Fail Of The Week: The Little Ultrasonic Knife That Couldn’t

We all know the feeling of an idea that sounded great when it was rattling around in our head, only to disappoint when we actually build the thing. It’s a natural consequence of trying new stuff, and when it happens, we salvage what we can and move on, hopefully in wisdom.

The thing that at least semi-defeated [This Old Tony] was an attempt to build an ultrasonic cutter, and it didn’t go well. Not that any blood was shed in the video below, although there seemed like there would be the way [Old Tony] was handling those X-Acto blades. His basic approach was to harvest the transducer and driver from a cheap ultrasonic cleaner and retask the lot into a tool to vibrate a knife rapidly enough to power it through tough materials with ease.

Spoiler alert: it didn’t work very well. We think the primary issue was using a transducer that was vastly underpowered compared to commercial (and expensive) ultrasonic cutters, but we suspect the horn he machined was probably not optimized either. To be fair, modeling the acoustic performance of something like that isn’t easy, so we can’t expect much. But still, it seems like the cutter could have worked better. Share your thoughts on how to make version 2.0 better in the comments.

The video is longish, but it’s as entertaining as any of [Old Tony]’s videos, and packed full of incidental gems, like the details of cavitation. We enjoyed it, even if the results were suboptimal. If you want to see a [This Old Tony] project that really delivers, check out his beautiful boring head build.

43 thoughts on “Fail Of The Week: The Little Ultrasonic Knife That Couldn’t

  1. This Old Tony is one of the best youtube channela out there, even f you’re not interested in mechanical engineering. His witty, absurd and unexpected commentary gets you laughing and rewinding back to check what he said.
    These days is not easy to find a youtube channel that is sponsor/patreon free. Most youtubers are in for the money, but not This Old Tony. If he sis, it’s not obvious.

          1. There was an entire Internet full of people sharing knowledge for nothing, back in the day. Much better than the money-grubbing crapfest we have now.

            Since the masses, and business came to it, there isn’t any more useful information or fun or interesting stuff. Just a lot, lot, lot, lot, lot more shit.

          2. I recall the marketing strategy for most business’s, especially those that invest in others business claims, patents, trademarks, etc. to first target what is compelling people to buy in even if a social service that is designed to be for the general public benefit. Then create some sort of seduction, coercion, manipulation and maybe even hostile takeover intimidation scheme to steal the business that was say for example not for profit or non-profit. Not cool man. I have to hunt down nuclear power and technology commercials of the 50’s selling the technology of the future to create free systems that like some plants there are excuses why they need to be torn out due to false claims. Like I was thinking hydroelectric damns with fish hatcheries. What happened to fish hatcheries for a hydroelectric damn ecosystem that produced more fish in a season than the natural resource in wild form created with stricter waste clean up solutions than the more toxic natural resource? Weird.

          1. As opposed to the fake jobs where people are still willing to trade money for skills, but you think those skills could be better used elsewhere?

            As someone who packed in my 9-5 engineering slave job to travel and earn money from odd-jobs that were previously beneath me, I love seeing the tall poppy syndrome at work. Literally at work too, because Id bet you’re all sitting in an office you hate as you bash out your criticisms.

      1. Having free ad-free content online and getting paid to make content are not mutually exclusive.

        The problem is that people want to first get paid for doing the work (patreon/sponsor drive), and then get paid again for letting people watch it (paywalls/subscriptions), and then get royalties if someone else finds new use for the work (copyrights), and then get paid tips/donations/admonies just because that’s possible as well.

        At this point it’s well to point out that ad money is not free money – it doesn’t come out of nowhere. It comes from the people who buy the products being advertised, which means you. Simply by adding to the view count you’ve increased the amount paid, therefore increased the amount you pay. The more ads there are in general, regardless of whether you personally view them, the more you pay, so the less money you have for spending at your own discretion.

        So, people in general try to extract the most money out of the least work by all possible means, by tricking you into paying multiple times, which turns the game of content production less about the content and more about the different means to monetize it. The content itself becomes almost irrelevant.

        This also means all the money goes to those who “cheat” in this way, and those who try to do honest business by offering real content for a fair pay don’t see a lot of customers going their way, because everyone’s already spent their money on the bullshit mongers – all those “Support me on Patreon!” types who are actually just panhandling.

  2. Apologies if this was covered in the “longish” video but i’d presume that the particular waveform of the ultrasonic transducer would be a factor in cutting efficiency, particularly with thicker targets. The more common waveform used by ultrasonic cleaners is triangular (more diversity of (Fourier) frequency components, it’s presumed), whereas for cutting solids I’d wildly guess (‘pun’ unintended) that a sawtooth might have more power to establish breakthroughs in more solid sub-layers …or not.

    1. Can you point to any references about the triangular waveform? I think your confusion comes from the frequency sweep (which is within 1-2 Khz of the resonant frequency which may be done via a triangular shape (Freq. vs time).

    2. TBH, no ;-) The transducer plus horn assembly is a highly resonant system, and will vibrate in a sinusoidal fashion at the fundamental frequency of whatever waveform you feed it, be it sine, square, triangle, etc. If there happens to be another, weird, resonance at a harmonic of the driving frequency, then things get interesting, but it’s pretty much nearly always sinusoidal.

    1. i was thinking the same thing
      I’m no expert at all but my gut feeling is the blades are too thin and flexible and probably absorb a lot of the vibration on their own instead of just “passing” them along to the cutting edge. also doesn’t mass in the horn play a crucial role in the transmitted power?

  3. It appears deceptively simple. Attach a knife firmly to a piezoelectric element. Add something to counter the vibration at the other end. Do some math and punp it full of the resonant frequency.
    In reality even the commercial ones are temperamental.

  4. “We think the primary issue was using a transducer that was vastly underpowered compared to commercial (and expensive) ultrasonic cutters, but we suspect the horn he machined was probably not optimized either.”

    If you’re just going to echo the conclusions in the video, you ought to at least credit those conclusions to the guy who did the work, instead of phrasing it like you’ve got some remarkable personal insight.

    1. Lower cost manufacturers do not bother with driving transducers at exact resonant frequency. They just FM modulate the driver around the resonance so they get a general changing pattern of sound peaks in the water bath. You have to disable this modulation and then manually adjust the frequency to what that gives you enough cutting performance. It most probably not be 40kHz. Normal epoxy will not work for you here. Try JB Weld metal epoxy and torque your studing to what the maker of transducer specifies. Also 60W will be enough for cutting. Harder materials will need abrasive south between blade and what you want to cut

  5. Why not just add a little vibration motor from an old smartphone or an “ultrasonic” toothbrush an try your luck with that? The effect on a cutting blade will probably be around the same as with Tonys experiment, not worth the effort in the end i guess…

  6. I was really excited to see TOT tackling an ultrasonic transducer, love his channel. I tried running a knife blade with my own horn, and didn’t have much success. The main problem seemed to be that the addition of the knife really killed the resonance, probably because there’s no longer a nice flat surface for soundwaves to reflect back from.

    Dismantling the trandsucer is also a no-no (the bolt tension is carefully adjusted to tune the resonant frequency during manufacture), as is holding it by the rear mass, since both ends must be free to vibrate.

  7. Reminds me of the vibroblade from Battletech.

    I wonder if he’d have more luck vibrating the blade electromagnetically? Devise a coil to vibrate it in the appropriate direction. Rustle up some appropriate mechanics, maybe something heavy and strong for the blade to be attached to. Perhaps a metal slug that goes in the coil, and the blade is screwed or welded or whatever to that?

    Certainly a coil would offer a lot more parameters to mess with. Physical size, distance, frequency, amplitude, all would have a lot more range to play with than a piezo sounder.

  8. I recall some 30 years Eaton gmbh brought put ultrasonic injectors for petrol cars relying on fuel pressure with a spring holding a ball in place at the ejection spray point upon some suitable seat material. When the ultrasonic drive was On it would cause the ball to oscillate at a close enough resonant frequency then emitting a very fine spray – various orifice sizes and spring tensions offered a wide range of droplet sizes and delivery flows depending on pressure and temperature of the petrol.

    The body of the ultrasonic injector was effectively a machined aluminium ‘horn’ of an 1/e^-x shape approach with the farthest part holding tightly a disc with the titanate media and a pair of electrodes. Reports suggested it worked very well but, iirc it was considered helpful for long term reliability to have an upstream cutoff solenoid or better a fuel pressure release when ignition off. I understand the project was shelved despite the fact the spray was superior to that from a conventional Bosch injector – fortunately any patents on the issue well and truly expired ;-)

  9. I thought it worked spectacularly well for a junkyard wars style encobblement! He said ‘angle of attack,’ and didn’t really follow up. That’s the secret. There’s a sweet spot where it cuts best. Stacked paper is essentially wood. It cut through that stack like butter AS COMPARED to a finger-powered knife. The paper was scorched, meaning a LOT of that 60W was being transferred — that was the scorch-equivalent of a 20-30W soldering iron. The clear blister pack plastic was the best. You KNOW how tough that is to cut cleanly. Do I need one? Yes. Can I or do I want to build one, no. But the vid answers the question of how much power I need and will it do the job which is an emphatic and thrilled yes. Great vid.

Leave a Reply to Noah AndersenCancel 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.