Lucy (81-110 Motherboard)

Recent History

Lucy is the name I have given the 81-110(no letter) board that surfaced in 2026. It is named after the famous Australopithecus skelton since it is a missing link between the Ferguson Big Board and the first Kaypro II boards.

It was found by Stargeezer in the Huntsville, Alabama area from an estate of multiple Kaypros.

I traveled to Huntsville in June of 2026 to pick up the machine and it now belongs in my collection.

Description

"Lucy" is a 81-110(no letter) board. Note the C21, R37, and R38 are silk screened for components that are not installed. The U88 was intended to be a 16 pin chip but was replaced by the FDC9216 which only has 8 pins so the footprint extends beyond the chip. There are sockets on the board for the chips not present on the 81-110A boards (U7, U40, U46 and U55). The sockets appear to be originally soldered at manufacturing. Both ROMS appear to be Intel D2816 EEPROMs. On other boards they would usually be EPROMs which are not electronically erasable.

The board's serial number is 1145 which is earlier than 1353 which was the earliest 81-110A board seen. The QC Sticker is dated 26JL82 which is after the three earliest 81-110A boards. The chassis that it is in does not have a serial number but it has a Version 2 small power switch and black handle mount covers implying that it would likely be #3200-3800.

Likely reason the 81-110A replaced the 81-110

Andy Kay said:
Although we were taking orders, I decided to hold up production for one month while we re-engineered the floppy controller logic to support the double-density format instead of the single-density format. Our engineers said it was simple to implement. I'm referring, of course, to single-sided drives. The availability of double-sided drives was still a year away. It was a good move because it gave us a competitive advantage over the Osborne computer that had just been released, which also was a transportable. Osborne never was able to do this succesfully.
(Computer Shopper, August 1989)

"The Kaypro II is a Big Board with a few modifications. ... They also replaced the 1771 single density floppy controller with a 1793 double density package."
(Micro Cornucopia, Dec 1983)

The change in the FDD circuitry is consistant with changing from single to double density drives. So the 81-110 was likely the single density version and the 81-110A was the double density version.

Best Guess at it's History

I think the most likely scenario for it's use would be that Kaypro was trying to fill it's back order of pre-release sales and was running out of the 81-110A motherboards so they decided to put the extra effort into adapting the 81-110 mainboards with a series of bodge wires to be able to support the double density drives.

The mainboard supply problem is supported by the following information from Andy Kay:

Anyway, we were also able to cope with these orders because we made all of our own printed circuit boards. We stuffed them, assembled them, in our facilities. We were so crowded that year that we set up assembly lines on the outside lawn for most of the winter. Thank God that we didn't have a tough winter that year. The costs were about the same if we had an outside company make our boards. So there were no financial advantages. We could control the logistics and quality better ourselves. At one point, however, when we couldn't keep up with the demand, we did contract some board making to outside vendors.
(Computer Shopper, August 1989)

Floppy circuit progression

A Ferguson Big Board FDD circuit

From a reproduction FBB

Lucy's FDD circuit

The bodge wires used to adapt to the new chip.

81-110A FDD circuit

From FBB to Kaypro II

Meanwhile, Non Linear Systems, a small manufacturer of digital meters, purchased two Big Boards. Lo and behold there soon appeared the KayComp computer. They used that name until the original owner of "KayComp" found out about Non Linear. So, the name became Kaypro II.

The Kaypro II is a Big Board with a few modifications (actually more modifications than Xerox made). Non Linear used eight 64K bit RAM chips instead of 32 16K bit RAM chips; this made the board smaller. They also replaced the 1771 single density floppy controller with a 1793 double density package [this happened between the 81-110 and 81-110 A]. They also removed the three ICs that handled the scroll in the video circuit [This happened between the 81-110 A and 81-110 B].

They left all the ports and port addresses as they were. In fact, the Kaypro II and 4 have two PIO chips. They are using half of one for the parallel printer interface and half of the other one for system control. They could have used just one but the BB I had two PIOs. (Stick around, we'll be using the other half of those two in the near future.)

They had to rewrite the monitor routines and the BIOS because of the changes. They had to support double density 5" drives and they had to handle screen scrolling in software.

The choice of double density was very important since it gave the Kaypro II a very definite edge over the Osborne - the 9" screen and 80 columns were also a very good choices - but the choice of software scroll was unfortunate.

Try using the Kaypro as a terminal and set the baud rate over 1200. Once the screen fills, it starts dropping characters each time the screen has to scroll. You see, the processor has to move every character in video memory during a scroll (which takes a while). Meanwhile, the Z80 ignores all about the characters - that are coming in the serial port.

With the hardware scroll, the video circuitry automatically shifts everything by one line and points the processor to the correct spot in video memory. The processor doesn't have to so much as raise an address line.

One of the big reasons that speeding up the Kaypro processor improves the performance so much is this silly scroll problem. The Z80 spends most of its time just rewriting the screen. If you purchase one of the many speed-up boards that slows down to 2.5 MHz when writing to the screen, you don't gain much.

Non Linear found it easier to produce systems than service manuals. So, it was not unusual to find a Kaypro dealer with a set of Big Board schematics, trying to translate the IC numbers from one system to the other and trying to figure out what was left in, what was left out, and what changed. Somehow it all seemed very familiar.

The evolution of the BB character set is also worth a look. The first BB I boards shipped by Digital Research had an interesting character set. The uppercase characters were very standard uppercase characters. The lower case characters were simply shorter versions of the uppercase. A screen full of text looked very strange - and if you had all one case or the other, you really had trouble telling which you had.

Well, a group of us at Tektronix had placed an order for 7 boards right when the first ad hit the mags. Ours was the first order out the door and I have board A000014.

Anyway, we didn't appreciate the character set, so I took a few minutes one afternoon (with the help of Randy Dietrich and Lynn Cochran) and came up with a real lower case character set. It wasn't very fancy - the y, g, and q were just plain ugly - but the screen looked ten times better than before, so everyone in the group asked for a copy.

I told Jim Tanner that I had changed the characters. He said they were working on a new character ROM themselves and he asked if I would exchange my character ROM for a copy of his when he finished with it.

So I sent him mine. Six months later I called him to find out what had happened to his. He told me not to worry, they had decided to ship Big Boards with my character ROM (he calls it version II).

I mentioned that I had been hoping to sell my character ROM as a way to help finance the start up of Micro Cornucopia. So there I was, my first new product for the BB I was being duplicated by the hundreds and shipped with every new board. (He sent me a BB I kit as payment.)

So I went back to work on the character set and cleaned up the tail draggers that looked so unique (and ugly) and advertised the upgrade in Micro C. The response wasn't earth shaking, but it has helped us get started.

Then one day a friend suggested I take a look at the Kaypro. Well, looking at the Kaypro screen was like homecoming. There were my ugly g,. y, q, f, t ... There was no doubt about it. I knew that within that Kaypro there lay a heart of BB I and dinside the character ROM were the remnants of an afternoon at Tek.

So I combined my new character set with the Kaypro Greek characters and produced the Greek PRO-CHARACTER ROM. I also left out the special characters altogether and called it the Clean PRO-CHARACTER. The Clean character set was for those who were being driven to distraction by the funny little greek characters (mostly folks into data communications with mainframes.)

I sent a copy of my character ROM to Non Linear (after getting a non-copying agreement from David Kay). They have upgraded their characters somewhat since receiving mine but they haven't gone all the way with it.
(Micro Cornucopia, Dec 1983)