Sifos Technologies PhyView PVA-3000, Reference Manual

Page: 12 / 134

PVA-3000 Reference Manual
December 2, 2019 Sifos Technologies Page
12
Physical Impairment crosses some threshold level that will depend on the type of defect or impairment. At that point,
the rate of errors will increase at some slope that is also dependent upon the type of impairment involved. As the
graph suggests, a transmit jitter problem may cause a very steep slope from zero to high Bit Error Rate while reduction
in signal power (or higher channel loss) may have a much more gradual impact. Summarizing, knowledge of Bit Error
Rate does not point to the nature of a defect nor does it really describe the magnitude of an impairment.
Limitation #4 : Packet flow does not directly measure
Bit Error Rate. There is a statistical relationship
between packet loss and BER and the effect of that
relationship, as described in
Figure 1.5 , is to
dramatically sharpen, or increase the slope when Packet
Loss is related to Physical Impairment level. Smaller
packets help soften this relationship a bit, however
Ethernet restricts the minimum packet size to 64 bytes
including packet overhead. This translates to 512 bits
for 10BaseT and to 640 bits for 100BaseTx and
1000BaseT owing to coding overheads. The BER
resolution problem is worsened by 1000BaseT where a
degree of forward error correction built into the physical
layer will hide lower bit error levels by reconstructing
partially damaged packets. This in effect increases the
impairment threshold (see Figure 1.4) and sharpens the
slope relating each Physical Impairment to Packet Loss.
Limitation #5 : When performing receiver testing, unless the degree of Physical Impairment is carefully manipulated,
it is hard to know whether the Ethernet port-under-test is capable of tolerating IEEE 802.3 specified maximum channel
impairment levels. Measuring zero Packet Loss at zero impairment provides no insight to this capability and therefore
represents minimal defect coverage. On the other hand, if a receiver is exposed to specified maximum channel
impairments of various types and Packet Loss is zero, then it can be determined that the receiver is indeed immune to
the worst-case types of channel impairments it can expect to encounter while in service.
Summarizing, packet flow testing has thrived because it is convenient, low cost, and because the traditional PHY
compliance tests are both expensive and laborious. The PhyView Analyzer was developed to provide much of the
direct physical layer test coverage associated with traditional PHY compliance tests while providing all of the
convenience of a fully automated packet flow test.
1.3. The PhyView Analyzer Metrology
The PhyView Analyzer is designed to make “pure” physical layer measurements. These measurements are made while
linked to an 10/100/1000BaseT port-under-test. Connections are made using high grade RJ-45 cables and plugs so
there is no need of any special fixturing or probing as is common with traditional physical layer test methods.
PhyView measurements and test capabilities can be classified into four different topics:
1. Link Configurations & Assessments : Basic analysis of port-under-test link capabilities and link status
2. Transmission Measurements : Direct measurements of the port-under-test transmitter function
3. Interface Measurements : Evaluation of electrical interface parameters
4. Receiver Tests : Assessments of port-under-test receiver performance characteristics
Each of these areas will be addressed in the following paragraphs.
1.3.1. Link Configurations and Assessments
Link configuration and assessment is the basic ability to determine link state and link configuration, and to control link
configuration. Since link-up and link configuration is essential to all other PhyView Analyzer measurements, the
instrument must have these capabilities. Link configuration features are presented in the following table.
Link Configuration Options Description
Link Rate Auto, 10, 100, 1000, 10/100 Forces specified or at auto-negotiated rate
Connection Mode
Auto, MDI, MDI-X Forces specified or link determined connection mode
Duplex Auto, Half, Full Forces specified or auto-negotiated duplex mode
Gigabit Mode Auto, Master, Slave Forces specified or auto-negotiated 1000BaseT gigabit
timing mode
Figure 1.5 Packet Error Rate vs Bit Error Rate
Bit Error Rate
Packet
Err
or
Rate
Large Packets
Small Pa
ckets
No
F
EC
F
EC
No
FEC
FEC
PER=1-(1-BER)
Bits/Pkt
Bit Error Rate
Packet
Err
or
Rate
Large Packets
Small Pa
ckets
No
F
EC
F
EC
No
FEC
FEC
PER=1-(1-BER)
Bits/Pkt