On Sat, 16 Mar 2002 22:41:35 +1100, "Hughy" wrote:

>That is quite extraordinary and alarming given that the poster is/was a
>Director of Technology.
>
>Although I am not familiar with 1394 comms, on other types of systems
>polling is done using short duration pulses (in the order of
>milliseconds or less) at relatively infrequent intervals. Consequently
>the duty cycle restricts average dissipation to milliwatts or less,
>which will not normally harm silicon junctions. Do 1394 systems have
>the potential to deliver higher powers than this?
>
>If firewire ports are indeed being burnt out, I cannot yet see how the
>polling can be responsible. Was there any further information given on
>the precise cause of the "burn out"? I wonder if the burnouts were a
>result of an unusual component failure on the capture card?

A guess:
Most PC FireWire cards use 4-pin to 4-pin FireWire
connectors (I hate those badly-designed-for-easy-socketing
plugs...!); Macs use 6-pin to-4-pin connectors, which
additionally supply voltage to power devices. Perhaps
inserting the plug backwards places the supply voltage on
a pin it should not be on? Or shorts it? FireWire is
supposed to be "hot plugable", and in a few years of
using it on PCs with Sony gear, I have found that to be
true...