Sifos Technologies PhyView PVA-3000, Reference Manual

Page: 19 / 134

PVA-3000 Reference Manual
December 2, 2019 Sifos Technologies Page
19
1.3.11. Receiver Verification Overview
Unlike transmitter testing, 10/100/1000BaseT receiver testing is
not generally addressed in 802.3 standards. 802.3 standards do
define the worst-case link (or cabling systems) environment for
Ethernet ports with specifications for channel insertion loss, return
loss, and crosstalk. These specifications are ultimately derived
from TIA/EIA 568 specifications related to Category 5/5e cabling
systems. Transmitted signal characteristics such as signal
amplitude range, clocking frequency, distortion parameters, and
timing jitter are also thoroughly specified in the standards and by
inference create additional constraints on required receiver
performance. Finally, allowances for added noise, or alien
crosstalk also affect receiver requirements. This collection of
receiver requirements is depicted in
Figure 1.15 .
The 10BaseT and 1000BaseT specifications also define explicit goals for bit error rate (BER) performance. The
100BaseTx specification, and its subsidiary TP-PMD specification make no mention of bit error rate targets. Section
1.2.5 above raised the additional issue that BER is not directly measurable and must be inferred from packet flow
measurements.
The table below summarizes specifications that are relevant to 10/100/1000BaseT receiver testing.
Link Rate Target BER Cabling System Rx Noise Insertion Target Packet Error Rate
1000BaseT 10
-10
Cat5e
-24dB @100MHz
25mVp-p
100BaseTx Spectrum
10
-7
(125 octet frames, no
account for coding ovhd .)
100BaseTx Unspecified
(UNH: 10
-8
)
Cat5
-24dB @100MHz
40mVpp
Band Limited 100MHz
Unspecified
(UNH: 1.5*10
-5
, 64 octet
frames)
10BaseT 10
-8
Cat3
-11.5dB @10MHz
300mVpp
Band Limited 15MHz
Unspecified
(5.12*10
-6
, 64 octet frames)
Summarizing, receiver test methodology is largely the responsibility of Ethernet port designers and implementers. The
task of modeling all of the possible signal and channel impairments while assessing bit error rate performance can be
daunting and the exact criteria for success (pass or fail) can be ambiguous.
1.3.12. Receiver Testing Metrology
The PhyView Analyzer offers two distinct measurement techniques for receiver testing:
1. Link Monitor : Receiver Measurements on stand-alone Ethernet ports
5. Packet Flow : Receiver Measurements on multi-port (bridging) devices such as switches, hubs, and repeaters
The Link Monitor was introduced in Section 1.3.1. This measurement resource assesses both instantaneous and time-
sampled
Link Status , as well as other link configuration parameters. In receiver testing, the key parameter is sampled
Link Status. Link Status can be one of three physical layer indicators:
1. Link State , that is link UP (linked) or link DOWN (unlinked) – available for 10/100/1000BaseT
6. Remote Rx Status
, a physical layer indicator from the link partner indicating Rx “OK” or Rx “Not OK” –
available for 1000BaseT only .
7. Local Rx Status
, a physical layer indicator from the test port receiver indicating Rx “OK” or Rx “Not OK” –
available for 1000BaseT only .
Link Status can be a single, instantaneous sample or can be configured to report a count of up to “UP” or “OK”
samples evenly spaced with sampling intervals of 20, 50, or 100msec, whereupon it becomes a Link Stability
measurement.
Generally, the criteria for link “UP” in 10BaseT is the continuous receipt of link test pulses (
see Figure 1.16 ). A
highly impaired 10BaseT receiver-under-test may or may not choose to continue transmission of link pulses meaning
there is no sure way for one link partner (or Test Port) to discern the receive condition of the other link partner. This
means that
Link Stability generally will not correlate linearly to a hypothetical port-under-test bit error rate or packet
error rate.
Figure 1.15 Ethernet Port Rx Requirements