Andy Luckman (AJL Electronics) wrote:

In article , Lee
wrote:


You can sometimes get away with slightly longer cables if they are
shielded.


The shielding may not be the problem. With any parallel data, long line
lengths can give rise to "data skew" where some bits of the byte arrive out
of sequence. I don't know how long the cable needs to be to see the effect,
but it happens.

Interesting, I realised the shielding was only effective against noise,
but I didn't realise that data skew was such a problem on ide cables.

Doing a bit of searching on this topic, it seems there are reports of
data skew problems even in some "standard" length cables.
Also, given the error correction used in modern systems, this sort of
problem may not be obvious.

I have noticed a slight difference in benchmarked speed using different
length/quality cables, but I've always put it down to coincidence.
Maybe there is something more to it...

Lee
--
Email address is valid, but is unlikely to be read.