Building The SWTPC 6800 Computer

The 6800 had very few parts for a com­puter. Eliminating the front panel board used with the Altair design was a big help, but the integrated 6800 chip family also required fewer support chips. The 9 by 14-inch motherboard came with all the sockets you would ever need, and they were very different from the Altair (S-100) design which used card­-edge sockets and very thin plated lands on the boards. The SWTPC design used Molexr connectors that were long metal pins that stuck up through the mother­board. The circuit cards had sockets which fit over the pins, providing a positive contact. SWTPC provided all the pins for each motherboard. The motherboard held seven 50-pin sockets for processor and memory boards, and eight sockets for the smaller interface boards. You could parallel another motherboard if you ever needed additional slots. The power supply was large enough to support the full compliment of plug-in cards, which originally was one PM board, 4K of static RAM, plus eight interface cards. The design of the motherboard made it simple to build, and although it was tiresome to solder in all of the socket pins, it did not require the close work needed to install S-100 sockets, which had twice as many pins per socket.

The MP-A Microprocessor/System Board (MP-A Board) was the primary logic board used in the system. It contained the 6800 CPU, the 6830 ROM, and the 6810 Scratch Pad Memory (128-bytes) for the ROM. The MP-A also mounted the crystal-controlled pro­cessor clock driver and baud rate generator, plus reset and other circuits. The beauty of the SWTPC design was that the lands on the cards were very, very broad compared to S-100 cards. This made it much easier to solder them, and prevented the dreaded solder bridges.

The original memory capacity of the SWTPC 6800 was a huge 16K of RAM. Each MP-M memory board had a capacity of 4K, but when you bought the computer system you got the board with 2K of RAM chips. You could buy the extra memory chips to fill the board, and you could buy extra memory boards. The memory board with 2K was $85 and the additional RAM was $45. Four memory boards fit into the motherboard for the total of 16K. Of course later, when 4K chips became available, you could expand the memory, since, like all 8-bit CPUs, the 6800 was capable of addressing 64K of memory. However, the 2K of static 2102 RAM consumed 0.75 amps of power! By this same scale of measurement, if we used the same kind of chips today, 640K of RAM would draw 240 amps of power at 5 volts DC, thus consuming 1200 watts of power. You would need a separate power line to run the computer, and you could not also run the stove in an average house.

At first glance, the SWTPC 6800 system did not look much cheaper than the Altair or IMSAI_they all cost about $475_but with the 8080-based computers all you got for that price was a bare­bones computer. No memory, no I/0, and no software. You only got four slots, and even they didn't have all the required connectors. Your $450 was only a down payment on a very expensive computer. With the SWTPC 6800 computer you got all the connectors, an operating system in ROM, and a memory board with 2K of RAM for $395. The extra 2K of RAM was only $45, and the I/0 board was $35. For $475, SWTPC sold you a kit for a complete operating com­puter. Of course, you could add to it, but your total final cost was nowhere near the price of an S-100 system.

Software for the SWTPC 6800

From the beginning, Dan Meyer and Gary Kay, his engineer/designer, recognized that the secret to the suc­cess of their computer lay in software. Having the operating system in ROM was a break for them, but more soft­ware was essential. Fortunately, there was an Assembler program available for the 6800 that could be adopted for their com­puter. SWTPC made it available to owners for $14.95, in either paper tape for Teletyper, or audio cassette format. This low pricing set the pattern for all SWTPC software. While MITS was charging $150 for BASIC, Dan Meyer set the price by the "K," 4K BASIC cost $4, 8K BASIC was $8 and 12K BASIC was $12! Although the SWTPC 6800 did not have Altair BASIC, they had a version written by Robert Uiterwyk that was one of the best cassette BASICs on the market.

The AC-30 Cassette Interface

The greatest need for the early com­puters was a reliable method of mass storage. The paper tape of the Teletyper was only available to those who were lucky, or rich enough to have access to such a machine, and they were a pain­fully slow method of saving programs and data. At that time, "real" computers used digital tape drives that cost thousands of dollars, or the new disk system recently invented by IBM. Computer hobbyists, ever inven­tive, discovered that they could record the tones of a modem on an audio cassette and save them. When replayed, they would recreate the ones and zeros of a digital data stream, and from that beginning came the cassette data storage method. The only problem was that each manufacturer had a dif­ferent recording method, and the tapes were not interchangeable. In November 1975, Byte magazine called a meeting in Kansas City to set a standard for the recording of digital data by audio cassettes. SWTPC attended and ac­cepted the resulting Kansas City Stan­dard of 300 baud data speed, with 24OOHz sine wave representing a logical one, and 12OOHz sine wave represent­ing a logical zero. As a result, SWTPC built a cassette interface unit capable of supporting two cassette recorders and able to control the motors of both the cassette recorders. The unit was designated the AC-30 and sold in kit form for $79.50. While this unit was used with the 6800 computers it never became accepted by any other system because no other company used the "Standard." Both Apple and SOL com­puters had a cassette system that was reliable at 1200 baud.

The PR-40 Printer

Low cost, high quality printers are the usual thing these days, but I remember when printers cost much more than computers. The Centronics 779 finally broke the $1,000 price bar­rier in 1977, and Epson was the first to offer a quality printer at $600.

Way back in 1976, SWTPC alone found a way to sell a really low-cost printer to hobbyists. Seiko made a print mechanism for cash registers that would print 40 columns and Dan Meyer obtained these printer mechanisms. His company incorporated them into a little printer that printed 5 x 7 dot matrix, upper case only, at a rate of 75 lines per minute. The print line was 40 characters wide on a roll of 3 7/8" adding machine paper, but it was enough for listing programs or short notes. Such a machine would attract scant notice today, but in those days a printer for $250 was a great bargain. The PR-40 was sold as a kit although the print mechanism was com­pletely assembled. The electronics had to be constructed, and the entire assembly mounted upon one of SWTPC's metal chassis.

This completed the full starting lineup for SWTPC, and they advertised widely that here was a computer system that almost anyone could build. I say almost because we did have some people who ruined their kits. One of them glued all the parts to the board and brought the mess in to be wired so it worked. Another burnt the motherboard by using a torch for soldering. One accountant had such trouble that he kept coming into my store. We became friends, and he eventually became my partner. Needless to say, from then on he kept the books, and my technicians built his computers.

At the big computer show in Atlantic City, New Jersey in the summer of 1976, Dan Meyer and his entire crew came to the Shoreham Hotel to show off their computer system. Dan was very proud of his complete line. He and all of the SWTPC crew wore tee-shirts emblazoned with "Altair Sucks" on the front. This was a little too much for the show management, who did not want to offend their largest exhibitor. Meyer was requested to remove the shirts. However, he had already made his point, and he got a lot of attention.

After the show, the SWTPC 6800 continued to do very well; however, things were changing in the industry. Floppy disks were rapidly replacing cassettes as storage devices. At first, when the 8-inch floppies came out, they were too expensive for the price range of the SWTPC customers, although other companies sold them to use with 6800 computers. However, when the 5 1/4-inch floppies became popular, South West immediately designed a system to go with their machines. This started problems that Dan Meyer never expected.

SWTPC had no problems with soft­ware until the first floppy disks were ready to be connected to the 6800 com­puters. Bob Uiterwyk, who had written SWTPC BASIC, had promised to produce an operating system and actually pro­duced a system called FDOS. There were problems with this system because it only supported sequential files and not random files. To those not familiar with disk files, I must explain that this deficiency meant that FDOS was nothing more than a cassette system used on a disk. You see, a cassette stores programs and data sequentially, like a row of ducks, and you have to search through the entire tape to find what you are looking for.

The normal disk operating system, with its random file system, rapidly locates data anywhere on the disk. This is the major advantage of disks over tapes, and without it the speed advan­tage of disks does not exist. Well, the lack of such a system hurt the sales of SWTPC systems just at a time when floppy disks were replacing cassettes all over the industry. Finally, Dan Meyer and his staff wrote a specification for a real DOS, and it was implemented by TSC under the name of FLEX. This single­-user system became quite popular and later was expanded to multi-user opera­tion under the name UNIFLEX. Another DOS often used with 6800 systems is OS/9, which will also run on SWTPC machines with the 6809 CPU.

The problem with the disk system also produced terrible strains in the dealer organization. When the system was selling well, Dan never shipped all the components we ordered. I don't know if this was because demand exceeded production or because he still sold equipment direct and the more computers the dealers sold, the greater the demand for add-ons, which often only he could supply. Our strategy was to order much more than we needed. SWTPC would cut our order, and we would end up with the quantity we actually needed. The slowdown in sales hit our store much later than it did others because a lot of our business came from overseas, but it did affect us.

Then I began to notice that we were receiving a lot of packages from South West in our daily UPS shipments. I called Kenny in and asked him what was going on. Did he have a large foreign order to fill? He told me that he did not have any unfilled orders and that stuff was starting to fill the storage cabinets. Then I called San Antonio, and they told me that they were filling back-orders that had not previously been shipped! I quickly took an inventory and canceled all back orders we absolutely did not need. However, the damage had been done, and I owed Dan Meyer more money than I could pay him by the end of the month. I called and told him my problem. Dan was not very pleased.

"I told you had to pay me at the end of the month, and there would be no extensions," he tartly snarled at me.

"Okay, but I have a cash flow problem right now. My taxes are due, and if it comes to a choice between paying my taxes and paying you, it is an easy decision for me to make. I will pay what I can now and pay you the rest as soon as possible. Just don't send me any more stuff."

"Don't worry, I won't. But from now on you are on a C.O.D. basis."

"That's your decision," I told him. "You won't find it easy to replace our store in New York."

So three years of close friendship went down the drain.

Later, I met one of his other dealers at a computer show. "How are you getting along with Dan Meyer?' I asked him.

"Dan is mad at me," he replied. "He shipped me so much stuff I couldn't pay for it."

I told him of my experience, and when we talked to other dealers, it was the same story! The story we pieced together was probably true, although I cannot completely vouch for it.

It seemed that what Dan had done was to repeat an old trick attributed to Henry Ford. He had a commitment to Shugart Associates for a large quantity of floppy disk drives, and they shipped them in every week. If he canceled the contract they would charge back the price of all his drives at the smaller quantity price. They might also sue him for violating his contract. He was not selling the drives and did not have the cash to meet his payments. So he scrapped together every part he could get and used them to fill all the dealer's unfilled back orders. He shipped them all over the country and thus built up his accounts receivable, which he could borrow against to meet his obligations. The trouble came when the dealers could not pay all the bills on time.

I do not know what happened next, since, with my credit cut off and sales of SWTPC stuff declining, I sort of lost contact with Dan. We continued to sell the SWTPC 6800, but with the growth of Apple II our sales were at a minimum. We finally all but dropped the line as far as selling new computers went.

This may have had something to do with Dan's later decision to get out of what he called "hobby computers" and concentrate on "business machines."

SWTPC did improve its products. It introduced a new, improved 6809 CPU, and both a line of 8" flop­py disk drives and 5 1/4" floppy disk drives. The powerful UNIFLEX operating system was introduced, and much-improved terminals: the CT-64 and the CT-82.

SWTPC had always been a kit com­pany, and they never attempted to pro­duce a factory-manufactured computer product. This was completely in tune with the hobbyist market that started this industry. However, only so many people want to actually build their own computer, and by 1978 most of them had already done so, and they were com­mitted to their particular type of machine. The advent of the Apple II, the TRS-80, and the SOL changed the market completely. You could now buy a better computer than you could build, and for less money. The market for kits collapsed.

SWTPC then offered their System B, a completely built computer system with two floppy disk drives and a ter­minal mounted in a desk. The system cost $4,495 and had 40K of RAM, and 1.2Mbyte of disk storage. It ran the FLEX operating system, and came with BASIC and Assembler.

I do not know how successful this system was, but at almost $5,000 1 doubt if many were sold. The industry at that time was offering much better value in CP/M systems. However the SS-50 Bus was still very strong in the 6800/6809 field, and many special-pur­pose computer devices were sold for in­dustrial purposes. Two other com­panies, GIMIX and Smoke Signal, built SS-50 bus machines, and they did very well for a long time. SWTPC eventually withdrew completely from the "hobbyist" market, and only built business machines and special purpose computers that were used in point-of-sale systems and other commercial applications.