[Arm-netbook] IC for analog and digital buttons (EOMA-68)

[Luke Kenneth Casson Leighton]

On Mon, Aug 11, 2014 at 10:00 AM, krasi gichev <krasimirr at gmail.com> wrote:

> You say Gbit Ethernet - but IC1T does not have this? So I, in the position
> of designer of base board, cannot rely on this because some cards with give
> limited speed?

 again, the end-users get what they pay for.  if they pay only $20 for
a CPU Card, they should expect to get $20 of functionality.  if
however they pay $100 for a CPU Card and it says GbE and USB3, and 4gb
of RAM, and octal-core, they again get what they pay for.


> I have another question, sorry if it is out of context here. I know PCMCIA
> was designed long ago - is it certified that it will run fine with 5Gbits
> signals USB3.0? Does it comply to impedance matching for all interfaces?

 give-or-take, yes.  the impedance of PCMCIA pins is around 100 ohms.
USB is 90 ohms.  i'd say they're close enough, and the distances
involved are very very short.

> How about EMI?



 this was discussed a couple of years back.

 to repeat that discussion: we'll find out soon enough :)

 again, to repeat what was mentioned back when this was first
discussed: if it turns out to be a problem there is the "gold card"
version (known as CardBus) where a metal shield is soldered at 10
different points onto the PCB, covering the PCMCIA connector.

 also what was discussed, and repeated now, is that unlike with the
original PCMCIA signalling the high-speed signals going over the
PCMCIA connector are mostly differential pairs with much lower voltage
swings.

 the only one that may be of concern is the RGB/TTL interface, which
will be operating at (in some cases) 75mhz.

 so, to repeat the conclusion of the discussion from two years ago: if
the EMI tests do not pass, we simply convert to the CardBus version
with the gold shielding.  however if it is not needed i do not wish to
pay the extra costs, hence the reason why we go with the standard
version for now.

l.