
A5191HRTNGEVB
www.onsemi.com
15
Figure 20. Characteristic of the Second Stage of the Filter
APPLICATION IDEAS
The A5191HRT takes care of the HART modulation. This
HART signal must then be superimposed on a 4−20mA
current loop. Below are some possible implementations of
both a master and slave transmitter.
Slave implementation
A simple slave implementation is shown in Figure 21. The
analog loop current is set by a DAC from the
microcontroller, while HART signals are added in from the
A5191HRT. The DAC can be PWM or sigma−delta
topology. To explain the operation of this circuit, let us first
look at an example where the DAC is not of a switching
topology. In this case, R2and R3can be 0 W, and C2, C3and
C5can be left out. As one end of R6is tied to local ground,
it can easily be seen that the voltage at the negative loop
terminal is negative with respect to the local ground.
Resistors R4and R5are then chosen so that in steady state
their common terminal is a virtual ground point in the
absence of HART signals, since the negative terminal of the
amplifier is also connected to ground. A similar principle
applies when HART signals are applied. So both amplifier
inputs are regulated to ground. A compensation capacitor C4
may be required depending on the operational amplifier
used. To avoid offset generated by bias current in the
operational amp, R3should be dimensioned to approach the
impedance seen by the positive terminal.
The amplifier will then determine the current flowing
through the loop by changing the base of a transistor in
emitter feedback configuration. The value for R7is
determined by the output range Vo,max of the amplifier used:
It is often recommended to take a value as large as
possible, so that noise effects are minimal.
Typically the value of R6is chosen equal to R7. The
voltage over R6and R7combined should however be less
than 12 V when the current setting is 20 mA.
Next, the values of R4and R5are chosen depending on the
most significant bit of the DAC.
When the DAC is not a switching topology, we can now
choose R1and C1. We have:
Where:
In practice, C1is chosen sufficiently small so that Z [R1.
For a PWM or sigma−delta output DAC, the circuit gets
a bit more complicated, as we need to filter away high
frequency DAC components, but leave HART signals intact.
If the bandwidth of the DAC is larger than 2.2 kHz, adding
C3introduces a low−pass filter to the loop that will remove
most of the switching noise.
Where Rpis the parallel circuit of R4, R5and R1.
If the bandwidth of the DAC is close to the HART
frequencies, an alternate high−frequency feedback path
must be introduced so that HART signals are not removed
by the low pass filter of the DAC. The exact calculation of
component values in this case is more complicated.
However, it is based on a similar principle, but now with two
summing junctions, for low−frequency and high−frequency
signals separated.
Resistor R3may be needed to compensate for amplifier
bias current. It is chosen so that its resistance is similar to
resistance seen on the positive terminal. Depending on the
amplifier used, it may also be required to provide a
compensation capacitance C4.