View Single Post
  #8   Report Post  
Old January 1st 06, 04:46 PM posted to rec.radio.amateur.antenna
Wes Stewart
 
Posts: n/a
Default using an HP 8405A to measure SWR ?

On Sat, 31 Dec 2005 18:14:04 -0800, dansawyeror
wrote:

This is the next chapter in the antenna measuring saga. Today's adventure is
trying to measure SWR with an HP 8405A Vector Voltmeter.

The measuring setup is a standard signal generator feeding a tee. One branch
feeds the A input of the 8405A and the other a 20 db directional configured in
reverse, that is feeding the out port. The assumption is the reflected signal
will be read correctly on the in port. The in port is then terminated or
connected to an antenna.


It would also be nice it you had a 6 - 10 dB pad between the generator
and the directional coupler (DC); located right at the DC. You want
the source match to be set right there and the A probe to sample right
there.


An open termination reads 180 degrees and a 50 Ohm termination reads 0 degrees
and 55 db down from the input. (I assume 20 of that is the coupler so I am
subtracting 20 from that reading - is that a correct assumption.)


No and no. Although an open isn't as bad a reference as some folks
think (in coax anyway), a better reference is a short. Then you know
that the reflection is 100% -180 deg. (Where "" means "angle of")
An open has a bit of fringing capacitance and a tiny bit of radiation
so it strays (sorry) from 100% 0 degrees.

The other thing you need to do is normalize all future readings to 0
dB. Because, as Owen notes, you're throwing away phase information in
your quest for SWR only data, then start thinking in terms of return
loss. Return loss is measured (or calculated) against a 0 dB
reference. So from the beginning, when you get a reading with a short
(or open) that ratio becomes your 0 dB reference.


Using the method to read the SWR on an antenna produces values very close to
those from several SWR meters used as controls. When the output is terminated in
25 Ohms (2 x 50 Ohm terminators on a tee) the reflected signal reads about 24.5
db. (The B channel won't calibrate to within 1/2 a db so readings are estimates)
Subtracting the 20 db bias yields 4.5 db. Within the error of the instrument
this is pretty close to an SWR of 2:1.


Are you saying that it you step the input power to the "B" channel by
1 dB, it measures a 1/2 dB change, or that with the same signal
applied to A and B they only agree within 1/2 dB?

If it's the former, then stop here and fix the instrument. If it's
the latter, you don't care; you are measuring ratios.

Is there an error in this logic? If the instrument is working well enough to
perform this calculation then it should support more complicated measurements.


There are errors someplace. A 2:1 SWR is a return loss of 9.55 dB.