[Amps] Windows executables for analysis of transmission lines.

Dr. David Kirkby drkirkby at ntlworld.com
Sat Sep 13 19:14:01 EDT 2003

Mark Chun KH6HPQ has kindly produced some 32-bit Windows executables for
the latest release of my software 'atlc' for numerically computing the
properties of arbitrary shaped transmission lines (including couplers).
The software's homepage is at:


Note the software is designed for cases where there are no analytical
equations available for solving your transmission line's geometry - or
they are just too complex to work with. It is pointless using this
software to find the impedance of coax, when you can do it in 1% of the
time with a pocket calculator. 

The software uses numerical simulation, so can be a bit slow, although
with modern machines it is not such an issue. 

Mark's binaries use the latest version 4.4.4 source code and are newer
than those previously made available some time back by Randall DuCharme
AD5GB based on version 3.0.7 of my source.  The Windoze binaries are
made from exactly the same source code as that used to build on UNIX or
Linux systems. 

Get the binaries by going to http://atlc.sourceforge.net/ then click
"Downloads", then click, "View all project files", then you will see
atlc-4.4.4_Win32.zip under the heading "atlc-4.4.4". Read the
documentation carefully. This is complex sofware. 

I think Mark tested the software under Windoze 2000, but they should I
guess work on other 32-bit Windoze systems. I've not myself checked the
software, but Mark assures me it passes all the 90 self tests accept
one. We know the reason for this is one failure is due to the differing
ways Windoze and UNIX handle text files, so is not a problem.

The UNIX source will compile on just about anything. I've tested it on
Linux PCs as well as workstations from SGI, HP, Dec, Sun, an IBM RS/6000
and a Cray supercomputer. Someone built an earlier release on a Sony
Playstation 2 !! They must be mad. 

As you will see from my signature, I'm no fan of Windoze, but since
enough people have asked me about Windoze versions, I have made Mark's
binaries available to all. 

Accuracy has been carefully tested:


on a number of 2-conductor transmission lines with a single dielectric
with errors or typically around 0.3 %. For couplers, only one geometry
has been tested, where the errors are typically around 0.6%. No error
has ever reached 1%. 

For multiple dielectrics, testing has been very limited due to my lack
of any exact analytical means. If anyone can advise me of any other
exact methods to compare the software against, I'd be interested. 

Note these binaries are command line driven, just like the UNIX version.
There is no graphical user interface. If any Java expert fancies putting
a platform-independant GUI on the front of this it would be useful. Any
offers of help ????

There is a old Windoze version with a GUI at


I would strongly advise *not* using it, but if you really can't work at
the command line and your geometry of interest is covered, that might
just be of some use. 

The day Microsoft makes something that doesn't suck is probably the day
they start making vacuum cleaners." -Ernst Jan Plugge.

Dr. David Kirkby,
Senior Research Fellow,
Department of Medical Physics,
University College London,
11-20 Capper St, London, WC1E 6JA.
Website: http://www.medphys.ucl.ac.uk/~davek
Author of 'atlc' http://atlc.sourceforge.net/

More information about the Amps mailing list