[time-nuts] indoor GPS test.

Hal Murray hmurray at megapathdsl.net
Wed Jan 23 13:17:33 EST 2008

> I have had no success getting anyone to loan me their GPS units. Does
> anyone have suggestions as to what inexpensive unit to purchase, which
> will give a good indication of whether a GPSDO unit will receive and
> process the signals? 

Which end of the "nut" scale are you on?  If you have to do the test, you 
might as well get the gear you will eventually use.

There are two types of GPS receivers.  One primarily tells you where you are 
and how fast you are moving.  It also tells you the time.  The other is 
primarily for getting the time.  You tell it where it is (or ask it to figure 
it out, aka survey) and it uses that to do a better job of telling the time.

There is a huge market for commercial GPS position receivers.  They usually 
come with mapping software which jacks up the price.  Most of them don't have 
PPS outputs.  Many of them use USB.

GPS chip sets have been getting better over the years.

The Garmin GPS-18-LVC is popular with people running ntpd.  You can get them 
for under $100.  It takes some soldering.  They need 5V.  You can get that 
from USB.  Note the LVC.   The other GPS 18 models don't have the PPS signal. 
 For ntpd, the PPS signal is usually wired up to a modem control signal and 
the kernel records the time when an interrupt happens.  (There are a couple 
of good web pages on how to wire them up.  I'll fish them out if you have any 
troubles finding them.)

I have one.  It just barely mostly works inside my house if I'm careful about 
where I place it.  (which makes it good at tickling error cases)  I have 
trees and neighbors but no serious RF shielding so my antenna position is 
probably "medium" on the wide scale of possibilities - better than many 
apartments but worse than most outside installations.

The GPS 18 has been on the market for several years so its chip-set is not as 
sensitive as some newer ones.

I just wired up a MR 350P from Global Sat.
It uses the SiRF III chipset which is much more sensitive than the GPS 18 
LVC.  Unfortunately, the PPS output is only 1 microsecond wide which probably 
isn't enough to get into a PC.  (But I haven't tried that yet.)  This also 
requires some soldering.  I cut off the connector.  (Their RS-232 adapter 
cable drops the PPS signal.)

So, depending on what you want to do, I'd suggest either the GPS-18-LVC or 
the MR 350P.

These are my opinions, not necessarily my employer's.  I hate spam.

More information about the time-nuts mailing list