Wire Rope ROMs (Was: back to the AGC, was Re: TTL 7400's Available)

Brent Hilpert wrote: Speaking of Wang Calculator ROM's, I've recently built a fixture that I'm successfully able to read Wang 700-series ROMs with. It's unfortunately a non-automatic system...toggle switches and TIL-311 HEX displays (can automate it later), and am in the (slow) process of dumping a known good Wang 720C ROM. The ROM strobe signal timing is very tight on the 700-series ROM...off even a little bit on the timing (pulse width), and the readout gets real inconsistent. Also trying to write a microcode execution engine in Perl to run the code, but there are lots of interesting timing considerations that require deep digging into the schematics (which is something that I have little patience for...basically, I'm not very good at it) in terms of the timing of all of the register transfers in the machine. On the surface, it looks simple, just a basic 10-phase non-overlapping clock (shift register), but there's a lot of combinatorial logic that derives a lot of weird timing from the basic clock phases. So far, my attempts to execute the code I've extracted so far lead to execution of illegal instructions (all zero ROM locations), or infinte loops, and the code execution doesn't seem to make much sense. There are also some microcode instructions that aren't documented, and having to dig through schematics to figure out what they do. Wang is also famous for purposefully putting errors into published schematics to throw off competitors who would use such schematics to reverse-engineer how the machines work. So, some of my problem could be that the schematics are not necessarily an accurate representation of the actual logic. Still some tinkering to do. Still, I'm very sure that the fixture to read the ROMs is working perfectly. All ROM locations return consistent results, and other than the emulator running into all zero ROM content (there are quite a few "unused" locations in the ROM), all of the instructions decoded thus far are "valid" in terms of the allowed values of the various microcode fields. The Wang 700, 500, and 600-series machines all used this style of ROM, so, in theory, the reader (other than hardware pinout variations and subtle microcode field diffrences) should be able to be tweeked to read the 500 or 600-series ROMs without too much work. The main goal of this is to be able to capture the microcode images for as many of the Wang calculators that use this type of ROM as possible, as this technology is really unweildly to troubleshoot and repair (and I have some ROMs that are known bad that I want to try to be able to fix), and also because the genius of Harold Koplow (now deceased) is buried in that microcode. I want to preserve as much of it as I can (and maybe figure out how he made these machines do what they do) as a legacy to him. Rick Bensene The Old Calculator Web Museum http://oldcalculatormuseum.com