CRC testing in video applications: The stages of CRC

CRC testing in video applications: The stages of CRC

Technology News |
By eeNews Europe

The Stages of CRC

Development Phase…
CRCs can be employed across many facets of the development phase of a video product to gauge the performance of the video signal chain; during thermal testing; during power supply testing; during the assessment of layout changes, during final software configuration changes; even during cable selection if a cable is to be supplied with the product.

Compliance Testing
Before a video product can be marketed as HDMI compliant and carry the HDMI logo, the product must undergo a series of stringent test at an officially licensed HDMI approved test centre (ATC). These tests ensure that the product meets all of the requirements set out in the HDMI compliance test specification (released in alliance with the main HDMI specification). One of the toughest tests conducted as part of this suite of testing is an analysis of how robust the video receiver is to jitter on both the clock and data channels.

Meeting the criteria outlined in this test are quite often challenging and video product design and manufacturers frequently send their prototype systems to for expensive pre-compliance checks if they do not have access to the ultra expensive equipment specified in the official tests.

The frame checker in the ADV7850 can be used as a low-cost substitute for early iterations of pre-compliance testing if the specified equipment is not available; it provides an insight into whether the receiver is correctly receiving and decoding the HDMI data (factors which can influence this range from whether the correct configuration writes are being employed, to power supply design).

If the specified equipment is available, the frame checker can still be employed as it provides a definitive insight into whether the receiver is correctly receiving and decoding the data. This level of analysis goes beyond that mandated by the CTS which mandates only visual checking.

HDMI Cable Selection
Many video product design and manufacturers, especially in the professional audio/video market, depend on HDMI cables to route video between system components. An HDMI cable is constructed using 19 conductors; the HDMI specification outlines five different categories of HDMI cable for varying speed grades.

Cables, due to their limited bandwidth, typically introduce a particular type of noise or jitter into the video stream; inter-symbol interference (ISI) jitter. ISI jitter is the interference between current and subsequent symbols. This "blurring" of symbols makes it more difficult for a receiver to decode and interpret the data.

Figure 5: An HDMI Cable

For example, in a video conferencing system, video may be routed around a room from a central console to multiple monitors or projectors via a series of HDMI cables of up to 30 metres in length (see Figure 6). HDMI cables at such lengths however can be a significant component of the system cost with prices ranging from tens of dollars to hundreds of dollars; video product design and manufacturers may choose to evaluate cables from low, medium and high cost suppliers. Cables suppliers can often justify the costs of the cables which they manufacture through the quality of those cables; video product design and manufacturers however must balance the quality and cost of the cables which they choose to supply with their products.

Figure 6: A Sample Conference Room Video Transmission Architecture

When evaluating such different cables, a CRC test can be employed to great effect; starting with the benchmark of a known good, trusted cable which provides stable system results, an evaluation engineer can compare how cheaper or longer cables impact on the CRC results – gaining an interesting data-point in determining the suitability of such cables.

A typical investigation could be as follows; a video product design and manufacturer is currently using a particular cable that has been characterised with a system and provides a reasonable and trusted level of performance; the data received from the cable is degraded to the extent that it impinges slightly onto the HDMI compliance test specification jitter tolerance mask at 1080p but not to an extent that would cause any issues in the ADV7850 recovering the data.3 Running the frame checker on the video received from this cable results in a pass on all channels for up to 255 frames.

Figure 7: Cable Performance at 1080p

If, in a product refresh, support for 4k x 2k (double the clock and data rate of 1080p 8-bit) is added to the hardware, the cable will have to be revalidated. Using a simplistic approach of checking the impact of the cable on the 4k x 2k data, a worrying result is recorded (see Figure 8).

Figure 8: Cable Performance at 4k x 2k

The losses in the cable are now causing a significant degradation of the signal; in functional tests, the ADV7850 may still recover the data but will all of the data still be fully intact? Will random or intermittent bit errors, which could contribute to triggering serious system issues such as HDCP snow noise (as discussed earlier) now exist in the data?

By employing the frame checker function of the ADV7850, the evaluation engineer can confirm exactly whether or not any data errors are occurring; the cable can also be subject to extended testing such as heating/cooling/bending whilst the frame checker is running.

Power Supply Testing
The power supply is one of the most important aspects of a design to be tackled and is considered by many to be one of the most challenging. Many factors can influence the quality of the power supply output; and the output of the power supply can influence some many characteristics of the system. Power supplies typically introduce a particular type of noise or jitter into the video stream; periodic jitter.

Designers must choose whether linear drop-out regulators (LDOs) or switch-mode power supplies (SMPS) are to be employed; what frequency switch regulators to use and what filters should be employed to suppress any harmonics making it through to the system; whether power supply planes are routed on a single or multiple layers; whether it is possible for the planes to overlap on adjacent layers; how the decoupling capacitor architecture is implemented – even the location, the size and the material of the decoupling capacitors selected; all of these elements and many more can yield a significant influence.

Figure 9: Power Supply Design Comparisons

CRC testing can be used when evaluating the impact of power supply design changes both on a single revision of a board and between multiple revisions of a board. By changing the decoupling capacitors employed on a system and running a CRC test after each change, a system evaluation engineer can benchmark which decoupling architecture assists in achieving the most stable system.

System evaluation engineers can also benchmark changes in subsequent revisions of prototype systems by running CRC tests on them as long as no other significant layout and schematic changes have been made. Finally, CRC tests can be used to access the impact of changes which may occur in the natural tolerances which systems may be subject to e.g., variances in power supply voltage levels through the tolerances of power supply components.

Layout Changes
The techniques involved in the layout of high-speed digital video signal chains are complex and deserving of a dedicated article themselves. The HDMI specification demands that traces are routed with a 100? differential trace impedance but achieving support for 4k x 2k video data introduces extra challenges; at such high data rates, signal integrity becomes a major issue with elements such as trace width and trace length becoming major factors which need to be considered.

The impact of layout changes between subsequent revisions can be accessed using CRC testing. By comparing the output of two subsequent revisions of a prototype board featuring some fundamental change e.g., a different board material (FR4 versus Rodgers), a different stack up, different HDMI trace widths, the impact of the change can be accessed.

Thermal Testing
Confirming that a video product operates correctly over its specified temperature range is a vitally important phase of evaluation. Video product design and manufacturers must ensure that the ambient temperature within their product does not exceed the silicon vendors specifications and that, across the ambient temperature range which the product will be subject to (e.g., 0°C to +70°C for consumer products or -40°C to +85°C/+125°C for automotive products), that the product’s performance is consistent and reliable.

Figure 10: Automated Thermal Testing

Often in this type of testing, a prototype system is placed in a temperature controlled oven which can be cycled over the product’s specified temperature range e.g., -40°C to +85°C. The output of the system is then observed to ensure that it is stable over the whole range of temperatures, video frequencies and video patterns.

This testing can be automated quite easily using CRC testing and allowed to run indefinitely. By automating the control of the oven, the video generator and the CRC analysis tool, an evaluation engineer can easily sweep the temperature, change video formats and patterns whilst monitoring the CRC results for frame after frame of video data.

If no changes occur when the video pattern and format remain constant, the test can continue; if a change in the CRC occurs when the video pattern and format are constant, the environmental variables should be recorded (e.g., temperature, video format, video pattern, etc.) and the testing can continue. This type of testing can easily be set to run overnight or over a weekend with the result being the unmanned completion of hundreds of hours of robustness testing.

Software Configuration Changes
Certain aspects of modern semiconductor devices need to be tuned depending on the prototype system into which it is incorporated e.g., clock and data relationships may need to be adjusted to accommodate particularly long or short signal routes. A CRC test can be employed in such circumstances to assist in tuning the available controls e.g., equaliser settings, PLL settings, clock and data phase relationships to provide the most stable system possible.

Manufacturing Phase
When a video product design and manufacturer needs to verify the consistency and correctness of its manufacturing process by inspecting all or a cross section of its finished products, CRC is an ideal tool which can be employed to gauge the correct soldering of certain connectors (i.e., HDMI), external passive and active devices (e.g., HDMI ESD devices), the correct soldering of the CRC processing device (e.g., the HDMI receiver).

CRC can be employed in final testing in a number of ways; a CRC could be implemented in the video product itself; a CRC could be implemented in a discrete end of line testing device (see Figure 11). Implementing a CRC in a video product would require a semiconductor solution in the video signal chain which supports the capability of delivering a CRC e.g., an FPGA or microcontroller.

Implementing a CRC in a separate device may reduce the BOM cost of the video product but requires the investment in a separate device. It does however offer the benefit of being capable of testing the stability of the entire system; the coverage offered by a CRC test solution embedded into the video signal chain is dependent on the location in the signal chain of that test solution; a CRC test located near the start of the signal chain may offer a low to moderate level of coverage; a CRC test located near the end of the signal chain may offer a moderate to high level of coverage.

The manufacturing quality control can then place acceptance and rejection criteria based on the results of the CRC test performed on the unit; sending failing units back for debug (a process which can also involve CRC tests) and sending passing units onward to packaging and shipping.

Figure 11: Manufacturing Test Flow

The CRC test is a powerful tool in an engineer’s armoury of system development, manufacture and debugging test methods. Although it cannot quantify the exact extent to which a system is being degraded by some problem (like bit error rate testing), it is more flexible (it can be run on any static pattern – no requirement to know the pattern in advance) and it is easily implemented.

3 Acosta-Serafini, P. et al., 2013. Apparatus and method for digitally-controlled automatic gain amplification, US20130033326 A1

About the authors

Michael Corrigan, Senior Design Evaluation Engineer

Michael Corrigan is a Senior Design Evaluation Engineer within the DVP group of Analog Devices. He qualified with an honours Bachelor’s Degree in Electrical and Electronic for Cork Institute of Technology in 2002. Michael joined Analog in 2007 having previously worked in EMC and GE security.

Joe Triggs, Applications Engineer

Joe Triggs is an applications engineer in the Digital Video Products group at Analog Devices (Limerick, Ireland). He has worked for the company since 2007 and is responsible for HDMI receiver, transmitter, transceiver and video signal processor
products. He earned his primary BEng degree from University College of Cork in 2002 before continuing to complete his MEng through research at the University of Limerick in 2004. He recently completed his MBA at the University of Limerick’s Kemmy
Business School.

If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :    eeNews on Google News


Linked Articles