Homemade SawStop Attachment Is Just About As Sketchy As It Sounds

TL;DR — when [Colin Furze] is your “safety inspector,” you really should be reconsidering your project goals.

Most of us have probably by now seen the SawStop brand of self-stopping table saw, which detects when something meatier than wood has the bad taste to touch the spinning blade, more or less instantly stopping it and preventing sudden traumatic amputations. It’s an outstanding idea, and we’d love to see the technology built into all table saws. But alas, SawStop saws are priced out of reach for many woodworkers, which left [Ruth Amos] to roll her own DIY version of the system.

It should be stated right off the bat that none of what [Ruth] does here is a good idea, and that everything shown is really just a proof of concept. The basis for her build was a somewhat flimsy-looking contractor-style saw, to which [Ruth] attached an Arduino set up to detect when something conductive touches the blade. She shares no particulars on the sensing method, but our guess is capacitive coupling. She then sets about experimenting with a series of above-table gizmos to arrest the blade, with limited success, plus all the attachments would make the saw essentially useless. But working above the table does make sense in the prototyping phase, and allowed her to figure out what wouldn’t work.

In the end, it was an electromagnetic clutch from an electric lawnmower that seemed to do the trick, albeit at the expense of heavy mods to the saw and a considerable increase in the system’s angular momentum. Nonetheless, the blade stops pretty close to instantly in the old hot dog test. It doesn’t drop the blade below the table, of course, and the hot dog is a little worse for the wear, but it’s still pretty impressive.

We’ve discussed SawStop’s technology before and why it isn’t perhaps as widely available as it should be, if you’re curious.

56 thoughts on “Homemade SawStop Attachment Is Just About As Sketchy As It Sounds

  1. I thought when they first came out that Sawstop was an attachment/retrofit to other table saws, and as they grew and developed/sold their own model they discontinued the attachment version.

    Do they license their technology? I know Bosch has a table saw that is supposed to have the same safety features as a Sawstop.

    1. I thought Sawstop sued (and won against) Bosch and are trying to mandate their technology (which to be fair actually does work though can have false and expensive positives) be a safety requirement despite only them (?) holding the patent? Didn’t seatbelts become standard even though Volvo patented the three point seatbelt?

      1. “Volvo then made the new seat belt design patent open in the interest of safety and made it available to other car manufacturers for free.”

        I’ve seen comments recently about SawStop lobbying to make their tech mandatory, paying youtubers to promote the idea etc. and a lot of folks pointing out that it works great in the lab but struggles with wet wood or humid environments etc. and of course trashes a saw blade every time it triggers, plus needs an expensive (and patented) new cartridge before the machine will work again.

  2. It would be cool if an electric brake could stop the blade fast enough – it would make it fully reusable.

    The rotational energy of a typical 10 inch blade at 4000 RPM is 600 J. Dissipating this in 100 grams of motor windings would raise temperature just 20°C, so the total amount of heat is not much of a problem. But stopping the blade before the next teeth hits means less than 0.5 milliseconds of time, for a peak power of 1 MW.

    With a 230V PMSM/BLDC motor it would take a current of 5 kA to accomplish this. If the motor windings have an inductance of 100µH, a voltage pulse of 2kV would be needed to get the 5 kA average current in a 0.5 millisecond pulse.

    So, it seems a 2 kV 2000 µF capacitor bank and some very beefy MOSFETs should do it. At $400 for just the capacitors it probably won’t come out cheaper than manufacturing costs for SawStop, though it could avoid the patent.

    1. Now this is the kind of homework I love to see laid out. A megawatt! Wow. I’d love to see someone try your capacitor bank solution, but something in me suspects that you’ll very quickly be chasing down dozens of other issues which crop up once you try moving that much energy around that quickly. If a lead melts somewhere or the motor winding is a bit degraded or an eddy current interrupts the flow, say goodby to your hand.

    2. “Before the next tooth” is a tall order. Ruth does remarkably well, way better than I would have though possible.

      The sawstop technique — throw something metal into the blade and absorb some of the energy in a breakaway bearing — is still very clever and probably failsafe b/c it’s mostly mechanical.

    3. I wonder if an alternative to firing a slug into the blade could be to use the blade like a disc brake and fire two large pads of friction material against the sides like a brake calliper, at least it would be less likely to trash the blade and may be more re-usable.

  3. Hey Colin would make a great safety inspector – he can do any and all the stupid things and is still here, and intact a decade later… That suggests any project you like can be done safely no matter how dangerous it seems when you make him your safety consultant (and operator).

      1. He spent most of the past 2 years since start of the quarantine building a bunker tunnel. There is possible a limited number of articles about digging earth appropriate for hackaday.

  4. The chopped sausage is an indicator of why the production sawstop not only stops the blade but retracts it too:
    The parts of our body most likely to be injured by a table saw are the hands, and particularly the wrists. When pushing a part through the blade, the time when you are most likely to slip and contact the blade is when you have reached near the end of the cut the the cut speed rapidly accelerates (because you are going cutting from with a significant portion of the blade length to a much smaller portion), which results in your hand moving more rapidly forward than you expect as you are now applying excess force to the workpiece. It’s also reflexive to keep fingers out of the blade path, and the way grasping with your hand works that means the wrist now moves closer to the blade path. A major artery lives a few mm under the surface of the wrist, so it only takes a remarkably shallow cut to sever it and cause an injury that is fatal in short order. At best, you are now left with one hand gripping the other wrist as hard as possible to stem arterial flow, and really hoping you set up voice-dial on your phone you can no longer manipulate to call for help.

    A saw that stops itself after killing you is not an effective safety measure.

    1. Oh no! The weekend shed project isn’t a finished commercial product ready for certification! The absolute horror and destruction it’ll cause to mankind is truly UNLIMITED.

    1. ^ this is an excellent reminder that there are multiple safety issues with table saws.

      Pretty sure the laws in the US changed a few years back and all new saws must include a riving knife, which does help mitigate kickback (esp. on rip cuts). It’s also the kind of safety feature that doesn’t get in the way (unlike a splitter or anti-kickback pawls) so people won’t be as tempted to remove it during use.


      1. Riving knife is pretty damn good.

        Anecdote time: I took a safety course here in Germany where the table saw had
        a) an overhead hood / blade guard / sawdust remover
        b) riving knife

        Between the two, there was almost no way to get yourself hurt at all. I saw people practically leaning into the side of the blade with a piece of wood, and it didn’t catch once. I have a bare-blade, janky, upsidedown circular-saw type setup in the basement, and slight deviations from perpendicular are instantly rewarded.

        Cross-cut sleds are great too, on the subject of putting the wood through.

        A table saw is a dangerous beast, but there are ways to tame it.

    2. While kickback can be shocking its not seeming at all likely to me to suck you hand into the blade, and if you are standing in the right place it should just fly past you harmlessly…
      Worst I’ve ever had was something flung back and bounced off the clamp on the sled towards my face. Plus it can only throw something heavy so hard. Very nasty but still nothing on putting your meaty bits at the spinning death machine part…

      I guess I can see your point though, the blade is a known and expected danger, kickback while it can be reduced is something that just jumps up at you out of nowhere. It may not be as nasty a result, but the shock factor is there.

  5. I have an extreme dislike for the “sawstop” thing.
    I won’t ague whether it works or not, that is not the problem.
    The problem with it is that is has been designed by 3 lawyers who have optimized it for making as much profit out of it as possible. “Safety” is abused here as a vehicle to make profit. It is not a goal, but a side effect.

    * The sawstop both uses expensive (USD100+) single use cartrides, and there is a high likelyhood that it wrecks your saw blade.
    * The saw is quite overpriced. It costs at least 3x more than othewise comparable saws.
    * The build quality is apparently quite reasonable, but it’s not on par with what you’d expect for a saw in that price range.

    Compare that with the Bosch REAXX.
    That also uses some cartridge, but it does not wreck your sawblades. It does not even stop the sawblade itself, but instead it just pushes the saw below the table very quickly. It also uses some kind of cartridge system. I think it uses cartidges that are designed for (similar to) car safetybelt tensioners. You can still buy the cardidges. A set of 2 costs around USD100, so half the price of a sawstop cartridge.

    Unfortunately they have been sued out of the market by sawstop (another reason I don’t like sawstop).
    However, It’s quite possible to make something similar to the Bosch REAXX yourself, and depending on the the way the height adjustment of your table saw is currently made, it may be relatively easy to do so.

    Or you make something in between. A big factor is reaction speed of the system.

    1. “Compare that with the Bosch REAXX.”
      Thanks for mentioning the name of the saw, your comment yesterday omitted it, leaving me wondering who built it.

      1. I’ve done some more reading and I I’ve found some more reasons why the sawstop thing is so expensive.
        They have over 90 patents related to saws. I’m not sure what a patent costs, but I think around USD10.000, so that’s close to a million USD invested in those patents.


        There is some more about the patent issue between sawstop / bosch in this link:

        And apparently:
        “The two SawStop patents that the ITC found Bosch violated will expire in 2020 and 2022.”

        So with a bit of luck Bosch will be back with their saw this year…

    2. Oops, forgot to add.
      The sawstop works with a capacitor that is discharged though a metal wire to melt it, and that further actives the mechanism with simple steel springs. The part that breaks costs maybe 5ct, but part of the design is also explicitly made to deform a big piece of aluminium, so “hobbyists” can’t repair the cartridges themselves.

      The more I think about it, the more I come to the conclusion that it is indeed been designed to be expensive, so those 3 lawyers can make more profit.

      I think there is a great opportunity here for an open hardware project. The design goals are:

      * Base it around a common and reasonable quality tablesaw.
      * Use the Bosch idea of only pulling the blade below the surface (stopping it quicly is difficult, expensive and not needed).
      * Use the detonator wire as used in the sawstop, Replacements then cost 10ct or so.
      * Make some build plans for this and put them on the ‘net.

      Note that patents prevents you from selling gadgeds, not from building them yourself. I’m not absolutely sure how patents relate to releasing build plans that use (part of) a patent.

      Some notes:
      * There are of course safety issues around a project like this.
      * But because only a piece of wire gets destroyed, everybody can do extensive testing of the system themselves.
      * It’s better then “nothing”.
      * Sawstop cartridges get activated quite a lot. From wet wood, to forgotten nails, or some aluminium foil on some isolation material you saw though. Apparently it even triggers on unreliable (generator?) mains voltage.

      1. Hilarious “It’s just lawyers!” conspiracy ranting aside:

        The breakwire is not the only consumable part. The catcher/deceleration block is consumable by necessity. Deformation of it is the fundamental mechanism by which the saw is decelerated and retracted in a failsafe manner. Without that, you need another energy sump within which to dump the saw energy within a handful of milliseconds, and that mechanism also needs to do so reliably and without active components (e.g. if you need to dump several hundred megajoules (a spinning tablesaw blade contains a truly extreme amount of rotational KE) into a braking motor, you now have an enormous energy bank that needs to be reliably charged on every spinup and discharged every time the machine is shut down, without failing, and a braking motor that needs to sit idle for years than reliably take a massive burst load, without failing).

        Just pulling the blade away alone whilst at full RPM is only partially effective, as your limb is descending from above due to the only reason your limb is there in the first place is because you have either been pressing down on the workpiece to hold it against the table, or fallen onto the table. You only need to descend a few mm into the tooth path to deliver a fatal laceration to a subcutaneous artery.

        You can see this in testing of the Bosch REAXX system:
        In both tests, the cut depth into flesh is the same depth to which the saw was exposed. The palm strike test, if conducted with the wrist rather than the palm, would have bee sufficient to sever an artery. The retract-only system simply is not fast enough to prevent fatal injury.

        1. > Blade can’t be retracted fast enough

          That depends entirely on the mechanism used to retract the blade, surely? Just because an existing system isn’t fast enough doesn’t mean it’s hitting a physical limit (except, you know, light speed.)

          1. The retraction needs to occur before the tooth that contacted skin and triggered the sensor can start cutting. By stopping the sawblade you are tackling that problem directly, but without stopping the saw that retraction needs to occur in a way that moves the cutting edge away from the skin. For most contact geometries that is not even possible (would need to ‘retract’ the saw upwards!) due to the individual teeth travelling along the rim of the blade ending up ‘descending’ at the forward edge. For the cases where it is (e.g. contact at the top of the saw arc and backwards form there) retraction speed would need to exceed blade rim linear speed, and exceed that speed starting from 0 with acceleration completing well before the next tooth reaches the same position.
            For a ~4000RPM 125mm radius blade, tip speed is ~50m/s. For a 50 tooth saw, there’s 300 microseconds between teeth passing a given point, so that would need an acceleration on the order of 170km/s (17,000 g). Not really feasible to move the saw assembly that quickly. On the other hand, moving an interceptor block into the saw blade can be done faster (it weighs much less) and uses the energy in the sawblade itself to move rather than having to fight the blade.

    3. Oh, it absolutely wrecks the saw blade when it hits- it fires an aluminum framework into the blade path, which slows the blade down at the same time a release on the module triggers and slams the motor and blade assembly down under the table. The instances I’ve heard of it happening render the blade either jammed into the module with no chance of recovery, and damages the tips on the blade as well.

      Is it overpriced? Very much so. Would I personally buy one? No. Too expensive for the contractor type unit, and TBH respecting the tool, following safety protocols when it’s used, and using proper PPE and push sticks/blocks/jigs all go a long way to preventing injury.

      If anything, kickback is a much more common thing to happen, which is why one uses push sticks to feed the material into the blade- I was making a cut for a project a year or so ago, and the material kicked back on me, resulting in the push stick hitting the blade; the material was flung off the table, the stick was flung out of my hand, and I very nearly had a ‘brown pants’ moment. I still use that push stick with it’s scar as a reminder to respect the spinny blade of dismemberment.

      1. > I still use that push stick with it’s scar as a reminder to respect the spinny blade of dismemberment.

        Indeed, I’ve done the same, though I think I need to make or buy a new one myself though, as a few nicks and it stops being a great push stick. Though it was the cheap plastic one that came with the tool in the first place, so one can argue was never a great push stick. Certainly not as good as the scrap of plywood and pine I’ve been using, and its always nice to have two, or more as you can benefit from different size and shape pushsticks depending on the job.

    4. A saw that costs 1/4 the price of a Sawstop (“3x more” means “4x the price”) model isn’t in the same quality tier as the SawStop.

      Blades can be repaired. The Maker/Hacker space I used to belong to had a SawStop cabinet saw and sends blades off for successful repairs almost every time the saw triggers.

      As far as “make as much profit as possible” well, that’s pretty much every business nowadays.

      1. Instead of speculation, do you have any data on how many false triggers there were and how many fingers (or other body parts) have been saved? (and over what time period / how much use, etc).

      2. I apologize, I was too quick about the price difference. I’m guessing an USD3400 PCS175-TGP236 from sawstop is somewhat comparable to a USD2150 Laguna Fusion F2. Still a significant difference, but not by far as extreme as I first thought.

        How fair is it to compare an USD1600 Sawstop Jobsite Pro to an USD630 Bosch 4100XC-10 ?
        As you already mentioned, price is only a part of the equation.

        I had not looked too closely at prices (my fault) because I can’t fit such a thing in my apartment anyways, and the workshop I can hobby with wood has an Altendorf F45, so there is not much reason for me to go buy a saw at the moment.

        1. I maintain a ramshackle arrack of 1940s through 1980s powertools, that look and sound scary, because if you gonna let one saw mollycoddle you, you’re only gonna turn around and drill through your hand, because you’ve got your brain no longer believing “ALL tools are dangerous”

  6. In addition to the electromagnetic clutch, what if it also shoved a big block of sappy wood into the blade? Maybe below the surface of the table. It could (maybe] save the blade from further damage and speed up the stopping.

  7. This is Yet Another Comment about goofy expressions on YouTube thumbnails.
    (As Charlie Brown used to say, “Bleah!”)
    It almost made me not to want to watch the video.

  8. I’ve never understood table saws, the table is the size of a cafe table for 2 and a whole sheet of wood has to safely sit on and be movable at all times. So it should be a garage filling thing the size of a car. A panel saw holds a full sheet stationary and the saw moves behind on tracks precisely with the sheet clamped on both sides of the cut, nothing to fly. The whole thing can fit beside a car in a garage it’s up against the wall.

    Yeah it probably won’t let you make those little cuts and such but those are where fingers get too close in the first place. That’s what a bandsaw is good for. Some in a hurry want to make that quick cut on such a quick but brutal tool, instead of cut accurately and in a reasonable time.

    Here hurry hurts.

    1. I’m down with that. For the big cuts, I use a circular saw and a guide rail. Less fancy than a full panel setup, and requires more floor space in use, but it’s cheaper!

      That said, how much would it take to DIY a real framed panel saw? Not much, I’d bet. Get your angles square and you’re done. Hmm… hobby-sized panel saw… Anyone ever seen one?

    2. Table saws are cheap, in part, because of how dangerous they are.
      My dad has a couple, both of which came free from the school he worked at, when they decided that they were too dangerous for kids to play with.

  9. I have a project that could be greatly enhanced with an Arduino sawstop like solution. It’s a plastic shredder, and it revolves much more slowly than a table saw so just shutting down power would probably be enough. I wonder if Ruth has made or would be willing to make her Arduino solution available to the rest of us?

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.