TLC3704-Q1 Datasheet, PDF (8/23 Pages) Texas Instruments – QUAD MICROPOWER LINCMOS VOLTAGE COMPARATORS

TLC3704ĆQ1
QUAD MICROPOWER LinCMOS VOLTAGE COMPARATORS
SGLS191 − JUNE 2004
PARAMETER MEASUREMENT INFORMATION
The TLC3704 contains a digital output stage which, if held in the linear region of the transfer curve, can cause
damage to the device. Conventional operational amplifier/comparator testing incorporates the use of a servo
loop which is designed to force the device output to a level within this linear region. Since the servo-loop method
of testing cannot be used, we offer the following alternatives for measuring parameters such as input offset
voltage, common-mode rejection, etc.
To verify that the input offset voltage falls within the limits specified, the limit value is applied to the input as shown
in Figure 5(a). With the noninverting input positive with respect to the inverting input, the output should be high.
With the input polarity reversed, the output should be low.
A similar test can be made to verify the input offset voltage at the common-mode extremes. The supply voltages
can be slewed as shown in Figure 5(b) for the V_ICR test, rather than changing the input voltages, to provide
greater accuracy.
5V
1V
+
+
Applied VIO
Limit
−
VO
Applied VIO
Limit
−
VO
−4V
(a) VIO WITH VIC = 0 V
(b) VIO WITH VIC = 4 V
Figure 5. Method for Verifying That Input Offset Voltage Is Within Specified Limits
A close approximation of the input offset voltage can be obtained by using a binary search method to vary the
differential input voltage while monitoring the output state. When the applied input voltage differential is equal,
but opposite in polarity, to the input offset voltage, the output changes states.
Figure 6 illustrates a practical circuit for direct dc measurement of input offset voltage that does not bias the
comparator in the linear region. The circuit consists of a switching mode servo loop in which IC1a generates
a triangular waveform of approximately 20-mV amplitude. IC1b acts as a buffer, with C2 and R4 removing any
residual d.c. offset. The signal is then applied to the inverting input of the comparator under test, while the
noninverting input is driven by the output of the integrator formed by IC1c through the voltage divider formed
by R8 and R9. The loop reaches a stable operating point when the output of the comparator under test has a
duty cycle of exactly 50%, which can only occur when the incoming triangle wave is sliced symmetrically or when
the voltage at the noninverting input exactly equals the input offset voltage.
Voltage divider R8 and R9 provides an increase in the input offset voltage by a factor of 100 to make
measurement easier. The values of R5, R7, R8, and R9 can significantly influence the accuracy of the reading;
therefore, it is suggested that their tolerance level be one percent or lower.
Measuring the extremely low values of input current requires isolation from all other sources of leakage current
and compensation for the leakage of the test socket and board. With a good picoammeter, the socket and board
leakage can be measured with no device in the socket. Subsequently, this open socket leakage value can be
subtracted from the measurement obtained with a device in the socket to obtain the actual input current of the
device.
8
• POST OFFICE BOX 655303 DALLAS, TEXAS 75265