On Thu, 16 Mar 2017 01:58:16 -0400, "Michael A. Terrell"
wrote:
Jeff Liebermann wrote:
Nope. Most of my test equipment is 50 ohms. However, I also work
with 75 ohm devices and antennas. Transformers do not have a very
flat frequency response. If I need a flat frequency response from DC
to many GHz, I use a minimum loss pad:
https://www.maximintegrated.com/en/app-notes/index.mvp/id/972
(...)
I have used some of the six hole ferrite beads to make transformers.
six turns, tapped at the fifth turn. The 1.2 turns ratio, squared is a
1.44:1 impedance ratio. These are the same beads used for CATV line taps
and broadband splitters.
They convert 75 ohms to 52 ohms. I put some into some old Blonder
Tongue inline attenuator housings with an F connect on the 75 ohm end
and a BNC connector on the 52 ohm end. (52.0833 Ohms) They were handy to
use a TV FSM as a 50 ohm RF millivoltmeter. Now, I have some precision
attenuators, and a Boonton 9200 RF Millivoltmeter ans different
feedthrough terminators.. 
I initially did the same thing using various iron and ferrite toroid
cores. They worked, but not if I wanted a really flat frequency
response (+/-0.5dB) over many octaves of frequency range. I had a few
adapters that I optimized for frequencies of interest, but getting it
fairly flat from 1.5 to 30 MHz (marine bands at about 4.5 octaves) was
rather difficult. Instead of fighting the problem, I switched to a
minimum loss pad. The low frequency pads ended up inside Pomona
aluminum boxes, while the microwave stuff was on microstrip stuffed
into a box made from brass or unetched PCB material.
Actually, it wasn't quite a minimum loss pad. I worked out the
numbers for a 50 to 75 ohm -6.0dB pad instead of -5.71dB, which made
power and voltage scaling, measurements, and calculations easier.
One problem with minimum loss pads. They develop feet and walk away.
I wasted about a day of company time building and characterizing a
collection of about 20 attenuators for my own use. Within about 8
weeks, they were all gone, probably "borrowed" by my co-workers.
Hint: There's no such thing as a precision attenuator in a lab that
works with transmitters. Eventually, they all get cooked. The best I
could do was characterize what was left of the attenuator after
someone accidentally transmit into it, and compensate for the changes
in loss and impedance.
--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060
http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558