Click here to go to our main microwave history page
Click here to go to our history of microwave software page
Click here to go to our main computer-aided design page
Click here to learn about rubylithography
Barbara Girard, the artist who drew this sheep, did a lot of rubylith cutting before BAA was developed.
New for February 2021: We will build this page in installments, first the history of the topic, second a full scan of the BAA users' manual (download at bottom of page!) and third, some example artwork generated in BAA, if we can dig some up (gotta look through some dusty old boxes in the Microwaves101 Warehouse). And if anyone has any knowledge of similar design tools from the late 1970s/early 1980s, please tell us about them and we will start by adding them to our history of microwave software page. And if you are bored by the history of microwave engineering, get back to catching Pokemon, you won't get any respect until you hit Level 40...
Bedford Automated Artwork (BAA) was a home-grown computer artwork tool for microwaves, created at Raytheon's Missile Systems Division headquartered in Bedford, Massachusetts. In many ways it was ahead of its time, you could simply command it to drop in a rat-race coupler, and specify the frequency band, and BAA created that piece. The "BAA" manual was a such a hoot, some of us saved it from the dustbin of history by storing it for 40 years. "Baa" of course is the sound a sheep makes, so that manual has some fun with sheep. If you live in Germany, where a "schaf" says "mäh", maybe this doesn't make a lot of sense...
BAA was principally the product of one individual, Tom Dowling, but he had help from Nunzio Cavallaro, Joe Hunt, and others. "BAA" is an acronym for "Bedford Automated Artwork". An internal memo stated that the cost savings in the first year BAA was used was $182K, compared to rubylithography. That's almost a half-million 2021 dollars!
In researching this topic, these comments came from Nunzio:
Tom did most of the coding. He started it on the TRS-80 using Basic and then transitioned to FORTRAN on the mainframe. Myself, Larry Kent and Sarah Wong expanded the code capability. The manual was mostly written by Joe Hunt, I also had some input. The sheep drawings in the manual were drawn by Barbara Girard.
"BAA" was published at the IEEE International Microwave Symposium in 1982, in Dallas, under the title "A Novel Approach to Computer Automated Microwave Circuit Mask Design." BAA ran remotely from a Control Data Corporation (CDC) cyber computer. The user interface was a Tektronics graphics terminal, Tektronix 4012, the 'ol green screen. The final output would be a large magnetic tape that you would pick up at the computer room, and hand carry to the "Gerber Room", where a technician (often Ralph Murrow) would use a Gerber Photo plotter to generate the artwork. Typically it would be plotted at 10X and the artwork would be reduced to improve resolution. Remember, always tell the Gerber Room to plot your artwork with the emulsion side down! Next stop would be to the etch lab where Roger Gilbody or Vinny would etch your parts, which was a long walk up the hill to the "Components" building in department 7175. If it was raining or you were lazy, there was a little blue school bus that made the rounds of the campus about every fifteen minutes. It had been driven in a CCW circle so long that the springs on the ditch side were collapsing and the bus seemed to have a case of scoliosis. Then the panels would return to Phase 2 microwave electronics lab in Department 7645, and Charlie Band, Tom Maxon and others would dice the parts and build the circuits. For alumina circuits, a Disco saw was used, for Duroid (the standard for soft boards at the time) an X-ACTO knife would suffice.
You can explore the ghost of Raytheon's past by looking at this link to 180 Hartwell Rd. in Bedford, MA. The "Systems" building was built in phases, which retained the names "Phase 1", "Phase 2", etc. Each successive building used cheaper and cheaper construction techniques, and at some point trailers were used to house overflow staff. In the link, Phase 4 is in the lower left, and the Components building in the upper right. The Flight Test Facility is on the lower right, if you drag the screen you can see the Hanscom AFB runway below it. Very convenient for strapping dummy missiles onto Korean-war-era aircraft that had about a million pre-drilled holes from past experiments!
Link to Google map of Raytheon Bedford
180 Hartwell Rd. is a toxic waste site, which is the reason why most of the buildings have not been torn down. The property was transferred to the Navy in year 2000, and is now called "Naval Weapons Industrial Reserve Plant Bedford" while they are trying to restore it. Read about the project here (takes a while to download). Most of the pollution came from the Components building (etch lab, remember?) and the Flight Test Facility.
The main parking lot for the Systems building was across Elm Creek, via a small walk bridge. It appears that that lot had been reclaimed by the town for a housing development. If anyone ever tried to market water "bottled from the sparkling springs of Bedford Massachusetts", don't drink it. Maybe you could use it as a solvent to remove cosmoline.
The cited BAA paper included Harold Stinehelfer as an author, who was pursuing his own fame and fortune as a time-domain guru. The paper includes an explanation of how you can convolve frequency domain data into time domain, gate it, and come up with a "de-embedded" measurement. He went on to found his own company, "Made It Associates" to sell time domain products. It seems weird that the two topics were combined into one paper, perhaps it was because a frugal defense company would not want to have to send two people to a symposium when they could send just one.
Below is the free abstract from the 1982 paper provided on IEEE Xplore:
A time sharing program has recently been developed which simplifies the mask design process. The FORTRAN program called BAA (BEDFORD AUTOMATED ARTWORK) runs on a CDC cyber computer and provides a user friendly interface between the design engineer and the Gerber automated drafting machine and photo-plotter which generate the actual circuit mask. Users access BAA via a graphics terminal such as the Tektronix 4012. The BAA commands define a powerful language which allow the user to describe his desired microstrip or stripline circuit in consistent mechanical or electrical terminology. The user selects building blocks from a library of commonly used microwave components. Typical library elements which have been implemented include branch line couplers, ring hybrids with equal or unequal ‘power division, overlay couplers, interdigitated DC blocks, distributed RF chokes, mitred bends, and other components. In order to specify a circuit, each BAA element is described by giving the generic name (component type), parameters, specific name, and position of the component. The position is specified by three numbers representing the X and Y coordinates\ and the angular...
First names of the authors are listed below for posterity. All have since retired or passed away.
Tom Dowling
Jim Birch
Steve Temple
Steve Monaghan
Harold Stinehelfer
Nunzio Cavallaro
Alan Davis
The BAA Manual
Here it is! You can now download the full BAA manual, revision 2.0, from 1983. We were pleased to see that this hard copy actually includes the SEEIT manual as well, a welcome bonus to any microwave archiology project.
Bonus poster!
All of these people were involved in microwave design in the 1980s. The first person that is NOT in the picture to name half of them is eligible for a $50 reward!
Reference
T. Dowling, J. Birch, S. Temple, S. Monaghan H.E. Stinehelfer N. Cavallaro and A. Davis, "A Novel Approach to Computer Automated Microwave Circuit Mask Design," 1982 IEEE MTT-S International Microwave Symposium Digest, Dallas, TX, USA, 1982, pp. 465-467.