[X-HW] FW cables

[Alex]

On Friday, Apr 30, 2004, at 10:36 Canada/Eastern, John wrote:

> [...] Are cables really that "finicky"?

Yes, they are. A specific answer should come from someone with an 
intimate knowledge of the FW specs (i.e., not me), but the general idea 
is that a wire behaves a bit like a low-pass filter. That is, the 
higher the frequency of the signal, the more problems it has to go 
through, but, if one wants to send a lot of data very fast, one needs 
high frequency. Greatly simplifying, there are three issues, 
attenuation, interference, and reflection. A low-quality conductor 
(including poorly or improperly plated connector pins) has higher 
resistivity, and hence cause higher attenuation. Poor shielding allows 
interference either from the outside, or even from the cable's ground, 
if there's noise in the device's circuits. And impedance mismatch 
between the cable and the device creates an interface where reflections 
are generated -- and part of it is determined by the cable structure.

All this results in garbling of the signal; if the Mac can't establish 
reliable communication with the device, the device can't be used.

> I know the cheap clear cable works, because my iPod mounts and charges 
> fine from it. I also have a FW CD burner, which works fine with the 
> clear cable as well.

It's difficult to say why it works with one device, but not with 
another. Some devices may have higher tolerances for data loss; but it 
may be something else. For instance, a poor cable may allow 
interference internally from a noisy ground. If you connect it to a 
device with poorly-designed or -manufactured (i.e., noisy) circuits, 
the ground will interfere with the signal; but a high-quality device 
should have quiet circuits, and it won't put a lot of noise into the 
ground.

f