[Contact Me]
--={ [C64/128 Info]
[C128 Tower]

Valid HTML 4.01!

Increase your website traffic with Attracta.com
Welcome to my C128 Tower project page.

[System current state]

This page is intended to more or less document the building of my fourth C128 tower, with the final state shown to the right. This tower has since been disassembled so that its components can be re-used on one of my other C128's.

This was the final iteration of the tower project, which has lasted 12 years. Scroll down for more information and plenty of photos. Click on any of them to see the full-sized photo. You will need to enable pop-ups and Javascript for this page, so that the high-resolution image viewer can run.

For information on my previous towers, try these links:

Version 3, Built over the course of November 2005 to December 2009.
Version 2, Built over the course of August 2001 to April 2005.
Version 1, Built sometime in 1997 or 1998.

May 10, 2010

New LED/Power control board put into service.

[Actual prototype] Originally I had built a two-part LED/power/audio control circuit for my last tower, but now it bears a properly-made PCB thanks to the fine folks at Olimex. This board also incorporates a few minor changes that were made to the two-part board (where I adapted it and my internal cabling to be closer to industry standards). In addition, this unit has already had a few tweaks made to correct a couple of minor design errors. See the corresponding entry on my Projects page for more information.

September 3, 2010

Updates to the control panel.

[Closeup of the revised front panel!]
[A fancy new control panel bezel!]
When I made the control panel label seen in the images below, I did so by first producing an image using The GIMP, then printing it out and covering it over with clear tape. That looked okay, but I thought it could be better. Enter, Gene Buckle, with his laser cutter/engraver! In addition is Inkscape, which I'd been using off and on for a few years anyway, and which I used to re-draw the panel image into a vector format version suitable for use on the cutter (why I didn't go this route to begin with, I'm not sure). Shown here is the result, along with the rest of the front panel: a 1/16" etched and back-painted acrylic bezel that slips right over the switches like it was meant to be there all along. :-) As you can see in the upper image, the CD burner, FD-2000 mechanism, and the handle on the 1571 drive still bear the paint from the previous tower. No big deal, I'll sand then lightly and paint over them at some point. Also in the upper image, I had to paint out a scratch-like mark on the new bezel. It's just a lighting artifact, but it was easier to paint it out than to set up and re-take that shot. :-)

August 16 - 17, 2010

The braided sleeving I ordered from cableorganizer.com finally arrived today. I made a few other tweaks here and there unrelated to the sleeving, and snapped the control panel into its final place in the front panel (see above for the updated picture).

[Lots of braided sleeve!] Needless to say, I spent the rest of the day, the entire night, and part of the following morning sleeving all the unruly wires and cables in the tower. Some of it was difficult to do because, let's face it, this system was not designed with this idea firmly in mind or anything. Some parts were easy though. I still have one or two wires that need to be done, but that'll have to wait - I ran out of the size of sleeve I need! In one or two places, I had already used some sleeve I got from Al, so I left that stuff in place - why change what looks good? :-)
[Better FD2000 heat sink] My FD-2000 needed a better heat sink, so I took the opportunity to fashion one from slightly thicker aluminum than before, cut it wider, and wrapped it into more turns (it was a bit longer of a piece than before). I also replaced the regulator and used a bit of Arctic Silver Ceramique just to be sure (it definitely works!). While I was at it, I replaced the 100uF capacitor near it, touched up some solder joints in the area, and soldered the back of the power jack to eliminate a loose connection.

August 10 - 11, 2010

I built the control panel over the last couple of days. It is far better than the previous attempts, but I'm still not entirely happy with it. To improve on it would require a better printer and some way to cut nice, clean, round holes for the switch caps and LED windows.

[Control Panel Completed]

[Control Panel Under Construction]

In my last tower, I simply drilled a bunch of holes into a knockout panel to receive switches and LEDs. It was functional, and looked okay more or less, but it didn't fit right and I knew I could do better. This time, I'm using two knockout panels, one behind the other. The beige colored knockout panel in the photo at the left is one I had left over from another computer, and is the "carrier" that holds all the switches and LEDs, while the black one, which came with this tower case, will serve as the actual faceplate.

I first cut the carrier plate down just enough to fit it inside the back of the faceplate, glued them together like that temporarily, then drilled the mounting holes for the LEDs and switches using my drill press. After that, I carefully broke the glue, set the carrier plate aside, and very carefully re-drilled and sanded the switch holes on the faceplate to make them big enough to let the black and red caps stick through. Along the way, I also filed off the Commodore logo and sanded the front surface smooth, as it would have created a bulge in the label. On the carrier plate, I mounted each switch snugly but not fully tight, and mounted the LEDs from behind with only the tiniest amount of superglue. Each LED got a piece of heat shrink tubing around the front to shade it from neighboring LEDs and holes. I then screwed five 10mm threaded standoffs to the carrier plate, checked the alignment of switch caps and the faceplate, and superglued the standoffs to the back of the faceplate, while still attached to the carrier.

After the glue dried, I unscrewed the carrier plate, added more superglue to the joints between the standoffs and the faceplate, and alternated between tightening the switches down and checking their alignment with the faceplate. Once assembled, the switch knobs stick out about 1/8", just enough that they sink into the faceplate slightly when fully depressed. The four empty holes are reserved for future expansion.

I drew the label image in The GIMP, starting from a 300 DPI scan of the faceplate as a model, which I then laser-printed onto plain paper. I cut the LED windows with a knife, covered over the front of the label with two layers of clear packing tape. To get the frosted look of the LED windows, I rubbed my finger across the back of the tape through the holes, which distorts the adhesive and fogs it up. Unfortunately, this has the side effect of leaving debris from my finger on the tape. I then applied one layer of double-sided scotch tape to the back side of the label carefully, but left the second-side liner on, then I cut the holes for the switches with a knife. Finally, I peeled the tape liner off and affixed it to the faceplate.

August 7, 2010

After taking a brief break and then doing a lot of measuring, test-fitting, and a couple of trials-with-errors, I've assembled the components.

[Closeup of Assembled tower] Mostly, the tower went together like I wanted it to, with the C128 mainboard at the back of the stack, 1571 controller in front of it, SuperCPU and RAMLink at the bottom of the case, CMD HD controller at the top, and the drives in the bays at the front where they belong. That said, I did run into a couple of snags because of the massive fan on the inside of the left-hand case panel, requiring me to move my FD-2000 controller out of the drive bay it was to go into; I've chosen to mount it in the lower-left corner of the case, just in front of the rear slots. So much for that rail I made earlier! I also took the opportunity to move my CMD HD's disk into one of the 3.5" bays just below the 3.5" drive mechanism. This has the benefit of freeing up the entire 5.25" bay where the control panel will go. Everything (including the case panel) seems to fit together well, except that I need to brace the Turbo232 cartridge seen near the bottom-center of the tower - it tends to work loose. The system still has a minor glitch or two, but nothing serious and certainly not like it used to be. As an aside, that little white wire sticking out the bottom-right is the case's side panel fan LED control wire.

August 1, 2010

I finally got the materials and parts together that I needed to start on my new tower. Mostly I was waiting on a new mainboard, as my previous one just wouldn't quit acting weird with some of my hardware. In an attempt to fix it, I brought it back to stock condition, but it couldn't even handle a stock 1750 REU, so I opted to replace it. Many thanks to Al Anger for supplying me with a working mainboard.

[Inside the empty case] I started with a brand new Commodore Gaming case, because they look nice and let's face it: They're Commodore-branded all the way down to the little feet on the bottom. They were sold for a while by a company that built high-end gaming PCs under the Commodore brand, but they appear to be out of production now. This one has been modified only slightly: I drilled thirteen holes for mounting studs and screws for the mainboard, 1571 controller board, SuperCPU and RAMLink, and removed the Firewire/USB/Audio panel from the front, which occupied one of the external 3.5" drive bays. I replaced it with the floppy mechanism for my FD-2000.
[C128 mainboard] The C128 mainboard arrived in fine shape, and received only minor modifications. I added a small header near the expansion port and soldered a handful of wires to the board to tap into the audio, video, 128-mode, and mainboard reset signals. I also added a 5-pin header for power in the spot Commodore already designated for this purpose. In addition, as you can see here, I added a Stereo SID adapter of my own design, with the second SID placed at $D700, and mounted the SuperCPU 128 MMU board in its usual place, with proper color-coded microclips to hook it up to the 8502. What is not visible in this photo is that I've also upgraded the VDC RAM to 64K.
[1571 controller board] The 1571 controller board also received a few minor modifications to tap into the reset signal, to add an 8/9 switch and a unit enable/disable switch, and to pick up the "1571 is device #8" signal.
[1571 drive mechanism] The actual 1571 drive mechanism received modifications to lengthen the motor control and sensor wires, and I clipped off the zip tie that holds the head cables down, which adds a little more length to those as well. The faceplate is from a standard-issue 1541 drive with the Newtronics mechanism, which fits the 1571 drive mechanism almost perfectly. The only modification needed was to clip off a small tab from the top to make room for the write protect sensor sub-board. Thanks to RedrumLOA for the tip (and the faceplate)!
[FD-2000 controller board] The FD-2000 controller board required some modifications as well, most of which had already been done during the construction of my other towers. Two of the LEDs have been replaced with headers, the main regulator got its own heat sink, the power switch header has been soldered into the "on" position permanently, and the power connector had to be replaced (the old one had gone bad). I also added mounting brackets. In this iteration, I have replaced those brackets with nice, solid aluminum rail so that the board can be screwed in from the edge like any other 3.5" device.
[CMD HD Controller board and drive] In my previous tower, I transplanted my CMD-HD controller into the shell from an old CD-ROM drive, but I screwed up slightly - I mounted the faceplate upside down, meaning I had to turn the shell upside down accordingly, which didn't work as well as I expected. For this iteration, I've corrected the faceplate, which means the CD-ROM shell is now mounted right-side-up and its contents are upside-down. Since the controller board sticks out the back of the shell, part of the bottom is exposed, so I added a piece of cardboard as a protective insulator. The label on the faceplate is a standard-issue CMD HD label, which I cut and trimmed to fit the faceplate when I built my previous tower. The hard disk itself is a standard-issue 4GB SCSI-1 unit originally meant for a Macintosh, and which "Eddie the One" had intended to give to Dave Mohr ("Lord Ronin"), to replace one which quit in his CMD HD. My CMD HD disk also started having problems, and since Dave has since passed away, Eddie gave it to me instead in Dave's memory. Many thanks, Eddie!
[Power supplies] The power supply received some modifications, most of which were done for my last tower. I bolted on a 10:1 transformer and a relay to be used for the C128's 9VAC supply, and fastened on an 800mA 9VDC wall wart next to it, to be used to power the RAMLink. I also replaced one of the Molex connectors with one that fits the C128 and C128-DCR power headers (same pinout on both, except that the flat C128 has an empty spot where the C128-DCR adds a 12VDC source). Inside, I tapped into one of the unsed 12v wires, and the live 110VAC mains input, which are used to power the transformers and relay. For this tower, I cut off the 4-pin CPU power connector, which I just don't need, and replaced one of the two floppy drive power connectors with two more standard Molex connectors. I also flattened out the mains prongs on the wall wart both for safety and to make more room around it.
[SuperCPU 128] My SuperCPU received extensive modifications in the past to join it with my RAMLink, and to allow for external switches, Turbo LED, and a separate "SCPU is enabled" signal (the green wire on the left). I had also added mounting brackets to the bottom edge of the unit. I've since separated it from the RAMLink, but the card edge still bears the scars from the two peripherals having been joined. The RAM card required some repairs years ago, which is how it ended up with those white wires and the 90 degree SIMM socket (as opposed to the usual angled socket). It carries 16MB. For this iteration, I moved the mounting brackets to face to the SuperCPU's right, so that it can be mounted on its right edge.
[RAMLink] My RAMLink's mainboard has had a few minor repairs here and there, plus a wiring change (on the back, not visible in this photo) to implement the "RL-Direct" hack. The RAMLink case bottom is unmodified, but will be used in a manner CMD hadn't intended. :-) Originally, the four holes on the sides were used by CMD to fasten the two halves of the RAMLink case together, but in this setup I'll be using the two closest to the LED board to mount the RAMLink to the bottom of the tower. Between these holes and the SuperCPU's brackets, the two devices will sit solidly together in the bottom of the case, on their right-hand edges. Another reason I decided to use the RAMLink's case bottom is that it acts as a heat sink for the main regulator. On the LED board, three of the LEDs have been replaced with headers, as has the "Enable" switch, to move them to the control panel. The JiffyDOS switch has been replaced with a jumper fixed into the "on" setting. The RAM card is unmodified, and carries 16MB.
[Turbo232] My Turbo232 cartridge is an interesting story. A Turbo232 normally bears a connector on the back to receive an RS232 cable, but I replaced that connector with a ribbon cable ending in a connector and backplate, similar to what is shown here. The card edge is solder-coated because I used to have a SwiftLink bolted to the Turbo232, with lots of wires running between the two card edges, and a little glue logic to move the SwiftLink to a different address. At one point, I did put the connector back onto the edge where it used to be, but there isn't enough clearance inside this tower case to allow for that, so I returned to using the hard-wired cable. I no longer have the modified Swiftlink.
[Various cables] Most of the other internal cables remain unchanged from the last tower. At one point, I rebuilt the expansion port extension cable to eliminate any noise it might be picking up. The new cable is shown here at the bottom, although it didn't fix the problem I was working on. Also, the User Port mounting plate and bracket at the upper right are relatively new. I used them in the previous tower, but only briefly before I disassembled it. The molex connector on the far left, with the DIN and barrel connectors, is used to power the CMD HD and FD-2000 controllers. I also replaced the audio and video cables on that four-RCA backpanel in the middle, with cables originally intended for those purposes.
[Keyboard cable] Only one cable couldn't be directly re-used: the keyboard extender. In my previous builds, I've always used DB25 connectors at both ends, which fit into the C128-DCR. I then carried this over to my second mainboard, a flat C128, which I had modified to replace the keyboard header with a proper DB25 connector to mate with that cable. For this iteration, I left the mainboard's header alone and instead built an adapter/extender using a DB25 connector and the cable from a dead flat C128 keyboard.