Anatomy Of A Cloned Piece Of Hardware

What would you think if you saw a bootleg of a product you design, manufacture, and sell pop up on eBay? For those of us who don’t make our livelihood this way, we might secretly hope our blinkenlight project ends up being so awesome that clones on AliExpress or TaoBao end up selling in the thousands . But of course anyone selling electronics as their business is going to be upset and wonder how this happened? It’s easy to fall into the trap of automatically assigning blame; if the legit boards were made in China would you assume that’s where the design was snagged to produce the bootlegs? There’s a saying about assumptions that applies to this tale.

Dave Curran from Tynemouth Software had one of his products cloned, and since he has been good enough to share all the details with us we’ve been able to take a look at the evidence. Dave’s detective work is top notch. What he found was surprising, his overseas manufacturer was blameless, and the bootleg board came from an entirely different source.

A clone of a clone!, we hear you groan.

Original Minstrel ZX80 (Green) versus bootleg Minstrel ZX80 (Red)
Original Minstrel ZX80 (Green) versus bootleg Minstrel ZX80 (Red)

The product in question is the Minstrel ZX80, a recreation of the Sinclair ZX80 home computer on a PCB in the same form factor as the later ZX81. It’s a project we covered back in 2017 upon its launch, and it seems that Dave has been quietly selling kits ever since. A reproduction of a particularly obscure 1980s home computer is hardly a high-volume product, so it was with considerable surprise then that he recently discovered an eBay listing complete with text lifted from his Tindie page offering it in a non-existent “Version 2.6” that he’d never released himself. Had that Shenzhen board house knocked out an extra run to satisfy the huge world market for bootleg clones of clones of 1980s computers? It seemed not, because though the boards were identical enough that they could be overlaid with nearly all components lining up it appeared from minor routing differences and a pair of surface-mount regulators in place of the original through-hole part that the bootleg item had been created afresh rather than simply produced as another run from the same Gerbers.

It’s all in the little differences.

The CAD view of part of the genuine board (top, yellow) versus the same part of the bootleg (bottom, red).
The CAD view of part of the genuine board (top, yellow) versus the same part of the bootleg (bottom, red).

Examining the images closely revealed some fascinating details. In the comparative image of a small section which we’ve  placed on this page they can easily be seen. The silkscreen layer reveals a completely different set of footprints between the two boards, for example the decoupling capacitors and the reset button. There are also text differences, not just the missing BASIC commands and the orientation of text relative to components but in the text itself. “NTSC” versus “NTCS” next to the jumper, for example.

In most cases aside from the already-mentioned voltage regulators the components are identically placed and the tracks follow very similar paths. In some parts of the board there are extra tracks on the bootleg version that do not appear on the original, but the circuit derived from the original ZX80 with a few small modifications remains the same. The autofill copper however has significant differences that can be readily seen all over the board, it becomes clear that a different autofill algorithm is being employed working to its own tolerances.

It seems likely then that the bootleg had been created by hand in a CAD package, with the original serving as some kind of template. There remains a possibility that the Gerbers could somehow have been intercepted and imported into a package, but following some detective work on the part of Dave and others a different story emerged.

The path to the bootlegger

In April last year a customer who had bought a Tynemouth Minstrel ZX80 posted about it (Google Translate) on the Russian-language ZX-PK Sinclair enthusiasts forum. They put up a link to a high-resolution scan they had made of the board, and it was this scan that was used by another forum member who decided to make their own version (Google Translate). The translation isn’t exactly on-point, but it seems that he did indeed create it by hand in his CAD package before producing his first batch of boards.

What is clear from this tale is that the copy had nothing to do with any of [Dave]’s supply chain, so the popular theory that Gerbers sent abroad are sure to be bootlegged does not hold water in this case. We’re not saying it doesn’t happen, merely that it didn’t happen here. That an enthusiast with an eye to a quick buck would go so far as to directly clone the Minstrel from scratch is something of a surprise, as the task would require significant talent and expertise to achieve. We’d expect someone capable of that work to have no problem creating a product in their own right, perhaps a far more obvious target in this space would be a direct clone of the ZX80 itself with its PCB backing for the membrane keyboard. Either way it has been an interesting case to examine, and should you desire a ZX80 clone of your own we suggest you look at the real thing rather than a bootleg.

58 thoughts on “Anatomy Of A Cloned Piece Of Hardware

  1. One of the electronic devices my employer sells has been cloned (poorly) and shows up on That Auction Site for less than what we sell ours for. When the clones break down, we get them sent in for repair, and we do nothing except laugh and send them back to the sucker that got such a great deal.

          1. For those who don’t know, RMA= Return to Manufacturer Authorization
            Most mfgrs will not accept any package sent to their repair depot without an RMA number clearly printed on the package/label.
            Of course, once last year a local customer appeared at the loading dock with one of our products and said “Fix it!”
            B^)

  2. What about the UK101? A clone of the OSI Superboard, it was never clear how they got away with it.

    The Dragon 64 was a clone of the Radio Shack Color Computer, I don’t know if there was some negotiation there, though the CoCo was fairly defined by a Motorola application note.

    Where would the world be without the massive Apple II clone field? You could get anything that way, a whole industry and stores to cover the market. And then a bit later, the IBM PC clone world came along, using infrastructure from the Apple II clone industry.

    I realize this isn’t quite the same thing, the ZX-81 is no longer a major computer and cloning divides a tiny market, ut there is precedence.

    Michael

    1. With reference to the Dragon being a clone of the Coco, I’m afraid that was not the case. Motorola created a reference design for the 6809 CPU and associated chipset which was used by Tandy and DragonData.

    2. According to wikipedia, Dragon 64 is quite far from a Coco clone.

      I happen to own one, though i don’t have a power supply (stupid weird voltages) and i’ve never started it and don’t know if it even works. I need to do that some day, when i have 40 hours a day.

      Damn, the wikipedia page says there were new Dragon 64s for sale in 2017. If only had i known.

  3. So, this is a clone of a clone? Why not be more enterprising and just use a single FPGA to make it all? I guess there are copyright issues in cloning a Z80? Why not an all SW emulator like the one at https://www.aptanet.org/eightyone/? Should be lots of those around for that and other stuff, too (e.g. Apple II?). I’m not going to recreate the Radio Electronics Jonathan Titus’s Mark-8 (see https://en.wikipedia.org/wiki/Mark-8) with a blazing fast 125 kHz clock Intel 8008, home-made 2 KB SRAM plus hex keypad entry and LED display that I built back in 1976 while in college. Sure, it was fun doing that then, but way too much work now. I have more productive things I can do now, that are way more exciting, too. Sorry if I popped someone’s balloon.

    1. You’re not popping anyone’s balloon, you’re just making it clear that you wouldn’t find stuff like this interesting. That’s fine. Other people do, though, and for many of us the very point of projects like this is that with old-school through hole technology you can really *see* how the hardware fits together, It’s easy to modify, easy to expand, and overall just more fun. The fun part is the important bit here! Some folks find the soldering part more fun, others prefer a more code-based challenge. There is definitely room for everyone in the retrocomputing world.

      Indeed, there are people like Grant Searle who have both produced reference designs for people wanting to build their own Z80 machine (or 6502, or whatever) but have also done the same thing on an FPGA. Why choose between one or the other when you can have both? :)

    2. FPGA’s have the irritating characteristic of having too many pins, and not work at 5V. Or they have enough pins, but not enough lut’s. Or they are 5V tolerant, but only affordable for the rich and famous. FPGA’s are great. But it turns out that often it’s cheaper to just recreate the computer with off-the-shelf available chips. And more fun as well.

  4. So what exactly is the problem here? One enthusiast created something by cloning the original, another one created something by cloning the not-so-original? Both do sell their clones. I bet the russian clone would have been done from an original instead of a clone if one would have been available. Obviously the person is able to pull this off. The only thing that is wrong here is not saying where the “original” clone came from and shamelessly copying the description.

    To be honest I did the very same thing a few times with free designs: When I would prefer another package for my components I wouldn’t re-develop the whole board but only the sections that mattered to me. Only difference is I never intended to sell anything.

    1. I am not terribly surprised that an enthusiast actually did this either. They are not the only ones as entire corporations have been doing this for decades.

      Just seems better to collaborate than do all this extra work from scratch though but nobody ever wants to share their designs. But then people get upset when things like this happens to “their designs” where somebody is accused of “stealing” a design, which isn’t entirely their original design to begin with.

      But then again it’s not like Sinclair invented the PCB or electricity either. Nor is this the only copy of the ZX80 out there either.

      What territory specific legal protections are there aside from trademark law here anyway? Copyright doesn’t endless protect things like a technical statement of what the device does or supports. Patent law is likely not relevant. They even put in the time to develop their own version as well.

      1. I wonder if it wasn’t simply a case of “I don’t speak the same language as the original cloner, so asking permission would be difficult…

        “But my EDA software, I already know that. I can sit down with the layout editor and copy this RIGHT NOW and not even have to wait for email.”

      2. In some cases – not projects for sale, but more or less small private projects – it is just extra effort to prepare a design for sharing. Extra documentation and photos are necessary. I designed a small inrush current limiter for transformers. I don’t have a website, so there is also no good reason to do more documentation work. But theoretically somebody could find the PCB on OSHparks website and ask me about it. In that case I would happily share things like BOM or schematic.

      1. Thanks! It wasn’t terribly clear to me from the article that this was the case. Although on a second read I see the sentence about lifted writing and version which implies it’s the same name.

        That sucks. Considering the work involved in cloning it you’d think someone would want to punt it as their own product.

  5. A former employer has its components routinely ‘cloned’. A cloned component that is typically connected to AC mains flamed out and burned out a residential garage. The local fire marshal (an idiot) attempted to force this former employer into the subsequent investigation until we produced documents where the (well-known California supplier) had been previously informed of their counterfeit and asking them to remove the company name and link to our web site. Their local DA was also certified dick-head whom was too stupid to know that he was stupid.

    Cloned parts can be very bad news. Do not cheap out on stuff connected to hazardous voltage levels or hazardous energy levels. On second thought, this is HAD, connect whatever you want to AC mains…

    1. Unless you live in a 230V country. We’re a little more paranoid about plugging dodgy stuff into the mains here because while 110V can certainly offend, it’ll only kill if you’re pretty unlucky. 230V packs a real wallop by comparison (don’t ask me how I know this).

        1. Remember that fuses/breakers exist. Often you can disable just the circuit you’re working on. And check that you’ve pulled the correct fuse. And check that no-one can reinstate the power without you knowing.

          1. There are times when you have to work on live circuits when trying to find a fault. Also, until the meter says the circuit is dead, keep that spare hand in a pocket, especially as wrongly labelled fuse are not unknown.

  6. This has been happening for a long time. In the mid-80s I worked for a small company in Central Indiana. We built a variety of things, but the bread and butter were on-stage smoke/fog generators and pest-control foggers. They both used pulse-jet engines with long tubes to eliminate thrust or electrical heaters around a tube. They sold fairly well, especially to municipalities for mosquito etc. fogging. It wasn’t people in China that cloned our products and sold them, it was people in India. Some of the clones were very good copies, some were a disaster. Our legal department was looking into it, but it didn’t look like there was much to be done. I don’t know how it went, I left for a job with more pay and less of the toxic workplace going on at the factory. It was basically Dilbert before Dilbert.

  7. Apple sued anyone who copied their design, ever heard of the Pineapple, or Laser(got one!)?
    IBM had enough brains not to go after cloners, and their design, copied a gazillion times, is now running the desktop world

    1. It’s not that IBM were clever – IBM would have been quite happy to have a monopoly on sales of IBM PCs – it’s that the first clones were reverse engineered from the IBM BIOS in a “clean room” environment – one team built a spec which was then implemented by another team who had been vetted by lawyers and had never seen or had access to the original code. As the BIOS was the only proprietary chip in the machine – indeed, one of the big design points of the IBM PC was that it was built using off the shelf parts – cloners could then design their own hardware around it.

      Apple cloners almost always just copied the ROM, which was much harder to defend. The other point is that the law on copyrighting computer code had simply not been tested very much until Apple took the cloners to court. It’s much clearer these days than it was back then.

  8. “particularly obscure 1980s home computer”?!

    It was the first sub-£100 computer, and single-handedly kickstarted the home computing revolution. That’s a landmark in history, not an obscurity.

      1. Hm, how the times have changed – the ZX80 was $200…
        The only US competitor at the time seems to be the CoCo (which was considerably more powerful…) @$400.
        I guess the VIC-20 (still better than an ’80) @ $299 the following year is the closest equivalent. http://oldcomputers.net/vic20.html seems to agree, claiming it was the first computer to sell over a million units.

    1. Yes and no. I think you are remembering the ZX81. The ZX80 wasn’t available for long, didn’t sell in huge numbers, and didn’t kickstart anything. The machines developed from it, ’81 and Spectrum, very much did.

      1. There were (iirc) ~70k ZX80s. It was the first ‘affordable’ computer IMO. Yes, the ZX81 was much more mass-market, but the ZX80 had made it possible. By the time the Spectrum arrived, things were in full swing.

    2. If anything, what the ZX81 or the Spectrum did was kickstart the UK software industry (specially videogames). Those machines were largely irrelevant outside the UK and some close countries. The same goes for Commodore, Apple and Atari, but for the US. Other parts of the world had their own “revolution starters”. I for one learnt with a Sony Hit-Bit.

  9. The point here is that the Russian forum version is a clone of Tynemouth’s clone with only minor changes, rather than a clone of the original ZX80 PCB. Whereas the Tynemouth version is a new PCB
    See this https://fjkraan.home.xs4all.nl/comp/zx80/ for photos of the original Sinclair ZX80 PCB.
    The ZX80 can be thought of as the prototype of the ZX81, The ZX81 had a lot of the logic moved into a ULA to make assembly easier as well as improving the design for reliability.

      1. The ZX81 had “SLOW” mode, where the computer would only execute programs and read the keyboard during the vertical blank period. During screen drawing, the CPU was dedicated to that, since it was a part-software video system.

        The ZX80 could be upgraded with a ZX81 ROM but still wouldn’t support SLOW. It had minor differences in the video circuit. This meant that ZX80 programs would lose screen sync (it just went fuzzy and lost the picture) during keyboard read and other things. You’d only get a stable screen when the computer wasn’t doing anything.

        During FAST mode, the ZX81 acted just like the ZX80, with the screen rolling around helplessly, but allowed BASIC to be executed 100% of the time.

        Machine code was different, programs could use the system’s built-in video “driver” with their own machine code, just a matter of interfacing it right.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.