[time-nuts] Something better than a Thunderbolt?

Eric Lemmon wb6fly at verizon.net
Fri May 4 00:47:00 UTC 2012


Mark,

I suggest that the Thunderbolt E is significantly better than the original
Thunderbolt.  The Thunderbolt E has 12 channels, is more sensitive, and is
more stable.  The double-oven OCXO appears to be immune to ambient
temperature changes.  It also requires only 24 VDC to operate.  Compared to
my two original Thunderbolts, the E model wins easily.

73, Eric Lemmon WB6FLY
 

-----Original Message-----
From: time-nuts-bounces at febo.com [mailto:time-nuts-bounces at febo.com] On
Behalf Of Mark Sims
Sent: Thursday, May 03, 2012 5:03 AM
To: time-nuts at febo.com
Subject: [time-nuts] Something better than a Thunderbolt?


Recently Sam managed to poke and prod a Trimble/Nortel GPSTM (NTGS50AA)
enough to wake it up out of its slumber and be recognized by Lady Heather.
The NTGS50AA is a version of the Thunderbolt done for Nortel.  It has some
interesting features (like hot-upgradable firmware,  single 24 or 48V power
input, cheaper than a tbolt,  etc.  It also has a few warts...  no TSIP
command documentation being the main one and a few commands are definitely
different than the Tbolt.
The wakeup technique is rather crude and can take a couple of minutes (shout
a particular command into its ear until it wakes up).  Trimble's software
manages to get it talking immediately.  Duplicating the commands that
Trimble sends does not seem to work.  Once it wakes up, it stays awake until
you power cycle it or run Trimble's software.

I purchased one of these units from an Ebay seller in Old Cathay (around $70
or make offer plus $30 shipping) to see what it would take to add support to
Lady Heather.  My unit came in a week or so later.  I hacked a 48V power
connection (literally) onto the board and powered it up with a wall wart.
After some futzing and puzzling over the proper ribbon cable orientation
between the main board and front panel board,  I got the unit woken up using
Sam's technique and puzzled out the commands to make the oscillator
disciplining (time constant, damping, dac gain, etc) work.  The old survey
location was in a sketchy Guatemalan smuggler's haven border town at what
looks like a private residence.

After running it a while,  it became apparent that it works better than the
Thunderbolt.  The temperature sensor does not have those glitches that
plague the tbolt.  The receiver has a bit more sensitivity.  And, best of
all, the oscillator is pretty much immune to external temperature changes
(the Tbolt oscillator makes a good thermometer).   The reported OSC and PPS
rms errors are exceedingly low... you have to actively thermally stabilize
the Tbolt to approach these numbers.  Hopefully this quality extends to its
phase noise, etc spec.  It would be interesting to see what thermally
stabilizing the unit would do...
 		 	   		  





More information about the Time-nuts_lists.febo.com mailing list