[GPS_Standard] preferred GPS - Trimble modules, GPS tips

[SP2IQW]

Hi Bert and Group,

I'm sorry for a bit long text but I would like to share my expierience 
gained when I have worked with the GPS receivers.
Working with AVL systems, I have designed a few boards with the latest 
Trimble GPS modules. A new types are very interesting as a low cost, low 
power 1pps engines.

Timing GPS receivers are more expensive so in my opinion for typical HAM 
radio applications up to milimeter waves most important is to get a good 
low noise (D)OCXO.  Please remember that short time stability as well 
low phase noise depends mainly on quality of OCXO. Added GPS support 
takes care about long time stability only.

First module I have used for GPSDO application was the Trimble's Lassen iQ.
Sized 26x26x6mm, features timing accuracy of ±50 nanoseconds when valid 
position fixes are being reported.  Very interesting is power 
consumption 100mW which makes it ideal type for portable application.

The next module from Trimble launched on the market is Copernicus II.
Thumbnail-sized 19x19x2.54 mm, less than 94mW power consumption with 
increased sensitivity is specified as ±60 ns RMS (in static mode, please 
note explicity stated this type of operation) and ±350 ns RMS (in indoor 
aplication, static mode -145dBm signal). But don't use indoor mode to 
discipline your standard !

The most promising is the latest Trimble's Condor receiver. It is 
available in different shapes for easy use as an upgrade in an old 
design PCB's (as for example replacing Lassen-iQ). Very promising is ±25 
nanoseconds PPS specification, close to dedicated timing models 
(although without T-RAIM functionality).

All these types as well as mentioned by Stig OZ3XO Resolution-T are 3.3V 
CMOS output devices. There are many ways to interface 3.3V CMOS logic to 
5V CMOS compatible input of a PIC microprocessor. In my implementation 
of the iQ receiver I have used two inverters - the first, the 74HCT 
series and the second, the 74HC series with 1k pullup resistor between them.
In another place I have used common base transistor (BC847B/2N2369A) 
shifter, with base biased at 1V by 2.2k/560 voltage devider decoupled by 
.1uF capacitor, emitter connected to output of the GPS module and 
colector to 5V level logic input with 1.5k - 2.7k pullup resistor. It is 
rather tricky so better is to use logic level shifters.
Limited rise time is not a critical case as eventual jitter is 
considerable lower than 1pps granularity. Important is to get proper 
voltage swing.

One cent about PPS accuracy. Nicely looking numbers as for example ±25ns 
or ±60ns don`t mean that pulse by pulse it will be within this 
specification. It describes statistic value which is just of our 
interest in GPSDO application.
The granularity of PPS pulses depends on the internal GPS clock. For a 
16MHz clock the garnularity is about 67ns so it is the worst case pulse 
position innacuracy or half of this number in some cases.


Regardless of which GPS type you choose there are several tips you have 
to keep in mind to get most from the GPS receiver:

- provide good visibility of satellites, it is worth to place an active 
antenna on the roof and go down with signal with low loss cable.

Low cost antennas are terminated with a RG174 cable which adds 1dB/m 
attenuation. Shortening this cable to 20-50cm and expanding lengh with 
10-20m of good quality eg. RG8-series cable will still provide a good 
signal (use SMA or N connectors).

- don't think that a window position will give you reliable GPS 
operation!  GPS constelation provides at least 6 satellites in your area 
but don't guarantee that 3 satellites will be "visible" from your window.

GPS receiver, configure as follow:

- use static navigation mode (Dynamics > stationary)

- limit use of low elevation sattelites by setting "elevation mask" at 
least 5-7deg

- use only sattelites with good signal by setting the Signal Level Mask 
(C/N) over 30dB-Hz  (Trimble use different units AMU - set it for 2.2)

- configure PPS output as "Fixed-based" in other words, the PPS pulses 
will be available on the output only when the GPS has a valid navigation 
solution.
   Without synchronizing to satellites 1pps is taken from the standard 
grade crystal RTC oscillator.

- as a default in most GPS receiver modules, the precision timing 
information is carried on a rising edge of PPS pulse, so no matter how 
long it is.

If you have doubts check and eventually change GPS configuration using 
software provided by manufacturer.

- the performance of a GPS module depends on the availability and 
accuracy of navigation data as sattelite ephemeris and almanac. To keep 
them valid during a power failure as well to keep real time clock (RTC), 
a backup battery should be used. Some data are transmitted every 12.5 
minutes but valid for a few hours or even a few days so the
reacquisition of position and later of precise timing is much more 
faster when battery backup is used.

If you really need better timings, it is worth to look for timing models 
as for example an old Motorola Oncore Timing moduls.
Less known are Trimble timing receivers mentioned below but with the 
exception of Thunderbolt model, don't expect you will get them as a 
second hand.

Resolution-T,  ±15ns (1σ)
Resolution-SMT (same size as Copernicus, probably reasonably priced), 
±15ns (1σ)
Mini-T GPS  - small child of Thunderbolt with 10MHz OCXO on board,  
±15ns (1σ)

I hope these GPS timing tips will be useful. They apply also to the 
timing modules but their specific consideration (T-RAIM functionality) 
are not covered here.


vy 73, Michal
sp2iqw



W dniu 2010-06-02 18:54, Bob Bownes pisze:
> So, do folks have a preferred GPS module to discipline clocks? Clearly
> some are better than others. I'd like one that will output in NMEA so
> I can use it to drive some other things as well, but other than that,
> it has become clear that the Rockwell MicroTrack TU00 just isn't going
> to cut it as it only locks to 4 satellites, has quite a bit of jitter,
> doesn't hold a very good 3D lock (+/-100m just isn't good enough for
> me...)
>
> Thanks,
> Bob
> ______________________________________________________________
> GPS_Standard mailing list
> Home: http://mailman.qth.net/mailman/listinfo/gps_standard
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:GPS_Standard at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
>