ESD. Two serious issues: GPS-ext-antenna and Touchscreen outer resistive foil

[joerg]

Am Di  4. März 2008 schrieben Sie, Andy Green:
> Hi Joerg -
>[...]
> Can you post about these two issues to the kernel list... 
> [...] I think they need consideration.
>
> Best to make it clear it is about ESD and there are two separate issues.
> 
[...] 


>
> >>> Touchscreen is a good point, although again I think it is inherently a
> >>> high impedence path.
> >>
> >> For GTA01: U6003 U6004 have no protection (maybe built in?)
> >
> > They are just bipolar transistors with base series resistor, should be
> > ok.
> >
> > http://www.rohm.com/products/databook/tr/pdf/umd6n.pdf
> >
> > For GTA02 that is a worse situation :-/  they just travel direct to the
> > CPU ADC input.  Datasheet does not discuss characterization of these
> > inputs.

First of all it should be checked whether there is a shielding plane covering 
the outer resistor foil, which would make this a null-issue. However i don't 
think so, each conductive layer takes some % of transparency.

TS is connected via flat ribbon? Maybe something can get spliced in there.


> >
> >>> GPS I will look at later, I guess there is some matching network at
> >>> least between it and the chip.
> >>
> >> Nope, nothing in the path there, just phantom, which is L and so is
> >> useless.
> >
> > Yes you're right, GTA02 is the same.  I find the RF switch datasheet
> >
> > http://www.ncsd.necel.com/microwave/english/pdf/PG10218EJ01V0DS.pdf
> >
> > It does not discuss ESD but shows +26dBm "max input power" which is
> > 400mW.  I would just blow it off if we only support internal antenna but
> > it is easy to imagine bad things happening with external antenna in a
> > car on a hot day, etc.  Hum.
>

As mentioned in prev posting, this might get fixed with right type of 
connector (make GND befor signal)


jOERG