#cpm — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #cpm, aggregated by home.social.
-
Huh. I kept making EEPROMs and having them not work when installing them in the H108A expansion unit's ROM capsule... but the DAKCOM ROM in there works fine every time.
Tried my EEPROM in one of the main unit capsules and it was fine. The DAKCOM ROM is the only 8K one in my set, so I found another 8K ROM in the collection, put it on the EEPROM and it works in that slot.
Makes me wonder if one or more of the lines on that socket needed for >8K ROM is hosed. Hhhmm...
-
Luckily, the Epson mini-DIN serial cable was one of the items still with this #px8 when it was found. I checked the pinout with my meter and, as luck would have it, the first 25->9 serial cable I dug out created the right pinout for the MH-20 simulator.
A bit of dorking around later (including pulling the Zork binaries off my IMSA-8080) and here we are.
Next up: making some #px8 EPROMS while I wait for the replacement belt for the tape drive to get here next week.
-
I took a brief detour when I thought of this by firing up my lone Windows laptop (which I pull out grudgingly for less wholesome tasks). I'm not tainting my M4 Max with java.
Of course, after driver hell for the USB serial, the java app on Windows exploded every time it tried to use COM6. Typical.
On a whim, I decided to try it on my 2015 iMac running MacOS Mojave, already installed ancient java and to which this USB serial is usually attached. Worked like a charm.
-
I really would like to have it running entirely on the machine such that I could use it as the untethered (now with homemade batteries!) laptop that it should be... rather than dragging an eleven year old iMac around on an umbilical.
BUT... it checks the box: Zork is running locally off media the #px8 perceives to be legit. Not a terminal session to one of my other machines.
-
So far, I'm stymied on burning Zork to EPROM - none are large enough (even a spanned pair). The cassette drive, even with a 60 minute cassette, is too small and the heat death of the universe would take place before it loaded anyway.
What I did find was the java-based Epson MH-20 simulator, which offers emulated Epson floppy drives over a serial connection.
So, a 2015 iMac is currently attached by serial to my #px8 hosting a disk image I made with cpmtools.
-
-
Amidst the EPROM shenanigans, I got sick of looking at the dust / faint mold sandwiched in the layers of the display. I took it apart, cleaned it all carefully and replaced the two caps I found along the way.
The worst part is the decorative metal strip on the front. The screws are under it and the adhesive is nuts. No matter how careful you are, you're going to slightly nick something or rumple the metal a bit. It will never smooth out. :(
Before and after shots.
-
My Zork dreams, however, are not yet dead. I have another idea to try. :)
-
I wrote to Fred Jan in the Netherlands who hosts the tremendously useful archive of PX-8 information and where I was reading about the ROM formats:
https://electrickery.nl/comp/px8/index.html
Fred wrote back a bit ago with a very detailed reply to my various questions. Short version: 1M EPROMs on the #px8 are a no go as things stand at the moment... but perhaps not impossible with some dev work he was doing previously. The OS seemingly won't handle it currently.
-
Via OneROM, I found out about the 231024 EPROM, but there's little to be known about them out there.
I've been running the PX's PXPROM2 tool on my IMSAI8080 replica (very convenient!) and making PX ROM images today, but so far I've not managed a 128K version that the PX-8 will read.
There are so many things I could get wrong in making this image, though, that I can't be sure if that's my issue or if it's just not possible to use on the #px8 at all.
Mysteries abound!
-
In my ongoing quest to run Zork on this thing, I keep hitting capacity walls.
The enticing thing is that this website:
https://electrickery.nl/comp/px8/proms.html
suggests that 1M (128kB) EPROMs do work in the "multi function unit only and PX-4 External RAM Disk". I have a HS108A "Multi Unit 64" attached. But it might be saying that's only with a PX-4? The "and" in that sentence is doing a lot to keep me hopeful.
Also, a 1M 28-pin EPROM is a unicorn, it seems. But OneROM does it!
-
So having a little fun with the #px8 today on the EPROM front.
Here is Microsoft BASIC running for the first time on this unit. For added fun, it's working off a 28-pin OneROM rather than a normal EPROM.
I 3D printed molex carriers using a 0.2 nozzle at 0.08mm and they help hold the OneROM in place, but it's finicky and too tall to close the cover. Handy for testing, then committing the final ROM image to a real UV EPROM and folding its legs into the carrier properly.
-
Actually, maybe I found it: EPROM type 231024. Now, whether it's compatible is the question...
Stumbled on this thanks to the OneROM Visualizer: https://onerom.org/visualizer/
-
Can anyone square how I'm seeing a claim that I can use a 1Mbit EPROM in the add-on unit of my #px8 (HS108A), but it's a 28-pin socket? Is there a 128KB (E)EPROM out there that isn't 32 pins?
Referring to this statement on the last bullet item under "Configurations supported for PROMFO2 or PXPROM2" about 2/3rds down the page:
https://electrickery.nl/comp/px8/proms.html
128KB might make my Zork dreams come true. :)
-
I do enjoy that if you hit play on an unmounted tape, it will audibly play the saved data. :)
Edit: decided not to murder my Mastodon instance with a video. Here a YT link instead:
-
Found Jeff Birt's video on restoring a PX-8. In the process, he refurbs the tape drive including a belt replacement. I've ordered the same belt (all of $4.95) and will do additional gear greasing work to mine once that arrives. Maybe it will be more reliable after that!
Meanwhile, have a few fun ideas to test with ROMs and serial stuff... but that's not for tonight. :)
-
Everything back together and seemingly functional. I even “formatted” a tape, copied STAT.COM to it and then very slowly executed it successfully.
Oddly, even if properly unmount the tape, I can’t remount it. I just get an error. Obviously reading and writing worked from the above test. Guess I’ll have to dig into that a bit.
But quite happy to be at this point. Now for tape and ROM and software tinkering!
-
Oh yay. More batteries, but perhaps no damage. I opened the Communciations Module add-on wedge for the first time tonight. It has a 4.8v four cell pack in it as well and a little green corrosion on the connector. The module does work, though.
I need to look into what all it does beyond the modem and the added ROM drive that explains the need for the battery.
Flip switch housing on the back has seen better days. I think that and the case screws are just rust.
-
Installed. No magic smoke and the scope shows all the signals I would expect in the right spots.
Time to reattach the LCD and close up the case to see if it all holds together.
-
I checked pins 15/16 for the LCD and, for the first time, they were a solid -15v. The LCD worked fine once reattached.
So I have mixed emotions here. Perhaps this was a bad solder joint that just manifested suddenly while the unit was running on my desk last week? Maybe Q34 is funky and will go south on me again without warning?
No components replaced, just ultimately reflowed the hard way. :)
I'll plunge ahead, but with trepidation mixed with new knowledge.
-
With ZD20 (really ZD10) seeming fine, I put it back in place (after ordering a kit of zener diodes I realized I need to have on hand) and focused on transistor Q34. It's a tiny SMD version and an in-circuit test seemed inconclusive.
So I removed it, carefully testing it on the bench without accidentally flinging this grain of sand into orbit. Tested fine.
So I carefully soldered it back into place. I routinely checked pin 7 again and suddenly it was ~2.8v.
-
Based on a theory @diver300 had, today I focused on the Zener diode ZD20 associated with transistor Q34. First off, I couldn't find ZD20. Period. I finally traced the circuit per the diagram and the diode that did exist there is actually ZD10, labeled with a 20 on its orangey surface. I also noted that C15 in the diagram is actually C16 on the board.
I took ZD10 out and a diode test passed. I then put a resistor and ~22v on it and breakdown was the expected ~19.8vdc.
-
I spent a TON of time probing the board with the scope and meter, staring at the schematics, hassling @diver300 with my "know just enough to get it wrong" questions... but, honestly, enjoying the problem-solving-driven avenue toward learning more fundamentals. I actually do enjoy these issues since it gives me a vehicle to learn.
-
When the LCD quit working suddenly last week, the first thing I spotted was a missing -15vdc on pins 15/16 of the finicky ribbon cable that attaches said LCD.
The next point of focus was pin 7 of IC 14D, a hex inverter, where the square wave was there, but low. That resulted in pin 6 being constant high. That starved the constant-voltage circuit of the sort of see-saw mechanism (my take) between two key transistors that ultimately generates the -15vdc.
-
We're back in business on the #px8 and the LCD is working again! With a lot of patient suggestions and insight from @diver300, I learned a lot more about various fundamentals and have a better (albeit not at all rock solid) sense of the behavior of the primary DCDC circuit where I felt pretty sure my problem originated.
The concerning bit here is that I didn't end up finding an outright bad component, but I'll detail findings in the next few posts on this thread.
-
The 1/3 AAA batteries arrived today. I’ll weld four in series soon to make a drop in replacement for the existing internal backup battery.
Right now I don’t want to add any variables to the mix while I try to solve the dead LCD issue that cropped up. The circuit I’m tinkering in does, in fact, also feed this same battery for charging… so, set that opportunity to add chaos off to the side for now. :)
-
If it helps, I have put the full technical manual in PDF format at archive.org:
https://archive.org/details/px8_technical_manual
Poking at things, I will note that that R55 and R54 in the path to pin 7 both test fine at 10K (in circuit). I removed and replaced C27 between them with no change in behavior.
My entire understanding/assumption that pin 6 should just be barfing out the inversion of pin 7 could be wrong... so I appreciate any feedback to set me straight. Learning! :)
-
I wondered if someone more experienced in #electronics could give me some thoughts.
Per the attached description, the pattern going into pin 7 of hex inverter 14D should be coming out of pin 6 inverted. I have a pattern on pin 7, but I'm just getting a flat 5v on pin 6. Is there a minimum voltage for the pattern arriving at pin 7 for the inverter to work? I've swapped the IC and same deal.
What might influence this behavior? Scope shots w/details in the alt text.
-
Having hit a bit of a wall for the moment (partly due to lack of SMD components on hand matching a few that might be in question), I decided to pose my question on the VCFED forums where I've seen some other good PX-8 discussions from recent years.
I just wrote it up and posted if anyone wants to follow or read the details and offer (here or there) undoubtedly better informed thoughts on the matter. :)
Once the post is approved:
https://forum.vcfed.org/index.php?threads/epson-geneva-px-8-lcd-stopped-getting-15v.1258602/