Well, I’m making progress on my composite video output.
I changed up the schematic a bit to suit the irregular video output of the H89, and quality of the video signal has improved significantly. That said, there is still plenty of work to be done.
Every so often, the signal going to my Emerson television will become erratic and funky, but most of the time it looks alright. I’m also getting glimpses of data on the attached Hamlet 302A waveform/vectorscope module, but they flash in and out way too fast to be of any use.
I took a closer look at the horizontal sync signal from the H89 on my 2445B, and I’ve also determined why I’m losing that entire line of text. The NTSC standard dictates that the horizontal blanking period should by 4.7μs, but the one I’m getting looks to be about double that, if my oscilloscope’s width measurement tool is to be trusted. I just have to figure out how the hell to cut it in half (the part of the signal that sits close to the middle of scope’s screen.
But, on the positive side of things (no pun intended), I’ve sorted out the flutter on the v-sync side of things, by way of a Mylar capacitor. I’ve also replaced a handful of the NOR gates with Schmitt triggers which have cleaned up the individual pixel signals, making inverted text much crisper and easier to read! In general, the text has been sharper since that change.
In related news, I’ve ordered a replacement set of MC1489N’s to drop into the spare terminal logic board I have. For some reason, both the 1488 & 1489 chips were absent, but I had a pair of spare 1488′s from my initial H89 repair work months prior. If my composite video signal board is successful, I’ll need to find a power supply and keyboard module next to make a solid self-contained terminal unit.
Lee Hart was more than happy to show me his Zenith Z19 terminal. Without even a second thought, he opened up the case to show me the goodies inside. What I didn’t expect was it to be equipped with a Northwest Digital Systems Graphics-Plus board, greatly improving its capabilities.
It has built-in menus, a screen saver (that just stops sending graphical data after a period of inactivity), and several different graphical modes including 80x24, 80x49, 132x24, 132x49, and even a Tektronics mode! You can change the baud rate within the menu, reverse the video on the screen, pick your word wrap settings, and much more – all from within a software menu. Normally to change many of those settings, you open the case and change a few DIP switches.
And to get all this capability, you just drop this board into a standard Z19 or H-19 terminal, and hook them together. It should be noted that the Z19/H-19 terminal is its own Zilog Z80 based computer, but with a criminally low amount of RAM. Either way, I need to get my hands on a Graphics-Plus board
Lee and I discussed Heathkit/Zenith terminals for awhile, including my H-19 project. He was very encouraging about it, and gave me a myriad of suggestions and pointers. He told me that it would be easy to implement PS/2 keyboard functionality on an H-19 if I just rewrote the ROM, since many people have already given Z80 based computers the ability to read PS/2 keyboards. I’m guessing that burning custom ROMs is a trivial task for him, since he treated it like an everyday thing to do. I seem to recall him saying “If you can find a way to add more RAM…” then the implications would mean a stand-alone computer, or at least a more powerful terminal.
The whole time, Lee’s attitude towards me implied that I could do anything, even though he only had known me for about an hour. Maybe describing my composite video mod impressed him. He also clued me in to some secrets about the H-89 when it comes to floppy disk drives… seems that my hunt for a Siemens FDD-100-5 isn’t necessary to restoring my H-89 to functional status. We barely even talked about COSMAC ELF 1802 membership cards, despite that being his specialty. I did let him know I had purchased one at VCFEX though.
Thus ends my Vintage Computer Festival Midwest 10 photos. I hope you liked what you saw, and learned something too.
Commodore Z & Hero Jr. at Vintage Computer Festival East XI
I exhibited my Heathkit Hero Jr. robot and H89 at VCF East XI last weekend. Hero Jr. was fully upgraded for the occasion, completely decked out with all the bells & whistles. I set up and ran programs on Hero Jr. throughout the day from his master cartridge, and programmed simple BASIC sentences for him to parrot back on command.
Unfortunately the loud exhibit hall made hearing Hero Jr.’s synthesized voice a bit of a challenge, even with the volume cranked up. Plus it interfered with his ability to measure distances with his ultrasonic sensor, rendering him incapable of navigation. Demonstrations were well received by all sorts of folks, and a few people even programmed in custom phrases for Hero Jr. to speak. The most popular trick was him singing Daisy, as it’s a staple of computer tunes.
Near the end of the exhibit time on Saturday, Hero Jr. started acting odd. He wouldn’t access ROM routines correctly, and began to freeze up whenever I selected a specific mode. So I powered him off, put him on the bench, and opened up all the panels. After powering him back on, he barely got out a few words before ceasing to function as a tiny puff of smoke rose up from his power regulation and sensor board. And with that, Hero Jr. was down for the count, and I had won the “first to smoke” award, more commonly given out to Apple III’s at VCF East’s.
I have successfully logged into my Raspberry Pi via the H89! I feel so accomplished right now.
However, it seems that my ’T’ key is not functional, so I can’t use many important commands, like “history”. I ran pipes.sh, but it looks really weird. And lynx doesn’t even look right, which makes things… awkward. I had better install links while I’m at it to see if there are any differences. I should also install that image->ASCII converter for maximum lulz.
An authentic H19 keyboard arrived in the mail today. After a bit of cleaning up, I hooked it into my fledgling terminal, along with its new 9V power supply – one of sufficient wattage. It took some mild troubleshooting to bring it to functional status, but it works!
The video signal still needs a bit of tweaking, but I would say that looks like a functional and viable terminal. I love this sense of accomplishment. Next I’ve got to find a suitable enclosure for the guts, and fashion some more permanent connectors.
Today I scored a Heathkit H-89 computer from 1979 (also sold as the Zenith Z-89). It’s a Z80 based computer that was typically available as a kit machine. It came with an external double 5 ¼ floppy drive that had no manufacturer’s markings on it whatsoever, as well as two boxes of disks for it.
So far, it seems to be in good enough condition where I wouldn’t be worried to fire it up. I do want to connect up one or all of the floppy drives first, since the internal connector appears to have been undone.
Upon pulling away from the curb to turn around in the culdesac outside the estate sale I bought it at, I let out a very enthusiastic “YEAAAAAAH!” to the point where I was heard by an old couple who were crossing the street. They turned and looked at me like I just insulted John F. Kennedy or something, and they even did it again once I made my 180 to head out the way I came. Hopefully my thumbs up let them know that I wasn’t pissed but ecstatic
I performed a myriad of tests tonight, including checking the voltage on the CPU board in a few places and confirming clock and logic signals. A +5VDC line had a loose connector coming from the power board, probably causing some issues. That’s fixed now. I was going through clock signals in various places when I got to one that was correct, and I decided to do a system reset. Then magically, to my surprise an H: prompt appeared on screen. I started a memory test, and continued doing my diagnostics with the scope, then shut it down, closed it all up and turned it back on.
And it still booted to an H: prompt! So I did another memory test (seen here in progress). I was also able to get it to perform a computational check successfully. However, I think my floppy disks have not aged well, and unfortunately I have had no luck getting HDOS to boot from a floppy. Oh well… I should probably test out my external floppy drive at some point too.
Supposedly, my null modem connector may not be correctly wired, hence why I’m getting no throughput in or out of the terminal. The next step is to hook the oscilloscope to the send/receive lines to see which one is which. Hopefully then I can get the terminal feature finally working in lieu of software.
I made a small makeshift table out of a board and those two Ultra 10’s I brought home today – perfect size for the H89 too!
Oh, and the letter ’T’ works again. I disassembled the keyboard as much as I could, tested it profusely with the continuity tester, and even resoldered the ’T’ key’s pcb mounting points, but I’m not sure specifically what fixed the problem. I tested out irssi with moderate success to log into #joysticks. I also visited retro.hackaday.com to see how it looks for myself on classic hardware.
Somebody on a Heathkit computer user group was kind enough to send me a pair of test floppies for my H89. What makes them special is that they are hard sectored floppies.
One is Heathkit’s HDOS 2.0, and the other is Quikstor’s QS-CP/M. Now I just need to reconfigure my terminal board and see if I can properly boot off these disks. If not, I’ll have to figure out what is wrong with my floppy drive.