(Sorry, ultra-long-winded answer follows... I did not have time to trim it down)
SHORT ANSWER
I am thinking more and more now that the
root cause is that the last device in the chain just before the HDMI/Cat5e baluns is not putting out the full power necessary on the +5V line (pin 18 on a HDMI Type-A connector) for the HDMI sender-balun to properly apply the necessary pre-emphasis on the HDMI signals to run over long Cat5e links. The
far end receiver is powered by a DC adapter so that
its chip has power and to provide the +5V on pin18 to
its downstream device. But the SENDER balun does not have a +5V DC power plug.. Nor should it need one. The HDMI spec says that the upstream device should provide that power..... but maybe that upstream device is NOT providing that power (or enough of it)...
LONG ANSWER
Well in theory, those opinions talking about EMI and ground loop issues (the two are extremely closely related) I believe are probably correct-- HOWEVER, I do NOT believe they are the
ROOT cause, which is why using shielding Cat-X cables don't always fix the problem. In my case, I am stuck with what I have, running a new (HDMI) cable would be
extremely difficult and costly. So the whole shielding issue is a moot point for me.
In my mind there are one or two things big possibilities..
(
DISCLAIMER! I am not a pro on the HDMI spec! This is all just speculation on my part!).
On the one hand you have differential signaling (LVDS) being used on some the data lines. On the other hand you have NON-differential I2C communications going on for the DDC link..
At least this is my understanding. So the DDC link does not benefit from the differential techniques that the other data lines use. And since DDC is used in the HDCP handshake process, it's left to be more vulnerable over long runs. This *could* be helped by using shielded Cat5e/Cat6 cables, but again is not the root cause (in my mind)... Now it is true that I2C communication is extremely slow (Almost at a stand-still in comparison to the other channels in an HDMI link), and that's probably the reason that it is not differential (less costly).
So that issue is probably not the root cause either.
Now, the second aspect is this +5V signal that's carried on pin 18 of a type-A HDMI connector. Reading various forums over the past couple of days looks like not all devices put out the max allowable power draw. The simple net result is that any downstream device that needs power to do its job (and expects it per the HDMI spec) is set up to be a disappointment if the upstream device can't power it sufficiently. Certainly active (chip-based) HDMI/Cat5e baluns need power for their chips to do their job. They act as a buffer/isolator/repeater device.
But as with any electronic device, if power is 'starved,' then the device will work sometimes, but not other times. It becomes extremely "situational dependent." (hmmmm.... beginning to sound familiar, isn't it?)
In the baluns I use (Arkview), only the far-end (TV side) receiver is powered by a DC power adapter. I always wondered WHY the source-end adapter was not powered. Foolishly I thought that the far end device was simply supplying power to both sender and receiver. But that wouldn't really make sense-- Cat5e isn't really made for power distribution (though it CAN be used for it... just look at the PoE standard). So in retrospect I
now believe the sender-balun is simply expecting to draw
its power (because the HDMI spec says it can) from the upstream HDMI device. But not all upstream devices may actually provide the balun what it wants/needs (or maybe just not as cleanly as it should?), resulting in the sender-balun in some situations not having enough power to properly do its job..... The sender balun is being starved, which not only yields for a much WEAKER signal going through the Cat5e wires, but also makes those signals (and the transmitter device itself) much more vulnerable to EMI spikes.. even static electricity from a hand-to-doorknob-static-discharge). This *could* be the cause of these HDMI - Cat5e baluns not working the way they should with some devices, and working flawlessly with other devices. It sure is lining up with my understanding of electronic devices and what happens when you starve them of their full necessary power..
In fact, glancing quickly at an HDMI/Cat5e extender at Amazon finds this review.. sounds familiar.. the first comment listed below also sounds familiar.. the second comment I quote below is about the 10th time in the past couple of days that I have heard that bit of information, which is where I began formulating the above hypothesis.
Item:
Tripp-Lite (Model P167-000) HDMI over Cat5 Active Extender Wall Plate Kit
Review:
Amazon.com Customer ReviewQuote:
By John A. Gorrilla (August 4, 2009)
I just got off the phone with Tripp Lite and am sorry to report the HDMI over Cat5/6 extender will not work through a receiver. No explanation from Tripp Lite other than it has to be plugged into a source directly.
There does not seem to be anything about this in the product literature unless I just missed it, but since I am sure I am not the only person who has dreamed up this configuration, I thought I would pass along the information.
CommentsQuote:
Timothy M. Vann says:
I had the same problem.
The tripplite would work with the PS3 plugged directly into the tripplite like this:
PS3-->Tripplite-->TV
but it would not work going through the receiver like this:
PS3-->Pioneer VSX 94 Receiver-->Tripplite-->TV
and
Quote:
J. Kolb says:
Hi, i was very excited by my recent discovery of a solution to this problem so I thought i would share. Apparently, alot of receivers do not comply to the HDMI specification requiring 5V of power over one of the pins. On a hunch, i bought this from amazon for $15 Cables To Go 42223 RapidRun Digital HDMI Voltage Inserter (Black). They are also sold on other websites.
This piece solved my problem and now I have pretty much flawless picture!
I had the problem on two receivers, yamaha 663 and brand new onkyo ht-rc180. It fixed it on both! Definitely a much cheaper solution that buying a new balun.
Please note, i also instead a viewhd 2 port powered splitter, which gave me a picture but not without significant "sparkling" of the picture. Using the voltage inserter, i get perfect picture, although every now and then I notice a 1 frame flicker of grey fuzz, which may very well be the source i have hooked up, but i havent connected any other devices yet.
Looks like I've convinced myself to buy one of those HDMI Voltage Inserters....
cheers,
..dane
