[time-nuts] GPS Interference Question

Jim Lux jimlux at earthlink.net
Fri Sep 30 14:01:23 UTC 2011


On 9/30/11 6:44 AM, Jason Rabel wrote:
> Please don't start a political discussion, I just have a SIMPLE TECHNICAL QUESTION that I'm hoping someone can answer (and I don't
> think it was discussed in the past).
>
> In all the stuff I've glanced over about L2, they talk about better filtering for GPS modules to eliminate the interference issue.
> Likewise they talk about potential replacement of existing hardware and such because of poor filtering that overlaps into other
> frequencies...
>
> My question is...
>
> To filter out the L2 signal, would an actual GPS receiver have to be replaced / modified?

No
>
> Or would a more simple and cheaper alternative be to get a new antenna (with fancy filtering) to replace my existing roof-top
> antenna and expect all my old equipment to be happy? Is that technically possible or does it have to be on the receiver?
>

Yes
for that matter one could do fancy antennas that place a null on the 
interfering source.


> Or even something in-line where I wouldn't have to replace either my antenna or receivers... It would just look like an attenuator
> that you stick somewhere along the cable?
>
Yes

---
everything is a compromise, of course.  A filter that rejects frequency 
A doesn't usually have zero loss at frequency B; ditto fancy antenna 
designs.

If the interfering signal is strong enough, you won't be able to have an 
LNA right after the antenna, because the LNA will saturate.  (this is 
the potential problem with your "in line" alternative)

The real issue is that these solutions take care of a fairly small 
fraction of the total GPS receiver population (e.g. it works ok for 
fixed timing receivers, but not so hot for size, power, weight sensitive 
applications)

They're also, potentially, fairly expensive in the aggregate ($100 of 
filters for 10 million GPS receivers is getting into serious budgetary 
area. And there's a lot more than 10 million receivers out there... I 
could believe 100 million, but probably not a billion)



More information about the time-nuts mailing list