Are Patent Claims Coming For Your WS2812?

There are some components which are used within our sphere so often as to become ubiquitous, referred to by their part number without the need for a hasty dig through a data sheet to remind oneself just what we are talking about. You can rattle a few of them off, the 555, the 741, the ESP8266, and so on.

In the world of LEDs, the part that most immediately springs to mind is the Worldsemi WS2812 addressable LED. This part consists of three LEDs in red, green, and blue, all in the same package with a serial interface allowing a chain of individually addressable multicolour lights to be created. We’ve seen them in all sorts of places, and if you don’t recognise the part number then perhaps you will by one of the names they’re sold under: Neopixel.

Yesterday we received an email from our piratical friends at Pimoroni, the British supplier of all forms of electronic goodies. Among their range they have a reasonable number of products containing WS2812s, and it was these products that had formed the subject of an unexpected cease-and-desist letter. APA Electronic are the manufacturer of the APA102 addressable LED (which you may know as the Dotstar), and their cease-and-desist asking for the products to be withdrawn from sale rests on their holding a patent for an addressable multicolour LED. We’d be very interested to hear whether any other suppliers of WS2812-based parts have received similar communications.

US patent number 8094102B2 is indeed a patent for a “Single full-color LED with driving mechanism”, which does look a lot like a WS2812. But as always, such things are not as cut-and-dried as they might first appear. The LED in the patent for example relies upon a clock line for its operation, while the Worldsemi part doesn’t. I am not a lawyer so I’d hesitate to call this a baseless and speculative move, but I suspect that there will be plenty over which the two semiconductor companies can duke it out in the courtroom.

It’s fair to say that a large part of the ethos of our movement shares something with that of the world of open-source, so news of legal manoeuvres such as this are never likely to go down well. We’re small fry in this context and our commercial influence on APA102 or WS2812 sales will be minimal, but inevitably APA’s standing in our eyes will be diminished. Companies such as Pimoroni are not the target but a piece of collateral damage in a battle between manufacturers.

Whether the patent has been violated or not can only be decided by the courts. It is not uncommon for patent holders to go after companies selling the “infringing” products in hopes that rather than risk a costly court battle, they simply adhere to the demands, in this case buying parts from APA and not from Worldsemi.

So, if you rely on addressable LEDs, watch out! There may be trouble ahead.

Header image: Tristan Robitaille [CC BY-SA 4.0].

115 thoughts on “Are Patent Claims Coming For Your WS2812?

  1. If there’s a Gofundme to cover legal costs or something, let us know. I, for one, haven’t got anything WS-containing from Pimoroni yet, but I know they have quality products that deserve to be available.

  2. To figure out if something could be in violation of a patent, do not read the summary. Read each claim, and see if the claim is exercised in the part in question. Then figure out if there is prior art for that claim.

    This is obviously a move to scare the part manufacturer into purchasing a license, by closing their distributors and dealers with FUD so that they can’t sell the product. The dealers and distributors should have their lawyers contact the lawyers of the part manufacturer, and proceed on advice of counsel. If your legal counsel advises you to keep selling the component for now, you are in general protected under law in doing so. If you go forward without the advice of counsel, you are not protected.

      1. Said slightly differently, proceeding under advice of counsel (who of course must document a non-infringement argument in writing for you) gives you a good-faith reason to continue selling. It does NOT mean, however, that you aren’t infringing (in which case you could be impacted financially). It simply prevents the plaintiff from winning multiplier (often 3x) damages against you IF a court determines that you did in fact infringe. At least, that’s been my experience with patents etc in 33 yrs in the hygiene business. :-)

    1. You may be protected against willful infringement (it could be evidence that you had a good-faith belief that you were not infringing). But I believe a manufacturer could still be liable for the patent-holder’s lost profits, and/or a “reasonable royalty”.

      1. Yes, simple damages for a licensing fee that the court thinks is reasonable could be assessed. It’s not always worth the plaintiff’s costs or time to even go after you for that. A lot of the time, the defense attorney will present the amount that could be recovered and the costs of recovery, and then the parties will settle for some fraction. But (as I know well) there’s always the idiot.

  3. Reading the patent, every independent claim requires a “clock input” and a “clock output”. If the chip does not have these features, it is not infringing, and they can safely point this out and keep selling these parts. Looking at the datasheet, this seems to be the case. Without a clock, there is no problem. Even if they had a clock “input” but not and “output” they would be perfectly fine.

    They should have their attorneys tell this patent troll to pound sand.

    1. Even the WS2813 seems to dodge this by adding a second data line in lieu of a clock line. I thought this was an ugly hack at first, but it seems it may have been a way of dodging this patent instead…

    2. In addition to all the repeated talk of clock input and clock output in the claims, the primary claim also states power input *and* power output – neither the WS2812 nor even the APA102 meet that requirement as they only have power inputs :-)
      Additionally, the primary claim infers power to the controller IC through the said “power input” which also supplies the LEDs – the WS2812 has two power input pins where one is for the LEDs and the other for the controller IC, so again avoiding the claim definitions.

    1. I think that’s the point though. If there is a hint of merit behind the claim, a company like Pimoroni would have to fight this at great expense. Most companies won’t do that because it’s expensive and there is uncertaintly about the outcome. In this case if distributors think they are in the right they can’t just stop selling WS2812 products since they are so pervasive.

      Bruce’s advice above to seek legal counsel as a means of working through this is excellent advice.

      1. Speaking as an inventor named on a bunch of patents with a fair amount of experience with the USPTO (on the Commercial>>Advance portfolio, if anyone remembers that VCR product from the 1990s) I’d bet you’d find it very difficult to get an attorney to issue that opinion for you, due to the potential exposure of the attorney in issuing that opinion. Unless it was super, super, super obvious, and maybe not even then.

        1. With all due respect, they should ask their lawyers for a finding of non-infringement. There are a lot of complications in issuing this, but good lawyers know the right expert to hire, and understand both how to protect their clients, and how to protect themselves. Exposing an attorney to lawsuit or sanction for this – well, don’t even try. It’s very unlikely to succeed.

          1. As I understand it, a finding of non infringement is the summary judgement of a judge, not a lawyer. Is that not true? I’ve never gotten very far in any infringement case, we’d always settled early on. And my comment wasn’t about trying to bring an action against an attorney, I was saying an attorney might be reticent to issue an opinion due to his own exposure. We did on occasion need to hire expert witnesses to reverse engineer potentially infringing devices and the tactic worked well.

      2. Western distributors need to consdier stop sell, meanwhile Chinese distributors can carry on selling regardless and enjoy the protection of their government from external claims.

    2. I’ve already decided. I’m putting APA Electronic on my list with FTDI, John Deere, SCO, et al.–companies I won’t do business with due to their abuse of intellectual property law. I’ll certainly never make a difference to the bottom line of any company on that list, but my conscience is clear.

    3. This totally affects the end users. If two competing products from two companies suddenly become one sold by just one company, you can bet all your horses the part price will skyrocket.

  4. Considering the fact that Pimoroni also use APA102 LEDs in some of their other products and that WS2812 are not pin-compatible with these, it would seem to be a pretty stupid email…

    * Why annoy a client who is already buying your components in volume?

    The answer is probably that some grunt at APA or in some law firm is trying to troll up some business and hasn’t a clue what they’re doing.

    1. I was thinking the same thing and if they have a problem with products containing the WS2812 then they should take it up with Worldsemi and not vendors who happen to use both products.

      1. This! I would think that the responsibility here lies on the part manufacturer, not the distributor. If they wanna actually commit PR suicide… err… I mean proceed with legal proceedings, then they should be going after the source. It’s not the distributor’s responsibility to decide whether the part being bought infringes. it’s the responsibility of the part manufacturer to fight this to show their distributor their part is legit.

        Until APA gets a judge to say “this part may not be imported or sold in X country”, then the distributor has no obligation to cease selling, and a claim against the distributor seems outright shady.

        Only thing this guarantees, is APA might be going on the DO NOT BUY list, just like FTDI.

        1. The problem is that there is something called “contributory infringement”:

          35 U.S.C. § 271(c) Whoever offers to sell or sells within the United States or imports into the United States a component of a patented machine, manufacture, combination or composition, or a material or an apparatus for use in practicing a patented process, constituting a material part of the invention, knowing the same to be especially made or especially adapted for use in an infringement of such patent, and not a staple article or commodity of commerce suitable for substantial noninfringing use, shall be liable as a contributory infringer.

          1. Except no court has decided that. Until a court does, they have no obligation to comply. If a court says it infringes, then yes, they should cease selling the product, and then if they don’t… Well, APA can file a REAL C&D then… Not this fakeout.

    2. I’m short on time, and didn’t have the opportunity to finish reading all of the comment herein, but the few cease and desist orders I’m privy to were ALL sent via snail mail. Emails can be faked, and as such are unreliable. Especially considering that this was delivered through some sort of form (I saw that mentioned somewhere), this is most likely just an attempt to boost public attention to the company of the “claimant.”

  5. I’m not even sure if it is an official cease and desist. It shows that it came in through the contact form, and directs them to a Google link to the patent. It doesn’t even call out the patent number, other than the reference within the link. Now, I’m no patent lawyer, but that looks like the least professional cease and desist letter you could possibly mash into your keyboard.

          1. You actually can serve someone by email, and write a little testimony that you did so serve them. Service by web form, and not listing the patent number? Perhaps not sent by a legal professional.

    1. The way it’s written, I kept expecting to read that the patent rights are held by a Nigerian prince. I have no expertise in legal matters, but have read a fair few phishing scams. This smells like one.

  6. I wonder if they could use anything like these as examples of “prior art” in order to undermine APA’s original basis for their patent claim:

    Addressable LED architecture (2007 / 2008)

    Single wire serial protocol for RGB LED drivers (2004 / 2005)

    Tile Lighting Methods and Systems (2004 / 2005)

    IP-addressable light-emitting diode (2001 / 2002)

    Multicolored LED lighting method and apparatus (2000 / 2001)

    1. I agree. I’m often upset by the moronic inability of patent examiners to examine patents when big companies come along. I’m not sure if there are bribes involved or if the examiners just don’t have the time under their fixed fees to argue against limitless corporate patent attorney badgering, but it is noticeable that with big companies utter tosh gets granted. Try filing a patent yourself and the examiners fail to see inventive steps in entire new processes, but, in one of my favourite examples, they found that Nespresso were inventive in the novel method of sealing a coffee capsule by using an o-ring. To be inventive you have to do something that is not obvious to someone skilled in the art, so apparently people skilled in the art of sealing wouldn’t think of using an o-ring, one of the most popular methods of achieving a seal out there.

  7. Lawyers are the gold standard for Bullying. If your making a product and bought pieces in good faith then tell them you will not cease and desist but will file counter suit naming both the company and their attorneys for attempted fraud, torturous interference in trade etc. since they are infringing upon your propitiatory product design and right to do business. They are trying a end around play to force your supplier to sell out to them cheap. It should also be noted that until a court of law decides their is patent infringement you are not a party to the infringement.

  8. On another note: I only found patents from them for CN, US, and DE. How can this be considered a “global patent”? There doesn’t seem to be one in GB, where Pimoroni is located.

  9. I believe that there is not one “global” patent… As I understand it one must file (and maintain) a patent in every jurisdiction that you wish legal protection. So… If you have a US Patent and fail to elsewhere, then your patent is only enforceable in the US. As for enforcement…it is not automatic… you, as the patent holder, must bring legal application to those companies/individuals that you claim infringe on your patent.
    Given the firm is based in the UK, a US patent would not be enforceable within the UK unless a patent was filed there as well. The only potential issue I could see is that if the US patent is valid and upheld, they may be able to restrict the firm’s ability to sell into the US.

    1. Patents are only valid in country they were received in unless there is special treaty that covers this, there’s no treaty between US and EU as far as I know, but there is treaty within EU so patents are EU wide (EPO).

        1. Really dave, is it that clear cut that initial dispute is criminal and not civil, surely it only becomes criminal after a judgement for damages if at all and even then not at discovery/directions in terms of early stages of the case etc ?

      1. Close, there are separate national and EU wide patents. One does not always imply another. One can hold a national patent infriging EU wide one and as long as the holders don’t battle it out they both are valid in the nation in question.

  10. All this is just more evidence that the very idea of patents is stupid.

    Patents are nothing more than a mechanism so the idealess can “own” an idea. They should get psychological help, not legal help.

      1. Actually, people don’t get to own ideas under the law. They get to own a process, a machine, a manufacture (way of making something) or a new composition of matter (a formula). These things are all implementations.

        I could also teach you about copyright, trademark, and trade secret, but maybe Hackaday should fly me to their next conference before I do that :-)

    1. Patents actually can be very valuable in protecting someone with actual novel/innovative ideas to protect their invention during the time it takes to get the product tot market, work out the kinks of the first generation and win back the cost of development. If this protection is not there a big corporation with a large engineering department could put something better to market faster once they get a whiff of the new design, and the original inventor is screwed.

      The problem is nowadays there aren’t enough patent examiners to keep up with the endless flow of bullshit patents brought by large corporations and idiots alike, claiming all sorts of totally non-inventive cruft. Since examiners are overworked many patents get granted that have no right getting granted and the system breaks down.

      Don’t even get me started on the idiocy that is software patents though.

  11. Personally I’d offer to pull all products that use either APA102 or WS2812 until their patent suit has gone through. Are they actually suing the manufacturer of the WS2812 or are they just threatening to sue pimoroni?

  12. I don’t even understand what was “invented” in this patent. It says the LEDs, the serial interface and drivers already exist and are already used in an RGB configuration. Why is making it smaller not obvious to someone skilled in the art?

    1. Agreed. IP is one of those things that is so pervasive in our culture and way of thinking that it’s difficult to conceive any other way, but there’s a disconnect between how patents are supposed to work and how they do work. I don’t think we’ve followed intellectual property to its logical conclusion yet.

      1. No the rules are not fair.
        The rules should be there to give greater power to the individual above and beyond to the corporation.

        The corp has far far more power to weild in any argument with an individual. More money, more expertise, more bureaucracy, the ability to write off expenditure agasint tax and so on.

        Too many laws benefit corporations by default.
        They are not people yet they have too often inherited the same rights as people.

        1. So at what point would you specifically demonize a 1 person shop that hires a few people, sees some success, grows to a 10 person outfit, then hires 100, succeeds further, grows to 1,000 employees (and pays taxes to support roads, schools, etc), grows further …..

          The last time I checked, the ~100,000 employee company I work for is made up of 100,000 hard working, individual people who labor year round to make great products with high value for consumers (products which there’s a high probability of being found in your residence). And yes my “corporation” pursues it’s patent rights so that those 100,000 individuals can continue to support their families and the communities in which they live, work, and play.

  13. Is this a joke? Why would anyone take this seriously when it came through a website contact form? Except that Pimoroni now made it public that they received it, it should be completely ignored. Sounds like an APA salesperson is trying to drum up some business through a juvenile scare tactic. Does APA not realize the internet exists and actions like this can hurt you big time? Just mention FTDI drivers…

  14. I had a thought on how to deal with patent trolls:
    1. create an LLC company that has no assets
    2. intentionally violate as many patents as you can
    3. when sued, act as your own defense
    4. when found guilty, close the company down
    the troll has now spent lots of money that they wont get back

    1. How would you violate patents without assets?
      I’m pretty sure no one cares about a company that literally does nothing but think about other companies’ patents.

  15. Claim 1 of the patent already shows this to be a fake C&D: “1. A package structure of full-color LED (Light Emitting Diode), comprising: a power input, a power output, a clock input, a clock output, a serial data input and a serial data output…”

    One can stop reading after the “a clock input”, given that the WS2812 does not have one.

    While the patent has several independent claims, they all fail for the same reason.

    There’s one positive aspect to this C&D. I didn’t know about WS2812 earlier, but now I do. I don’t mind going from 6 pads to 4 yet maintain the same functionality so I think the next time I need these things, the APA stuff is off to second source status.

    1. That’s not the way it works. First you go after the small fish that cannot defend themselves then you use those victories to support your claims against the bigger fish.

  16. The patent claims all have this in common:

    1) Clock, data, and power lines
    2) Discrete current limiting resistors

    There are 3 protected embodiments of the invention, each claims cites a specific layout of the components in the package: driver chip, discrete LED chips, and discrete current limiting resistors “current limiting resistor are chip resistors with two electro-conductive surfaces.”

    The WS2812 does not have discrete current limiting resistors or clock lines, so the 3 protected embodiments of the invention are not the same.

    Looking at microscope shots of the WS281X, there is a driver chip and 3 LED chips, the current limiting appears to be done on the silicon die, not with discrete chip resistors.

    I would say that WS knew about this patent when they designed their LED and purposely designed around the patented claims.

    It would not surprise me one bit if WS already had lawyers review and produce documents to protect them against this patent.

  17. I think this issue might be a little more complicated. They’re not a “patent troll” by your traditional metrics – they’re a decently sized company that manufactures a real product that people buy and integrate into their products. And judging by their history and patent issue dates, they were most certainly first. Also, APA is the Taiwanese company, and Worldsemi is a Chinese company – one country is a little better known for copying and undercutting new ideas.

    For most current hobbyists, the WS2812 is better known, probably in part because they’re cheaper, and in part because they have a more convenient interface. Actually, I even vaguely remember a time when both of these two generic-sounding foreign manufacturers were both putting out product to US hobby shop. But suppose APA came first back in the late 2000s, and really was the first company to design a digitally controllable LED with LED an controller in the same package, which at the time was a novel and IMO non-obvious idea. Would this be any different from making a hypothetically perfect iPhone clone and, say, adding a microSD slot and headphone jack, both of which would be very valuable features and might sell better?

    The bigger complexity is that I found that APA has a much more extensive set of patents in China (not sure if that means anything) and Taiwan, some of which may indeed cover the WS2812 ( Now, legally, this doesn’t mean much, but having dealt with patent filings before, it is exceptionally expensive to pursue international patents, and many people opt to only select one fairly important one to file. They may have picked the wrong one – but you can think of it this way – Worldsemi knows that patents aren’t exactly well-enforced in China, so they see which ones are enforceable for their international export markets and work around them.

    Now, the most awkward thing about this whole patent thing is timing. APA could have sued Worldsemi years ago, when the idea was fresh and early in its patent term. I am not an international (or domestic) patent lawyer, so it’s entirely possible this case has been running in the Chinese or Taiwanese courts for a decade, and they’re just now getting to doing something about international IP.

    This runs into my personal opinions on how the patent system can be improved – I think patents need to have variable terms depending on market. If this article came out a decade ago on Hackaday, when APA first just came out with their product, and WS2812 was the clone, perhaps there would have been a little more sympathy for APA, because an RGB LED with controller in one package was a pretty new idea. (Or perhaps not, because WS2812 was probably cheaper, and had a nicer interface). But at this point, the WS2812 has sort of turned into a hobby standard and kind of the brand name, so it’s hard to like APA over WS. Neglecting whatever issues may have delayed enforcement actions on APA’s part for a good decade, I feel a patent like this, which would not have needed more than a few years perhaps to recoup development costs and lose its novelty, should have expired by now.

  18. Just a marketing stunt – ignore it.
    1) Cease and desist must be delivered by registered mail, so there is proof of notice.
    2) Muat have a (non-redacted) signature.
    3) Must include a (non-redacted) means of responding.

  19. The more I hear about this the more it sounds like an out and out scam. Maybe we should all get together, figure out the various methods involved in scamming and file for “global patents” on them and then we can sue them if they try to scam us without an appropriate license…

  20. I recently found out about the Inolux IN-PI554FCH.

    Very similar to the WS2812B but:
    – Orientation marking indicates pin 1 as it should (and not pin 3!)
    – The company behind it looks more “serious”?
    – It can be sourced from known/big distributors like digikey ( and mouser/arrow (although in bigger quantities) which is a BIG plus since you know what you’re buying exactly and you always get the same thing (I’ve heard many horror stories about dodgy neopixels).

    They may be a bit more expensive, but I think the are a very good option for serious projects.

    Has anyone tried them out?

  21. One of the sketchy, but useful pieces of advice I was given was don’t talk about patents in writing. The reason being people’s memories are terribile and if you say “analyze” something and put it in an email and say something like “I think we might infringe” but continue selling anyway — WHEN and not if you are subpoenaed ALL of your emails/sms/chat etc can be subpoenaed. The mere admission of maybe can change the context from “I didn’t know” to intent to violate which can make damages huge. If you do want to talk patents in an email then you always have to side on the ultra cautious boy scout response where you assume everything everyone tells you is truth and act accordingly.

  22. APA’s website has clearly been written by someone who’s about to go patent trolling:

    “As the original manufacturer of the patented integrated IC LED”
    “APA-102, APA-104, (Detail Patent)”

    Methinks thou does protest too much, APA Electronic Co.

    As others have said, chasing outfits like Pimoroni seems odd when they should be suing Worldsemi 1st, surely?

    1. It also may look the company more serious. If they invested this much in patent protection you can be quite sure that they will invest in a lot of development in this type of product. It is a common marketing tactic, to the point that ‘useless’ patents are acquired just to be able to say you have a patent.

      In some countries you can obtain the patent for free or for very little cost and sometimes even without prior art search, until you ‘activate’ it. This allows you to market about your product having a patent without spending much in case you think the novelty is obvious and other non-infringing implementations are trivially made, or there is already a lot of prior art, or you will not go further in this line of products to warrant paying for the protection.

  23. What a shame, I actually quite liked the APA102 because they have a clock line – much easier to use in non-realtime scenarios (i.E. RasPi+Python). But I don’t think a want to support bullies like this.

  24. I read the claims. Each claim has prior art.
    * There already existed color changing LEDs -> prior art to multi-die packages of controlling ICs, LED dies and current limiting resistors
    * Using a shift-register to connect multiple ICs in series -> so much of prior art
    * APA-LEDs use a 2-wire SPI-protocol with data and clock whereas the WS281x uses a single wire protocol

  25. “Are patent trolling and false patent and copyright claim laws coming to your justice system?” Would be a better question.
    It’s a rampant form of abuse and it’ll only get worse until SOMETHING is done to discourage it..

    At least they still have firefighters trying to put out fires I suppose, look at the bright side eh.

  26. Assuming that the C&D order actually came from APA, which is doubtful, considering the way it was written and delivered, they should consider the case of The Flintstones. Hanna-Barbera’s Flintstones series was a pretty blatant rip-off of The Honeymooners. At one point, Jackie Gleason, creator and star of The Honeymooners, wanted to sue Hanna-Barbera for copyright infringement, but his lawyer asked him if he really wanted to be known as the man who killed The Flintstones. Gleason relented, and both series retained their popularity.

  27. Is it really patentable to “integrate” existing components into a single module? I mean, I’ve got some MaceTech “ShiftBrite” modules from 2008, which use an Allegro A6281 serially-controllable 3 Channel LED Driver chip whose datasheet says 2006…

    1. Yes. The patent system allows you to patent a machine, and of course the machine is in general composed of components which already existed and perhaps had been invented long ago. I’m not saying it’s good, or nice. That’s just how it is in law today.

      1. However that old stuff you found might be prior art. Hold on to it. I doubt we’ll be called upon to help these guys with prior art, but sometimes that happens.

  28. Solve it with a product notice:

    Due to a patent claim by XXX, we are pleased to offer a inferior replacement part at a premium. You will need some minor design changes to use this inferior replacement part.

    All you need to do, is:

    redesign your board to use a clock line, which the the deprecated part does not need.

    redesign your board to use a licensed proprietary protocol.

    Don’t worry, the new part costs MORE, and is proprietary. To meet demand, we forecast a 10 week lead time for qty > 1 part.”

    Remember, licensed parts cost more, so they are better, even though they involve you spending more money.

    We are pleased to offer no support for this part, and no warranty.

    You may write to XXX for the datasheet, which may or may not arrive.


  29. E-Mail Template
    Feel free to adapt and sent it to:

    Dear ladies and gentlemen from APA Electronic Co Ltd,

    due to your recent attempts to scare suppliers of WS2812 LED products by threat of legal actions [see 1] I like to inform you that:

    * I removed any of your products from my professional electronic designs.
    * I removed any of your products from my research projects.
    * I removed any of your products from my lectures for bachelor and master students.
    * I removed any of your products from my personal projects.
    * I added your company to my exclusion list of parts manufacturers.

    With kind regards


  30. It’s really nice that these companies are being promoted with all this advertising but what about all the companies here in the US like Adafruit and such who sell the same items?

  31. Weird an I always thought it was the other way around that WorldSemi was the first with those LEDs.
    For those who who are on about the clock signal on the APA102, there are also the APA104 that are the same as the WS2812 and are actually used primarily now by Adafruit for NeoPixel products instead of the WS2812 simply because they perform a lot better.

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.