Instructions for the MATCH.EXE resonant Z-matching networks.

The program MATCH.EXE for PC-type computers is used in a Windows GUI
environment to design the 14 NBMATCH networks that ARRL has been
publishing for many years in such publications as the Handbook and the
UHF/Microwave Experimenter's Handbook. These very useful networks are
described in the 2005 - 2009 ARRL Handbooks in Chapter 14 and are included
in the Broadband Transformers PDF file on the 2010 Handbook CD. 

The new program corrects some errors that were found in the original 
BASIC programs, NBMATCH.BAS and NBMATCH.EXE. It is also much more 
convenient to operate. All 14 networks have been verified, using the 
new MATCH.EXE.

MATCH.EXE is a Visual Basic V5.0 program. It is about 26K in size, and
requires the file Msvbvm50.dll (1324K) to reside in the Windows\System
folder. This is a standard file (public domain) for Windows 98 and is
also on the '98 System compact disk. Both files should be downloaded
from the ARRL website for use on '95 Systems.

The usage is pretty much self-explanatory. Select a network, input the
values of RS, RL and F0. The value of Q (selectivity) is requested,
which must be greater than the minimum value that the text window
indicates. Networks 13 and 14 have a fixed value of Q which cannot be
modified. The L and C values are calculated, after which two courses of
action can be taken: 1) Modify RS, RL, F0, and Q or 2) Select a
different network. In all cases the main text window gives instructions
for the next action.

The ouput of MATCH.EXE is the three L and C values for the network,
based on user inputs. These values can then be transferred manually to
such programs as ARRL Radio Designer or SPICE or other analysis programs
to get frequency and phase responses and also attenuation due to L and C
component losses.

An improvement to the program involves the Q value. One measure of Q is
the selectivity, which is defined as approximately the ratio of the center
frequency to the 3 dB bandwidth. The Q in MATCH.EXE is a parameter that 
is used for network design purposes and does not in every case represent
the actual selectivity. The improved version asks the user to input the
actual Q (selectivity) that is desired.  The program output gives a value
of Q that is the actual value. If the *actual* value is different than 
the *desired* value, modify the input value in the proper direction.
The L and C values are then correct for the desired selectivity.
Often the difference is not critical, but if the true selectivity
is important, this procedure is helpful.

The displayed Q value pertains to a network that is resistively loaded 
*only* at one end. This is a standard procedure for networks of this kind.
If the network is counjugately loaded at both ends the Q is one-half the 
singly-loaded value. If this other end is loaded, but not conjugately loaded,
it is best to simulate and revise the network iteratively, using MATCH.EXE, 
to get the required selectivity. Also, the amount of loading of the network
has a small effect on the resonant frequency that should be checked by 
simulation.

A "singly-loaded" network means the following:

A) For networks 1,2,3,4,9,10,11,12,13,14, RL is very large
B) For networks 5,6,7,8, RS is very small.

Questions regarding the program can be e-mailed directly to me:

William E. Sabin W0IYH
1400 Harold Dr., S.E.
Cedar Rapids IA 52403
w.sabin@mchsi.com