Jump to content


  • Content Count

  • Joined

  • Last visited

1 Follower

Profile Information

  • Gender
  • Location
    Minneapolis, MN, USA

Profile Fields

    Software development/engineering

Recent Profile Visitors

3553 profile views
  1. Yeah, that was my reaction too. The nearby megaphone might be a bit over the top, though.
  2. Well good to hear that! I was about to go down the rabbit hole of pondering how soldermask can degrade over time and allow humidity to breach the surface. Also on another note, I'd add that there have been SE boards that were exposed to much more humidity than this one and have a correspondingly more corroded mounting bracket. But apparently many of those still work fine.
  3. Ha ha, probably the same experience Bill Atkinson had back in the day trying to explain "digitized photographs" on an Apple II to his wife. Why would you take a perfectly good video and ruin it like that?
  4. Yes and no I guess you could say... I've heard rumors that someone not on 68kMLA has indeed designed one but never manufactured and tested it. It would be nice if we could track down them and their work-in-progress, though.
  5. Also it's worthy of note. The two traces you've pointed out specifically are RA5F and RA6F. These are easily exercised regularly by 8KB memory block accesses, so testing with 1MB of RAM is plenty. If you're not worried about the RP1 traces on the rear side of the board, more RAM should work just fine.
  6. These traces routing between RP1 and RP2 and the RAM SIMMs are the RAM address lines. The filters themselves are inserted to properly condition signal reflections on these high-frequency lines. The current-rating of the full-width traces as printed is actually quite high, around 2 amperes... far more current than is actually used. Although 39 ohms is a relatively low resistance, still you'd have to have quite a bit more trace degradation before this is an issue. 0.1mm width should be fine. Another note, looks to me like there is soldermask on the surface of the trace impurities
  7. Probably not, most likely they tweaked their existing reverse engineered Mac Plus Brainstorm PALs using technical specifications of the Macintosh SE. F.Y.I. the replica RTC is still available for those who want to use it, currently it's program-it-yourself, or place a custom-programming order with Digi-Key, etc.
  8. Definitely that is a card that was first and foremost designed for PCs and merely retrofitted to Macs. You almost certainly don't need that much circuitry to support the terminal emulation given the computational performance and timing capabilities of the Mac, but it would be essential for the PC.
  9. Regarding source code matters you were wondering about on the GitHub page... Here's my take. Vintage software development with an SCM encompassed a centralized version control system like CVS... translating this to the modern era, using a serial terminal connection to a Linux box with Git and synchronizing files via ZMODEM is a quite workable setup that I use. Of course AppleTalk shares, ftp, and telnet work just as well, it doesn't really matter. Understandably at this point in the project, you're mentioning the main challenge may just be the time it takes the progen
  10. Just a wild guess, the "PC" suffix chip could be a binary counter, similar to the ones seen on full-fledged video expansion cards. Considering the other chip is a 16-DIP rather than a 14-DIP, I'd be suspicious that it is a flip-flop rather than a simple hex inverter or quad-NAND gate. In any csae, those two chips together can provide just enough glue logic to do some simple signal conversion relating to the sync signals and serialized video data.
  11. Yep it's definitely compatible with Molex SL, you could even just do 1/10 inch pin headers if you didn't care for the polarizing latch. When taking into consideration my experimental measurements thus far, hypothetically that is correct that you don't need any additional components to get a nice image. The thing is that in the original connection setup, there was a full-blown video breakout board that could plug into the DE-9 connector. I'd prefer not to encourage people to do away with the vintage panel connectors in case the original video breakout board ever turns up. Unless,
  12. Now that sure is interesting, I would have thought the line drivers would more just melt in a puddle on an overcurrent overheat. But that's silly upon further thought... it's not thermoplastic casing, it's epoxy. So disintegration like cracking would be more likely than melting to liquid.
  13. @codyNC90 Only the GLUE is unique to the SE/30. It's fairly easy to implement the logic for a replacement, the main problem is that it is impossible to buy a modern CPLD/FPGA that operates directly at 5 volts and has enough pins, so supporting adapter junk is required to surround the primary logic, and that in turn can interfere with PDS expansion cards if stacked vertically. Also, only the GLUE needs trivial modifications to increase CPU & RAM bus speed, assuming you want to keep all low-speed peripherals as-is.
  14. Ooo, that's interesting, looks like MegaScreen introduced a later revision of the MegaScreen SE that nixed the solder-in coprocessor option. Of course, because if you wanted to add the coprocessor by then, you'd just buy the SE*M and plug in the expansion board. Exact same video controller chip and still doesn't have any ROM on the board by the looks of it, so the drivers should be compatible. I've been meaning to send a design to PCB manufacturing to get one of my own adapter boards, and of course that would mean I'd get three boards and wonder what to do with the two others. S
  15. Regarding incorrect RAS/CAS timing, looks like if I correct the timing of S0 and S1, that should be worked out. Many thanks to @H3NRY's redrawn schematics for also including some signal timing diagrams!
  • Create New...