HDMI.Org published ver 1.3 of the HDMI specification in June 2006. This
version adds higher display resolution formats, including 1,080p, Deep
Color formats with serial data rates of up to 3.4Gbps.
HDMI
1.3 is the most significant upgrade yet in the
specification that has become the de facto standard interface for
high-definition (HD) devices that include DVD players, HDTV and the
newest gaming devices. HDMI 1.3 defines the latest HD A/V technology
that began hitting the consumer market late 2006, and continues to roll
out this year.
To ensure reliable information transmission and interoperability,
industry standards specify requirements for the network's PHY. HDMI
Compliance Test Specifications (CTS) define an array of compliance
tests for the HDMI PHY. Last March, HDMI.Org published an updated CTS
(CTS 1.3b) document that incorporates new tests for the capabilities
defined in HDMI 1.3.
 |
| Figure
1: These are the logical links of HDMI TMDS signaling and core tests
required by CTS specifications |
.
Test challenges
High-quality test and measurement instruments, and solutions for
high-speed serial data are fundamental to ensure compliance with the
HDMI CTS 1.3b and the development of successful products that implement
the HDMI 1.3 specification.
Figure 1 above shows the
major elements of the HDMI transmission system - source, cable and
sink. While it is recommended to perform as many tests as possible, the
core electrical tests are critical for compliance.
Per HDMI CTS 1.3b specifications, the complete compliance test
solution includes real-time oscilloscopes, sampling oscilloscopes,
signal generators, differential probes and test fixtures. However, the
key is how to choose the right test platform to simplify the latest
HDMI designs and to ensure compliance with the latest test
requirements. Engineers need to integrate several instruments into a
complete, flexible and cost-effective solution.
A capable, integrated solution enables fast eye-diagram rendering
and jitter verification, required to accumulate huge waveforms within
as short a time period as possible.
Also, signal generators need to operate in a closed loop to
automatically perform the needed complex cable and sink tests using
"one-button" compliance test software resident on an oscilloscope.
 |
| Figure
2: Shown are HDMI source compliance test results. |
Real-time oscilloscopes
A complete test solution for HDMI CTS 1.3b includes high-performance
hardware and comprehensive software for design and test engineers in
consumer electronics, semiconductor and cable manufacturers.
In terms of the key instrument, to test HDMI 1.3 signals need a
minimum 8GHz real-time oscilloscope in which HDMI compliance test
software is running. Figure 2 above is
an example of HDMI 1.3 source test results using HDMI compliance test
software.
Meanwhile, to ensure adequate representation of signal
characteristics, the CTS 1.3b specifies a minimum oscilloscope record
length to acquire the data signal.
This ensures that at least 400,000 unit intervals (or Tbit) are
accumulated for building the eye diagram. With 16M/20M record length,
at least 400,000 unit intervals (UI) can be captured for
lower-resolution signals, and over 2.6M UI for higher resolution
devices. Considering the bit rate up to 3.4Gbps, the minimum
oscilloscope record length is above 16M with high-speed sampling rate
(above 10GSps).
At the nerve center of any transmission system is the clock jitter.
The jitter test checks to ensure that the clock signal is not carrying
excessive jitter - e.g. duty cycle jitter is an excellent method of
assessing deterministic jitter.
The CTS also defines the margin to be +10 percent from the nominal
50 percent duty cycle. It's important that the variance in duty cycle
is measured over a large number of acquired signals.
 |
| Figure
3: This clock jitter test used DPO technology. |
As per the CTS, a minimum of 10,000 triggered waveforms are required
for test purposes. Trigger re-arm rates of the oscilloscope then take
center stage. Nominally, oscilloscope trigger re-arm rates are of the
order of about 100wfms/s.
This can mean unacceptably long acquisition and test times.
Fortunately, there are sophisticated techniques such as FastAcq on
Digital Phosphor Oscilloscopes (DPO) that enhance the trigger re-arm
rates and deliver over 300,000wfms/s.
Figure 3 above demonstrates
the clock duty cycle test using the FastAcq technology. Notice the
richness of information that ensures clear and compelling measurements.
Signal generators
One of the most critical characteristics of a sink device is its
tolerance to specified levels of jitter in the signals. The standard
defines the limit as 0.3*Tbit. Specified amounts of jitter are injected
in steps (from low to high jitter) into the transmitted Transition Minimized Differential Signaling
(TMDS) signal until the sink device fails to recover the
signal. The amount of jitter that the sink device is able to tolerate
is compared against the limits for compliance.
Several measurements are carried out by injecting a specified amount
of jitter. Three measurements are performed over two test cases: (a)
data jitter frequency at 500kHz and clock jitter frequency at 10MHz,
and (b) data jitter frequency at 1MHz and clock jitter frequency at
7MHz.
When selecting the right equipment for the HDMI CTS test setup, it
is important to understand the necessary aspects that must be
addressed. TMDS signal generation plays a pivotal role in the Sink
tests.
 |
| Figure
4: Shown is an integrated sink testing system based on a closed-loop
mechanism |
.
The key challenge for a signal generator is to provide a full
complement of highly accurate signals and the ability to precisely
control their parameters.
For performing minimum differential sensitivity tests, a resolution
of 20mV is required. The intrapair skew test requires precise delay
settings down to sub-picosecond resolution.
The jitter tolerance test assumes larger challenges, since both
clock and data jitter need to be varied. Generating jitter frequencies
on the order of 10MHz requires a combination of signal generators.
<>Since margins are tight, precise control is required on jitter
amplitude. With various parameters to be adjusted and tight margins,
this test tends to be extremely complex and can take a very long time
to complete. Figure 4 above
shows an example of an automated sink test system.
The DPO connects to the DTG5334 (TMDS signal generator) using a GPIB
cable and to the AWG/AFG (jitter insertion signal generator if
required) using a GPIB USB-B cable.
HDMI design and test engineers can control signal generators in a
closed-loop mechanism to automatically perform the needed complex cable
and sink tests using "one-button" compliance test software resident on
the oscilloscope. The closed loop mechanism shrinks test time greatly
and eliminates nonlinearities of test setup.
Differential probing system
It's also critical to have a flexible, versatile probing system with
high performance. Meanwhile, many of today's high-speed serial data
standards use differential signaling on multiple lanes that are
challenging to measure simultaneously on a single oscilloscope.
What is needed is an oscilloscope that can simultaneously acquire up
to four high-speed differential signals with the use of four
differential probes. As an added benefit, the inputs on the probes
connect to high quality 50 ohm terminations that offer industry-leading
return loss, a critical specification in compliance testing as
frequencies increase.
This differential probe provides a common-mode DC voltage input to
the termination network. The termination voltage can be supplied either
externally by the user or internally by the oscilloscope. In addition,
there is also an automatic mode that senses the common mode voltage of
the input signal and automatically sets the termination voltage to
match.
Sampling scope
Differential transmission lines used in achieving fast data rates are
very sensitive to impedance matching. Consequently, impedance
characterization is a very crucial test in HDMI compliance testing. The
through-connection impedance has a limit of 15 percent variance to its
100 ohm specification. The impedance at termination needs to be tighter
as the margins are only 10 percent of its characteristic value of 100
ohm.
Meanwhile, engineers need to verify cable key characteristics, like
intra- and interpair skew and crosstalk. Time Domain Reflectometry
(TDR) is a powerful and accurate tool for measuring impedance and
length in interconnects.
While fundamental concepts of TDR are relatively simple, a number of
issues must be considered to make accurate measurements, the foremost
being the ability to perform true-differential TDR.
Evan Sun is Market Development Manager, Asia Pacific Region, Tektronix Inc.
