Michael Knudsen @mk@bsd.network

I changed the inverter logic to use a 74AHC1G14 which should be more than fast enough. I went for an SO-553 package which was kind of a pain to solder, but it went ok. I didn't finish soldering, but hopefully I can test tomorrow.

> Barring any [] or timing issues

Well, it didn't work, because it seemed that the BC847 was much too slow. It took 1.88 us before the base voltage (blue line) starts to climb noticeably , and it takes 380 ns to reach 5 V, and the CE pulses have a period of around 1 us.

The result is a nonfunctional, garbled display. I guess my board keeps driving the data bus for much too long which breaks both execution and DMA transfers to the line buffer. Maybe it also affects character ROM lookups, but I'm not sure.

Hvis man tager de dansker-røde briller af, skriver Variety, at filmens åbning har skuffet.

Det er simpelthen så useriøs en dækning af kulturstof at gøre det op som en vunden fodboldkamp. Det er vildledende enten bevidst eller af inkompetence, og jeg ved ikke, hvad der er værst, og det er simpelthen så dovent.

\o/

After fiddling with this some more, I now have a complete dump of the four ROM images on my replacement chip, and it is a complete match, so the contents match and the order should also be correct.

Barring any electrical or timing issues, this should bring my #VT100 back to life. However, it's way too late in the day to start yanking chips from the motherboard.

I'm also a bit concerned about damaging the ROM sockets if I used the wrong type of legs.. if anyone out there has any pointers about this, please speak up.

#electronics

Well, it didn't, for multiple reasons, one of which was me, and the other was some fucker putting the wrong value resistors in the 0402 10k. I'm not sure if my dodgy soldering nuked the transistor or if something else is at play, but the inverter wasn't really of the inverting persuasion.

I haven't checked everything yet but at least I'm getting something back which matches one of the ROMs at the expected offset. There are a few bit errors, but I suspect the "ROM" isn't seated correctly in the socket.

If this even remotely works it's a frigging miracle.

I've now programmed an EEPROM with the original #VT100 firmware.

The original firmware is scattered across four 2kB ROM chips. Noone appears to sell compatible EEPROMs in that size anymore, so I've merged the four 2kB images into a single 8kB image.

I ended up reconstructing the image myself, because the only source for a combined image seems to have been 404'd. Doing so is easy, but you need to know which order the chunks go in.

I had a source which confirmed the order I thought was correct, but to be really sure I also checked the bytes in the image against those in the disassembled VT100 firmware here: https://www.pcjs.org/machines/dec/vt100/rom/VT100.txt

With any luck I can assemble one of these boards tomorrow and then put the resulting "ROM" in my fw dumper to verify that it works.

The #vt100 replacement ROM PCBs showed up about a week ago, and they look pretty good. The scariest part was dry-fitting a DIL-24 socket to find out if the footprint for those pins was correct.

Yesterday I received the programmer adapter needed¹ to program the EEPROMs, so hopefully one of these days I'll have time to populate the boards.

¹ Technically, it isn't strictly necessary because I designed the board so I can drive the WE# and OE# pins directly which enables programming, but since I'd bought it I want to test and actually use the programmer.

#electronics