[TowerTalk] Measuring sma relays at UHF

[jimlux]

On 8/13/18 5:58 PM, Norm n3ykf wrote:
> Hi all!
> 
> Playing with my toys.
> 
> Using an Anritsu S113b which does not have direct voltmeter readout.
> These numbers were calculated by making a return loss measurement and
> converting it to relative phase with their Line sweep tools software.
> 
> I went through this whole process (took about 5 times longer) prior to
> my trip to 5J0 and realized that mistakes had been made. Took a pair
> of crossed linear 10 element diddled by a Tohtsu spdt 12v instead
> Worked like a charm from AOS to 5-7 degrees above the horizon so that
> I could switch polarities. What a blast. Used my homebrew rotator.
> 
> https://www.youtube.com/watch?v=1sDP6lAiWGw
> 
> Please ignore my 2yo little monsters. They can and do make s9+ noise.
> 
> Trying to get the coax for a cp antenna correct this time. Waste no more coax..
> 
> I have two Transco SMA 24V relays and an SMA jumper, consisting of two
> elbows and a straight piece.
> 
> First order of business was to calibrate all the test gear. Note how
> long the adapters attached to the N connector were as well as noting
> the length of the test cable.
> 
> Got some interesting results:
> 
> RYA NC: 232.2 degrees
> 
> RYA NO 231.8 """"""
> 
> RYB NC 231.3""""""""""
> 
> RYB NO 231.0 """"""""""
> 
> 
> SMA jumper to connect Common->NC to NC->common 281.3 degrees.
> 
> Does this look correct? How does such a small package contain such a
> lengthy RF path?


wavelength is about  70cm, so 1 cm is 360/70 about 5 degrees in free 
space. More like 7-8 degrees in coax.

what was the reference plane for the calibration of the analyzer?  The 
spread among paths is <1 degree which is plausible for a small relay. 
I'd expect the path length to be a few cm.  You might compare the 
measurement you get with a barrel connecting the two test cables and the 
relay connecting the two test cables (e.g. COM-NC).

a couple feet of coax would add about 200-250 degrees of phase.

> 
> Norm n3ykf
> _______________________________________________
> 
> 
> 
> _______________________________________________
> TowerTalk mailing list
> TowerTalk at contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
>