GSM-noise "buzz" issue

[Joerg Reisenweber]

Am So  21. September 2008 schrieb Andy Green:
> Somebody in the thread at some point said:
> 
> | But if it's something more complex, such as the GSM module contaminating
> | GND itself, then you'll have to be able to do proper GSM signaling in
> | order to be able to reliably reproduce the problem.
> 
> I don't think so, I could just make a call and "reliably reproduce the
> problem" for the duration of the call when I force it to transmit.  It
> seems the customers are the same in not needing expensive gear to make
> the problem come.
> 
> This seems to be a pure frequency-dependent coupling issue, especially
> since Joerg tells us he can force coupling to this net by external GSM
> phone transmission nearby.
> 
> | (BTW, I also don't get the buzz in Argentina, where GSM uses the
> | 1900MHz band.)
> 
> You would think it also depends on output power from our RF section; it
> might be worth trying with "piece of wire" antenna to force it to use
> highest transmit power if otherwise your base station is close.

Using "piece of wire" might serve well to spoil SWR of transmitter, thus 
reducing antenna gain. So it's correct BTS will advice MT to increase 
xmit-power until level seen at BTS receiver is within limits. But this 
doesn't pan out to induce higher energy from "piece of wire" to the sensitive 
part over the air, as the field around "antenna" (=wire) is actually 
attenuated down by antenna/SWR-loss and exactly compensated by increase of 
xmit-pwr, so in the end we see exactly same field strength around DUT. ;-)
Only thing that changes is power drawn from battery and burnt in RF-amp of 
transmitter (actually it's the part of RF-energy that's reflected on 
mismatched foot-point of antenna and flows back to amp where it's absorbed in 
the power-transistors)

> 
> If it isn't that you are using low power, that's pretty interesting, it
> couples at 1.7GHz but not 1.8 - 1.9GHz then.  It seems something we are
> missing is frequency response of the radiated coupling to the victim net.
That's why I asked for it some mails ago, so we had an idea which frequencies 
are the ones GTA02 is sensitive to.
Though the result wouldn't probably change anything about proposed fix, which 
is to place beads/EMI-filters there on JK4401, and please do it as soon as 
possible.

/jOERG
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 194 bytes
Desc: This is a digitally signed message part.
Url : http://lists.openmoko.org/pipermail/hardware/attachments/20080922/dc484cc5/attachment-0001.pgp