DPRG mailing list archives: [DPRG] GPS and ground planes

Sun May 4 13:07:52 CDT 2008

I should have added that the sizing of the ground plane referenced if for a
standard monopole configuration antenna (such as a mag mount.)  Because
these are passive patch antennas and the wave is circularly polarized, the
value I posed before for size is a "worst case" => *GUESS*.  I just thought
to throw that out there before someone calls me on it, or before anyone
heads out and buy material.

-Kenneth

-----Original Message-----
From: dprglist-bounces at dprg.org [mailto:dprglist-bounces at dprg.org]On
Behalf Of Kenneth Maxon
Sent: Sunday, May 04, 2008 9:09 AM
To: Jeff Sampson; Kenneth Maxon
Cc: dprglist at dprg.org
Subject: RE: [DPRG] GPS and ground planes

A short connection from as close to the antenna ground reference as you can
get it.  If one has a commercial hand held GPS then the DC return of the
unit (lowest battery connection) will have to suffice.  In your situation
with an OEM module, you can get much closer, like the grounded FRO chassis
directly off of the edge of the module or PCB.

David Anderson sent me a photo of his solution on jBot.  I think his ground
plane was just the robots DC return potential that was common to the GPS?
(Going from sketchy memory there.)   The part from his experience to share,
is that he used a rater largish chunk of U-Channel extrusion (appears to be
aluminum) and set the whole radio down into it.

Following is a snippet from David's e-mail...

      I found I had to do this on jBot to get reasonable GPS data, and also
      mount the GPS rather high above the rest of the robot platform.
Here's
      a picture:

      <http://geology.heroy.smu.edu/~dpa-www/robo/Encoder/imu_odo/jbot_imu_o
do_02.jpg>

      that shows the GPS (yellow and black device) mounted in a grounded
aluminum
      channel which shields it on 3 sides, which is what I eventually had to
do
      to get a reasonable performance.

On the command set side, I have been issuing the commands out to the GPS to
turn the DPGS lock, I just have never had a lock.

In the bigger scheme of things, GPS is a "nice to have" for my robot and all
of the other sensors are really designed to get it around.  I spent and
inordinate amount of time debugging this issue to no-avial before I had to
stop working on robotics for a few months.  I am now re-engaged as my other
obligations have passed, however likely I'll not spend a large portion of
time on this aspect for a while since one or two other systems still deserve
attention.   In the past week I have been working on the long range
telemetry portion of the robot with a high gain directional antenna,
amplifiers, radio and of course filters.   Lots of fun to play with.  :)

-Kenneth

-----Original Message-----
From: Jeff Sampson [mailto:jsampson at pobox.com]
Sent: Sunday, May 04, 2008 1:52 AM
To: Kenneth Maxon
Cc: dprglist at dprg.org
Subject: Re: [DPRG] GPS and ground planes

Kenneth Maxon wrote:
>
> Hi Jeff,
>
> I had many problems with a SMIR(III) based GPS receiver.  Adding a ground
> plane helped some of the problems but not all.  Additionally, making sure
> that I used a very short extremely low bonding resistance connection
> improved the signal returns.

Short connection from where? From ground plane to the chassis? Or
between the ground plane and the receiver case? I also have a SIRF III
(the EM-406A from Sparkfun) that I could ground the case. I'm not happy
with the performance of that one yet. Maybe I just haven't played with
it enough yet.

> GPS has a primary wavelength of just under 7.5", so a ground plane that is
a
> full wavelength or larger will work well.  (bigger is better w/diminishing
> returns.)   If you have the room to make a round plate with a 15" diameter
> you will not get much better than that.   I haven't had time to go out and
> solve this one for a circularly polarized wave, I just did it
> experimentally.   BTW, In my solution I ended up using a much smaller
piece
> than that (square ~7").

I might try a square or rectangular plate. At the moment I am just
carrying the GPS. So I have more options in what I use. The 15" circle
may work on the robot when I get that far.

> If you care to see, a few months back (late fall / early winter) there
were
> a number of messages back and forth on this list for reference with some
> great input from several group members to help track down some of the
> sources of error.

I knew there was some discussion, but I wasn't sure what to search for
or what the topic was. I'll check it out. I remember some discussion
about ice.

> My system still suffers from excessive wandering in the GPS signal but
that
> wandering is now almost always contained to ~50ft.  On a rare occasion it
> will still wander up to ~100ft away from actual position, however this is
> rare compared to the previous errors of 280ft that would happen several
> times over the course of a days worth of captures...
>
> As a side note, I also have a Garmin CSX60 and it has identical
performance
> issues.   I took the Garmin on my last trip to Savannah with me and inside
> my hotel room running for 2hrs capturing data it never wandered more than
> ~30ft (15ft radius) without any ground plane... Since I am only 250miles
> south of you here in the upper mid-west, I wonder if there is something
> geographically related to it?

I did notice that that my few short tests did have a whole bunch of
wandering when I first started my path recording. Either I did not wait
long enough to get a good fix or the internal software required that I
move before it decided on its calibration. Possibly it was assuming too
much gain because it had not seen me move yet. I'll play with that again
  and see which is causing it.

> Question: On the SMIRF(II) chipset that you have been using, have you ever
> been able to get a lock on the WAAS signal?  On the SMIRF (III) I have not
> been able to receive it and do not know if it is my location or the way I
am
> interfacing to the chipset.

A year or so ago when I played with this I saw a flag which inidated
DGPS, which I assume was WAAS. I haven't looked at the data this time
around on either eceiver. I'll check on that.

> -Kenneth
>
>
>
> -----Original Message-----
> From: dprglist-bounces at dprg.org [mailto:dprglist-bounces at dprg.org]On
> Behalf Of Jeff Sampson
> Sent: Sunday, May 04, 2008 12:11 AM
> To: dprglist at dprg.org
> Subject: [DPRG] GPS and ground planes
>
>
> Hi,
>
> I played with my GPS. It is a Rikaline USB GPS module for a laptop
> computer and has the Sirf II receiver. I noticed the resulting path that
> I recorded didn't look like I expected. It was close, but looked sloppy.
> So I stuck it to a piece of steel sheet, about 8"x9" and ran it again.
> (This is one of those that has a magnet to stick to the top of your
> car.) It worked much better the second time.
>
> Does a GPS receiver prefer a ground plane? Or did I just have a better
> fix the second time? If a ground plane is preferred, does anybody know
> the recommended size of the plane?
>
> --
> Jeff Sampson
> http://tcrobots.org/members/jsamp.htm

_______________________________________________
DPRGlist mailing list
DPRGlist at dprg.org
http://list.dprg.org/mailman/listinfo/dprglist