Wednesday, May 9, 2012

Intermittent GPS...

Ok,  I finally got sick of it happening and I decided to take down my "PBJ" (Peanut Butter Jar) GPS antenna and figure out why it was that I kept ending up with "zero" satellites in view.
Figure 1:
Homebrew GPS antenna, in use since 2003, affectionately
referred to as the "PBJ Antenna" since a glass peanut
butter jar (and its
lid) are used to house the antenna.
Click on the image for a larger version.

Years ago (2003, I think) I obtained a surplus HP Z3801 GPS receiver.


Well, they became available for an affordable price (about $250, I think) and with it, I could use the GPS system to obtain a very accurate source of both time and frequency.  Of particular interest to me was that it was capable of producing a 10 MHz output that was accurate to enough to hold a clock to within a second every several hundred thousand years.

I don't really need a clock that is that accurate, but a source of stable, accurate 10 MHz was useful in testing and calibrating other frequency sources - such as the local oscillators of microwave transverters.  A year or so ago, I put a low-power small form-factor PC on this receiver to permit monitoring of its status as well as crunch away at the NOAA weather satellite images that spun by from overhead spacecraft.  Adding this computer also made it easy to provide a local NTP (Network Time Protocol) server that came in handy when operating some of the narrowband digital modes such as WSPR:  Having a local time server that I could query very frequently was helpful in keeping my operations synchronized with everyone else's...

More recently I added a Trimble Thunderbolt GPS Disciplined oscillator - a much smaller, lower-power box that did much the same as the as the old 'Z3801 - but with a really cool user interface program (more on this on a later date) and I was annoyed when, shortly after installation, I saw that both it and the Z3801 were reporting frequent dropouts of the GPS signal.

Actually, this had been going on for a while, but it seemed to happen only rarely and for a short time, but now, as the season was warming up, it seemed to be offline about as much as it was on.


Taking it down, I poked, prodded and "Ohm'd", but nothing was obviously wrong.  During the poking and prodding, I might have cracked a surface-mount bypass capacitor or two (mounted "tombstone" style to the circuit-board ground plane on which the antenna's circuitry was built) so I replaced these just to be on the safe side, reflowed a few solder joints, lugged it back up onto the roof and connected it.


For a day or two, anyway...

After a day or so of behaving itself, it became worse than ever - and then the weather got bad for a week.

Last Saturday was a nice day and I'd just gotten back from a breakfast meeting of the Utah Microwave Group and was going to meet K7RJ in a few hours to work on his 10 GHz transverter so I went back on the roof and retrieved the antenna once again - this time, managing to do it while the receivers were reporting a loss of signal.

This time, I was "lucky" in that I noticed right away that the voltage on the output lead of the MAR-6 MMIC preamplifier was "wrong" and the bias voltage on its input lead was zero.  Using a jeweler's loupe, I stared at the connection where the MAR-6 input connection was made with the UT-141 coax from the turnstile elements and spotted a tiny flake of metal.  A few minutes of surgical unsoldering, cleaning with a straight pin and careful resoldering (to avoid the installation of another metal flake!) resulted in proper voltages in all of the right places.

Taping things back up and running back onto the roof I reinstalled the antenna and upon my return to ground level, I was gratified that the both GPS receivers now showed that they were in the process of re-synchronizing to the satellites.

Since then, everything has continued to work as it should...  I think...


As things like this turn out, after a week or so of flawless operation, the GPS antenna once again became intermittent, so I hauled it back down to the work bench.


The second time I finally replaced the MAR-6 MMIC since I'd replaced/resoldered everything else and that seemed to fix the problem for good!

Since then, both my old HP Z3801 10 MHz Disciplined Oscillator and my "newer" Trimble Thunderbolt have been happily locked to the signals from the GPS and providing accurate, stable 10 MHz references.


This page stolen from

No comments:

Post a Comment


While I DO appreciate comments, those comments that are just vehicles to other web sites without substantial content in their own right WILL NOT be posted!

If you include a link in your comment that simply points to advertisements or a commercial web page, it WILL be rejected as SPAM!