Cincinnati Commodore Computer EXPO

Many apologies for the lack of postings, but I was preparing for the  Cincinnati Commodore Computer Club EXPO (C4 EXPO), which happened over Memorial Day weekend, and I am just now getting caught up on postings.

Two weeks prior to the show, I shifted all focus to ensuring newly produced products would be available at the show.  X-Pander3 expansion boards and 64NIC+ network cards were in production, but delivery had been delayed a week or so.  This seems to always happen prior to a show, without fail.  I spent evenings obtaining manufacturing and shipment statuses, making last minute parts purchases, and packing for the show.

The week prior to the show was spent traveling to Ohio, taking in some sights (Kings Island, for example), and spending time with the family.  Since we were already on the road, I rescheduled last minute product shipments from China to arrive at the hotel before the show.

Though not the only reason, it is true I offer products for sale to finance these trips.  I can enjoy catching up with old and online friends and relaxing and enjoying the hobby.  This year was no different.  I arrived mid afternoon on Friday, with the show room already open and a few folks setting up systems.  Over the years, I’ve learned how to pack lighter, so my tables didn’t take too long to setup.  Then, I spent time finishing some products for the show and chatting with the early arrivals.

I called it quits early Saturday morning (1AM or so Central), getting back to the room before 9:30AM.  Traffic was lighter than usual, and seemed light all day.  Sales were nominal, but steady.  JiffyDOS sales were hampered by issues with my Willem programmer (after I returned home, I determined the configuration switches were faulty), preventing me from flashing some of the JiffyDOS versions.

Jim and Oliver working on project
Jim and Oliver working on project

Traffic stayed light all day, which is a bit disappointing because this is the second year traffic has been light.  By afternoon, sales dissipated and I spent time fighting with the programmer.  Oliver VieBrooks (Six of Style) came over late afternoon and we brainstormed on project ideas.  Thus began the prototyping portion of the show.   Given that sales had stopped, I decided to start on a project right then, and continued through the normal annual Saturday “C4 EXPO Golden Corral” dinner period (to be fair, I probably would have went, but we’d already enjoyed GC as a family earlier in the week).

Later in the evening, I worked on a project for Leif Bloomquist. Since the Fall ECCC ’09 show, Leif and I have been working to bring his VIC-20 MIDI cartridge to fruition. The main issue with the original design is the UART utilized, a MC6850 that is hard to source and pricey. I suggested rewriting the drivers to use the much more common 16X50-style PC UART ICs. I’d source a batch of them, but still needed to wire up one to Leif’s board so he could test.

By Sunday, I was out of energy, having worked on two prototypes the previous night.  Thus, I packed up and left without much fanfare.

Seeing friends and talking hobby is always nice, but I feel the Memorial Day slot for this show hurts it quite a bit. Traffic was light, as previously noted, and sales were likewise light. Surprisingly, the Louisville crowd didn’t show, which hurt attendance more.

24 Pin and 28 Pin Headers Arrive

Finding economical 24 pin and 28 pin IC headers for adapter boards like ROM-el has been a significant challenge. DigiKey and Mouser stock on forked post headers, which are mainly designed for attachment of wires, not soldering to a PCB daughterboard. Aries Electronics offers the correct product, but they are almost $2.00 per piece, which is not sustainable.

Recently, I ventured into bulk purchasing of some components, like the 22/44 .100″ connector for C64/C128 cartridges, and the 6/12 .156″ connector for the C64/C128 cassette port. Given the success, I recently ordered 1000 each of some 24 and 28 pin IC headers, at much better prices. Of course, since I have no immediate need for so many units and I suspect others have the same issue sourcing these items, I’m soon placing them in the store.

Thus, if you’re in need of 24 or 28 pin IC headers, machine pin, low profile, check the store in a few days.

JiffyDOS Production Units

JiffyDOS Production Unit
JiffyDOS Production Unit

After many months of preparation, JiffyDOS production ROM overlays are finally ready for distribution. The system utilizes the ROM-el FLASH-based ROM emulator, allowing the units to be fully assembled before programming.  Units should be available shortly for sale in the online store.  Click on the picture for a larger view.

ROM-el testing complete

23XX and 23XXX ROM-el
23XX and 23XXX ROM-el

Aries was backordered on the required 24 and 28 pin headers needed to finish testing, and the order just came in this past week.  Thus, I finally had a chance to solder and test the units.  Testing went well, and I can program the units with my Willem programmer.  Originally designed to hold a 29EE512 64kB EEPROM, I found a good price on Atmel AT49F001 128kB 5V Flash, so I made the necessary adjustments and tested with the new memory.  Testing went well, so I am releasing the design to production.

I have noticed one issue that I need to address.  Some CBM units have an RFI shield that doubles as a heat shield.  To perform the latter function, metal “fingers are stamped out of the shield that press on the top of the ICs.  I need to ensure the metal shield/heat sink does not contact any of the pins on this unit.

ROM-el in production!

I’ve initiated phase 1 of ROM-el production, scheduling the PCB design for production and shipment.  ROM-el will be completed in 3 phases:

  1. PCB manufacture and shipment of 2 bare boards
  2. Manual assembly and testing
  3. Release PCB for assembly and shipment

This is typical for new designs, for the following reasons:

  • Given the low cost of production boards, it’s cheaper to make a production board run instead of an initial prototype board run.  In the best case, the board is operational and no prototype board costs are incurred.  At worst, the board must be “spun” once more, but the total cost does not exceed a prototype + production run cost.
  • Shipping 2 boards ahead of time permits verification of the design before assembly.  The boards are shipped at no cost by “piggybacking” them on a previous order that is nearing completion.

Normal turnaround is 10 business days (2 calendar weeks), so I expect boards around July 12th.

To save costs in creating the SMT stencil and the boards themselves, I “merged” 4 designs into one set of files for the PCB house using Gerbmerge.  I struggled to install the program in Windows, which was unsucessful.  The application has some dependencies, one of which requires a C compiler.  The Windows install of the dependency assumes Visual C++, which I do not own.  After wasting a few hours, I installed on the local Linux server, which was trivial.  The application is very straightforward, and I was able to create a minimal merged set of files in a few minutes.  Thus, I will be receiving 100 units of:

  • ROM-el 2364
  • ROM-el 23128/23256
  • 2364Adapter (simple DIP style 24-28 pin adapter
  • 6540Adapter (CBM-Hackers folks asked for this design, which has not been tested)