On 6 Nov 2006 06:26:53 -0800, "Denny" wrote:

A signal generator for driving the coax... Put a 3dB pad on the output
of the generator to stabilize the impedence it sees...

Hi Denny, and student,

This pad should exhibit a 50 Ohm load to the source, and a 50 Ohm load
to the line. This is usually a T-Pad.

At the load end
of the coax is a resistive load for terminating the coax in its
characteristic impedence...

This is good advice, which, unfortunately is thwarted by the advice
for:
An oscilloscope pick up across the load for
seeing amplitude and phase...
Wrong. At frequencies above 50MHz (or 200MHz for the best of
equipment), you should always use a 50 Ohm load to the input of the
O'scope - NOT a probe, or link.

I normally drive the coax at a frequency
that makes that coax length an odd 1/4 wave at the test frequency...

Odd Eighth wave is much better as transformations will be far easier
to measure - and easier is indicative of available accuracy. Quarter
and Half waves do offer convenience when you terminate in opens and
shorts, but with other loads (other than the characteristic Z of the
line), Eighth wave is preferable.

As for Half or Quarter Wave choices, short the far end. Open
terminations are problematic. Measuring a open/short termination
transformed as a short is preferred. Hence, the best configuration is
a shorted Half Wave line.

Then I trim the coax for the voltage minimum...
As far as phase batching the the individual coax pieces after they are
cut to frequency I would pick one coax piece as the master..
I would put a splitter on the output of the generator and drive this
master coax from one side of the splitter and also drive a test coax
off the other side of the splitter... At the load ends of the of the
coax pair I would use a dual channel oscilloscope. or vector meter,
etc., to compare phase between the load on the end of the master coax
and test coax, repectively......

This is taxing the resources of equipment at 1GHz. The method is
informative, but it does not reveal the accumulation of error (and our
student hasn't actually expressed how accurate these matches need to
be).

Now, having said that let me comment that unless you are using hard
coax, I suspect at GPS frequencies just routing two flexible coax
cables around the innards of a piece of electronic gear will introduce
enough internal mechanical distortion to cause phase variations at the
load ends... I may be wrong but that is my suspicion..

Think of quarter wave radius sweeps.

73's
Richard Clark, KB7QHC