Text preview for : 5991-2375EN Using Microwave Switches When Testing High Speed Serial Digital Interfaces c20140905 [28 part of Agilent 5991-2375EN Using Microwave Switches When Testing High Speed Serial Digital Interfaces c20140905 [28 Agilent 5991-2375EN Using Microwave Switches When Testing High Speed Serial Digital Interfaces c20140905 [28].pdf
Back to : 5991-2375EN Using Microwa | Home
Keysight Technologies
Using Microwave Switches When Testing
High Speed Serial Digital Interfaces
Application Note
02 | Keysight | Using Microwave Switches When Testing High Speed Serial Digital Interfaces - Application Note
Many high speed digital interfaces
Lane0 +
use multiple lanes to achieve the
throughput requirements of their Lane0 -
systems. These systems present Lane1 +
validation and characterization Lane1 - LaneTest +
challenges because of connection LaneTest -
requirements, including single-
ended as well as differential LaneN +
testing and more lines than
LaneN -
oscilloscope channels available.
Solving this problem with a
switching network will lead to Figure 1. Two multi-port microwave switches can be used to switch high speed
lower test costs, higher reliability, differential lanes
and unattended testing, which
managers desire for productivity.
This application note addresses There are two primary use cases that will be addressed in this application note. These
the use of these switching are identified by the nature of the device under test and, as you will see, determine
networks, their calibration, use different steps in the calibration of an intervening network.
of calibration, accuracy that is
achievable, degradations that Case #1
might be encountered, and
Device under test is an integrated circuit, and the objective is to measure to the pins
other considerations such as
of the device. A printed circuit (PC) board is used as test fixture. It breaks out signals
de-embedding test point access
to test to standard RF connectors such as SMA.
adaptors and 2-port versus 4-port
characterization.
Switching Use Test board
Cases Addressed
L0+ L0-
Comments about Case #1 L1+
Device L1-
The PC board should use the best
under
materials possible and may (or
test L2+ Lane test +
may not) have the RF connectors Lane test -
L2-
close to the pins of the target
device to be characterized. In the
latter case, there is enough loss
to warrant a measurement of the
s-parameters of the paths to the
RF connectors and to comprehend Figure 2. Test device is an integrated circuit on a test board
them in the compensation (or `de-
embedding') process.
03 | Keysight | Using Microwave Switches When Testing High Speed Serial Digital Interfaces - Application Note
Comments about Case #2 Case #2
The performance of this kind The device under test is a `system' where the output is through a defined digital
of device usually includes the connector such as HDMI and DisplayPort. Such systems require a testpoint adapter (test
standard digital connector mated fixture) to break out the lanes to the oscilloscope.
pair so the device's performance
is measured at the virtual point
on the oscilloscope side of that
Testpoint adapter (TPA)
mated pair. The performance of Test device
the testpoint adapter may be so
good as to be ignored, common
to all measured devices by stan-
connector
Standard
dard decree, or de-embedded
if the s-parameter file for it is Lane test +
Lane test -
available. The s-parameter file
should comprehend the whole
fixture except the connector and
connector pins. The connectors'
performance for these digital Figure 3. Test device has a digital standard connector, not an RF coaxial launch
standards does vary consider-
ably, so the mating connector of
the adapter