Editor’s Note: This “How it Was” story is told by Aubrey Kagan, who is a professional engineer with a BSEE from the Technion-Israel Institute of Technology and an MBA from the University of the Witwatersrand. Aubrey is engineering manager at Emphatec, a Toronto-based design house of industrial control interfaces and switch-mode power supplies. In addition to writing several articles for Circuit Cellar and having ideas published in EDN and Electronic Design, Aubrey wrote Excel by Example: A Microsoft Excel Cookbook for Electronics Engineers (Newnes, 2004).
Before Dot and Tape (see below) there was Rubylith. This consisted of a thin red film stuck to a mylar film. Generating a PCB layout was done in the negative, removing the red film where there was to be track. The technique required a completely different mindset to today’s “positive” approach. It was really good for earth planes, but I am not sure if double sided layout was possible and certainly a track between IC pins was impossible. I was only introduced to this at the start of my career in 1976 and never actually worked with it. I don't really recall how one created IC and transistor pads. I can only assume that they used some template placed on the Rubylith and exposed the area where the metal would be. The finished piece of Rubylith was the negative of the layout (and in 1:1 scale) and was placed over the photo resist and then exposed to light. There was no photographic step needed.
The author "Then" (left) and "Now" (right)Dot and Tape
Dot and Tape refers to the product used to lay the patterns of tracks and IC/transistor pads on to mylar sheets. There were several manufacturers although the only one that I remember was Bishop Graphics. The tape was self adhesive and black with a kind of crepe texture and was supplied on a roll of about 2" diameter when fresh and about 1" when finished. It was available in many widths. The dots were annular rings made from the same material as the tape and also came with different inner and outer diameters. They could be used for transistors and ICs, but were mostly just for vias. IC and transistor pads were also available pre-made. You could get them in 1:1, 2:1 and 4:1 scale if memory serves. Often the IC pads had the track between them already in place. Dot and Tape was applied with the aid of a fine, and very sharp, Xacto knife.
An alternate technique mentioned by David Ashton in his comment to Max’s How it was: CAD, CAE, and EDA
article was to do the layout on a single layer using two different colored tapes, which meant that alignment was less of an issue. The photographic process used filters to separate the solder from the component layer.
I was working with Dot and Tape while still a student circa 1973, so my experience in fact pre-dated my knowledge of Rubylith that was shown to me by some old-timers once I started work. When the 1973 Yom Kippur war started studies were suspended and I got a job at Elbit Computers (then associated with Control Data Corporation) verifying PCB layout to schematic by going over every trace one for one, a practice I still do to this day. The layouts were only done with black tape and required 3 or 4 layers. The 3 layers consisted of the solder and component sides, silk screen, and sometimes a fourth layer which was only the pads and vias to improve the alignment. Aligning the layers on top of each other needed targets to ensure good registration during photography. Layout registration was often aided by special punches that made coded locating holes and then the matching holes in the sheets were fitted on to metal templates.
All work was done on a light table to facilitate seeing through the layers. The first layer on the light table was an accurate grid, and sometimes the fourth lay of just the pads was actually placed on a grid sheet (light blue if I remember) and the photography filtered out the grid.
Of course the layers had to be thermally stable or expansion due to heat could cause problems on the layout. The drill information often was a copy of the pad layer (or solder layer) with letter coding beside particular holes and a table was provided with the hole sizes versus the letter coding.Computer-Aided Design (CAD)
I started laying out PCBs in Autocad. There was a package called Autoboard II from the Great Softwestern Company. The schematic symbols had to be generated with Autocad scripts and it got so that I could compose them in my head. Once the schematic was created you had to run an external program on the file to extract all the relevant information. As today you then had to place the components on the board and then run the autorouter. There was no “connectivity” and moving any component either in the schematic or the PCB required each segment of the connection to be edited.
I started working with an IBM PC XT running at 4.77MHz. This meant that a “pan” from one part of the screen to another could take up to an hour, so you made absolutely sure that you were complete before moving on to the next section. This was further compounded by the use of a monochrome monitor (because it had better resolution) although switching between layers was much faster than a screen regeneration. To improve performance I first installed an 8087 maths co-processor in the PC which made a big difference and then I replaced to 8088 with a NEC V20 which wasn’t anywhere near as significant. At the time NEC wasn’t available in South Africa for political reasons and I remember whilst on vacation driving around half of Sydney, Australia trying to find one. This must have been around 1986, because I remember I was in Australia at the time of Chernobyl.
The router seemed to be quite good, once you got it to run without errors and I would re-run the autorouter until I got about 95% success. Each routing attempt took about 10 hours and I would run it overnight. It was always very frustrating when there was a problem and the routing terminated, especially when it was a dumb error.
When the PC AT arrived it ran at 8MHz, a noticeable improvement. I bought a clone that was capable of running at 8MHz or 10MHz. The clone had some compatibility issues and so I installed IBM ROMs, but PC DOS used to check the frequency and if it was greater than the 8MHz, it would freeze. The frequency selection was done through a jumper. I installed a changeover switch across this jumper and would boot at 8MHz. Afterwards I would flip to the higher frequency, but with no electronic synchronization this would often lead to the PC crashing. But the25% increase in performance was vital. Adding the 80287 maths co-processor again improved the performance significantly. Autocad also allowed the simultaneous use of colour and monochrome monitors (with a suitable graphic card) and that also helped although it did not allow different views.
At the time there were no mouses (mice?) and input was done through a digitiser. The stylus of the Bit Pad One (one of the multitude of makes) had a pressure sensitive nib to indicate a point selection and to this day I still miss using it for graphics input. It is far more intuitive that the mouse.
Output was to a plotter (the Autocad output to a dot-matrix printer only came later, and anyway you could only use the printer for the schematic). The schematic was easy enough with fibre pens and colour helped with some debugging even as it does today. They could work at speed and posed no problem except when the fibre tip split and the lines became very ragged. However the PCB plots had to be done on suitable material and with an ink pen using Indian ink. The idea was to replace the dot and tape technique and so there were 3 or 4 plots that emulated the original approach. Keeping the flow of ink going was the bane of my life at the time. If the pen blocked in the middle of the plot, all the time was wasted. In order to use the partial output (reducing the possibility of the pen blocking even earlier) I developed a technique to produce the plot a second time by going over the existing output without a pen, up to the point that the pen was needed. I would then pause the plotter and insert the pen and continue. The repeatability of the plotter was quite good so there wasn’t a noticeable kink in the track. The pens were best stored wet in a special container filled with a special cleaning liquid.
I only had an A3 plotter and so I had to do larger boards at a 1.5:1 scale instead of the usual 2:1 or 4:1 and on one occasion I even had to break the plot into two parts and then tape the pieces together!
Funny how much stuff I still have around. I still have the paper, the pens and the cleaner, the digitiser and the plotter – any museum want it?
From the above it may seem that I have specialised in layout, but that is not the case. In the early years of my career from 1978 till 1989 I ran my own company (in South Africa) and to economise I produced my own artwork. Since I was never very neat, I got into CAD to improve the quality of my work. As the industry progressed into Gerber files, pad stacks, flashes, multi layer and many other buzz words my career took a detour into systems engineering in 1990 and when I returned in 1993 I decided to leave the board layout to experts. I have worked through schematic and library entry in PCAD and am now starting with Altium and I don’t really miss the old days.Click Here
to see other articles in this "How it was..."
series...Editor's Note: It would be great if you took the time to write down short stories of your own. I can help in the copy editing department, so you don’t need to worry about being “word perfect”. All you have to do is to email your offering to me at max@CliveMaxfield.com with
“How it was” in the subject line.I can post your article as “anonymous” if you wish. On the other hand, what would be really cool would be if you wanted to add a few words about yourself – and maybe even provide a couple of
“Then and Now” pictures showing yourself as a young engineer
("Then") and as the hero you've grown into
If you found this article to be of interest, visit EDA Designline
where – in addition to blogs on all sorts of "stuff" – you will find the latest and greatest design, technology, product, and news articles with regard to all aspects of Electronic Design Automation (EDA).
Also, you can obtain a highlights update delivered directly to your inbox by signing up for the EDA Designline weekly newsletter – just Click Here
to request this newsletter using the Manage Newsletters tab (if you aren't already a member you'll be asked to register, but it's free and painless so don't let that stop you [grin]).