QL workstation, how I did it :-)

QL Box & QL Work Station

Writing 30 years after about things made 30 years ago (…) this suddenly made me aware of the whole amount of time that has passed. After Giorgio has asked me to write about QL Work Station, to revive the memory of how that object originated and had been built, I had to take up everything I archived and kept, and review it calmly for some days. Magazines, brochures, manuals and some photos and video footages are that archived objects.


Remembering those years recalled me also images of the typical QL-sites which I was used to frequent.

Places like “Nuova Newel” in Mac Mahon street, where I purchased my QL and where it was possible to find plenty of hardware add-ons and softwares, not only for the QL, but also for almost all popular personal and home computers.

Entering that premises it was possible to see flanked – inordinately – Spectrums, QLs, Amigas, Commodore 64s, VICs and various Ataris, as well as a number of magazines and software cassettes, floppies and, yes, also microdrives.

Places like “Hex Electronic” at 16 Jenner avenue, another popular reseller located in a four-windows shop, who sold much of what was available on the market at that time for Sinclair computers, including Sandy interfaces. There I bought the SuperQBoard and Thru-Con-Ram expansions I installed in the first version of my QL-WS.

A third decisive place for me was “Delta Elettronica” in Valparaiso street. It had nothing to do with the QL, but being one of the most well-known retailers of electronic components in Milan, it had the same function that today the specialized online sales websites have. An extremely well-equipped place, well known, easily accessible and able to quickly provide even parts that are difficult to find. Through that shop I was able to find some of the components that I needed to make Paolo’s QL-Box first, and then my Work Station.

In a big city nothing remains the same for a long time; so today, in the corner of the building where Newel was, now there is a doors/windows retailer and a mover. Instead of “Hex Electronic” there is a diner bar, and in place of “Delta” a bank branch. Faded physically but not in my memory, I kept many of their price lists. In Italian lire of course.


The QL was not my first computer. In 1985 I bought a Sony HB 75P MSX with an attached SDC 500 cassette player. Like all Sony products, they were ergonomic, aesthetically neat and awfully expensive. At first the curiosity and the novelty of that programmable electronic object intrigued me, but its great limits became immediately evident. I used it for a year and then I sold it. Through an ad on “Secondamano” [translator’s note: “Second hand”], I found a buyer to whom I sold it, getting a good part of what was needed to buy the QL. Of that machine I still have, well preserved, several compact cassettes containing the first programs I wrote in MSX Basic.


Concerning the QL-Box, the first partial hardware transformation for a QL carried out by Paolo Vanni and me, here is my brief description about its origin as it was, with a few years in advance, the prototype of the WS.

Paolo bought his QL (AH ROM) in 1984, not long after it had been presented and marketed. Almost immediately, it also purchased the memory expansion and the disk interface for the brand new 3.5 inch floppy discs recently introduced by Sony; but also realizing the mess that flying cables and external hardware generated once connected around the computer.

At that time we had been knowing each other for almost ten years, and it was common for us to be at one’s or the other’s home. After some months watching him using his QL surrounded by power supplies, floppy drives and ugly as well as unsafe wires and flat cables, I suggested him to gather all that mess in a single container. Paolo liked the idea. Knowing my experience as a plastic models enthusiast, together with the capability of shaping and working plastic panels with precision, he asked me to create the new container. We decided together the shape and arrangement of its contents. Among the various types of plastic available, for greater workability and lower cost I chose PoliVinilCloride as the constituent material of the panels of what would become a parallelepiped-shaped black case. Making it with Plexiglass, a material of greater value but also more expensive and fragile which I utilized at work every day, seemed to me a little exaggerated.

Inside the case we inserted the power supply for the QL, the one for the disk drive and the drive itself, limiting the cables connected to the computer and around the desk to the power supply wire and to the flat cable for the floppy disks (in addition to the monitor cable, obviously). Since “the box” would have remained next to the QL, I conformed to its aesthetics by painting the six sides with an opaque black spray painting, and engraving through a pantograph the front panel with the characteristic horizontal lines of the Sinclair products along its entire length, adding the QL-BOX logo. On the same panel I fixed two switches also. The first for turning on the QL and the second for the floppy drive.

Having substantially worked only on the aesthetics of the QL + Peripheral system, without intervening on any electronic part, there were no problems, and once completed the assembly everything worked without difficulty. It was 1985. My same MSX year.


My Work Station, a couple of years later, was a bit different matter. Although I found appreciable the work done for my friend, that “compaction” level was not enough for my needs. Based on that experience, I realized that we could go further and, together with the peripherals, put the entire computer inside the container, leaving only the keyboard outside. The same practical form factor that was now widespread in PCs.

In my previous Sales Management article I’ve already described part of the technical problems concerning electronics, which I had to face and solve in order to finally realize the Work Station, therefore I will not repeat it. Here I dwell on the part concerning the construction of the case structure and the keyboard.

As you can see from the attached photos, the panels of the case are held together by four screws each. Self-tapping screws that attach to an aluminum corner frame. The stability of the assembly is obtained by the combination of the frame and the rigidity given by the 5mm thickness of the PVC panels once screwed to it.

The power supply units are insulated and fixed on a little plastic panel each, and risen from the lower base of the container. The QL PSU, known to be subject to overheating, was removed from its case leaving it free to dissipate its heat. Being the most thermally critical element it was placed so that it had space around itself, providing in the same place ventilation grids on the side and top panels. Grids were also present above the heatsink near the microdrive units, and beside and above the Gold Card. The QL and Gold Card boards are fixed to PVC shelves, in turn fixed with aluminum corners cut to size. The two floppy drives are side by side, below the QL motherboard, and they are also screwed onto little PVC panels. The reason why power supplies and floppy drives are not fixed directly to the lower base of the container, but on separate bases (the floppy drives also are placed on a rubber support) relates to the need for air circulation around the whole component, besides neutralizing the transmission of vibrations to the rest of the computer, separating each unit from the main structure.

The front and rear panels of the container are the only ones to follow the aesthetic horizontal lines pattern of the original QL, and were obtained by pantograph milling. Eight rectangular holes have been made on the front panel. Two of them for mounting the power switches (QL and floppy drives), two for the floppy drives buttons, two for accessing the microdrive mechanics, and two for the small rectangular LEDs of those microdrives. The microdrives, just for the record, stopped working already during assembly inside the case. Which I didn’t worry about so much, given the presence of the much more reliable and capable floppy drives. The “Missoni” reset button, located to the right of the machine, can be pressed through one of the ventilation slots on the right. Data and video communication connections have been moved from the motherboard to the rear panel DE9, DB25, Din, RCA and Jack sockets through extension cables. A mains filter and a fuse are cascaded upstream in order to electrically protect the entire machine. The filter is integrated into the panel socket where the power cord plugs.


I had to face the bulk of the construction process when I made the keyboard (peripheral device whose manufacture was absolutely unexpected). Initially I was going to buy a keyboard and then connect it to the computer. In a well-known flea market located in the Navigli area here in the city, I got for small change a physically intact but non-functioning keyboard from an XT PC. I opened it and replicated the circuitry layout of the QL keyboard. At first it seemed to work. A shame, though, it generated electrical signals slightly different from those expected by Intel 8049 processor (appinted to manage the keyboard), crashing it and causing to generate an uncontrolled amount of characters on screen when pressed the corresponding key. Anyway, I used that keyboard for a while, trying to understand if it would be possible to fix the problem, but, in the absence of more complete information, I realized that it was not worth losing more time.

Unable to use the XT keyboard, and not without some hesitancy due to the fair complexity of the task, I decided on my own motion that the best solution was to manufacture from scratch the entire (evil) peripheral, while maintaining the color and design of the QL and its new case. At home I had another keyboard (until then used only for testing) got by disassembling it from a bulky computer terminal from the 70s. One of those old-styled with high and thick keys. Not very elegant keys, but I was sure they worked well and without the bouncing effect of the XT keyboard.

I began the work of general design and shaping of the pieces using the same type and thickness of PVC utilized for the case (this time white instead of gray). It took several days to individually build all the pieces necessary for the “body” assembling. However, making the outer shell was only a part of the job. Inside that shell a solid frame had to be inserted, in order to support the mechanical parts of the keys, keeping them firmly in place once positioned. It was then necessary to make all the electrical connections between the keys themselves, to replicate the mapping of the original QL keyboard and, last but not least, to make two cables from scratch (again!). The first cable to connect the keyboard to the computer through female DB25 connectors on both ends. And the second, positioned inside the case of the computer itself, to carry the signals coming from the keys to the motherboard, from another DB25 (male) fixed on the lower base of the main case, in correspondence of a rectangular opening made for this purpose.

The solderings made in those days were really numerous, and each one checked immediately with the multimeter to be sure it wasn’t a cold one. I made the last twenty solderings of the series on two small copper-plated bases, shaped, engraved and adequately thinned before, in order to be inserted in the slots on the QL motherboard, in place of the thin ribbon connectors of the (infamous) original keyboard membrane. The icing on the cake was “discovering” that the keycaps of the original QL were perfectly compatible with the X-shaped keycap pillars from the recycled XT keyboard.

It was not the only icing on the cake. The QL has a Caps Lock button to enable upper case letters. As known, it doesn’t have a LED signaling its condition (has it been pressed or not?). One of the buttons recovered from the aforementioned terminal was double-acting. At the first pressing it remained down. At the second one it came back up. I placed it at the Right Shift position. Thus, even without a LED, when that key is pressed, it signals the activation of capital letters remaining visibly lowered. Completed and perfect!


At the end of 1988 I finally had my new QL ready (and above all working). Almost two years had passed from the first evaluation on how to realize that transformation after the QL-Box prototype, until the concrete accomplishment. Two years spent putting together information, finding the necessary components and building it a little bit at a time. During that time I asked some information about the bus-cable to “Hex Electronic”, whose technicians warned me about the lability of the signals on the QL expansion port. One of the reasons why all QL interfaces were connected to the computer without extensions. For other questions, I contacted “Delta Elettronica”. Meanwhile, a professional electronic friend of mine also gave me some help.

The WS had two versions. The first, based on the Sandy expansions and interfaces and with two 720K floppy drives, was active from 1988 to 1991. The second and final one, based on the Miracle Gold Card and the attached 4 Mbytes floppy drives, was a modification of version 1 in 1991. The different size of the Mitsubishi floppy drives, compared to one of the previous two (the Panasonic JU-363 drive), resulted in the visible resizing of one of the two relative apertures on the WS front panel. Since the beginning the size of the case was designed to also contain a 3.5 inch HDD, when it would have been possible to connect one to the QL; and, for the record, after skipping the standard ST disk interface from Miracle Systems, I was tempted to take Nastasic’s QBide when it was presented. I gave up not only for a reason of costs, but also because, after all, the ED floppies I could use with the Mitsubishi drives attached to the GoldCard, proved to be adequate for my storage needs.


Who knows the evolution of the QL and of all its hardware that different manufacturers have proposed on the market since the early days of its commercialization, with hindsight, will have noticed that I could avoid to make some changes, or that some components I made and “invented myself” for my QL Work Station were already available at the time of its realization. To be clear, I’m talking about expansion buses, keyboard interfaces, various third-party keyboards, enclosures, and so on. Why didn’t I use / buy them? Well… first because, as I wrote elsewhere, the Internet was still science fiction, QL Wiki was far from existing, and the only information sources to learn about the existence of those parts were, apart from word-of-mouth, British magazines which here in Italy, although present, were not so abundant. Secondly, even if I had knowledge of their presence, those add-ons, to be bought in sterling, were not so cheap. If you had been in my shoes, and knowing to be able to make what you need saving money (in addition to having fun and learning in the meantime), what would have you done?

Two more words about the pictures accompanying this text. Some were taken at the time of the initial construction and in the first years of use, but most of them are recent, and document the appearance of case and internal components layout. I hope they are sufficient and explanatory of the object and the work done on it.

I knew right from the start that radically modifying a computer designed inside a “keyboard” would have been a very difficult and demanding job, even if at first I didn’t think to push myself that far. What is certain is that after having faced and completed it, I had the satisfaction of using the product of that “madness” for many years and with profit.

Antonio Gareffa