eIN
RS
–
+
(e1)
(e2)
IS
5
6
I3
I4
R3
R4
1.25kΩ
1.25kΩ
+VCC
A1
+VCC
A2
+VCC
8
D1
(e1)
IB1
–In3
eIN
IB2
VPS
+In4
(e2)
100µA
IO
7
+
Q1
+VCC
eL
+VCC
A3
–
RL
2mA
I1
R
1kΩ
1
R
2
52.6Ω
I2
IO
Voltage Controlled
Current Source
10
IREF1
11
IREF2
2.5kΩ
Ω
IO = 4mA + (0.016 + 40/RS) eIN, eIN = e2 – e1
FIGURE 1. Simplified Schematic of the XTR101.
and minimizes thermal feedback to the input stage. Also in
such applications where eIN full scale is small (<50mV) and
tions shown in Figure 2. Thus the 2N2222 will safely
operate below its 400mW rating at the upper temperature
of +85°C. Heat sinking the 2N2222 will result in greatly
reduced accuracy improvement and is not recommended.
RSPAN is small (<150Ω), caution should be taken to consider
errors from the external span circuit plus high amplification
of offset drift and noise.
2. TIP29B in the TO-220 package. This transistor will
operate over the specified temperature and output voltage
range without a series collector resistor. Heat sinking the
TIP29B will result in slightly less accuracy improvement.
It can be done, however, when mechanical constraints
require it.
OPTIONAL EXTERNAL TRANSISTOR
The optional external transistor, when used, is connected in
parallel with the XTR101’s internal transistor. The purpose
is to increase accuracy by reducing heat change inside the
XTR101 package as the output current spans from 4-20mA.
Under normal operating conditions, the internal transistor is
never completely turned off as shown in Figure 2. This
maintains frequency stability with varying external transis-
tor characteristics and wiring capacitance. The actual “cur-
rent sharing” between internal and external transistors is
dependent on two factors: (1) relative geometry of emitter
areas and (2) relative package dissipation (case size and
thermal conductivity). For best results, the external device
should have a larger base-emitter area and smaller package.
It will, upon turn on, take about [0.95 (IO – 3.3mA)]mA.
However, it will heat faster and take a greater share after a
few seconds.
ACCURACY WITH AND
WITHOUT EXTERNAL TRANSISTOR
The XTR101 has been tested in a circuit using an external
transistor. The relative difference in accuracy with and
without an external transistor is shown in Figure 3. Notice
that a dramatic improvement in offset voltage change with
supply voltage is evident for any value of load resistor.
MAJOR POINTS TO
CONSIDER WHEN USING THE XTR101
1. The leads to RS should be kept as short as possible to
reduce noise pick-up and parasitic resistance.
Although any NPN of suitable power rating will operate
with the XTR101, two readily available transistors are
recommended.
2. +VCC should be bypassed with a 0.01µF capacitor as close
to the unit as possible (pin 8 to 7).
3. Always keep the input voltages within their range of
linear operation, +4V to +6V (e1 and e2 measured with
respect to pin 7).
1. 2N2222 in the TO-18 package. For power supply volt-
ages above 24V, a 750Ω, 1/2W resistor should be con-
nected in series with the collector. This will limit the
power dissipation to 377mW under the worst-case condi-
®
XTR101
6
BB [ BURR-BROWN CORPORATION ]
