Re: Shootout of Substitute SG504 heads = a better head.
David and Ancel,toggle quoted messageShow quoted text
My interests in the SG504 are slightly different. After my first attempt at
measuring the flatness of David's head it was pointed out nicely but firmly
that my methodology had some serious flaws. RF is not my area of expertise.
It would be accurate to say I don't have any area of expertise at all since
I have avoided becoming an expert in anything in favor of becoming adept at
as many things as possible (a jack of all trades).
Several years passed after my first attempt at measuring the flatness of
David's head when George Steen gave me one of his original SG504 heads. At
that point it became possible for me to do a better (still not perfect in
some people's eyes) test of David's head by comparing it to George's head.
By comparing the two I was able to eliminate many of the flaws that plagued
my first attempt. I learned a great deal from my earlier mistakes and was
able to make more accurate measurements the second time around.
Some might wonder why it would be worth concerning ourselves with
differences that are less than 0.1dB. The answer is simple - because we can.
Between the first and second time I did this test I became a lot more
rigorous, or less sloppy if you prefer, in how I made my measurements. The
criticism of my first attempt made me a better experimenter.
Having three different versions of the head, designed over a 40+ year
timeframe, is an opportunity to see what progress we have made in RF design.
Ancel produced what appears to be a great design in about a month. I would
guess David took longer when he did his about 10 years ago. I am sure George
took much longer 40+ years ago.
I would not be able to design anything like this on my own. But I'm learning
a great deal from testing what they have done.
Dennis Tillman W7PF
Sent: Sunday, March 20, 2016 5:22 PM
Subject: RE: [TekScopes] Shootout of Substitute SG504 heads = a better head.
The build I have for your head uses a gold plated SMA launcher. Mine uses a
modified Nickel/brass BNC launcher. As you can note from the pic the SMA
launcher requires a SMA coupler to connect to the minicircuits SMA
attenuator whereas mine requires an BNC-SMA adapter to achieve the same.
Given that the SMA coupler I used is from Kirkby Microwave as part of the
cal kit for the VNA, I doubt it could be at fault as it has been quality
checked before shipping.
That just leaves the SMA launcher itself as compared to the BNC launcher,
the SMA launcher is not likely to be worse than the BNC launcher which is on
my head. SMA carries a much higher frequency spec. than BNC. The 0.2dB of
loss at 800Mhz looks like some kind of resonance dip as it improves
afterward. That implies some parasitics at work.
There is also the issue of the 10dB cross talk reduction gained by using
different schottky detectors in the new head. It is possible, that at that
frequency, the cross talk between the monoblock combined ref. detector and
the actual detector disturbs the leveling. But that's just a guess.
All in all though, I am satisfied that the design I have is more than
adequate and that both designs do the job.
I think rather than concern ourselves about 0.1dB improvement here or there,
the improvements in the DIY nature of the latest design matters more.
Given the availability of the components used in the new design as well as
the simple build construction any decent home workshop can economically turn
one of these new heads out and I figure that adds longevity to the Tek SG504
instrument lifetime and that's what the forum is all about.
Posted by: mosaicmerc@...