Notices by Tube❄️Time (tubetime@mastodon.social)

here's a handy guide to electronic components.

My coffee table book, Open Circuits, with beautiful cross-section photos of electronics, is 25% off through the 12th. Use the super-secret discount code "SHOWOFF". What, you already have the book? BAM--you just thought of a friend who would enjoy a copy.

nostarch.com/open-circuits

i've been tinkering with this board which emulates an IBM 729 tape unit. (some examples of which can be seen behind me in the other photo)

it's been a slow bringup process since i want to be very careful about connecting this to a real 1401. this weekend i've been checking out each of the level shifters on my board (which take 3.3V CMOS logic levels and convert them to the IBM logic levels). the most accurate way is to check it with the real SMS cards that the 1401 uses to drive and receive tape unit signals.

in my new design, i've incorporated everything inside that 19" rack, including power supplies, into a single circuit board. even the termination resistors, seen here.

the original emulator uses two ribbon cables that go from the emulator to one of these waffle connectors.

the original design was built around 2005-2008 or so when the team was trying to fix the 1401 and the other peripherals, including the 729s. it is a bunch of hand-wired boards in a 2U 19" rack mount enclosure.

appears the transformer card has a problem. can you see it?

the read bus circuit is a bit more complex. the big orange block on the right-hand SMS card is a transformer.

oh yeah IBM logic levels are weird. the tape units use "P" and "N" logic levels. the 1401 itself uses these levels to talk to the tape units but internally it mostly uses "T" and "U" levels. there are even more logic levels shown in the official quick reference here.

the 729 is a little unusual in that, during a read operation, it just amplifies the raw flux transitions coming off the tape. the signals go into a shared analog read bus that enters the 1401, and circuits in the computer turn the flux transitions back into binary data. naturally, i have to emulate those analog signals, which i do using some analog pulse shaping circuits tied to microcontroller GPIOs.

got the full complement of cards to implement the read channel for 1 bit. it seems to be working.

i fixed a computer this week. this is the 1960s IBM 1401 at the Computer History Museum in California.

hmmmmmmm. FM radio card maybe

Sencore! (I'll add alt text to these images later)

it's the last Silicon Valley Electronics Flea Market here at West Valley College! 🧵

mystery device.

this'll really ring your bell

oh wow an Armatron

some newfangled computer with a colorful case