Purchase
Less than two years after buying my Motorola MEK6800D1 evaluation board, I was ready for a system that was more expandable, and had more widespread support.
I bought my Southwest Technical Products (SwTPc) MP-68 6800 Computer System as a kit in June 1978 in its minimal configuration, which cost me C$504. Like the newer 6800/2 version, my chassis has four DB-25 cutouts, the processor card was the MP-A2 instead of the MP-A, and the monitor ROM was SWTBUG 1.0 instead of MIKBUG. But I'm quite sure my system came with the older 4KB MP-M memory board (with MP-MX) instead of a partially populated MP-8M board, an MP-C serial interface instead of the MP-S, and bare aluminum trim on the cover instead of black, so maybe it was an in-between model?
Over the next few years I expanded the memory, added a floppy disk drive system, switched to the 6809, and made various other changes. This was my main computer system for about 6 years.
Final Configuration
My SwTPc system's final configuration was:
The Wangco floppy disk drives (with case and power supply) and Gemini-10X printer were sold several decades ago, and the Cybernex terminal was junked many years ago (I kept the keyboard), but I still have the manuals for all of them.
(Click on a picture for a larger version.)
 |
 |
 |
 |
| Pictures of my SwTPc 6809 system. (465-1123KB .jpg) | |||
Configuration Changes
Configuration changes over the years (not in exact chronological sequence):
The 4KB MP-M board would initially have been addressed at $0000-$0FFF. Having only 4KB in the system at first was a step down from the 8KB (or 16KB) I had available on my MEK6800D1 system.
The Mini-FLEX operating system required RAM at $0000-$2FFF (or up to $6FFF) for user programs, at $7000-$7FFF for the operating system, and at $A04A-$A07F (in a 128 byte SRAM on the MP-A2 CPU board) for the system stack. As a result, when I installed the MF-68 hardware and Mini-FLEX software, I would have moved the 4KB MP-M board to $7000-$7FFF and installed the new SSB 16KB board at $0000-$3FFF.
The FLEX 9.0 operating system requires RAM at $0000-$2FFF (or up to $BFFF) for user programs, and at $C000-$DFFF for the operating system and system stack and SBUG-E monitor program. Typically the I/O was moved from the original $8000-$9FFF addresses to $E000-$EFFF, but could remain at $8000 if SBUG-E was modified, which is what I did. As a result, I would have removed the MP-M board, installed the new DRC 32KB board at $0000-$7FFF, and relocated the SSB 16KB board to physical addresses $A000-$DFFF, which SBUG-E split into two blocks and remapped to logical addresses $8000-$9FFF and $C000-$DFFF. When I later replaced the SSB 16KB and DRC 32KB SRAM boards with the Micom 64KB DRAM board I had to disable two 8KB blocks (physical addresses $8000-$9FFF and $E000-$FFFF), so ended up with the same 48KB RAM at the same physical addresses remapped to the same logical addresses.
I also have a SwTPc MP-N MM57109 calculator board and a partially assembled Microware SC1-B serial card that someone gave me, and I bought a Percom SS-50 bus prototyping card, but never used these. I still have the MP-A2 card, but the MP-C, MP-M, SSB SRAM, and DRC SRAM boards ended up in the garbage several years ago before I realized they might be worth real money in the future.
I even "rented" out my SwTPc system to the comptroller (CFO) at the company where I worked, in return for him paying for half the cost of the DynaCalc spreadsheet program. He used it for many weeks to learn how to use a spreadsheet.
My sister and I ran the RMS database software on the SwTPc system to generate several thousand mailing labels for the Pacific Coast Computer Faire Association (PCCFA). My brother wrote a major term paper on solar energy for a university class on this computer.
Moving On
After a few years of using the upgraded SwTPc system, I decided I wanted a more modern, more compact 6809 system, that supported 64KB RAM with I/O and ROM occupying as little as possible of the address map, and could handle double-sided, double-density floppy disk drives. This led me to design and manufacture and sell my ST-2900 Computer System.
I used the SwTPc system to develop the ST-MON firmware, the FLEX Conversion Package, the first version of the OS-9 Conversion Package, and all related user manuals and brochures for the ST-2900. After an ST-2900 system became my main computer, I still used the SwTPc computer, but only to program EPROMs and drive my printer.
More Information
A very valuable resource of information on all SwTPc products is Michael Holley's SWTPC web site. Thank you, Michael, for gathering and organizing all that information, including the interviews. His web site was originally at www.swtpc.com/mholley but disappeared at the end of October 2019. As of October 2020 it was back at a new address, www.swtpc.org/mholley, but less than a year later the new site was down, too. Fortunately most of this information is still available in several other places:
I have parts of two TSC FLEX newsletters, No. 3 and No. 4.
Restoration
In May 2020 I started restoring this computer. Since it hadn't been powered for many (maybe more than 25) years, reforming the large electrolytic capacitors in the power supply seemed like a prudent course of action. Physical inspection of the four capacitors didn't show any signs of bulging or leaking electrolyte. I removed the 91000µF 15V Mallory capacitor from the chassis, then used the same reforming circuit for it as I had used for the MEK6800D1 power supply. After 24 hours of reforming, the capacitor settled at a leakage current of 189µA when supplied with +12.15V thru a 10kΩ resistor, which, although acceptable for such a large capacitor, was a bit higher than I had hoped for. But I didn't finish the restoration at this time.
In September 2021 I finally re-installed the 91000µF capacitor in the chassis. Since the three smaller (2200µF 35V) capacitors for the +/-16VDC supplies are soldered to the MP-P power supply board, and the wires between that board and the transformer are soldered at both ends, I unplugged the motherboard power connector from the MP-P board, then used a variable autotransformer to slowly ramp up the AC voltage to reform the capacitors in-circuit over a period of 2 days.
A few days later I used my oscilloscope to measure the ripple on the three unregulated
voltages while under load:
+7/8VDC: 1 capacitor @ 91000µF, load was 10 resistors @ 15 ohms
+7.29V to +7.68V, ripple = 390mVp-p
average load was 5.0A (50% of rated capacity of transformer)
+16VDC: 2 capacitors @ 2200µF, load was 2 resistors @ 100 ohms
+16.98V to +17.61V, ripple = 630mVp-p
average load was 345mA (69% of rated capacity of original transformer)
-16VDC: 1 capacitor @ 2200µF, load was 1 resistor @ 100 ohms
-17.30V to -18.03V, ripple = 730mVp-p
average load was 175mA (35% of rated capacity of original transformer)
I'm not sure how this compares to when the system was new, but I consider it to be acceptable, as long as the +7/8VDC supply isn't fully loaded.
On October 25, 2021 I successfully booted the FLEX and STAR-DOS operating systems on it from floppy disks, for the first time in over 23 years!
This SwTPc system is now FOR SALE.
On June 22, 2024 I exhibited this system at the Retro-Computing Expo '24 at the Anvil Centre in New Westminster, BC (Canada). I didn't have any floppy disk drives attached, so booted the 6809 FLEX Plus 10TM operating system from a disk image file on my laptop computer via serial port B of my modified DACIA board, running at 38400bps. This also required porting my DSKdrv driver to the SwTPc. Both serial ports on the SwTPc were connected to the laptop computer using my new USB-XR-RS232 modules.
Other Notes
** I have now verified most of the modifications I made (and didn't make) to the MP-B motherboard:
"FLEX" was a trademark of Technical System Consultants (TSC).
"6809 FLEX Plus 10" is a trademark of David C. Wiens, dba Sardis Technologies.
"OS-9" is a registered trademark of Microware LP.
Last revised 2024-Jul-28 11:28 PDT.
Copyright 2019- David C. Wiens.
