The predetermined attenuation curve for HDMI High Speed cables allows them to function at the higher bandwidths under Rev 2.0.
As always there are good cables and not-so-good cables. This is why we put such a huge emphasis on getting cables that do in fact make it well over the complete specification, including bandwidth. This just proves it.
The folks at HDMI knew what they had under Rev 1.4 and used this built-in anticipated attenuation curve as a means to bring back the massive loss that takes place as we move up to the higher frequencies. But now you have to look back on your past installs and pray these products do in fact follow these curves so that they are functional under future revisions.
However, what happens to all the active cable products? The HDMI group did announce in one presentation that it was highly unlikely “extenders” would not support the over-3Gbps Rev 2.0 features.
Let’s face it: there are tons of these products in the field, especially in our sector. Long-distance hauls are mandatory, and over time we’ve all learned what can and cannot work. So let’s look into these and see how they fare.
Most semiconductor companies that make the electronics used for such devices design their products to what the specification calls for and that, historically, has been an aggregate of 10.2Gbps. As any good chip manufacturer would, they build the part to support the spec and no more.
This is done from both a cost position and to limit risk for any instability of the part. It’s like audio amplifiers; many are built to do their job within a given bandwidth of 20kHz. You will not find many that go over this. When they do, they have to build in some precautionary measures so the part does not oscillate and become unstable, which costs even more money. These are just good engineering practices.
So now back to HDMI. These same type of silicon manufacturers make a lot of the products we use every day in the HDMI world including switch devices, equalizers and clock recovery; and then you have the FPGA guys (Field Programmable Gate Arrays) making up there own concoction of parts. These guys are smart and try to make parts that work with a high yield percentage.
Equalizers have been used in devices we call extenders for most of our long-distance applications. These have become a major part of the integrator’s life since HDMI has always had limitations on distance. These distances are inversely proportional to the system’s bandwidth. The higher the frequency gets, the shorter the length of cable.
Shown are few response curves of one TMDS channel (in this case D2) on some competitive HDMI active cables. Yes, Houston, we have a problem. All but one product here supports the entire bandwidth of 5.94Gbs. Cable 1 is a response curve of an active 5-meter cable with a part that was never intended for an extended range over 3.4Gbs.
However, notice it passes through the limit lines at just over 2.4Gbps, not quite making it to the entry-level 4K60 requirement of 3Gbps.
Cable 2 is another 5-meter cable, but look at the response curve difference. It makes it out to about 3.6Gbps, over a gig better. We examined both products and found that the active part inside each was identical — so why the big difference? Remember, you only get what you pay for. Cable design plays a big role here.
Then there’s Cable 3, which is off the chart. The reason is because the particular part used in this active cable product was made to extend out to 6Gbps … pretty incredible, huh? It is well above the limit lines where bigger is definitely better.
Article originally published on CE Pro.
Leave a Reply