The 1394-1995 Std. specs. the Cable Power "Maximum output current per port" at
1.5 Amps and the "Maximum output voltage" at 33 Vdc (per the 1394a
spec.), The std. goes on to state:
"If a node uses cable power, it shall meet the following requirements:
a) It cannot use more than 1W of power after a power reset or after being
initially connected from the bus (transition from all ports unconnected
to any port connected). Any additional power usage shall depend on
receiving a link-on PHY packet."
The PCI-to-1394 adapter card itself does not exceed the PCI specs. The adapter
is taking 12 V off of the PCI bus and providing it to the 1394 bus via the 1394 connect.
-----Original Message-----
From: Lame Brooks-G14738 [mailto:brooks.lame@motorola.com]
Sent: Thursday, April 06, 2000 1:16 PM
To: PCISIGList (E-mail)
Cc: Fockler, Joe
Subject: RE: In-rush current specification (?)
Regardless of the PCI spec, isn't that 4.5A on 12V for 600uS way over the
1394 limit? 18mJ/3mS=6W; 4.5A*12V=54W. The 1394 card design can not be
expected to handle that surge; so, some of it will get back to the PCI bus.
-- BrooksL
> -----Original Message-----
> From: Fockler, Joe [mailto:jfockler@ti.com]
> Sent: Thursday, 06 April, 2000 08:42
> To: 'pci-sig@znyx.com'
> Cc: 'Kidwell, Gary'
> Subject: In-rush current specification (?)
>
>
> Is there a maximum in-rush energy limit for devices
> hot-plugged into PCI expansion
> cards? The 1394-1995 spec states:
> "The inrush energy shall not exceed 18 mJ in 3 ms."
> but I don't see any inrush spec. in the PCI spec.
>
>
> We've seen some devices that exceed the 500 mA tolerance on
> the PCI 12 V rail
> when hot-plugged into a PCI 1394 adapter card. This in
> itself doesn't seem to be
> a problem, but we have seen one PC system that gets reset
> when a 1394 hard drive
> is hot-plugged and the inrush current peaks around 4.5A (with
> a 600 us duration).
> It looks like the 12 V rail drops and causes the POWER GOOD
> signal to get deasserted.
> There is a 750 mA slow fuse on the PCI card 12 V trace.
> Is the Power Supply Rail Tolerance of 500 mA for the 12 V
> rail a "Maximum" current?
> All insight and comments are greatly appreciated.
> Regards,
> Joe Fockler
> Texas Instruments
> 1394 Host Controller Group
> 12500 TI Blvd., MS 8731
> Dallas, TX 75243
> (214) 480-6064 phone
> (972) 761-6954 fax
> email: jfockler@ti.com
> Visit our Bus Solutions website:
> http://www.ti.com/sc/1394
> http://www.ti.com/sc/pci
>