Re: HOW2 test HDMI cable?

[email protected] wrote:
> Years ago I used a HDMI cable from a RaspberyPi to the TV.
> Later I had a problem and suspected the HDMI cable.
> Now I want to revive the Rpi, but get no signal on the TV.
> How could the laptop [Win8.1 & linux] test/confirm the cable?
> TIA.
> 

At my local computer store, it's $8 CDN for a 3ft HDMI,
and $9 CDN for a 6ft HDMI. Drive down and pick up at the door
(can't go into store now).

Picking up cables online, might be cheaper for the cable,
but shipping will wipe out your saving. You'd need a
shipping service you'd already paid for, to improve
the economics.

Nobody at retail, should really be giving you a break.
You pay the bricks and mortar premium that way. But if I
need it, I can get a known good cable about ten minutes
from here.

*******

You test cables with an ohmmeter. Select pin one, touch
the second ohmmeter probe to pin two, then pin three,
each time verifying there is no short circuit. There would
be a large number of tests required. That is, unless you
know something of the orientation and internal
construction of the cable, to rule out some of the
possibilities.

A "breakout connector", which is a connector with a
patch panel PCB on the end, makes it easier to probe
the individual wires, but without as much eye strain.

On the other hand, you want to know that there is
continuity. That pin one on one end, joins to pin one
on the other end. Some testing jigs, what they do, is
they put the wires in series, by means of a wiring
pattern on the female connectors on the test jig.

ohm here   ===> 1-----12-----23-----3 <=== ohm here

This is a quick way of verifying all N wires, by putting
the N wires in series with one another. If the ohmmeter
reads zero ohms at the two "test points", the cable has
continuity.

But that doesn't cover incidental short circuits. That's
what the other test was for, for detecting whether pin one
is shorted to pin two or not.

You might be able to damage an HDMI output on a computer,
if some driver pads got shorted to the +5V on the cable.
Graphics cables sometimes have power on them to run
the EDID EEPROM in the peripheral end. And the common
mode voltage range of the color guns, is not wide enough
to handle a short to +5V. Now, I doubt something
like this has happened, and you probably need to
select the port on the TV to see the signal.

If the cable has visible damage, if the foil around it
is damaged, the cable has been bent and kinked (which crushes
the dielectric plastic), these are not good things, and
could cause a signal failure.

The signal propagates as differential pairs, a (+) wire
and a (-) wire. The signal on one wire in the pair
"moves in the opposite direction" of the other signal.
This gives, in effect, twice the amplitude at the
receiving end, but gives the unenviable task of the
receiver converting the waveform to digital 1's and 0's
at gigabits per second rates. Now, at the end of the cable,
is a flyby termination.

               RX-
                |
   -------------+-----X
                     100 ohm resistor
   -------------+-----X
                |
               RX+

The resistor does two things. It prevents reflections of the signal
off the right hand side of that picture, being sent back at the
transmitter well off on the left hand side of that diagram.

But the termination can also be "sensed". The transmitter knows
whether a TV is connected. It only agrees to put a signal
on the cable, if the 100 ohm load can be seen electrically.

In the Display control panel of the computer (Windows laptop),
you would check and see if a "second monitor" is visible in
the control panel. The graphics have Mirror, Span, and
other options. You don't want Mirror, because if the
laptop panel is 1368x768, the TV may not like that as an
input signal. If you Span two monitors, the first one
is allowed to be 1368x768 while the other one declares itself
as 1920x1080. Then you can test via looking at the TV
set for that new signal.

      Span      +------------+
                |  TV as a   |
      +--------+|  monitor   |
      | Laptop ||            |
      | Panel  ||            |
      +--------++------------+
           1           2

Generally speaking, all the 100 ohm loads would have
to be open circuit, to stop the laptop from thinking
a display was present. But the laptop also reads the
EDID EEPROM in the TV set, and if that connection was
not there, then the output resolution choice might be
1024x768, instead of 1920x1080, and you'd have a hint
the laptop is not seeing all the info it needs to work
good.

You can get some subtle hints, when the laptop only
offers 1024x768 on its end, and the TV screen remains
black. Perhaps a signal is not being sent, because
the laptop is not happy, or a signal is being sent and
the TV set is "all in a huff" because the food being
served to it, is not to its liking.

HTH,
    Paul