[time-nuts] GPS splitter measurements

Azelio Boriani azelio.boriani at screen.it
Thu Oct 11 18:16:57 EDT 2012

Please take care that the S14WI is "smart": if it doesn't sense a DC
current on the antenna port then it reports that to the user ports and I
think it can even power off the internal amplifier, this can account for
the observed loss instead of the gain.

On Thu, Oct 11, 2012 at 11:20 PM, Magnus Danielson <
magnus at rubidium.dyndns.org> wrote:

> Dear fellow time-nuts,
> I have fooled around with the network analyzer on a set of GPS splitters
> in order to illustrate the differences.
> The goal has been to measure:
> 1) Gain (input to port 1 and input to port 2)
> 2) Group delay (input to port 1)
> 3) Port isolation (port 1 to port 2)
> I have measured these devices:
> A) GPS networking passive splitter
> This is a passive broadband splitter.
> B) GPS source S14WI
> This is a ruggidized "smart" 4-way splitter.
> C) Symmetricom 58536A
> A "smart" 4-way splitter.
> I used a 5 V supply on a spare port for the B and C splitters.
> Please see the attached files.
> For each I used a 1.1-1.7 GHz wide-band sweep, and I did a 1.57542 GHz
> 102,3 MHz wide sweep. The later sweep range was chosen as 10,23 MHz is the
> basic unit for GPS modulation, so you can see how many multiples the skirts
> cover. Narrow-band C/A only is only 2.046 MHz wide, where as Narrow-band
> C/A+P(Y) is usually around 20.46 MHz wide. Further, for the narrow sweeps I
> focused on the group delay read-outs. This is not an ultra-clean, all
> calibrated, full compensated measurement.
> The GPS Networking splitter has a fairly flat gain with about 4.5 dB loss.
> The isolation is about 25 dB in L1 and 20 dB in L2. Group delay is very
> flat at about 1.1 ns. This is expected. I didn't power this one, as I knew
> it was passive.
> The GPS Source S14WI splitter I expected to have gain, but found the loss
> to be about 20 dB in L1 and about 26 dB in L2. The isolation is about 48 dB
> in L1 and about 58 dB in L2. Group delay is a little rocky, but is around 3
> ns.
> The Symmetricom 58536A splitter I expected to have unity gain, but found
> loss to be about 14 dB in L1, and L2 is suppressed to 54 dB loss, as
> expected. It is clear that this is a filtered product, and this is also
> stated in the datasheet. The isolation for L1 is around 28 dB, which is
> high from the expected 50 dB. The group delay clearly shows to peaks at
> 26,129 ns and 21,889 ns, which translates to Q values of about 124 and 104
> respectively (withdrawing 1 ns for "electrical delays").
> This is the results of a quick-and-dirty measurements, leaving many
> details done sloppy, like unused ports left open instead of terminated, not
> following up on the gain structure misses, not doing full calibration etc.
> It's to show what some modern available splitters perform like. I really
> don't have the time to play these games this week, but I wanted to get
> started at least.
> I hope you enjoy the results never the less.
> Getting 50 dB isolation is possible. It seems open ports remains an
> isolation issue thought.
> The Symmetricom 58536A provides filtering, leaving a quite noticeable
> group delay. It does have the benefit of providing improved suppression of
> side-band signals.
> So, you milage will vary. I hope this little lab-report encourage you to
> look more at your splitters, measure them and get your hands dirty.
> Cheers,
> Magnus
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.

More information about the time-nuts mailing list