Read on Twitter
Say “Hello, World” to my new Zenith ZL-1 MinisPort laptop from 1989 with an 8088 running 8Mhz, 1MB RAM, and a backlit LCD display. Weighs just under 6 pounds.
The removable NiCd battery pack also provides the charge port connector. There’s a slot next to it for an optional modem (hint: there’s a TTL serial header in there for COM2).
There’s some curious branding on this particular machine. Lumisys Lumicam is printed around the screen. Any ideas what this is? A film printer? Additionally, the original Zenith logo has been removed. All that’s left is a ghost sign.
The Zenith ZL-1 MinisPort was the only computer to use a 2” floppy disk. I don’t have any so here’s a 3.5” floppy next to the drive for comparison.
I did manage to get the Zenith MinisPort to boot (though it constantly complains the CMOS battery needs replacing, even though I replaced it). But here’s the BIOS and MS-DOS 3.3 booting from ROM! Its rather hard to photograph this screen.
A couple of oddities: first, I can’t save BIOS settings to CMOS (reliably) like the clock and boot order. Curiously, the clock/date fields seem to add 32 to everything I enter. Enter “00” it will change to 32. Enter “2020” it changes to 5252. Odd, right? Why is this?
Another oddity with the MinisPort is under MS-DOS, If the screen reaches the bottom and begins to scroll, the character case on the whole screen switches from lower to upper case.
Opening the Zenith MinisPort to have a look. The bottom cover comes off easy enough revealing the motherboard. I’ve annotated my best guesses. Can anyone confirm the BIOS RAM location?
Looking at the MinisPort internal power supply and there are signs of leaking capacitors.
Removing the three SMD caps first. Looks pretty awful.
Removed the large surface mount caps. They leaked too. The legs were bent in opposite directions making it very difficult. Strong dislike!!
Started cleaning up the MinisPort PSU board with solder braid after removing the caps. The SMD pads cleaned up nicely. Maybe there's hope yet?
I also reflowed all solder joints that were corroded with plenty of flux. Next, I used a scratch pen to clean corrosion on copper traces under the solder mask. Thankfully limited to the surface. It’s looking pretty clean now.
Painted on new green overcoat on exposed copper traces. Ready for new capacitors.
Installed new caps and put the inductors back in. Looks decent.
Good news is the PSU works and powers the MinisPort! Bad news is it still behaves strangely. There’s an internal memory test and it revealed a memory error at address B800:0000, bit 5. Any ideas how to map that to a bad part on the board?
Scoping. I think I need a logic analyzer next. Any recommendations?
Stymied by the memory problem on this board, I decided to get a second one. Luckily it has 2MB of memory and the modem card. Unfortunately the power supply also needs recapping and restoration. Solder shouldn’t be white and fuzzy.
The caps are actively sizzling and leaking onto the board as I remove them. But they’re out. Now to clean up.
Lots of flux and solder braid later and it’s looking much better. There was one casualty : this SMD inductor came off as I was reflowing it. I cleaned up the bottom looking for pads but it disintegrated. Just two fine wires left.
Under a microscope, I added some adhesive backed copper squares to both sides, folded the wires onto them and soldered on Kynar pigtails. It still reads as an inductor!
Scratch pen on a few areas of corrosion under the solder mask. Not as bad as the first one.
Recoated the bare copper and added new caps. Glad I bought extras. The tiny inductor is hot glued to the side of a capacitor temporarily until I get a replacement.
And it works! And this one has no memory issues like the first one. Here’s the BIOS and MS-DOS 3.30+ booting from ROM.
I also got a matching external 2” floppy drive and some disks.
Here’s a stack of 8”, 5.25”, 3.5”, and 2” floppy disks for comparison.
Here’s a stack of 8”, 5.25”, 3.5”, and 2” floppy disks for comparison.
Back to the first Zenith MinisPort with the memory issues, probing the video memory bit 5 and the output enable with a logic analyzer. I think something isn’t releasing the shared data/address bus on AD5.
This 74HC245 at U41 acts as a buffer to separate the data bus from the shared address/data bus. It connects to both the video and I/O sections where both exhibit the same quirks. My first time replacing a SMD part. Hope this works.
I'm using this ChipQuik kit which is a low melting solder and flux. The idea is it should stay molten longer so you can lift it off. http://www.amazon.com/dp/B0019UZP7I/?tag=deflix
Fluxed all the pins, added some ChipQuik solder, couple of quick passes with the iron on both sides, a nudge and wow, that was quite easy!
Clean up with solder braid and alcohol. Looks great!
Crap, the new parts I ordered are too big? I’ve clearly got a lot to learn about SMD. But, lemons to lemonade, I’m bending the pins. I’ll replace it later with the proper package.
Crap, the new parts I ordered are too big? I’ve clearly got a lot to learn about SMD. But, lemons to lemonade, I’m bending the pins. I’ll replace it later with the proper package.
New 74HC245 installed. It sits a little higher than the old one but still low enough that it would fit if I left it. Now to see if I broke it further. 
You guys, that fixed it!!!! CMOS holds, no weird screen problems and it passed the extended memory test. I’m amazed that worked.
