AN273
5. Test 4: Receiver Off-Hook Test
GR-909 requires that the SLIC be able to differentiate between an off hook handset and a resistive fault. The dc
impedance presented to the SLIC by a typical telephone varies with applied voltage; therefore, this nonlinearity
may be exploited to determine whether a telephone is present.
By measuring the TIP lead in the TIP-OPEN linefeed state at two V settings, the two dc impedances R (refer
OC
TR
to Equation 3) may be compared. If these two measurements differ by more than ~50% and they are within the
expected range, an off hook telephone may be present. If a resistive fault is present, the variation will be within the
resistance measurement accuracy, typically 10–25%.
This test may be performed with or without R installed. If R is not installed, refer to Equation 5 to calculate
SH
SH
R
.
TR
5.1. Test 5: REN Test
The Si321X can detect relatively small capacitive loads on the line. Line capacitance can be measured by either
measuring the time constant of a decaying voltage on RING when in the OPEN mode (single-slope conversion) or
by detecting longitudinal current when a subthreshold ring signal is applied. From the measured loop capacitance,
the REN load may be estimated to within ± 0.5 REN.
5.1.1. Single-Slope Conversion
After charging the RING lead to V
in the FORWARD ACTIVE linefeed state, if the linefeed state is switched to
RING
the OPEN state, the voltages between the TIP and RING leads will discharge through their series output resistance
(see Figure 1). By measuring the time delay from V = upper threshold to V = lower threshold, the capaci-
RING
RING
tance can be calculated and the REN load estimated. This is ideal for detecting very small REN loads.
Because high REN loads present such a large capacitance and, therefore, very large time constants, it may be
more advantageous to implement a subthreshold ringing method. It is only recommended that the single-slope
method be used for loads less than 0.75 REN.
5.1.2. Subthreshold Ringing Method
The best method for estimating REN load when the load is >0.75 REN is by applying a small amplitude ring signal
to the line and correlating the power dissipated in the linefeed pullup (Q2) to a known REN load. This requires the
user to calibrate his hardware/software to known 1, 3, and 5 REN loads. This calibration determines the slope and
offset of a straight line that correlates the REN load to the Q2 power dissipation. Once calibrated, future measure-
ments utilize curve fitting to estimate the REN load.
Figure 7 shows the relationship between Q2 power dissipation and an applied REN load for a typical Si321X hard-
ware design. By calibrating the hardware and identifying two straight lines, REN can be measured with ~± 0.5 REN
accuracy.
Q2 Power
Best Fit Line
Measured
# REN (applied)
Figure 7. Typical Q2 Power vs. Applied REN Load
Rev. 0.1
7