Post #4384472

2026-08-04 22:08 UTC

It really wasn't doing a whole lot, it bleeped, PAGE cleared the screen, and not much else So I consulted the service manual, I initially looked at the keyboard, in theory in LOCAL mode or with TX/RX connected we should hear a bell similar to the one we hear at startup when CTRL + G is pressed, but alas nothing, checking the serial line on a scope, and sure enough, nothing! My first point of call was the keyboard, whilst PAGE was doing something, this existed outside of the keyboard matrix and was a switch which connected straight to the minibus backplane which contained all the control electronics Oh yeah - this machine is build before the advent of microprocessors, it's combinational logic boards on a passive backplane all the way down! A similar construction to contemporary minicomputers! After spending hours looking at the keyboard to understand it, I was rather blown away, this keyboard uses a 7 bit counter (using a 4 bit counter IC with some flip flops to extend it), feeding the lower 4 bits into a 4-16 line decoder which pulsed each row, and the upper 3 bits into an 8-to-1 multiplexer which scanned each column, when a key is pressed and the decoder pulses the row whilst the multiplexer reads the row, the output of the multiplexer goes low, stopping the clock going into counter, and the 7 bit number on that counter - it's the ASCII value of the key you pressed! #tektronix #tektronix4010 #vectorgraphics #terminal #vintagecomputing

Replies (1)

@crashtestdev@woof.tech 2026-08-04 22:36

As cool as that was, we also learned the keyboard is totally fine, so the issue is elsewhere I then looked more into the minibus backplane, and what a better way to get used to something than making it into a KiCAD symbol and schematic to hopefully make a single extender! https://forgejo.crashtest.dev/jasonalexander-ja/tek-minibus-kicad (I didn't know about the 3.96mm pin pitch standard and I spent hours with board and a vernier and managed to get to 3.97mm... which I am proud of) This did allow me to better understand the architecture of this though, and specifically I thought I should look at the 4.9mHz and 614kHz clocks signals produced by the TC-1 logic board, 4.9mHz was OK (after I stopped checking with a 10mHz Tek 222 and used a more powerful scope), but 614kHz... completely still Turned out to just be a dead U9 7437 quad NAND on TC-1 which droves the line by pulling it down, sadly all I had to hand was a 7400, which is the same chip but rated for less current required for driving a line, so I soldered in a socket and placed it in, and what ya know, 614kHz came riiiight up, probably wont last, but with a replacement on the way and the chip socketed it certainly works for now! As for what effect this had, we now had serial communication in line mode!

Open ##4385201