Recently in Retrocomputing Category

Apple III+ Clock Maintenance

| No Comments

It’s been a long time since I posted anything regarding retrocomputing, partially due to lack of time in general, but also because many of the projects I’ve undertaken as of late have been on more modern hardware. Now that the holidays are coming around, I have a little more time to play with some of my old toys.

One project that’s been in the wings since last August is  a new clock kit install for my trusty Apple III+. The Apple III’s current clock chip generates a fault during diagnostics, and the three cell AA battery holder suffered from leaky battery syndrome. I’d done a quick and dirty repair to it after taking the system under my wing a couple years back, but it certainly could use a replacement.

The upgrade kit itself is NOS (new old stock) and contains the battery holder, wiring harness and the clock chip itself — an MM58167AN. Also included is a motherboard connection clip/yoke that isn’t used on a III+ since the clock connector is integral to the newer design.

Most of the hassle of the install involves removal of the system’s expansion cards (in my case, a set of Titan III+IIe emulation boards with attendant cabling), the motherboard base plate, and lastly, a series of cables on the motherboard itself (speaker, keyboard, disk, video interlace, clock and power). I took several photos of the system during disassembly to ensure everything wound up looking the same during reassembly. 

All in all it was a fairly painless procedure, and everything worked as advertised once the job was finished.

I hope to be able to finish a few more tasks like this one over the next few weeks and will post again once they occur.


Retrochallenge 2012 Winter Warmup

| No Comments

I came across a post from 'urbancamo' regarding this year's Retrochallenge 2012 Winter Warmup, and figured it would be a great way to kick off a new entry for 2012. So without further ado:

Guys 'n' galls,

Just a quick note to tell you that the Retrochallenge 2012 Winter Warmup competition will be running in January 2012. The website is now up and running and already providing interest and amusement. The 'competition' is open to anyone using retro computing kit, it's a great community to be part of and there is always lots of interesting projects going on.

It'd sure be nice to see some SGI kit getting a workout this time around.

More info at our neck of cyberspace:

Regards, Mark.

I don't know if it's too late to actually enter at this point, but it certainly isn't too late to follow along! I'm hoping to start work on some retro projects this month as well.

More retro-show annoucements! Checking the N8VEM mailing list this past weekend, I spotted a posting for the Seattle Retro-Computing Society's inaugural meeting (text posted below). Would be neat to see some SGIs there in addition to the usual crowd of retro machines.

I am pleased to announce... the Seattle Retro-Computing Society's inaugural meeting, on Saturday, June 25th, 2011!

Do you do any of the following with old computers near Seattle?

+ Use them
+ Collect them
+ Play games on them
+ Write programs for them
+ Develop new hardware for them
+ Help other people do any of the above

If your answer was "yes," as I expect it will be if you're reading this near Seattle, then the SRCS is for you! We exist so you can show off your awesome stuff, bounce ideas off of fellow enthusiasts, and be inspired by one another's achievements, plans and aspirations.

No idea is too big or too small, and we're not picky about what flavor of vintage machine you prefer! Come on down and tell us about it!

The meeting is graciously hosted by the Living Computer Museum, a relatively new organization which is building a computer museum in Seattle's SODO neighborhood. There will be refreshments, presentations on various vintage topics of interest, and enough table space & power to set up anything you may want to show off!

For further details, please see our page at
Hope to see you there!

Mark Wickens has posted an announcement for the upcoming DEC Legacy Event in the UK.  He's invited the SGI, HP and Sun communities to participate as well.

Hi guys,

Just wanted to make an official announcement on Nekochan for the DEC Legacy Event. This is the second year of the event, and following on from last years great success I would like to extend the invitation to attend not only to those interested in DEC hardware but also in SGI, HP, IBM and Sun hardware as well (or any other kit that would be of interest).

Given the size of the venue I am not entertaining extending this invitation to our 8-bit friends as I believe this is a community better served by VCF-GB which I am informed will be making a repeat appearance in 2012.

The event is hosted over two days in the beautiful Lake District of the UK. Last year we had many interesting talks and presentations, including a very memorable talk by Stephen Hoffman via video-conference to the USA about his time spent at the mill. It's a very unique opportunity to meet up with fellow enthusiasts. This year we are planning on being a lot more hands on - expect to see a multi-machine/OS DECnet cluster being created for example.

The cost is GBP 15.00 for the two days, although you are able to book for either Saturday/Sunday separately as required. Most of this cost goes to hiring the venue.

Feel free to ask me any questions. I am about to spend a little time getting the website more up to date with people and machines that will be attending. If you would like to make a presentation or do a demonstration (or would like to volunteer someone!) that would be great - just let me know.

I would love to see a good representation from the SGI, HP and Sun community there!


Mark Wickens

Prototype Apple Disk III

| No Comments

It’s been a long time since I’ve posted something, but not for lack of things going on in both the SGI/IRIX and general retrocomputing world. I’ve picked up numerous machines, books and peripherals over the past few months, but most of them haven’t been tested. I figure it’s probably not too interesting just taking photos of the machines without actually doing anything with them first, but if anyone wants to see them let me know and I’ll post a few.

There have also been a few interesting IRIX software releases that have come about recently that I’ve not personally tried. Powering up my Tezro to try out software during the summer is just an invitation to misery — it easily boosts the already uncomfortable indoor temperatures by an additional 10 degrees in short order. Thus, most IRIX hacking will need to wait until the fall. I’ll post up some links shortly regardless, with the caveat that I’ve not personally tried the software myself.

The following are some photos I took the other day for the RetroMacCast website of a prototype Apple Disk III drive I picked up last year with a lot of Apple III parts, peripherals and software. It’s basically a Disk II chassis with a prototype Disk III analog card and some interesting Apple “racing stripes” on the sides. There are also Apple inventory stickers complete with “Property of Apple Computer Inc.” on the top and front of the unit.




For comparison, here is a production Apple Disk III analog card:


North Star Horizon

| No Comments

One of the retrocomputing projects I’ve been working on lately is finally taking shape — the restoration/construction of a North Star Horizon S-100 bus computer system.


I’ve been building this system up from parts, as the 1977 vintage Horizon chassis — though complete with a battered wooden chassis cover, HRZ-MB-3 motherboard and power supply — had been stripped of its complement of S-100 cards and diskette drives.


The HRZ-MB-3 motherboard supports up to twelve S-100 cards (only six are populated with connectors on my unit), and has a built-in serial terminal interface at the rear. I was able to track down a Horizon ZPB-A2 Z-80A Processor Board, Horizon RAM32 Memory Card and Horizon MDC-A4 Micro Disk Controller from additional eBay auctions. For the diskette drives, I went with Shugart SA-450s which are close to original spec (SA-400).


Next steps involve locating a new fuse holder for the power supply as the existing holder’s retention cover and fuse are missing. I also need to locate an appropriate floppy cable and operating system diskettes.


Perhaps later down the road, I’ll attempt to rebuild or repair the wooden chassis cover as it is in pretty poor condition. The original Horizon chassis cover was constructed of cheap veneer covered plywood (kind of a seventies thing), so repair would involve stripping the old veneer away, filling areas with wood putty and then applying fresh new veneer. It might just be easier to build a new cover to similar specifications out of quality wood, though it would be arguably less authentic I suppose.

Anyway, I’ve had fun putting this together so far; kind of a treasure hunt locating all the parts to form a mostly complete Horizon. Hopefully I can get it to a state where it will actually run once again!

Update: I noticed another NorthStar Horizon chassis has shown up on eBay. This one is missing the motherboard which may be a show stopper for many, but does have the Shugart SA-400 floppy drives, drive cable and a complete power supply. Kind of interesting - not sure why the Horizons that turn up these days are stripped, but I have seen a number of loose Horizon boards as well (including a bare motherboard last month) so it may not be too difficult to rebuild if one is so inclined.

Epson PX-8 Geneva


While catching up on older episodes of Earl Evans’ always entertaining Retrobits Podcast, he mentioned a company selling new, still-in-the box Epson PX-8 “Geneva” laptops. My interest in the machine was piqued after listening to additional shows which discussed both the PX-8 in general as well as covered his quest to get a virtual floppy system up and running for the machine.


Amazingly enough, though these podcast episodes were recorded nearly three years ago (in the summer of 2006), Star Technology continues to stock the new-in-box PX-8s for $99 each. Around the same time, I watched a used PX-8 go for $140 on eBay, which sweetened the deal.


I bought a unit, along with a new Software Library package containing Portable WordStar, CalcStar and Scheduler for an additional $15.

Earl provides a great overview of the machine in his podcast, which is of course recommended listening. But in a nutshell, the PX-8 Geneva is a Z80 compatible, CP/M 2.2 based laptop with an 80 column by 8 line LCD display. It sports a respectable 6-8 hours of battery life on a single charge.


Commercial software is provided by way of ROM modules that plug into a bay on the underside of the machine. Out of the box, the unit comes with CP/M utilities and BASIC on ROMs — and as mentioned previously, Portable WordStar, CalcStar and Scheduler are available in ROM format as well. User storage is provided via an internal microcassette drive, but an external cable can be built to interface with a PC-based virtual floppy emulator for more robust storage.


It’s a neat little machine! The display is surprisingly usable - having the full 80-columns does make a huge difference. The keyboard also has a nice, solid feel. It’s certainly a neat way to experience a vintage CP/M machine that doesn’t take up a lot of space, and it’s certainly fun breaking the seal on a machine that was “new” in 1984 … in 2009.

TRS-80 Model 4P RAM Upgrade


Last month I mentioned my desire to upgrade the RAM in my TRS-80 Model 4P to 128KB. To that end, I ordered a TRS-80 Model 4 RAM upgrade kit from computernews80 ($28, available via their eBay store).


The upgrade comes as a nicely packaged kit of eight Tandy-badged 8040665 64K x 1 DRAMs (essentially 4164 DRAMs with a Tandy part number) carefully inserted in a anti-static foam block. It also includes detailed installation instructions covering all revisions of the Model 4/4P series as well as a 5.25” memory test/diagnostic diskette.


The installation process was quite easy on my gate-array Model 4P and only took about ten minutes - I spent a little more time as I took this opportunity to clean the case plastics in the bathtub. In a nutshell, open the case, remove the logic carrier pan and simply plug the eight chips into the empty sockets on the logic board. Next, move the memory enable jumper from the E2/E3 pins to E1/E2.


And that pretty much sums it up - the additional 64KB works, allows for RAM disk capability as well as extended RAM for certain applications that can take advantage of it. At $28, the price may seem a bit high given current RAM prices for modern machines, but you do get a very nicely put together kit with authentic Tandy DRAMs. I couldn’t help but smile as I upgraded a roughly 25 year-old machine with “new” RAM (with 1984 date codes) in 2009. Fun stuff!


Next up I hope to find a source for CP/M for the Model 4 - it will be fun to see how the experience compares to other luggable CP/M machines of the period such as the Kaypro or Osborne.

eBay and Zip Chips

| No Comments

Ah, the Zip Chip. For those unfamiliar, the Zip Chip was a popular CPU accelerator solution for the Apple II series of machines, which bumped the default 1MHz CPU speed anywhere from 4MHz to 8MHz depending on the upgrade chip model. These chips are understandably still in demand and devilishly difficult to source. Unfortunately, when one does run across one of these treasures, it may turn out not to be what you’d bargained for.


I’ve come across a couple of these on eBay, and actually wound up buying a “phony” one last month (the other was also a phony, but the auction was more transparent to that fact). To be clear though, it wasn’t intentionally sold to deceive. Likely, the scenario played out along these lines:

  1. Twenty years ago, an Apple IIe owner buys a new Zip Chip. He installs the chip in his Apple IIe and removes the old 1MHz CPU, placing it in the same foam block the Zip Chip was housed in for safekeeping. He then places the old 1MHz CPU, manual, and diskette back in the original Zip box and stores it on the shelf. (Same thing I would have done.)

  2. Over the years, the Apple IIe is sold, passed on or otherwise disposed of with the Zip Chip still installed. The original Zip Chip box finds its way to another owner, probably mixed in with other Apple II software, and is placed on eBay. Hey, it’s got a CPU in the box - it must be complete!

  3. Gullible individual (me) buys it, finding a bog standard 1MHz CPU (a Rockwell 65C02 338-6503 pulled from a stock IIe).


In closing, make absolutely certain the Zip Chip in a given box is the real thing - there’s a very high probability it contains nothing more than leftovers from a previous upgrade. It’s becoming more and more apparent that the best way to get a Zip Chip is to find an Apple II with one already installed.

Power to the Model 4P

| No Comments

And another TRS-80 roars to life!


Like the Model III I resuscitated last summer (see “Power to the Model III”), I was able to bring up a Model 4P by swapping out a bad power supply. This particular unit was purchased on eBay for $11 before shipping, with the understanding it had some sort of power issue. As a nice bonus, it arrived with an extensive cache of diskettes which were not mentioned in the auction description.

The unit itself is in good shape; there are the typical minor scuffs, scrapes and yellowed plastics you’d expect to see on a 25+ year old machine, but nothing outright broken or missing. The keyboard’s feel is fantastic, probably the best out of the TRS-80 series.


The Tandy power supply was replaced with a new, old stock Antec 65 watt unit from computernews80. This is a drop-in replacement, and takes only a few minutes to install.

Once everything was buttoned back up, the unit booted without complaint, requesting a ROM diskette in multiple languages. I was able to bring up TRSDOS 1.3 and its Model III ROM and then softboot (using the reset button to keep the soft ROM in memory) the Model III versions of MultiDOS, LDOS, etc. This soft-ROM (Kickstart?) arrangement allows the Model 4P to run CP/M, as a physical Model III ROM isn’t intruding on address space.


After playing around with this machine a bit, I can certainly see passing my Model III on to another hobbyist - this little guy should fit the bill for any Model I/III/4/4P software I come across, and do so in a smaller footprint. My next step is to upgrade the RAM to the full 128K which looks pretty easy to do. More fun to come!

Running With ADTPro


A while back I mentioned my desire to get ADTPro running and transfer some disk images back and forth from my Apple II/III machines. I’d planned on ordering a serial cable for the IIc and IIgs to that end.

While I did actually order that cable, at some point after its arrival I shuffled it into the chaos that is my apartment - once I finally decided to try it out with the Apple III CP/M diskette it was nowhere to be found. Fortunately, neither the Apple IIe or III+ require the problematic DIN-5 serial connector, so I was able to bootstrap ADTPro using a standard 25-pin serial cable, leaving the IIc and IIgs out of the fun for now.

I ran the server software on my MacBook Air using a USB to serial converter. The bootstrap process was performed on the IIe by first booting ProDOS, selecting the Super Serial Card via “IN#1” (slot 1), and sending a command string which echos serial input to the console (first hitting CRTL-I then entering “14B” at the SSC prompt). At this point, the client software is streamed in and executed. Once the client is up and running on the Apple II, you simply write out a client boot diskette so you won’t need to go through the bootstrap process again (unless you’re entertaining guests or something). As a bonus, the ADTPro client diskette is also bootable on the Apple III which saves some work.


After creating the client boot diskette, I transferred an image of the Apple III CP/M diskette to the MacBook Air and uploaded a copy to the Asimov FTP archive. I’ve also mirrored a copy locally on FTP just in case.

Tandy CoCo 3

| 1 Comment

Last week I once again braved the streets of San Francisco to pick up an interesting bit of retrocomputing hardware - a Tandy Color Computer (CoCo) 3 for $20.

The lot came complete with a CoCo 3 console internally upgraded to 512K, a CRC Computers Inc. Disto Mini Controller 1 floppy interface, an external drive box with a 5.25” diskette drive, joystick, Metric Industries Model 104 serial/parallel interface, several ROM cartridges, diskettes and books.


I’ve never owned a CoCo personally, but I did have a lot of fun playing around with my uncle’s CoCo 2 system back in the eighties. Booting the machine evoked fond memories of keying in programs out of computer magazines and saving them on cassette, which is something I actually enjoyed doing back in the day.

However, I’d never actually used a CoCo with a floppy subsystem before and had to do a bit of Googling to unearth a few needed commands. I became a bit confounded at first, as none of the disk based commands I tried actually seemed to work.

I soon discovered that the switch on the Disto controller toggles between two different BASIC versions, and was initially set to standard Extended Color BASIC. Once the switch was in the correct position, Disk Extended Color BASIC became active and the commands I’d located online actually worked.


My attempts to boot Flight Simulator II still met with failure. Unlike some retro systems, the CoCo 3 does not perform a disk seek (autoboot) on power up - even with Disk BASIC active. So how to boot a piece of commercial software without a standard directory structure (the ‘DIR’ command returned gobbledygook)? Back to Google again, and this time came across a YouTube video of a chap booting up Flight Simulator II on a CoCo 3 emulator (linked below). Thankfully, he’d recorded the entire boot process, including the magic boot command - “DOS”. Once this arcane word of power was entered the drive sprang to life, booted OS-9 and finally Flight Simulator II.

I haven’t done much more with the machine just yet though I did boot up the original Flight Simulator I diskette just to see the difference. One thing I’d like to mess around with a bit is OS-9/NitrOS-9 itself. I actually owned an OS-9 based MM/1 tower about a decade ago, but at the time was less into the retro scene and wound up passing it on without really diving into it. The CoCo is quite interesting in that it’s default OS is a real-time, process-based, multitasking, multi-user, Unix-like operating system.

Stepping Up To RAMWorks III

| No Comments

Last month I wrote about the acquisition and repair of an original Applied Engineering RAMWorks card. Since then, I stumbled into a good deal on a 1MB RAMWorks III and decided to upgrade.


The card is very similar to the original RAMWorks - it even uses the same RAM chips, just more of them. Also like the RAMWorks, it can support an optional daughter card to increase memory capacity - this time up to 3MB. These daughter cards are virtually impossible to find though, so it’s pretty much limited to the installed 1MB.


My RAMWorks III arrived with rows of pretty metal/ceramic RAM chips installed. Regrettably, one of the chips failed the diagnostic and had to be replaced with a plastic package chip from my spares (an extra 256Kx1 I’d purchased during the RAMWorks repair last month). Now the card passes diagnostics and I’m able to successfully create a 1024K CP/AM RAM drive.


My Platinum Apple IIe is starting to really shape up into a nice system. I added a DuoDisk not long ago, and I also have a hand-held scanner with interface card standing by that I need to try out soon.

Korean Apple II Mockingboard Clone


I’ve been poking around for a Mockingboard sound card for my Apple IIe for a little while now, and the one thing I’ve learned from the experience is that despite robust third party software support, the dang things are rarer than hen’s teeth.


In recent years, GSE-Reactive apparently offered a limited run of homebrew Mockingboards, but those dried up a while back as well. Fortunately, it looks like another homebrew Mockinboard source has surfaced, this time hailing from Korea.

I scored one of these boards recently, and had the opportunity to try it out in my IIe with Electronic Art’s venerable Music Construction Set. The boards come stamped with a URL: My knowledge of Korean is non-existent, so you’re pretty much on your own as far as navigating the site - but it does appear to be geared towards Apple II enthusiasts.


After installing the card in slot 5, I plugged a set of headphones into the on-board mini-jack output and loaded up one of the Music Construction Set demo tunes. The card worked wonderfully, outputting stereo six voice polyphony as per a Mockingboard Sound II.

Make no mistake, a Mockingboard isn’t exactly the pinnacle of synthy goodness, but it’s several orders of magnitude better than the clicks and sputters belched forth from any stock, non-GS Apple II.


The seller appears to be offering a stream of the cards on eBay, but not at a fixed price. I wound up paying $46 before shipping for mine - you may (or may not) be able to do better depending on demand. For a home brew, limited run card, $46 seems pretty reasonable and I’m quite happy having made the purchase.

I finally took inventory of the cards installed in the IMSAI PCS-80/15 chassis I received last month. Here is the breakdown:


  • Processor Interfaces Inc. Cartridge Disk Controller (this is apparently a hard disk controller)
  • IMSAI MIO Rev 2
  • CCS (California Computer Systems) Model 2718A 2 Serial + 2 Parallel (two of these installed, only one routed to the back panel)
  • JTM 16K RAM (two of these installed)
  • Artec Electronics 32K-100 Rev 0
  • CCS Model 2422 Multimode Floppy Controller
  • CCS Model 2810A Z80 CPU

Looks like quite a bit of serial/parallel duplication - the IMSAI MIO and two of the CCS 2718As. Not sure if this is because one or more of the cards are squirrely; I’m just going to have to play around with them when I get a Variac.

It is really sweet that it has a Z80 CPU though, as that opens up the possibility of using a HarteTec Super I/O Controller. I also have a Vector Graphics Inc. 64K RAM board I can drop in there if need be.


Now I need to chase down documentation for these cards. Unlike the IMSAI 8080, this guy came bare. However, I can’t complain for the price!

Another cool thing to add might be the S-100 Bus Probe currently being offered on eBay as a kit or a pre-build. There’s a YouTube video of it in action (linked below):

IMSAI 8080


Earlier this week I picked up another IMSAI - this time the classic 8080.


This is really a special machine in multiple ways. Not only is it in wonderful cosmetic condition (save for one minor issue I’ve since corrected - more on that in a bit), it came complete with the original documentation binder, which contains warranty cards, schematics, instructions, etc. not only for the IMSAI parts, but every third party card as well. In addition, every single sales receipt for every component in the system is tucked into the binder’s rear pocket - all dating from the late 70s from various historic shops around the Silicon Valley (The Byte Shop, Anchor Electronics, Sunnyvale Electronic, Recreational Computers, The Soldering Iron, etc.). It’s a wonderful piece of history.


The computer was purchased from Stewart Chiefet, the creator and host of The Computer Chonicles. That alone is pretty darn cool!

The minor cosmetic issue I’d mentioned earlier was that it was missing three of the blue paddle switch handles (two Address Data toggles and the Examine/Examine Next momentary contact) Amazingly enough, the Fischer-Freitas Co. - which has owned the IMSAI trademark since 1979 - still sells replacement parts for the 8080, including the paddle switch covers! While ordering the paddles, I also bought four zinc mounting screws for the top cover as those were missing as well (the top cover was just resting on top when I got it). As you can see from the photos, I’ve already installed the new paddles and screws and they are absolutely spot-on perfect. Special thanks go to Thomas “Todd” Fischer for continuing to offer replacement IMSAI 8080 parts - very much appreciated!


I haven’t powered the machine up just yet as I still need to pick up a Variac so I can slowly bring the power supply online; don’t need any unnecessary explosions.

Here is the list of cards currently installed in the machine:

  • IMSAI MPU-A Rev 4
  • IMSAI MIO Multiple Input/Output board
  • Solid State Music Prototype Board (this currently contains a wire wrap implementation of a “Video Display VII” card according to included schematics)
  • Ithaca Audio 8K Static RAM
  • Thinker Toys SuperRAM 16K-A
  • Solid State Music MB6A 8K RAM Board
  • Solid State Music VB1-B Video Interface
  • IBEX 16K EPROM Board
  • Tarbell Elect MDL-1011A floppy controller
  • Active Terminator No. 106B (CompuKit/Godbout)

N8VEM Completed

| No Comments

The N8VEM booted for the first time today! I got the last couple components in from DigiKey, and was able to get it running during lunchtime.

The only thing that isn’t working is the power LED. I probably put it in backwards, and my attempt to reverse it didn’t work out. Not critical though - I can live without it. The computer itself is working great, including my freshly burned EPROM.

Now I need to start on the backplane and bus monitor boards - need to order some more parts!

(Yes, I do have a better power supply arrangement on order - this is just something I cobbled together to test with.)


Test Prototype Monitor Ready


F: PIP      COM : ASM      COM : STAT     COM : DDT      COM
F: ED       COM : LOAD     COM : DEBLOCK  ASM : RTC      COM
F: VDE263   COM : VINST263 COM : R        COM : BBCBASIC COM
F: W        COM : WSMSGS   OVR : WSOVLY1  OVR : WSU      COM
F: MBASIC   COM : XM50LB3  HEX : XM50LB3  COM : XM       COM
F: BCLOAD       : BRUN     COM : L80      COM : UNLOAD   COM
BASIC-80 Rev. 5.21
[CP/M Version]
Copyright 1977-1981 (C) by Microsoft
Created: 28-Jul-81
30776 Bytes free

ADM-3A Screen Restoration


I’m happy to report that the attempt to refurbish the ADM-3A’s screen went absolutely beautifully!


After watching a series of informative videos on YouTube stepping through the process, I was pretty confident that it would be a fairly easy task.


I followed the gist of instructions laid out in the video, with the exception that I didn’t preheat the lens with a work light beforehand - the CRT was small enough that it proved easy to heat up using just the gun. I also used bamboo skewers to poke underneath the lens to coax out hunks of PVA as I went, which worked out nicely.


Once the lens was removed from the CRT, I used Goo Gone to clean up any remaining PVA, and Windex with newsprint to clean up streaks and fingerprints. I applied bits of double sided foam tape around the outside of the CRT to hold the lens and keep it elevated from the CRT face, and after fitting the lens in place, sealed with packing tape around the perimeter to guard against any foreign matter entering between the lens and CRT. I then reinstalled the CRT assembly and it looks as good as the day it was made - no signs of burn in and perfect image quality.


Now I can concentrate on the lower case ROM project; I have the EPROM burned and ready, just need to wire up the shim socket. Hopefully I can get that completed later this week.

Apple IIgs CFFA


I finally took a trip down to Fry’s yesterday afternoon and picked up an adequate Compact Flash card for the CFFA - something I’ve been meaning to do for weeks now. They didn’t really have much of a selection to choose from, but I found a 2GB Kingston card for $15 which does the trick.


This morning I installed the new CF in place of the 16MB evaluation card, created a few 32MB ProDOS partitions and installed GS/OS 6.0.1 on the first. Everything went smoothly, and the machine boots up just fine. A pretty sweet setup all in all! I’d considered going with an Apple High Speed SCSI card, but this is really the superior option - it’s fast, quiet and has plenty of capacity to go around. Plus it’s pretty easy to read/write the CF cards on a modern machine for file transfer.

Yesterday afternoon I purchased a sheet of acrylic from Home Depot and cut a new monitor shield for my Neo Geo arcade machine. What an improvement that makes - it was somewhat tiresome looking through all the carved initials on the old one. I think I’m just about ready to move into the painting phase - I should order up some side art vinyl and a monitor bezel to get ready for that.

EPROMs Galore


I finally picked up a modern EPROM programmer. I actually have an old Hippo programmer designed for the Atari ST, but the lack of documentation makes it tough to use (plus it can’t handle 32 pin DIPs). Anyway, Willem programmers are amazingly inexpensive on eBay - the 5.0C I picked up only came to $42 including shipping.


I spent the early hours this morning installing Windows XP on an old laptop (my desktop lacks a parallel port) along with the Willem software. My first target - a fresh 27C080 from Jameco, which arrived a couple days back along with my other N8VEM parts. Why the new 27C080? There’s an enhanced ROM on the N8VEM web site which includes Wordstar and other CP/M goodies, and I didn’t want to sacrifice my known good, pre-programmed 27C080 should something go awry with my own experimentation.

It took me a couple of tries, but in the end I got a successful, verified burn. I’ve gone ahead and installed the new ROM in the N8VEM.


The Willem burner will also come in handy when the 2716s arrive for my ADM-3A lowercase project - the 2102-1 RAMs have already arrived and are installed and waiting. Which reminds me, I need to pick up a heat gun this weekend and see about that CRT lens …

Recapping and the N8VEM Taking Shape

| No Comments

I’ve done quite a bit of soldering over the past couple of days.

Yesterday, I dove into the monitor in my Neo Geo cabinet and performed an extensive capacitor replacement. I’d recapped it somewhat a little while ago and that really cleared it up quite a bit, but it was still blurrier than I would prefer. This time I replaced every capacitor on the chassis, including the capacitor on the neck board.


What a difference it made, too. In fact, the new neck board capacitor completely threw the color balance out of whack as it had been tuned against the old, dried out electrolytic. After adjusting everything, it’s pretty darn nice - great color and it no longer blurs after it’s been powered up a while.

I’ve also nearly completed the N8VEM single board computer. In fact, I’d be performing the smoke test right now, but Digi-Key seems to have neglected to include a couple of the components in the shipping box even though they show up on the invoice. I put in one last mop-up component order this evening, so hopefully I can finally smoke - er, power up - the unit in the not too distant future!


Starting on the N8VEM SBC

| No Comments

I took the first tiny step into assembling my N8VEM SBC this morning - soldered two carbon composition 1K resistors into the board. Kind of funny - these two resistors were second sourced from Allied Electronics; Digi-Key was out of stock and wasn’t going to get more until something like next January. Anyway, they arrived first so I stuck them in.


Hopefully the rest of the parts from Digi-Key and Jameco arrive soon so I can really tear into it - I’m chomping on the bit! Once the SBC is completed and verified working, I’ll start ordering parts for the ECB bus and bus monitor boards.

Lear Siegler ADM-3A

| No Comments

Well I finally got a hold of a ADM-3A terminal. This is one of those toys I’d been looking for a little while, and one finally popped on on eBay for a reasonable price ($29.95 before shipping).


The reason the unit went for such a low price is due to the condition of the screen - it suffers from “CRT cataracts”, a somewhat common malady on older CRT equipment. My understanding is that the glue layer (polyvinyl acetate or PVA) that binds the lens to the CRT breaks down over time and causes patches and specks which make it difficult to use the screen effectively.

There are a couple methods to repair this condition. One option is to remove the lens with a heat gun, clean and reattach. The entire CRT can also be replaced; new tubes can be had for around $50.



The ADM-3A arrived in its original packaging, which was quite a surprise. All the original manuals, warranty cards, etc. were included and in great condition. The ADM-3A itself wasn’t so lucky however - the unit and the plastic bag that contained it was smeared with a nasty, brownish liquid.

Indeed, the entire terminal was awash in this strange oily residue. I have no idea what it is - it almost looks like something that might leak from a battery, but the ADM-3A doesn’t contain anything like that. I opened the machine and couldn’t find anything that looked like a source - it was all over the place. It took me about an hour to clean it up, blotting at the logic board with a microfiber cloth.

Of course at this point the CRT cataracts were the least of my worries, and I was a bit too flustered to bother to take pictures of the oily mess. It’s kind of interesting in a way - the unit shipped from Houston and the auction closed right after Hurricane Ike. Whether it’s related, I have no clue - but I was concerned that the unit may not have even survived at the time.

PICT1484.JPG After the cleaning, I decided to give the unit a try. Upon power up, it beeped and a cursor appeared in the home position. I toggled the local mode DIP switch and typed some text on the screen. Yay!

The keyboard works great and the CRT actually seems pretty decent underneath the lens cataracts, so it may be worthwhile to try removing/cleaning the lens first. I figure if I break it I can always fall back on the new CRT route.

The other thing to look out for is a lowercase ROM - the socket is sadly empty. There is a site with some guidance on producing a lowercase ROM utilizing an EPROM and an adapter board - I’ll try contacting the author and see what can be done to get one going.

All in all, I’m happy now that the unit is cleaned up and functioning. The fluid bit is really weird though - hopefully I don’t sprout a new limb or something.

Applied Engineering RAMWorks

| No Comments

This morning I managed to get my Applied Engineering RAMWorks card up and running in the Apple IIe Platinum.


A couple weeks back I picked up the RAMWorks card (with 256K installed), a set of floppy media with utilities and diagnostics, and an additional 256K in HM50256P-15 DRAM chips - all from different eBay sellers. Once the three packages arrived, I was ready to give it all a try.


I carefully installed the DRAM chips, inserted the card in the IIe’s auxiliary slot and booted the diagnostics. The excellent diagnostics utility goes so far as to display the layout of the card on screen and marks any bad DRAM chips for easy troubleshooting.


Sure enough, I had one bad chip and hadn’t bothered to order an extra. I performed basic troubleshooting: moved the chip to a different socket (failed in new socket) and tried a replacement DRAM from a GS RAM card (passed). I then put everything aside while I waited for a couple more DRAM chips to arrive.


Now that everything is up and running, I can allocate a RAM drive under CP/AM which is nice. It’s too bad the original RAMWorks can’t reach a full megabyte though - it requires an unobtainium daughter card in order to perform that feat. Should still be adequate for my purposes though, and it was kind of fun snapping some chips in the sockets. What can I say - I’m easily amused.

I have lots of projects going on at the moment - I usually like to save the blog entries until I’ve completed something so I don’t get too boring with progress updates. Unfortunately I have so many irons in the fire right now there aren’t a lot of completed projects to report on.

Here are some of the various things I have going on at the moment:

CFFA card in Apple IIgs: The card has arrived and is currently installed in slot 7. I performed a test install of GS/OS 6.0.1 on the supplied 16MB test card and it boots right up, but I don’t want to head too far down that road until I grab a bigger CF card.

BOS on Apple III: Recently obtained BOS (“Bob’s Operating System”; replacement for Apple SOS) from Washington Apple Pi. I plan to install this on the ProFile drive in the near future.

ADTPro on Apple II: Have the cable, just need to sit down and try moving some disk images back and forth between an Apple II or III and my Mac.

TOP-DOS on Atari 800: I’ve received the original OS diskette and manual - need to test drive it.

Power up IMSAI 8085: Already talked about this machine a couple days ago, but it’s certainly something that’s still on the schedule as well. The 8” drives on this beast dovetail nicely with the 8” diskette project I’d mentioned earlier for the TRS-80 Model II; I’d like to be able use these drives to create boot disks.

Create new keyboard cable for TRS-80 Model 16b: Yeah, this is another recent acquisition. My 16b doesn’t have a keyboard and the Model II’s keyboard won’t work as-is since Tandy reversed the cable layout (the Model II has an integrated keyboard cable and the 16b has the keyboard cable on the keyboard itself). I’m going to order the parts to make a new cable that should allow the Model II keyboard to work with the 16b.

N8VEM SBC build: I’ve ordered the bare boards for the N8VEM single board computer and ECB backplane. Once they arrive I’ll need to order the components and start assembling it. I’m really looking forward to this project!

I have a bunch of other projects of varying size out there in addition these, but this is getting too long as it as. I’ll stop making excuses for not getting something done and cut into something I did work on today.


I played around with the Kaypro II a bit this afternoon. As mentioned earlier, this was a freebie I’d picked up last week. I may have given the impression it was 100% functional due to my glee over the IMSAI acquisition, but the truth is that despite powering up and requesting a boot diskette, the drives themselves failed to seek.

So I opened the machine and poked around a bit. Everything looked pretty decent except the large power bus connector on the power supply seemed to be backed off the pins just a tad. I reseated it and tried the machine. It booted CP/M 2.2 off drive A: without issue. Easy fix!


Also happy to report that both the CP/M 2.2 and MEX diskettes work. I also tried the CP/M diskette in my Kaypro 2X but it failed to read it. Not a huge surprise as I’ve heard it needs a special version of CP/M despite being a “2” (it’s apparently more like the “4” than the “II”; Kaypro was rather annoying with their somewhat mentally imbalanced machine version scheme. Heck, their first machine was the II!)


And that’s all for this weekend - I’ll have more during the coming week.

Enter the IMSAI PCS-80/15

| No Comments

Craigslist can be an amazing resource - you just never know what will turn up.

While browsing through the vintage computer listings on Craigslist this afternoon, I stumbled upon a free Kaypro II system. It was by just chance that I found it; I normally I search on keywords like “vintage computer”, “museum computer” or a mixture of brands (Atari, Commodore, etc.) For some reason I tried “Kaypro” this time, probably because I hadn’t gotten any hits on anything else for a while.

The Kaypro II post was already three days old. I figured it was probably long gone since it was free, but I wrote anyway:

“Hi - I don’t suppose the Kaypro is still available?”

The response came back quickly. “Why, yes it is! :) Let me know…”

After a brief phone conversation for directions and the best time to meet, I drove down to Los Altos to grab it.

When I arrived, the seller already had the Kaypro II set up on a bench outside his garage. He powered it on and showed me the boot screen.

He mentioned his wife had thrown out his diskettes at some point, but the Kaypro II CP/M 2.2 system and MEX114 (ModemEXecutive) communications diskettes were still loaded in the drives. Sweet!

“So, why did you want the Kaypro?” he asked.

“Well, I’m a CP/M fan - I’m always looking out for fun hardware to play around with,” I responded.

“And you didn’t want the S-100 system?”

I froze. “S-100 system?”

“Yeah, I advertised an IMSAI S-100 system on Craigslist too, but before you wrote no one had contacted me about either one. I offered the S-100 to the Computer History Museum, but they told me they already had several examples. Do you want to look at it?”

“Oh definitely, that would be great! Are you asking anything for the S-100 stuff?”

“Oh no, I just want to get it out of here. I have too much junk in my garage as it is and it would be a shame to take it to the dump.”

We headed back behind his garage where he had the IMSAI stored. It was a massive machine with a hand built, 8” drive tower. He mentioned that this system had belonged to his father in law and that he had hand built the memory cards.

So what started as pretty cool if low key mission to pick up a free Kaypro II turned out to be an amazing find. I’ve wanted an S-100 system for nearly twenty years, but they always seem to go for “Captain Insano” cash. Even the reproduction kits are extremely expensive. I’m still giddy over this!


When I got the IMSAI home, I did a little bit of exploration. Inside, there’s a label and Post-It note which seem to indicate it was once sold at a hamfest or swap meet: “Make offer - all or parts” and “Most docs avail.”. There was also a S-100 card and spare oscillator in an anti-static bag tucked in front of the installed cards, which may have been purchased at the same time and simply stored. The power cable was also stored inside the system.


The S-100 backplane is a massive 19 slot Thinker Toys Wunderbus. I tried pulling some of the cards to get a better look at them, but they’re pretty well stuck. I don’t want to break things this early in the game so I’ll come back to those. From what I can see, there are a couple 32K memory cards, I/O, floppy controller, CPU, etc. - around eight cards aside from the one in the bag.


The drive tower consists of a combination metal/wood chassis with three bays. The bottom two bays hold Shugart 8” drives, and the top bay is a storage compartment which also hides some control switches. It currently contains the unit’s power cord. The builder added some extra moulding to the door to make the compartment resemble one of the Shugarts. The whole thing just reeks of geeky homebrew retro goodness.



I’ll explore the system more as time allows - I don’t want to jump in without cleaning it throughly and making certain it won’t explode when powered on. I’ll also need to track down software for it which should keep me busy for a while.

Fun stuff though; I’m psyched!

Installing Atari Rainbow TOS


This morning I finished up installing Atari “Rainbow” TOS 1.4 in my Atari 1040ST.

I say “finished up” as the process began a couple weeks ago when I’d discovered a set of Rainbow TOS 1.4 upgrade ROMs on eBay. The auction stated to specify whether I required the two ROM or six ROM version. Not knowing this information off hand, I needed to rip the 1040ST apart and find out.

Getting to the TOS ROMs was surprisingly difficult. On the 1040ST, they’re located deep within the bowels of the computer, underneath the power supply. To get to them required a complete disassembly of the system, removing numerous screws, RF shields (with little twisted metal locking tabs), connectors, the power supply, etc . Once I’d done all that and confirmed I needed the six ROM set, I set the whole pile of parts aside until the ROMs arrived - it was a lot of work tearing it all down and I didn’t feel like reassembling it and doing it a second time a few days down the road.


Yesterday afternoon the new TOS ROMs arrived. They shipped in the original, unopened package just as dealer would have received directly from Atari. In fact, the supplied documentation stressed that this was a dealer, rather than end user install. Being that this is 2008 and all, we’ll throw that advice to the wind.

Since the machine was already gutted, swapping the chips out was straight forward, but figuring out where all the various sized screws went two weeks on was a bit challenging (yeah, I know - should have taken notes). I also had a bit of clear plastic sheeting left over - looks kind of like it was meant to isolate the power supply from the RF shield, but I couldn’t puzzle out how it was supposed to fit. No matter - off to the wind with you, odd bit of plastic!

Once everything was reassembled, the machine booted into TOS 1.4 without complaint. The “About” dialog on TOS doesn’t give version numbers like most other operating systems, but you can verify the version by the copyright dates. The old TOS 1.0 I removed was copyrighted 1985, and the new TOS 1.4 in 1989.


The Atari ST is kind of an odd duck in that the entire operating system is supplied on ROM rather than diskette. So why bother to upgrade to this newer version? Primarily for the MS-DOS diskette read/write/format compatibility - this will make it a lot easier to swap files back and forth between a modern internet capable machine and the ST. There are also other sundry features and bug fixes rolled in which makes this a worthy upgrade all around.

Neo Geo Upgrades


Over the past couple weekends, I’ve been working on refurbishing the console and controls on my Neo Geo arcade machine.


The machine had obviously seen some hard use in arcades over the years, and the existing Happ sticks/buttons and Cherry microswitches were corroded, sloppy and just not very attractive. One neat thing about this cabinet though is that it has an all-metal console - much like a Japanese style “candy cab”. This makes it fairly easy to mount Japanese joysticks and buttons which are primarily designed for metal or plastic consoles rather than the more common wood consoles on machines here in the west.


I hopped over to Lizard Lick Amusements’ website and ordered two Sanwa JLF-TP-8T joysticks with GT-Y octagonal restrictor plates and Seimitsu LB-39 “bubble tops”, a full set of gorgeous Seimitsu PS-14-KN pushbuttons, and a bunch of .110 solderless quick connects to wire everything together (Happ controls use .187” terminals so these needed to be replaced as well).

Next, I headed to Arcade Overlays and ordered a slightly customized Neo Geo console overlay. You’ll note in the “before” shot of the console, the original Neo Geo field was positioned well above the controls. I had Arcade Overlays print the replacement about 4” down from the top so the controls actually fit on the red/white striped field.


Once the controls arrived it was modification time. The Seimitsu buttons are 30mm rather than the approximately 28mm on the old Happs. This meant that the mounting holes in the console had to be ground out to allow the new controls to fit. A Dremel tool with a grinding attachment (and my brother) was used for this purpose.

I also needed to fabricate new mounting plates for the Sanwa joysticks. The existing console has integrated mounting bolts for Happ style sticks which don’t match up with the Sanwa mounting plates. I did a little research online and found a fellow who mentioned he’d fabricated new plates using junction box covers sourced from a local hardware store.

Following his lead, I headed down to the local Home Depot and poked around their selection of junction box covers. I found some that were even better than the ones the blogger had used - these already had the center hole for the stick pre-punched, and were of the correct dimensions so nothing needed to be trimmed away with a metal saw. The only modifications necessary were the eight mounting holes for the stick and plate.


My brother took the plates home and returned the next weekend with both plates drilled and nearly ready to go - I had to widen the console mounting holes a bit so the plates could travel down the bolts completely.


Next, the wiring harnesses on the Sanwa sticks were integrated with the existing JAMMA harness, and the .110” terminals installed on the buttons. Finished!


There’s still quite a bit left to do on this machine, but getting the console done was a huge step forward. I think maybe my next project will be to get a new arcade monitor and plexiglass. In the meantime though, Matrimelee with the new controls is pretty sweet!

ICD Flicker Free Video

| No Comments

I took some time yesterday afternoon to try out an ICD Flicker Free Video 2 I’d recently picked up.


The Flicker Free Video 2 is a small card that plugs into the Denise socket in older Amiga systems. In addition to the board itself, it arrived with documentation and a driver diskette.

The install process was rife with problems. More than likely unrelated - but annoying nonetheless: a capacitor exploded on my Newtek Video Toaster immediately after applying power to the system.

I also encountered nearly insurmountable problems actually mounting the Flicker Free Video to the Denise socket. Small cutouts on the card, presumably designed to allow it to fit around large motherboard components (such as capacitors), just do not align in the slightest fashion to the topography of the Amiga 2000 motherboard. This made it extremely difficult to mount. I was able to get it to work somewhat by adding a second socket to the motherboard which provided a bit of elevation, but it was still a really dodgy install.


Once “mounted” (and after the smoke had cleared from the Video Toaster), I was able to get a really nice picture on an attached LCD monitor. After reassembling the A2000 however, the FFV ceased to output a usable signal. While troubleshooting this, a pin snapped on the makeshift socket booster I’d cobbled together, so all further investigation is now off the menu until another socket arrives via Digikey. On the bright side, the Amiga still seems to work normally using the traditional monitor output.


Knowing that ICD actually sold this card as an A2000 compatible part, I really wonder what they were thinking when it came to the design of the board - it’s just crap. Maybe it aligns okay on an A500, but given a choice, I’d recommend an external scan doubler solution over this one. There’s much less pain involved.

The Mysterious APX-Z80BD-1

| No Comments

This is kind of an odd one.

Recently while browsing around eBay, I ran across a vendor selling large quantities of Apple II compatible CP/M cards for $15.99 each. I already had an Advanced Logic Systems (ALS) Z-Engine card which works just fine, but I figured for the price it couldn’t hurt to grab another. Maybe I could even convince it to work with the IIgs!

The oddity started a couple of days after I’d made the “Buy It Now” purchase. I’d already received confirmation from the seller that the card had shipped. Out of the blue, eBay sends down a “SHMO Notice” advising me that the auction had violated eBay policy, had subsequently been removed, and I was “not required to complete the transaction”. Sure enough, all signs of the auction had been scrubbed from eBay’s database. Personally, I saw nothing in the auction’s description or content that looked out of place - other than a reasonable price for retro Z-80 card.


Despite the doom and gloom, the card arrived this afternoon safe and sound. In addition to the card itself, it shipped with a three-page typewritten manual describing basic installation and use, titled “Z80 CP/M Board (APX-Z80BD-1)”. A quick search on Google for more APX-Z80BD-1 information or history yielded nothing.

The APX-Z80BD-1 could be a limited run, hobbyist produced card; there’s no obvious branding or marking other than a silkscreened “Z-80” label. It appears to be a well made, nearly identical clone of the Microsoft SoftCard in both layout and function. Most of the chips - including the Z-80 CPU - are manufactured by GoldStar, while the card itself is silkscreened as being made in Taiwan.

It also has an interesting row of four DIP switches which unfortunately aren’t documented in the supplied manual. I later found these to be a replica of the SoftCard’s DIP switches, and most likely have the following functions (from the Microsoft SoftCard Software & Hardware Details manual):

  • S1-1: Address offset (when off)
  • S1-2: Z80 DMA enable (when on)
  • S1-3: Non-mask int. (when on)
  • S1-4: Z80 interrupts (when on)

The normal operational setting for all four switches is ‘off’.


Another hint that this may be a fan produced project is this entry from the final page of the documentation:

“A CP/M DOS is not included with your board, as this would raise the price of the board tremendously.

Should you find a need for the CP/M DOS, check with your local or a large Apple computer club library. Almost all clubs have a few, if not many, public domain CP/M diskettes that have the operating system on them.

The board you have just bought is MICROSOFT compatible. That means the normal CP/M programs that run on a normal Microsoft board will see and use your Z80 CP/M board as a replacement.”

But most importantly, does the card work? Absolutely! I installed the card in my Platinum IIe (using slot 4 as recommended in the manual), and booted Applied Engineering’s Apple CP/AM 4.1 successfully, just as with my current ALS Z-Engine. I was also able to get the card running on my IIgs as well, though I had to first set the system speed to “slow” in the Control Panel in order to boot CP/M without freezing.


I’ve left the APX-Z80BD-1 in my IIe and installed the Z-Engine in my IIgs; the APX-Z80BD-1 is a lot longer than the Z-Engine and I worry that it might smack into the somewhat bulky TransWarp GS in slot 3 and short something out.

For a low cost, compatible CP/M card for the Apple II, you just can’t go wrong here. As mentioned previously, I’m not sure what the malfunction is with eBay, but Atlaz Computers does have them listed directly on their web site for $15.00 each if anyone’s looking for one.

Briel 4MB GS RAM


This past weekend I set aside a bit of time to install some additional RAM into my IIgs in preparation for a new CFFA and GS/OS 6.0.1 upgrade in the coming days.


I’ve been shopping around for a 4-8MB RAM card to replace my old school 1.5MB Applied Engineering GS-RAM card for a couple months, and finally pulled the trigger on Briel Computer’s 4MB GS RAM card. There are a couple existing vendors of GS RAM cards out there at the moment, but Briel’s offering can be had for $55 via their eBay store, less than a third of the 8MB Sirius RAM IIgs’ hefty $179 price tag. Considering most software won’t take advantage of more than 4MB RAM (not to mention disabled DMA for anything above 4MB), the allure of having the maximum possible RAM in my GS just wasn’t compelling enough.


The Briel card is quite cute, opting to use what appear to be standard SIMM modules rather than the discrete RAM chips found on most GS RAM expansions. This greatly reduces its footprint.

Installation was a breeze - I simply popped the top of the GS, removed the old Applied Engineering card and shoved the new Briel card in its place. A quick reboot and a glance at the memory control panel confirmed the new RAM was fully recognized and operating as expected. Fortunately, there do not appear to be any conflicts with the existing TransWarp GS card.


I’ll report more once the CFFA arrives - that’s when the fun should really start for this particular machine.

Happy 1050

| No Comments

In addition to playing around with retro computers, I enjoy installing various aftermarket upgrades which push the technology in interesting directions. A few such upgrades have started trickling in over the past few days, and more are still on the way. This evening I decided to crack into the installation of a new Happy 1050 upgrade for my Atari 1050 floppy drive.

The Happy 1050 is a small circuit card containing a new 6502 processor, PROM, RAM and support logic which plugs directly into the 6507 socket on the 1050 drive’s logic board. It provides enhanced functionality, including true double density operation, sector buffer cache, accelerated transfer speeds and MS-DOS floppy read/write support. The Happy upgrade was originally produced and sold by Happy Computers in the mid-eighties, but new units can still be purchased from AtariMax today.


The card shipped in an anti-static bag which also contained a Happy R7.1 software diskette, a letter with download URLs for the Happy manual/installation guide, and a page of advertising for other AtariMax products. The installation guide and Happy software manual PDFs must be downloaded from the AtariMax website - no soft or hard copy is provided.

The installation procedure was straight forward and nicely documented in the installation guide. I opened the drive chassis and removed the drive mechanism, logic board and RF shielding covering the CPU and ROM chips. At this point the old Tandon PROM and 6507 CPU were pulled from the logic board, and the Happy card installed in the 6507 socket (the logic board’s PROM socket is left vacant). Finally, after checking and double checking the cable connections (there are seven cable connections between the drive mechanism and the logic board with no orientation keys), I closed up the chassis.


I fired everything up and was greeted with … nothing. The power LED on the 1050 lit up, but the drive did not seek and was otherwise silent. As far as the Atari 800 was concerned, there was no drive attached at all - it “happily” dropped into the Memo Pad.

I became a bit concerned, having read the warnings in the manual on the various and sundry ways I could potentially fry the drive by not connecting the cables in the correct orientation or location, or by improperly installing the card itself (missed or bent pins, oriented backwards, etc.). I’d checked my work throughly - the only thing that stood out as a potential problem was that the pins from the Happy 1050’s onboard RAM chip protruded from the underside of the card enough that they hit the top of one of the existing chips on the 1050’s logic board. This caused the upgrade to sit in its socket at a slight tilt.

I removed the Happy 1050, and using a pair of pliers, carefully bent the row of pins on the bottom of the card off to the side. Once reinstalled, the upgrade seemed to fit into the socket a bit more securely. I did a test run without the drive’s cover installed just in case.

Presto! The drive sprang to life and booted the Happy 7.1 diskette. The software correctly identified the drive as a Happy 1050, and the diagnostics routines passed.


In closing, if you decide to pick one of these up you may need to mod it slightly if any of the chip pins aren’t allowing it to fit flush in the socket, either by bending them aside or trimming them down. Other than that caveat, it seems to be a great little upgrade and I’ll be playing with it more as time allows.



This past Friday, my latest SGI workstation arrived - a rather nifty Tezro.


I purchased the machine on eBay, knowing going in that there were a couple of caveats. For one, the seller didn’t know much of anything about the machine, so details such as the amount of installed memory, number and speed of CPUs - or whether or not it even contained a hard drive - were completely absent. The other rather large issue was that it was clearly missing the IO9 PCI card, which provides essential services such as the TOD clock, NVRAM, SCSI, IDE and console. This is probably the one thing that kept the final price so low - there was undoubtably a certain level of uncertainty as to whether or not the machine was even fully functional.

However, a little knowledge of this type of system can go a long way. I already had the IO9 situation covered as there were two of these cards installed in my Origin 350 system - you only need one if you’re running two bricks NUMAlinked together. The other helpful tidbit was that the high quality photos provided in the auction description clearly showed an eight PCI-X slot system, which meant at the absolute minimum spec, I was looking at a dual 700MHz machine. Knowing this, I bid based on how much such a system would be worth to me, and if I scored something better - well, that would be pure gravy.

This turned out to be a sound strategy. What I received in the end was a quad 700MHz machine with 3GB RAM, 18GB hard drive, DVD-ROM, DMediaPro DM3 and a Dual Channel Display option (DCD) on the V12 card.

There was a little bit of work involved getting it all up and running. First, the IO9 needed to be installed. I found the system’s IO9 retention bracket floating in the bottom of the case, which actually was quite nice - now the IO9 is securely installed just like a factory install. All the drive cables were intact, though the cable end at the SCSI drive bay was unplugged. I had to fiddle around in cramped quarters to get that reattached.

The boot drive was a stock SGI firmware 18GB Cheetah, with a base install of 6.5.23 and no root password. The drive was installed in the wrong SCSI bay; it needed to be moved down to the first (lower) bay. Finally, the side access panels were installed backwards with the push button latch releases facing towards the front, which makes them completely inaccessible. It was a bit tricky figuring out how to remove the panels in that configuration - I wound up tugging on the metal locking tabs with needle nose pliers to release the latches.

Once all the gremlins were chased down, cables reconnected and everything tightened back up, it was good to go!


Finally, I moved over several items from my Fuel, including PCI cards (SAS/SATA, U320, FireWire and an extra gigabit network card), hard drives and an additional gigabyte of RAM. So far I’m loving it - I’ve been playing with it as much as possible over the past couple days.


Atari 800

| No Comments

While on vacation, I made certain to periodically check the Bay Area Craigslist, just in case some interesting hardware bubbled up in my absence.

Sure enough, a couple things surfaced. The first: a complete NeXTstation mono slab system (including the laser printer) - free for pickup. I was apparently too late to capitalize on that one though as my e-mail inquiry garnered no response.

The second was an Atari 800 (listed as a 1200XL), with 1050 diskette drive, manuals, programming books and software (original diskettes for Zork, Atari DOS, SynFile and PrintShop, with BASIC and AtariWriter on cartridge). I grabbed it just this morning from San Francisco (my second Atari system from that city), all for $25. I can honestly say I’m getting better at navigating around San Francisco. Scads of retro computing hardware seems to crop up there, so I’ve been getting a lot of practice!

Upon testing the machine, I only ran into a couple of minor snags. The system came with an old-school RF converter for video output, which isn’t really something I can use on anything modern. Vexingly, there are no Atari monitor cables on eBay at the moment.

Fortunately, the DIN to composite cable I’ve been using on my TI-99/4a also generates a usable picture on the Atari 800. The image quality should improve considerably if I can source a Atari DIN to S-Video cable. I’ve fired off an e-mail to a vendor that’s apparently offered them in the recent past. (Update: I’ve been informed they will have the cables back in stock the first week of September.)

After temporarily resolving the video problem, my first functional test was to boot up Zork from diskette as shown below:


It was at this point that I discovered the keyboard was largely unresponsive. I wound up exercising each key individually before they began to work reliably. It took a bit of time, but all the keys are now fully operational.

Here’s a shot of Atari DOS 2.0S - it reminds me a lot of the Apple ///’s SOS with a menu based approach rather than the “traditional” command line shell. From what I’ve been able to gather online, I apparently have an early 1050 drive which shipped with Atari DOS 2.0S instead of the newer 3.0 (or even 2.5) supplied with most 1050s.


I’d never really played extensively with an Atari 8-bit machine before. I find this to be one of the coolest aspects of retrocomputing - it affords me the opportunity to toy with hardware I’d only read about while growing up. I did have a childhood acquaintance (Aaron Osborne or Darryl Schroeder) who owned an Atari 400, but my somewhat fuzzy recollection is that he used it mostly as a cartridge based game system. I don’t recall him having a diskette drive, though he probably had cassette.

Coincidently, there was a recent Slashdot article on the early history of Atari which mentioned they’d fired “a young programmer named Bill Gates” who was apparently slow to produce a BASIC for this machine. Atari BASIC was ultimately produced by Shepardson Microsystems, Inc. (SMI) rather than Microsoft.


I’m looking forward to searching out interesting software for the 8-bit Atari and putting it through the paces. Earl Evans of the Retrobits Podcast had some glowing things to say about Star Raiders and the Atari 800 system in general, devoting podcasts to both (shows 11 and 82). I’ll need to put Star Raiders on the lookout list.

Eight inch disks are both a blessing and a curse in the retro computing world. They’re awesome in that for the time, they stored a generous amount of data - even a single density double sided 8” diskette could hold half a megabyte, which was impressive considering early microcomputer hard drives (such as Apple’s ProFile) topped out at only 5 megabytes.


The problem with eight inch diskettes today is not so much the disk capacity, or even obtaining blank media (it’s still surprisingly easy to find inexpensive media thanks to IBM’s support of the format long after the microcomputer industry had moved on). The issue is how to get new data on one of these beasts in this day and age.

Some back story is in order. One of my recent Craigslist fueled acquisitions was an absolutely beautiful TRS-80 Model II with all the trimmings. The Model II is a very rare backplane based Z-80 system in the spirit of the S-100 style machines - separate cards for disk control, CPU, video, etc. situated on a “dumb” backplane coupled to a eight inch diskette drive.

The “trimmings” I mentioned included everything from original media, manuals, cables, books - even an unopened box of original TRS-80 Model II diskettes still in their shrink wrap! Despite the glory of the haul (all for $20), I soon ran into a rather large issue.

The problem became apparent as I tried to boot the system from the original TRSDOS floppy. After several unsuccessful attempts, I discovered that the magnetic coating of the diskette was literally wearing away where the read heads had touched it. You can clearly see the ring patterns in the shot below.


Of course at this point I cleaned the drive heads, but it seems this phenomenon is more due to the age of the media than any abrasive residue left on the heads themselves.

Afterwards, I found a backup TRSDOS floppy in the morass of diskettes that came with the machine which did boot successfully. No problem - simply copy the working TRSDOS diskette to a fresh blank, right? Unfortunately, no - for some reason, even though the diskette seems to boot and operate just peachy, it won’t copy; errors out with a message that the boot sector can’t be read. It too is probably on its last leg, and I’m reluctant to use it much considering the same fate could easily befall it. The Model II is useless without a bootable OS diskette - it has nothing in ROM to fall back on.


Now we get into the nitty gritty. There is a way to get fresh TRSDOS floppies, even a newer version than the one I have now. Also, a CP/M by Pickles & Trout is readily available. The catch? The software is only available on disk images that must be downloaded and written out to eight inch disks via a PC using MS-DOS. Yuck.


I did put out some feelers to see if I could locate someone who had already assembled such a system, maybe to create media for S-100 systems (or even a Model II). Unable to find any leads, I resigned myself to assembling my own imaging system.

Some of the first parts have begun to trickle in. I picked up a couple new boxes of eight inch floppies off eBay for $15 per box of ten. I also purchased a throughly nifty eight inch drive adapter card from D-Bit that should make hooking up an eight inch drive to a PC a much easier task. Oh, and I bought a Shugart 801 diskette drive similar to the one found stock in the Model II.


What remains is to locate a PC with a compatible floppy controller chipset to do the grunt work and some sort of external chassis or power source for the eight inch diskette drive. As I work through these remaining issues I’ll post updates on them. Should be fun, and hopefully once this is all up and working I’ll be able to offer pre-made floppies to others so they won’t necessarily have to go through the same pain themselves.

And now I’m off on vacation for a week to take a break from the grind at work, so blog posting will most likely halt for a bit. Until then, happy computing!

Osborne 1

| No Comments

Over the past few days I’ve picked up a couple “new” machines - an Atari 1040ST in perfect condition with all manner of software and accessories (more on that one as I explore further - lots of things to look at) and just this evening, an Osborne 1.


I picked up the Osborne in Morgan Hill form a Craigslist seller with the understanding that the floppy drives weren’t working. After my recent experience with the TRS-80 Model III I figured it could just be a power supply. Since the machine seemed to power up okay otherwise I wasn’t too worried - whatever it was would most likely be an easy repair.

Once I got the machine home and started troubleshooting, I discovered a stuck key was blocking it from booting - it won’t perform a disk seek unless you hit ‘enter’. This also explained why the screen would flash a few times after power up - the stuck key was forcing screen redraws until the keyboard buffer filled up. Exercising the keys a bit cleared the stuck key and everything now checks out. I also took the opportunity to fully disassemble the unit and “de-bunny” it as Dr. Dave would say.

The machine also came with a small collection of diskettes in the right-hand storage compartment, including copies of Osborne’s CP/M and WordStar.

All in all it seems like a pretty cool machine, though the display is impossibly tiny - I imagine quite a few Osborne fans opted for an external display when portability wasn’t a concern. After using it for a bit I feel Kaypro certainly had the advantage on the suitcase form factor with a useable screen size.


Wanted to point out that edefault has purchased another set of PMC RM7000C-600T CPUs and is gearing up for another run of 600MHz O2 modules. You’ll need to supply your own RM5271 module for the conversion.

More info can be found in this thread.

I’ve one of edefault’s conversions in my clear cased O2 and I couldn’t be happier with it. The image below is of my own board - the “blue wire” mentioned in the thread can be seen on the module toward the bottom.


ROM Upgrades

| No Comments

One of the projects I’m working on as funds and time permit is the restoration of a Neo Geo arcade cabinet.

An upgrade I performed not long ago was the installation of a Universe BIOS ROM. This handy little upgrade enables all manner of cool features on the Neo Geo hardware and is really something no Neo Geo enthusiast should be without; if you already have an EEPROM programmer it’s even a free download. Not having such I beast myself, I opted for the pre-made chip.


The chip arrived on the same day as a new Kickstart ROM for my Amiga 2000 (though from a different source), so it became quite the chip swapping event!



Next on the agenda is to install new joysticks and buttons as the ones I have now are pretty worn (they work, but they’re a bit sloppy). For sticks I’m thinking Sanwa JLF-TP-8T or JLW-UM-8; still undecided on whether to go with bat or ball style handles. Undecided on the buttons too, though the Seimitsu PS-14-KN looks interesting.

As always, more to come!

Of Floods and Tritorn Final


As retro machines get older, preserving software becomes more important as the media isn’t often as long lived as the machines themselves. Diskettes can be destroyed by ordinary wear and tear, the environment they’re stored in, and even extreme weather such as hurricanes or floods. But sometimes not all hope is lost. Today we’re going to look at reviving flood damaged diskettes.


I’ve gradually been collecting “new” software for my Sharp X68000 XVI; it’s usually fairly difficult to find titles on these shores - most things need be imported directly from sellers in Japan. Earlier this year I found a series of auctions for X68000 software in the United States that had been in a basement flood and were being sold as non-working scrap for pennies.

I bought a few of these packages, and in most cases the water hadn’t made it into the sturdy plastic boxes that most X68000 software ships in, so only the paper slipcover inserts looked a little ratty. One title was throughly damaged though - Tritorn Final.

The game packages had been stored in boxes with biodegradable packing peanuts made of starch. When these peanuts get wet, they break down and become a sort of goo that gets into everything. The Tritorn disks, though dry when I received them, were completely fused in their jackets due to the starch.

I now wish I’d taken a few photos of the actual process - just didn’t think it of it at the time - but the first step was to get the diskettes wet again so they could be separated from their jackets without damaging them further. After a soaking in water, I used an Xacto knife to cut the welds at the top of the disk jacket, and while wearing gloves, carefully removed the inner mylar disks. Next, I throughly rinsed and placed the mylar disks on a clean, lint free micro fiber towel. Another micro fiber towel was placed on top to guard against contamination.

Once the mylar innards were removed, the disk jackets were throughly rinsed, then stuffed with bubble wrap to prevent the inner padded lining from bonding together while drying.


After everything was completely dry, I reassembled the diskettes, sealing the top seam with small drops of cyanoacrylate near the locations of the original welds. I also wound up placing tape across the seam as they were getting stuck in the disk drive.


The result of all this pain was a working copy of Tritorn Final as seen in the attached images. At this point if you have means, duplicate the diskettes onto fresh blanks. Unfortunately, some titles are copy protected which can often make this difficult.




(Note: I’d originally posted this in the forum, but it’s probably better suited as a retro computing blog entry. I may have one of two more like this I’ll bring over before we get into some of the newer stuff.)


One evening toward the end of June I ordered a veggie pizza and buckled in for some fun installing Apple III Pascal on my newly acquired ProFile drive. It was kind of an interesting process, so I figured I’d share it here.

The following is based on an OCR’d scan I did of the relevant entry in the ProFile documentation:

The Pascal Language System

Adding Pascal to Profile Drive

The following information is intended for those persons whose programming requirements are primarily based on the Pascal language system. The files making up the Apple /// Pascal System are supplied on three diskettes identified as PASCAL1, PASCAL2, and PASCAL3. The following table lists the system files found on each diskette.







To eliminate the inconvenience of having to repeatedly swap the Pascal diskettes to access the different files, you can transfer the files from the Pascal diskettes to your ProFile drive as follows:

Note: The Pascal system must be booted with your boot diskette in the built-in drive. You will find it easier and more convenient to perform the other steps of this procedure if you have an external drive such as a Disk III, or a Disk II for Apple ///, attached to your Apple ///. If you do not have an external drive, all of the steps can be performed on your built-in drive, but the operation will require more frequent exchanges of diskettes.

  1. Make copies of PASCAL1, PASCAL2, and PASCAL3 using the Apple /// Utilities diskette.

  2. Use the Disk Format Utility on the Apple ///’s Utilities diskette to format a new diskette called PROFILEPASCAL.

  3. Use the Systems Configuration Program (SCP) on the Apple ///’s Utilities diskette to add the .PROFILE driver from the ProFile device driver diskette to the SOS.DRIVER file of the PASCAL1 diskette as described in Chapter 3. Write the combined driver file onto the PROFILEPASCAL diskette as /PROFILEPASCAL/SOS.DRIVER. (It is necessary to write the file onto the PROFILEPASCAL diskette because there is not enough space on the PASCAL1 diskette.)

  4. Place the copy of PASCAL1 you made in step 1 into the Apple ///’s built-in drive and boot the Pascal system.

  5. Enter the Pascal Filer and perform the following transfer commands to transfer files from the PASCAL1 diskette to the PROFlLEPASCAL diskette:





  1. Remove the PASCAL1 diskette from the built-in drive and insert the PASCAL3 diskette.

  2. Enter the Pascal Filer and perform the following transfer.


  1. Remove PASCAL3 from the built-in drive, insert your PROFILEPASCAL diskette, and reboot the Pascal system.

  2. Place the PASCAL1 diskette in an external drive.

  3. Enter the Pascal Filer and transfer the SYSTEM.FILER from the PASCAL1 diskette to your ProFile disk drive using the following transfer commands:


  1. Quit from the Filer.

  2. Remove the PASCAL1 diskette from the external drive.

  3. Enter the Filer — this will execute the copy of the Filer which you just recorded on ProFile.

  4. Place the PASCAL2 diskette into the external drive and use the following Filer Transfer command to copy all of the PASCAL2 files onto the ProFile drive:


  1. Place the PASCAL3 diskette into the external drive and use the following Filer Transfer command to copy all of the PASCAL3 files onto the ProFile drive:


After completing the above steps, your ProFile drive will contain all of the necessary Pascal files, and PROFILEPASCAL will be the only diskette you need to boot the Pascal language system. This results in the following advantages to a Pascal user:

• faster program initiation

• minimum ProFile storage space is needed (approximately 175 kilobytes)

• convenience

• less shuffling of diskettes.

Detailed information on the Apple /// Pascal system can be found in the following three manuals:

Apple /// Pascal: Introduction, Filer, and Editor Apple /// Pascal: Program Preparation Tools Apple /// Pascal: Programmer’s Manual (Volumes 1 and 2)

Whew! It does work perfectly if you follow the instructions to the letter, but we definitely have come a long way in ease of installation. It’s just amazing how much complexity a presumably novice computer user was expected to perform 25 years ago.



Power to the TRS-80 Model III

| No Comments

A couple weeks ago I picked up a nice condition TRS-80 Model III from a fellow in Antioch via a Craigslist ad. The seller was also the original owner and had used the machine in high school. That’s usually a good thing to hear when picking up retro machines as they tend to be in better than average condition and often have more accessories such as documentation and software.


When I arrived he had the machine set up on a workbench in his garage. A prompt, “Cass?”, was on the display. Everything seemed to be in order, so I gave him his cash and loaded the machine, a couple manuals and a nondescript box of diskettes into my car.

When I got home I ran into a couple minor disappointments. For one, the box of diskettes was actually old MS-DOS software from the late eighties/early nineties. The other was that the “Cass?” prompt was due to the internal disk drives not being recognized - it was in fact prompting for a cassette. Indeed, no power seemed to be getting to the drives at all.

I confirmed this by checking for power on the floppy drive molex connector and traced it back from there to the 35 watt ancillary power supply that runs the floppy logic and drives.

My first thought was to simply recap the supply and go from there, but as luck would have it, I found a vendor on eBay (computernews80) that had new “old stock” 35 watt power supplies suitable for use in both the Model III and Model 4 at a reasonable price. While I was at it, I ordered a few additional items from the same seller - TRSDOS 1.3, LDOS 5.3.1 and LSDOS Systems Utilities/File Utilities.

Shortly after ordering those goodies, I discovered a couple TRS-80 web sites of interest. The first, and perhaps most important to those without the means to create floppies from disk images, is Ira Goldklang’s TRS-80 site. He offers a service where you can send in a donation and receive premade floppies of several TRS-80 operating systems. I was graciously provided diskettes for DOSPLUS 3.5, MULTIDOS 1.6, MULTIDOS 4.01, NEWDOS/80 2.0 and Super Utility 3.2. He also provided a couple additional diskettes with some nifty games software.

The diskettes from both sources and power supply all arrived on the same day. After a quick removal and replacement operation, I fired up the Model III and the drives sprang to life!


I was a little concerned at first as the Model III apparently doesn’t display an on-screen message if it can’t find a bootable diskette - it just sits there, screen completely blank. After inserting a bootable diskette into the drive and pressing the reset button, I was in business.

Below is a photo of the failed power supply - after removal it was pretty obvious that one of the capacitors is a little burnt. There’s no residual odor so I’m guessing the damage didn’t occur recently. I’m planning on recapping the supply and keeping it around as a spare.


One interesting thing about this particular Model III is that both floppy drives are in a single bay, leaving the top floppy bay free for something else. Free for exactly what, I’m uncertain - but hey, it can’t be a bad thing!

The games that Ira sent along are actually quite good; I haven’t played them all yet, but the ones I tried were fun, simple arcade games in the spirit of Space Invaders.

It should be noted that if you do have the means to create physical floppies from disk images, there are several websites that offer all of the operating systems I noted above - and more - in downloadable disk image format. This requires a PC with a compatible floppy controller and disk drive. I myself don’t currently have such a machine, though it’s becoming evident that I’ll need to put something together, particularly if I hope to get my Model II up to snuff. It craves Pickles & Trout. (The Model II itself is another story I’ll relate in a future entry).

Some sites offering disk images for the Model III:

Timm Mann’s TRS-80 Pages - Offers an incredible amount of technical software including LDOS, compilers, assemblers, utilities, etc. All of Misosys’ catalog has been released to the public, and there are a lot of awesome things in there I’d love to try out.

System 80 - More interesting things from games to word processing.

Dave’s Old Computers - Disk images for various vintage machines, including the Model III.

  • Note: If you do need to get back to the “Cass?” prompt (to use the built-in ROM BASIC for instance), just hold down the BREAK key while hitting reset.


Of TransWarp and Loose Oscillators


I had some (unexpected) fun installing an Applied Engineering TransWarp GS card in my Apple IIgs.

I bought the card on eBay after searching for a few months - most TransWarp GS cards were going for amazing amounts of cash. Fortunately I lucked out and got a hold of a semi-reasonably priced “Buy-It-Now” auction before the hordes descended and bid it into the stratosphere.

The card arrived late last week, and I eagerly unpacked it and took a photo for posterity:


I was pleased to note it wasn’t simply a stock card, but had in fact been upgraded with new parts (presumably) from ReactiveMicro - a 32K cache board, high-speed GAL set, etc. Of course, despite noticing the shiny upgrades, I failed to notice an oscillator was missing from its socket.

After repeated attempts to get the card working met with utter failure, I was stumped. Could it be the upgrades had essentially tailored it to the system it was removed from? Was my machine pickier about timing? After perusing the ReactiveMicro site some more, I noticed all the various oscillator speeds they offered. Then it hit me - where’s mine?

I shook the anti-static bag the card had shipped in* and out fell a mangled ECS-2100 35MHz oscillator, its four legs hopelessly crumpled. My attempts to straighten them only led to one breaking clean off.


After an order to DigiKey and a few days, the replacement 35Mhz oscillator arrived in a ridiculously oversized box filled with copious amounts of packing material. While installing the new part, I made certain to securely fasten it to its socket via a small tie-strap.


I reinstalled the card in the machine, now nervous that the repeated removal and replacement of the CPU may have caused excessive socket wear. After flipping the power switch once again, I was greeted with a cool warp sound effect and animated high-res graphics title screen. The card lives!

With all that aside, it’s a really nice card - the speed boost is noticeable just tooling around in the Finder. My next goal for the system is to outfit it with a compact flash “hard drive” card, either by ReactiveMicro or R&D Automation. I’d considered simply going with a Apple High-Speed SCSI card, but they tend to go for a lot of cash (about the same as a CF solution).

Once I have a mass storage solution in place I can finally install System 6.0.1 and have some fun :)

  • (You can actually see the oscillator still in the anti-static bag in the card photo above - it’s right below the ribbon cable!)

CP/M on the Apple III


To avoid cluttering up the forums with my retro computing fluff, I decided to reclaim my blog and actually use it for something - it’s been a year since I’ve touched it. I’ll still post the occasional IRIX tidbit as they come along, but right now I’m having a lot of fun with retrocomputing in general. For those that come here just for the IRIX news items, you should be able to collate them via the category options.

On that note, I finally managed to chase down the CP/M diskette for the Microsoft SoftCard III.

After buying a bare SoftCard III on eBay last May, I’d put out a call on the Vintage Computer Forum hoping to find a diskette/disk image/what-have-you, but came up empty. Unbelievably, a factory original diskette showed up on eBay, and I was even able to grab it for less than $10.


The diskette boots just fine:


The next task is to figure out a way to backup the diskette and create an image file. SOS didn’t want anything to do with it when I tried to duplicate it that way; ditto for the Apple IIgs (it just wanted to format it). I’m going to try ADTPro but first need to build or buy the proper serial cable.

If anyone has any suggestions on the best method to copy and archive this diskette, please let me know.

Resurrecting an Onyx2

Joe Gill recently posted a link on Usenet for his newly created weblog detailing his efforts on getting a Onyx2 machine up a running, complete with pictures and part numbers. The machine apparently isn't functional yet, but it's already shaping up to be quite a nice documentary!

"New" SGI Gear

| 1 Comment

The past couple of weeks have seen the arrival of quite a bit of interesting SGI related gear at the residence. One of the first arrivals was the ACARD ARS-2000FU Ultra SCSI-IDE bridge I mentioned last month. I have to say the bridge has been everything I hoped for; extremely easy to install and use. It's amazing how much storage capacity this gives you for such little cash and the performance is wonderful. I can't say enough good things about these ... in fact, I'm going to buy a couple more of them very soon (one for the Indigo(1) and another for the second bay in the Indigo2). After setting the SCSI ID on the bridge and connecting everything up it was a simple matter to clone my existing system drive over. I used the "Cloning A Root Disk" section of Ian Mapleson's Disk and File System Administration guide, opting for the xfsdump/xfsrestore method outlined in that document.

Another item I received this past week: a Maximum IMPACT TRAM upgrade kit I'd "won" on eBay. Installation was a bit more challenging since I'm not familiar with the Maximum IMPACT board layout ... I initially neglected to install a TRAM module on the boardset's middle tier which resulted in missing scan lines in textured surfaces. A quick search on the comp.sys.sgi newsgroups helped to clue me in and my next test was successful. I must mention that it took quite a bit of pressure to fully seat the TRAM module on the middle tier though ... the topmost TRAM was very easy to snap into place in comparison.

One of the most exciting new bits I received: a cable set for the Galileo Video/Cosmo Compress pair installed in my Indigo(1). I was finally able, at long last, to loop my Cosmo card through the Galileo as well as connect up my ABOB! I've had everything save the cables for about a year now but haven't been able to do anything with them aside from use an IndyCam.

I tried a few test captures earlier this week and am pleased with the results thus far; I've been sending the captured video over the Indigo's integrated 10Mb ethernet (slow!) to the G4 for playback and/or editing. I'm planning on experimenting with the Galileo/Cosmo some more this weekend, trying out some of the various quality related settings to see how it effects the end result.

Finally, as a result of having a nice video capture solution operating on the Indigo(1), I chased down a Phobos G130 100Mb ethernet card, brand new in box through B&B Solutions (via eBay, cost was $42 US before shipping). This should make moving captured video over to the G4 much less painful.

By the way, I have an extra ABOB for Galileo available if anyone is looking for one.

Reawakening the Indigo2


ars-2000fu.gifI mentioned a while back that I'd begun work on restoring my newly acquired Indigo2 IMPACT R10000 system; way back in April I'd picked up a Maximum IMPACT boardset from Pastech, 3.5" & 5.25" drive sleds and a 32x internal CD-ROM from Used Tech Dot Org (formally Aftermath Technologies), a 195MHz R10000 CPU from Reputable Systems and a 100Mbit 3Com EISA 3c597 card (essentially a Phobos E100) purchased from an individual seller. The system has languished since then; the current system drive is too small to do much with (6GB) and my attempts to locate the TRAM option to fully expand the Maximum IMPACT have met with absolutely no success despite a couple postings to comp.sys.sgi.marketplace pleading for help.

The system will once again be receiving some attention; I've just arranged to purchase a Video for IMPACT boardset (essentially Galileo (ev1) for IMPACT systems) and will be picking up at least one ACARD ARS-2000FU Ultra SCSI-IDE bridge to utilize low cost high capacity IDE drives as I happen to have two 60 GB drives I'm not using for anything else at the moment.

The ARS-2000FU is a small tray like device that attaches to the underside of a 3.5" IDE drive, transforming it into a SCSI unit which will still fit the drive sleds of SCSI based machines like the Indigo2. After searching online, I discovered Microland USA resells these units for $80.00 each.

I should mention that these upgrades have freed up a couple of components if anyone is interested, namely a Solid IMPACT board (SGI PN 030-0786-002; would be great as a second head if you have a 060-0027-xxx power supply) and a 175MHz R10000 CPU module (SGI PN 030-1433-002).

The Ancient Computer

NOTE: This entry is already obsolete as I was forced to retire the Indigo after a series of hardware failures left it severely crippled. Ryouko will be missed. My Indigo2 IMPACT, Mika (named after the lead character in the anime Geneshaft), is now running the site.

It's quite amazing really, running a web site (among other things) on a computer that's now ten years old. That's right, the Silicon Graphics Indigo Elan R4400 powering this website was introduced in early 1992 (then stocked with a 64-bit, 100MHz MIPS R4000 RISC CPU). I suppose you could argue that this particular configuration formally turns a decade old in 2003 (which is when the 150-MHz MIPS R4400 that now powers the mighty Ryouko was released as an upgrade option for the R4000), but either way it's far removed from what we now consider an effective service life.

The Indigo At Work: A rack of Indigo machines at Fermilab. The Indigo At Work: Rack of Indigos at Fermilab during the 90's.

I've given serious thought to replacing the machine with something more modern, either a newer IRIX system or perhaps even a Macintosh running OS X, but for some reason I can't bring myself to do it despite having hardware for both options available to me. I derive a sort of obscure pleasure knowing that old, "obsolete" hardware which goes for pennies on the dollar on eBay and increasingly populates dumpsters can still perform serious tasks and perform them well. How many webservers are still out there running on Indigos anyway?

I am incredibly impressed by SGI's continued support of the Indigo. To this day, the latest and greatest versions of IRIX still run on it, and add on cards like Galileo and Cosmo Compress still receive maintenance updates. I often wonder if there are individuals at SGI that have a fondness for the machine and keep the support alive for nostalgia, or perhaps there are still large customers running Indigos someplace (an intriguing thought!) Especially considering the Crimson lost support as of IRIX 6.2 and contained the same R4400 CPUs, and even further back, the Indigo R3000 lost support as of IRIX 5.3.

Perhaps more prophetic than intended ...

Perhaps more prophetic than intended ...

I now have a more powerful SGI machine, a fully loaded Indigo2 R10000 MaximumImpact which I could press into service if need be. But for now I think I'll keep plugging along with the machine that has served me well in the 20th century and continues to perform equally well in the 21st.

About this Archive

This page is an archive of recent entries in the Retrocomputing category.

Mac OS X is the previous category.

SGI News is the next category.

Find recent content on the main index or look in the archives to find all content.