART master.
Digital signaling can potentially interfere with analog signaling. A much worse problem is the analog signal interfering with digital
signaling. This is due to the relative size of the two signals. For example, a change from 4 to 20mA can produce a voltage change of as
high as 16V across the field instrument terminals, depending on the amount of resistance in the loop. At the same time, the instrument
may be trying to detect a HART (digital) signal as low as 80mV p-p.
Separating the two is usually done with a combination of low-pass filtering of the analog signal and high-pass filtering of the HART
signal. The HART physical layer specification (section 7.2) limits the analog signal to 16mA p-p at frequencies below 25Hz and
constrains the output above 25Hz to fall within a -40dB/decade slope, as illustrated in Figure 19. This means, for example, that a
sinusoidal output at 25Hz must have an amplitude (into the 500Ω test load) of less than 8V p-p. It also means that a 25Hz square wave
of 8V p-p would not be acceptable, since its harmonics do not decrease at a -40dB rate.
The required roll-off of the analog signal can be achieved by various means, including:
1. Analog filters
2. Digital filters (software)
3. Inherent filtering in the instrument sensor
Often it is a combination of these. Any convenient method may be used to insure that changes in instrument output current fall within
the specification. If the field instrument uses a D/A converter (DAC) to generate its analog output, the output steps of the DAC must be
sufficiently small or else the high-frequency content of the steps must be removed by filtering. When large changes in the DAC output
occur due to calibration operations, caution must be used not to have these DAC changes active during a HART command or
response. Any large and fast current changes that fall outside the HART specifications can cause unreliable communications with the
HART master or slave.
The analog signaling of typical field instruments is usually band-limited to the range of 1 to 10Hz so that the roll-off starts well below the
25Hz point of Figure 19. This can lead to simpler low- pass filtering. For example, a single-pole filter that begins rolling off at 1Hz falls
below the curve of Figure 19.
8.0 Miscellaneous
Because the HART carrier must be started and stopped, the receiving UART and processor must become synchronized to each HART
frame. During the synchronization period at the beginning of the frame, it is normal for some transmitted bytes to be corrupted or lost.
Various HART requirements have been devised to insure that synchronization will occur. They are:
• The transmitting device must send at least five preamble bytes (hex FF).
• The transmitting device must load the first preamble byte into the transmit UART within five bit times of starting carrier.
• The receiving device must recognize carrier within 30 bit times.
Correct start-up is based on the fact that, during the stream of preamble bytes, the start bits (applied to the UART) are the only 0 bits.
Everything else (stop, data and parity bits) including idle time is a 1. Although some extra 0 bits may be generated by the start-up
transient, after a short time only valid start bits will remain and the device will be synchronized.
The UART can cause a possible problem at the end of frame. Normally, it is necessary to wait until the last bit of the last character of
the frame has been sent until carrier is turned off. The UART usually tells you it's empty which is an indication that you should turn
carrier off. However, UARTs differ in when they indicate empty. Some indicate empty at the time of the last shift clock -- that is,
simultaneously with the stop bit being shifted out. But if carrier were to be turned off at that time, the modem would cease transmitting
and the stop bit wouldn't be sent. Therefore, it is important to determine which kind of UART you have. If it behaves as described, then
you should wait at least one bit time after the empty indication until you turn off carrier. Note that adding a 1 bit delay can't do any harm,
even if it isn't needed.
To maximize speed, a field instrument will often be designed to begin its response frame immediately after the master's command
frame. These frames are close enough together that there is not enough time for carrier detect to drop out. This brings out an important
point regarding carrier detect: Its sole purpose is to indicate the presence of carrier so that receivers know that they have sufficient
signal to work with. Carrier detect is not intended to reliably indicate the start or end of a particular frame. Start of frame detection is a
data link layer function and must occur through examination of the frame content.
9.0 Appendix
For detailed information on the HART physical layer specification and requirements, see HCF_SPEC-54.
Note the RC circuit connected to the INRESET pin in Figure 20 and Figure 21. This is a form of POR. Other circuit solutions are
possible as long as the reset timing requirements shown in section 3 of the A5191HRT specification are met.
AMIS recommends using the R and C values depending on the VDD rise and settling times, and application.
9.1.1. Sample Schematic for HART Slave Physical Layer Circuit Implementation
Figure 20: Sample Schematic for HART Slave Physical Layer Circuit Implementation
Note that the four resistors circled with a dashed line above in Figure 20 specify different values than those used with the LSI 20C15. In all other respects, the circuit topology
and all other component values are identical to the circuit used with the LSI 20C15.
For detailed information on the HART physical layer specification and requirements, see HCF_SPEC-54.
9.2.2. Sample Schematic for HART Master Physical Layer Circuit Implementation
Figure 21: Sample Schematic for HART Master Physical Layer Circuit Implementation
Note that the four resistors circled with a dashed line above in Figure 21 specify different values than those used with the LSI 20C15. In all other respects, the circuit topology
and all other component values are identical to the circuit used with the LSI 20C15.
In Figure 22 the current consumption data is assumes a 499kΩ bias resistor and that the A5191HRT is receiving and demodulating a
message. This is the worst case condition for current consumption.
10.0 References
(1) HART Communication Foundation Document Number HCF_SPEC-54, HART FSK Physical Layer Specification, Revision 8.1; 9390
Research Blvd., Suite I-350, Austin Texas, 78759.
(2) HART Communication Foundation Document Number HCF_TEST-2, FSK Physical Layer Test Specification, Revision 2.1 9390
Research Blvd., Suite I-350, Austin Texas, 78759.
(3) AMIS A5191HRT datasheet: http://www.amis.com/products/transceivers/hart_modem.html
For more information about AMI Semiconductor’s products or services visit our Web site at http://www.amis.com.
Devices sold by AMIS are covered by the warranty and patent indemnification provisions appearing in its Terms of Sale only. AMIS makes no warranty, express, statutory,
implied or by description, regarding the information set forth herein or regarding the freedom of the described devices from patent infringement. AMIS makes no warranty of
merchantability or fitness for any purposes. AMIS reserves the right to discontinue production and change specifications and prices at any time and without notice. AMI
Semiconductor's products are intended for use in commercial applications. Applications requiring extended temperature range, unusual environmental requirements, or high
reliability applications, such as military, medical life-support or life-sustaining equipment, are specifically not recommended without additional processing by AMIS for such
applications. Copyright ©2008 AMI Semiconductor, Inc.