[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[gpsd-users] ✘Use for old HDD platters

From: Gary E. Miller
Subject: [gpsd-users] ✘Use for old HDD platters
Date: Fri, 7 Sep 2018 10:00:42 -0700

Yo All!

TL:DR: placing an HDD platter under your GPS antenna helps.

I have always been fascinated by the guts of hard drives.  When one
dies on me I open it up, and marvel at the precision of the parts.  The
platters are works of art, perfect shiny (these days) disks.  I always
keep several on my desk to play with.

During a recent discussion about adding ground planes to GPS antennas it
occured to me that the 95mm disks in a 3.5in HDD might make nice ones.

u-blox recommends a ground plane between 50 and 70mm in diameter,
warning that larger than 100mm will start to degrade the signal.  So
95mm is in the ball park.

So I slipped on under a GPS, while watching cgps.  What with all the
other flickering going on, hard to say if things got better or worse.

To see any change I had to modify the gpsprof polar plot mode to
calculate the min/max/mean/sigma of the satellite SNR as well as seen
and used satellite counts.

The changes are now in git head.

I tried adding and removing a platter from under serial puck GPS, USB
puck GPS, and external GPS antennas.  No attempt was made to actually
ground the platters.  With the platter, sometimes the GPS used an
additional satellite, sometimes the mean SNR went up by 5% or 10%,
occasionallly both.

The test is simple.

1. start with plain GPS running gpsd with stable output

2. run gpsprof:
       gpsprof -n 12 -T pngcairo -f polar > bare1.plot

3. when that finishes, place platter under GPS, wait 30 seconds.

4. run gpsprof:
       gpsprof -n 12 -T pngcairo -f polar > gnd1.plot

5. when that finishes, remove platter under GPS, wait 30 seconds.

6. repeats steps 2 to 5, incrementing the plot numbers.

7. plot the data:
        gnuplot bare1.plot > bare1.png
        gnuplot gnd1.plot > gnd1.png
        gnuplot bare2.plot > bare2.png
        gnuplot gnd2.plot > gnd2.png

8. Compare your png plot images.

Re-running the tests like above helps eliminate the natural ebb
and flow of satellite usage from the experiment.

The improvement is not large, but it is real.  I have lots of disk
platters laying around, so I'll use them.

See attached for cherry-picked sample plots.

TODO: test ground planes of different sizes.
      test if grounding the ground plane helps.

Comments?  Suggestions?

Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
        address@hidden  Tel:+1 541 382 8588

            Veritas liberabit vos. -- Quid est veritas?
    "If you can’t measure it, you can’t improve it." - Lord Kelvin

Attachment: bare1.png
Description: PNG image

Attachment: gnd1.png
Description: PNG image

Attachment: pgpIlrfAV9l2B.pgp
Description: OpenPGP digital signature

reply via email to

[Prev in Thread] Current Thread [Next in Thread]