The Custom Integrator Show Installment 01E is live.  This one starts our series on HDMI.  Before we get started though, Ian provides a little more background on his experiences with his Zune HD.  Unfortunately, there currently does not seem to be any DLNA support as we hoped.  I also cover a little bit about the new announcements from the National Institute for Standards and Technology (NIST) regarding their Smart Grid efforts.  They just released their NIST Framework and Roadmap for Smart Grid Interoperability draft document at  I feel these efforts really impact those integrators working with control and monitoring systems that may need to integrate with the efforts of the utility companies.  Of particular interest are the sections covering the Home-to-Grid aspects.  There are several other documents detailing these aspects that we plan to cover in future installments.


Direct Download  Subscribe via RSS  Subscribe via iTunes  Subscribe with Zune

We initially cover some of the generalized aspects of HDMI like the bandwidth requirements and the technologies behind the wiring architecture.  We dive into these a little deeper in the next installment, but here are some of the more applicable terminologies related to HDMI.  These will be useful as we discuss them in Installment 01F.

Bandwidth — The aggregate bandwidth requirement for HDMI Version 1.3a Category 2 is listed at 10.2 Gbps. This value is the sum of the individual color, data, and audio lines contained within the HDMI cable. More HDMI “dynamic headroom” will be needed above that to take it to even higher resolutions. Low bandwidth decreases the available resolution and or the color depth.

Capacitance/Dielectric loss — High capacitance and dielectric loss causes video frequency response issues and corrupted data on the data communication channels (DDC). This is caused by the proximities of the individual wires within the cable to each other and the dielectric materials between them.

Current for the supply line — The current required to pass the Hot Plug and EDID (Extended Display Identification Data) information between the source and the display. A weak current causes this handshaking to fail.

The HDMI standard specifies a minimum of 55 mA drain on the voltage supply lines for handling the Hot Plug feature and the passing of EDID information between the source and the sink. Some of the “active repeaters” built into HDMI cables draw their power from this same line. If the amplifier circuitry draws 100mA or more, which most do, and is used with a weak source current, the handshaking used during the creation of the connection and in maintaining it likely will fail due to the fact that it is not a requirement for all manufacturers to provide the high current required by such devices.

Display Data Channel (DDC) Corruption
— EDID and HDCP (High-bandwidth Digital Content Protection for digital rights management) data are passed as part of the constant handshaking process between the source and the display. DDC data corruption is the largest cause of the connection having problems or not working at all.

Video quality can downgrade or even change in color if DDC is corrupted. With some systems, any DDC corruption can cause the unit to fail totally. HDCP needs to refresh about every 2 seconds. If the system does not get this refresh, it surely will die.

DDC-I2C (Inter-Integrated Circuit)
— This electrical bus supplies all of the intelligence to the system. The clock signals used for the data transfers must match the timing of the data precisely.

Delay — The timing of the serialized signals varies for the individual conductors within an HDMI cable. If these individual delays fall outside of the “wiggle room” for the specification, the clock timing fails to operate correctly and failures occur.

Eye Pattern — This is the popular “Cat’s Eye” pattern shown in ads for many of the better cables on the market. The pattern visually depicts the electrical characteristics of the entire data transmission: output voltage, frequency response, jitter, noise, timing, and even the probability of its working well.  It is a good indicator of the overall quality of a cable and how well it actually will work.

A “mask margin” is inserted into the actual signal reflecting the HDMI minimum compliance spec as a base for comparison. This margin is a percentage measurement. The key is to see how much performance over and above the spec’s margin is achieved and, in some cases, negative margin (or non-compliance).

The left photo shows a near perfect eye at 1080i — notice how wide the white space is. The middle photo shows the eye for the same length at 1080i for a different manufacturer. Although both products pass the eye test with flying colors, the right cable would be the first one to fail when pushing for higher resolution (1080p) or longer distance (or if it is installed improperly). The left cable would get a higher ranking than the left. If they cost about the same, which one would you buy? Needless to say, the cable on the right is “in the dirt.” It may work, but there is a high probability it will be unreliable.

Impedance — All wires exhibit resistance to the voltage being carried across it. To produce a perfect non reflective transmission line, load impedance must match the HDMI impedance of 100 Ohms. If there is not constant 100-Ohm impedance, the longer the cable, the more the performance of the signal is degraded.

Intra Pair Skew — The individual pairs of cables within the HDMI cable itself are twisted to eliminate noise and establish the operating impedance (balanced pair signaling). The tolerance of the twists drastically affects the timing of the signals running across them. If they are not completely symmetrical, one wire ends up being the slightest bit longer than the other. This means that the signal from one wire arrives the tiniest bit later than the other one, which could cause enough duty-cycle distortion to create bit-errors.

Inter Pair Skew
— The timing between each of the balanced pairs of wires within the cable. They all must be twisted, laid, and cut precisely to reduce any length differences.  Again, if the signal of one pair arrives even the tiniest bit later than another, it could destroy the signal.


Intra- and inter-pair skew: Some cables are over 300 thousandths of an inch off in timing, meaning the individual wires vary too much in length.

Jitter — Changes in the timing of the signals over time.  Due to tolerances and variances in the signal propagations, the signal characteristics vary, which affect the timing. Jitter is measured in picoseconds and has a very low tolerance.



2 thoughts on “The Custom Integrator Show Installment 01E”
  1. Excelent topic. Any thoughts on why HDMI cabels did not use an optical transport? Using copper seems to be problematic as amount of data increases, also would alow a thinner capble to be used (like a Toslink cable)


  2. I am sure that was a consideration. In fact, there are several optical-based HDMI extenders on the market. Some are pure fiber and others are a combination of fiber and copper. Opticis even makes an HDMI Matrix Switch for optical cables.

    I can think of several reasons why fiber did not go mainstream:
    – Cost – Although the cost of HDMI can be exorbitant, most of the fiber-based solution start at several hundred dollars for the electronics and several hundred more just for the cables. This also means a huge additional cost to the manufacturers if they were to build optical ports into their devices, which always is a major problem for any CE manufacturer.
    – Power – The sink termination needs power for Hot Plug Detect and EDID transfer. Note that the optical receiver components also need power, so that would increase the current drain. Adding a wall wart to the receiver side would have been very unpopular, although several fibber solutions have them. Using copper in with fiber has some possibilities, but there are distance limitations induced over long distances at the voltage and current levels currently used for those services.
    – Reliability – Let’s face it. The average consumer would have fiber cables destroyed in no time if they were the standard way to interconnect. Although they could have been armored somehow, that would have increased the cost and manufacturing requirements.

    With that said, I agree that fiber is a great solution for long distances if required. There should be no reason why integrators would not consider using fiber-based HDMI cables if the cost is not prohibitive and they have the ability to run them. There are fewer problems and limitations with fiber compared to using copper-based HDMI extenders. Fiber also provides an excellent solution for digital signage applications.


Leave a Reply