Re: Re: Impedances of SP/DIF (distributrion)

[Bill Taylor <bill-taylor@xxxxxxxx>]

m> I think you're right on the high vs. low issue... however....

>> The output impedance of the Sky+ box *SHOULD* be 75ohms otherwise
>> you would have an impedance mismatch as an SP/DIF input *SHOULD*
>> be 75ohms too.

m> That was how I understood impendance matching to work until I had a
m> long talk with Keith at UKHA2004.

m> After a bit of digging around I came up with
m> http://arts.ucsc.edu/EMS/Music/tech_background/TE-15/teces_15.html
m> which talks about the difference between output impedance and input
m> impedance.

m> To quote from that article:

m> "There was a time when output impedance (also called source
impedance)
m> was carefully matched to the input impedance of a connected device,
m> because that is the most efficient way to transmit power. However,
m> since a connection will work if the input impedance is higher than
the
m> source impedance but will not if the input impedance is too low (low
m> impedance requires more current, remember.) engineers tried to design
m> circuits with somewhat higher input impedance than actually
necessary,
m> and lower output impedance. (A low output impedance implies that lots
m> of current is available.) There is a limit to how high input
m> impedances can go; if it is too high, unwanted low current sources
m> like radio stations will begin to affect the circuit. The usual ratio
m> of input impedance to source impedance is 10 to 1"

I think two things are getting confused here; transmission lines and
power transfer.

At frequencies were the cable length is less than the wavelength of
the signal, impedances are unimportant for most purposes. Yes, source
impedance = load impedance gives best power transfer, but at low
signal levels this is unimportant. It is normal for audio signals to
be fed from a low source impedance to a high load impedance.

At frequencies were the signal wavelength is of the same order as the
length of line, impedance matching becomes important. If the source
and load impedance don't match the line impedance you will get signal
reflections along the line and this will result in addition or
cancellation at certain frequencies. At audio frequencies this only
matters to BT, 'cause they have lines that may be miles long, but at
video frequencies it becomes an issue with pretty short cables. SPDIF
operates at video frequencies so you should try to keep the impedances
matched reasonably well.

m> OK - let's ask another way.... Has anyone tried it????

When I wanted to distribute a couple of SPDIF feeds I bodged a Keene
S-video DA; I rewired it to use both the luminance and chrominance DAs and
used
the audio connectors for the outputs. It worked fine for my purposes.

I wouldn't even bother to try a serious mismatch with SPDIF, but the
output impedance of your Sky box should be low enough to work with a
resistive splitter as others have suggested, as long as the signal
level is high enough.

Bill