Ultrasonic Transducers - Measurements and Horn Design

[@FelipeSantos-sw4kk]

0:07 Introduction & Circuit 3:42 Effects of Loading 4:32 Horn Design & Initia Testing 9:56 Finished Horn, Different tips, Nodes & antinodes 19:16 Sinewave vs. Squarewave 22:41 Temperatura Effects 24:41 Conclusion 30:29 The End

[@irfanbutt7991]

Nice

[@irfanbutt7991]

I have 400 modal i whant horn siez please halp me

[@upendergupta7885]

nice for introduction to measure resonant frequency

[@ROBOROBOROBOROBO]

Damn 7 years later, I find this absolute gem on youtube. This is treasure, thank you for making this. I make an ultrasonic cake and pizza cutter, using transducers found in cleaners as my thesis project, I dont know if I can get enough power out of them, maybe I use multiple of them and make an optimal horn design before connecting the blade

[@TechIngredients]

Thanks for the video. It was well done and the descriptions as well as the demonstrations are valuable. Before watching this, I questioned the need to mechanically tune the transducer/horn system rather than simply designing the driver to operate at whatever frequency an arbitrarily sized system requires. This had been mentioned elsewhere in your comments. However, you were able to add the mass of a horn and with careful tuning, match the bare transducer's resonance frequency at the first fundamental. Your 13kHz starting point demonstrates that you are not tuning to a harmonic. We will be constructing an ultrasonic cutter and your work will make this much easier. Thanks again.

[@nitrousman8882]

I wonder if it might be useful to consider changes in material in terms of resonance at the working end of the horn ie going from aluminum to brass.

[@themonkeydrunken]

@Tech Ingredients - Are you still planning to build an ultrasonic cutter? I'm eager to watch your video on the matter.

[@DerSolinski]

Why is that I find often comments from either Dustin or you under underrated vids I watch. Are our tastes in video consumption really that similar?

[@alocin110]

Lindsy: This is one of the best video I have ever watched in my entire life. The subject matter is well explained and demonstrated. I had been looking for something like this for years to aid one of my research work. You have done a great job and I must say this video would help many experimenters. This was very informative, educational, and well demonstrated. Thank you again. Please keep up the good sharing. I liked your video.

[@3metrosarequipa57]

Very nice!!!! PLL control loop is what you need to auto track the resonance. i had it done... ... working on double PLL method now, there is much power in harmonics also, for cleaning applications mainly. Heat and load conditions change the 28khz nominal on impredecible ways, it is important in industrial scale aplications. regards from Peru.

[@DerSolinski]

Thank you, your explanations allowed me to get my ultrasonic welder working again with the customized tip. It auto tunes automatically and the 2x2mm radius reduction was enough to whack it out of it's operating range.(

[@imajeenyus42]

That's great to hear! Nothing beats just messing around with things to see if it works ;-) The only thing I'd add is to make sure that there is good contact between the washer and the other parts of the horn - if it's a stamped washer, the faces will likely have a bur, so give them both a sand on a flat surface to improve contact. Otherwise it might dig in and loosen over time. And ideally it should be made from the same material as the horn, but that's probably harder to achieve.

[@DerSolinski]

@@imajeenyus42 Should hold and its a quality washer so time will tell, part of an automated machine. Still a lot of other kinks to iron out before the entire thing is production ready anyway. But at last it spit out the first signs of life and parts. So progress, yay. Just hope its predecessor/sibling doesn't kick the bucket in the mean time. There is some stock left but orders are somewhat unpredictable 🤷‍♂

[@irfanashraf1238]

All the hallmarks of an excellent teacher. “Hats off Esquire ! “

[@Jacopo.]

thank you, you made me save thousands of €!! this video is made very well and you explain all the details very clearly

[@wayne5000]

This was well done. At the end, where you warm up the extension by torch, I think the change in your resonant frequency is an interesting phenomenon that is not easily explained. Temperature change can affect both density and bulk elastic modulus. Either can affect the speed of sound and thus affect your resonance frequency. There could also be some influence of thermal expansion and thus your geometrical constraints change. Being that this is material property dependent, the functional parameters must be well understood and does not apply neatly across all solids the same.

[@imajeenyus42]

Thanks! You're absolutely right - there will be multiple interconnected effects and trying to predict what actually happens to the resonant frequency (increase or decrease) is probably very difficult. I'm not a theory person so don't really know much about it ;-) I wonder if horn designers actually take temperature into account when designing a horn - e.g. a high-power horn under continuous operation might run at an elevated temperature, so you might want to design its profile so it's at the correct resonant frequency at that temperature, not sure. There's actually a vaguely similar thing with laser diodes - since the wavelength increases slightly with temperature, if you're trying to match the wavelength to say a narrow absorption line (e.g. pumping a NdYAG crystal), you need to specify what temperature the diode will operate at when buying it from the manufacturer.

[@zbigniewlipinski2963]

Thank you. I was asked by a friend for help in diagnosing an ultrasonic stack but the trouble was that I have never worked with anything similar, so I've been looking for knowledge in the net. Today Google recommended me your film and now at last things look clear. Very well explained - the best material on that topic I was able to find.

[@ats89117]

Very interesting video, but I'm not sure there is anything sacred about the resonance point of the bare Langevin transducer. I would think that what you are really interested in is maximizing either mechanical power output, or mechanical power output divided by electrical power input (maximum efficiency) in the final configuration, at whatever the resulting resonant frequency is. Another thing you can consider doing is using an impedance measurement device to measure the impedance of the transducer around the resonance point. In this case you will see the impedance of the transducer go from close to -90 degrees at low frequencies (pure capacitance), up to zero degrees at resonance, to above zero degrees above resonance, then back down to zero degrees at the anti-resonance frequency, then back toward -90 degrees until the next resonance is approached. Of course this is what you are showing on the oscilloscope, but it's nice to have it quickly generated as a graph. The real part of the impedance which represents the load's power dissipation can also be graphed. The Analog Devices AD5933 development board can be used to perform this testing and can be purchased fairly inexpensively...

[@seeigecannon]

Fantastic video. I have been playing around with an ultrasonic converter at home and the original probe end was cracked. I got titanium and made a new one, but was having a big problem getting the right frequency (I barely made a change on the narrow end and blew past my target). This is exactly what I was needing.

[@imajeenyus42]

Thank you, that's great to hear! They are really sensitive to adjustments, glad you got it sorted out.

[@gristlevonraben]

I literally feel ten times smarter after watching this. I'm going to check out your other videos to see if you have cut things with it. You give me ideas about how to create resonant waves in magnets to make them stronger. Very awesome. I have an unfinished video about how I believe magnets and gravity and stuff works. I should have sound added in three days. Anyway, wow, so awesome. I know your work seems non-consequential to many people who are not into this sort of thing, but actually you've opened the door to advanced resonant matching in materials that can go beyond ultrasonic. Even in the ultrasonic, I was watching your video and wondered if you could recreate the sonoluminescence effect of when soundwaves collapse water bubbles to create temperatures four times hotter than the sun, for microseconds! What if instead of collapsing water bubbles, you used it to collapse the side of a magnetic field that is nearly spherical? Anyway, great video.

[@imajeenyus42]

Glad you like it! Single-bubble sonoluminescence has been recreated by quite a few people, usually in a spherical glass flask with two opposing piezoceramic discs attached diametrically opposite each other. I think there was a Scientific American article a while back about this. I actually observed it myself once (the important word is ONCE!), but it is very hard to get the right water conditions (correct amount of dissolved gas). Don't know how vibrating magnets will affect things but it sounds an interesting idea.

[@cntrlengr]

Used your testing method to find resonant frequency on a new trial setup. Worked great! Good information and well presented!

[@kozlov-photo]

Great video lecture! Thank you very much. Perfect to quickly get the feeling of how to approach ultrasonic transducers if you are completely unexperienced.

[@astraymark244]

Hi Lindsay. Great video, clear and concise content. Came across your website many years ago when I too were trying to drill fine holes in ceramic, like you, but for miniature gas burners. Gave up at that time as didn't have access to a lathe. That has now changed 😀 Looks like I need to get a DSO now.

[@edgeeffect]

You can divide by 1 for as long as you like, do it again and again and again if that's what you want. ;) Excellent video... Keep up The Great Work.

[@lokeswararaod8296]

Excellent video. Very interesting and wonderful presentation. This is a very good demonstration of the principles for new designers of the horns.

[@danfrankovic3413]

Lindsey fantastic video very clear and explained well, thank you. I would like to build my own transducer and ultimately make a ultrasonic cleaner. Now I need to find how to attach a transducer and how to tune it using a stainless container and water as my horn. I subscribed hope to catch more of your videos. You're a good teacher. Winter Garden, Florida

[@timappenroth3639]

Hi Lindsay, awesome video. The whole topic is somewhat complex in its own way but fascinating. Two little question: 1. What is the white cable parallel to the transducer (a capacitor?)? 2. The maximum current draw and phase equality is not at the same frequency but it should be, right?

[@imajeenyus42]

Thanks! Well spotted ;-) It's actually a small high-value resistor (100k I think from memory) whose sole purpose is to keep the transducer discharged when not in use. Otherwise, what can happen is the transducer expands or contracts due to temperature changes, generates a large static charge on the piezo elements, and give you nasty surprise when you go to pick it up ;-) It doesn't affect the performance any. As for current & phase, yeah, unfortunately it's not a perfect world and the two situations indeed don't occur at exactly the same frequency. I _think_ as the system is loaded more, and the resonance broadens, the difference gets larger. In practice, you would probably choose to operate at the point where the current is in phase, to keep whatever inverter is driving it happiest, and put up with the slightly less efficient current draw.

[@LBCAndrew]

So this would explain they my ultrasonic cleaner lost a lot of cleaning power when I replaced the tub with an identical one, but went a bit overboard with the amount of epoxy? all the extra mass where it squished out the sides around both transducers is probably causing the resonant frequency to dip too much.

[@SupraNaturalTT]

This was exactly what I was looking for. Wanting to build my own ultrasonic homogenizer.👍

[@anullhandle]

Here from thought emporium and this old tony ultrasonic adventures. Good stuff.

[@imajeenyus42]

Thanks! I didn't know about thought emporium's video, will give that a watch now ;-)

[@Yorumcu63]

Best video about Ultrasonic Design in KZbin

[@2lefThumbs]

Very interesting video👍 Couldn't you speed up the tuning/reduce iterations by using your initial measured length and resonant frequency to calculate the actual speed of sound in your material, then recalculate the required length (rather than shaving off a mm or two and remeasuring resonance repeatedly)?

[@imajeenyus42]

You could certainly do that, it would make things a lot quicker. I shortened it progressively mainly because I wanted to see how the frequency changed, and I was worried about going too short!

[@klazzera]

Overall great channel and a great video on ultrasonic tool design. taught much more than anything i've looked up in the past week. This video explains the basics of horn design and sonotrode over an existing transducer, but i wonder how to make a transducer from a bare piezo crystal. Is the process same for tuning the back and front pieces in a transducer? I guess the resonant frequency of a bare crystal would be much higher since it's very short, it gets me confused. Do you have any idea on that?

[@trhiaxi]

First of all thank you very much. We need to use 3d software to make more complex horns. What would you recommend on this?

[@imajeenyus42]

I'm sorry, I can't help.

[@vishnudas9476]

Why do you need a horn ? This is very fascinating can you suggest any book to read up on this subject ?

[@Byzmax]

Excellent video on so many levels. Thanks for doing such a great explanation.

[@FelipeSantos-sw4kk]

Hey Lindsay. I'd like to congratulate you for the fantastic content you've made. That's some priceless informations about an important subject. I'm building a horn to make extractions into alcohol. Do you think it's better to have one node inside the extraction cup or more?

[@imajeenyus42]

Hi, glad you liked the video! To be completely honest, I'm really not sure what would be best for that. I think I just about understand the operation of the horns themselves, but I don't know much about actual applications unfortunately. I would imagine that it would be sufficient for the end of the horn to be immersed in the extraction mixture, since it's at the end where most activity takes place. Hope that helps.

[@FelipeSantos-sw4kk]

Good night Lindsay, I'll develop a research at Uruguay in this area. I'll keep you up to date as the study evolves. Thanks for your attention, have a nice day.

[@Kikutuca]

Nice explanations, methods and knowledges. Thanks for sharing. Congratulations and regards from Brazil.

[@vahabsolgi9953]

Great job Lindsay, thanks for the information you gave in the video, they were tremendously useful for me. Could you please explain how much the tightening torque is important in achieving the best results and how can I find the optimum torque? moreover is it better to use epoxy adhesive between the horn and the transducer?

[@imajeenyus42]

Thanks, glad you found it useful! If you're referring to the bolt that holds the two halves of the transducer together, this should NEVER be touched, since the tension in the transducer determines the resonant frequency. However, if you mean just how tight the horn should be attached to the transducer, then it isn't really too critical. I just tightened with a couple of spanners and it worked fine. They actually tend to tighten through vibration so can be harder to remove. There's no need to use epoxy - obviously you'd never get the horn off again!

[@vahabsolgi9953]

Thanks, Lindsay, one more thing, in the case of adding a booster to increase the vibration amplitude can one make a part which can be used instead of both booster and the horn!? or it is necessary for the booster and horn to be two separate parts!? in other words why they make the booster and horn as two separate parts instead of combine them as one part?

[@imajeenyus42]

You probably could make something that was a combined booster+horn, but I think the main reason they make them in separate parts is convenience - with a separate booster, you can easily use different horn types, rather than having to re-make the entire thing for each one.

[@benjameslari]

Really well explained. Very didactic, visual and easy. Thank you.

[@phild_up]

WHAT A GREAT VIDEO!! I love your explanations. Thank you so much for taking the time to share what you've learned. I feel like you've given enough information for me to run with my project! Great Job!

[@imajeenyus42]

Phil Dupuis Thank you! I'm really happy you found it helpful!

[@phild_up]

Hey Lindsay, do you know what direction the horn is actually vibrating in? If the center of the horn has less vibration, it would lead me to believe it's vibrating either A.) side to side or B.) up and down but NOT C.) in and out. Do you have any thoughts on this?

[@imajeenyus42]

Great question, and it's one that confuses a lot of people, me included! The short answer: go here and look at the first animated image on the right: www.sonoanalyzer.com/content/ultrasonic-horn-design-using-sonoanalyzer-basic-guide. The longer answer: both ends of the horn vibrate purely axially. The middle of the horn, although it doesn't vibrate axially, DOES vibrate radially - it "breathes" in and out. Not by much, if you look at the relative amplitudes on that simulation, but it definitely does. This caught me out when I first designed my horn - I thought that it didn't vibrate at all in the middle (no radial, no axial) so I clamped the middle rigidly to mount it, and found that it didn't vibrate. Because it does vibrate radially, there needs to be some degree of elastic "give" in the mount to accommodate this - the usual trick is to have a ridge on the horn that's sandwiched between two rubber O-rings.

[@phild_up]

Wow, the motion of the horn was unexpected!! So it does breath in and out. That actually works best for my project! Thank you for that link and your great help. I very much appreciate your responses and kindness.

[@phild_up]

Hello Lindsay, when you setup your transducer for any operation, what equipment do you use to power your transducer? Do you use an ultrasonic generator?

[@TomashPL58]

You have made my life so so soo much easier now. Thanks!

[@hraklisrevo408]

nice video. You play with the shapes , good to see what happens if we play with the heat

[@dantosrock43]

Hallo.. terimakasih atas penjelasanya.. saya mendapatkan jawaban atas riset saya disini.. bahwa pada probe tranducer piezo juga tidak dapat bekerja dengan gelombang kotak (saya sudah mencoba) dan saya baru faham bahwa ini membutuhkan gelombang sinus. Terimakasih sekali ini sangat membantu. Konten ini sangat istimewa❤

[@timpower9660]

Thanks for the video Lindsay. Have you every experienced that the resonant fequancy is relating to the input power? With my transducer, the higher the power the lower the resonant frequency

[@imajeenyus42]

I have seen that as well - probably because these things aren't exactly nice linear systems and funny things will happen at higher power levels.

[@nitrousman8882]

Excellent video! You’re a master at instruction. Have you’ve done a video on modifying resonant frequency of commercially available driver boards? Thanks. Doug

[@johnfish837]

A sliding, adjustable horn design would be very useful...Something that would allow one to quickly adjust resonant frequency and also conduct the sound energy efficiently...A section of telescoping tubing or maybe a split shaft with sliding clamps...

[@imajeenyus42]

Unfortunately, any sort of moving/sliding parts would likely totally kill the resonance - you need to have the horn acting as one solid mass of metal for a good sharp resonance. Any interface between parts of the horn would disturb the resonance and lead to lower efficiency.

[@johnfish837]

I thought that would be the big challenge...I was thinking sliding parts that could be clamped down tightly, similar to the joints already used in a horn...So you could loosen the tension on the clamps or bolts, adjust the length, tighten them back up and then check the resonance...It would sure be easier than re-machining the end numerous times, plus you could use the same horn with various different tools at the end.

[@xrayrep]

I don't understand how you can add material to the original transducer and arrive at the same resonant frequency as the unloaded transducer. I would assume that adding mass would always lower the resonant frequency. Also, why is it so important to attempt to tune the horn so that the original unloaded transducer frequency remains as before. I mean, so what if the resonant frequency changes if you can simply adjust the generator frequency to the new resonant frequency of the transducer stack? Make sense?

[@imajeenyus42]

Answer: harmonics! Yes, adding material (lengthening) to the transducer will lower the resonant frequency, but remember that the system will also oscillate at higher harmonics (2x, 3x, and so on times the fundamental frequency). Example: say you've got a transducer with a fundamental resonant frequency of 20kHz. Say its length, for sake of argument, is 10cm long, and assume it's a solid chunk of aluminium as far as sound waves are concerned. If you add 10cm more to the length, in other words by screwing on a 10cm long horn, you've doubled the overall length of the system, so the fundamental resonant frequency will be half, at 10kHz. But the first harmonic will now occur at 20kHz, and this is where we drive the system. The transducer will then still be operating happily at its design frequency of 20kHz. As to the importance of matching - piezo transducers are designed to operate most efficiently at a particular frequency. Lots of factors affect this - the overall length of the transducer, obviously, but also the dimensions/thickness of the piezo rings and the compressive force applied by the bolt. The compression introduced by the bolt ensures that the piezo elements never experience tensile forces which could break them, but this assumes that they are located at a nodal point. If it's operated at something other than the design frequency, they might be damaged. It's a complex subject, so it's simplest to just say the transducers are happiest and most efficient when run at their design frequency.

[@iamcoolerthanconnor]

Thank you for sharing the knowledge!

[@arthurprior4638]

This is an incredibly helpful and very well put together video. Following this method, do you have any idea what accuracy I would be able to tune a horn to? I am told I would need to get it within 50 hz to avoid damaging an ultrasonic generator. Is there a paper or any other kind of literature showing the use of a function generator and oscilloscope in a similar set-up for measuring horns?

[@imajeenyus42]

I might be wrong, but I'd question whoever told you that ;-) Most generators will automatically tune themselves to the resonant frequency of whatever load is connected - I've never heard of one requiring a very precisely tuned load. Yes, you want to get the horn resonance as close as possible to the transducer, but after that, it's up to the generator to then drive it at that resonance.

[@MilanKarakas]

Very interesting presentation. I will back to this topic later, since currently messing with my first 3D printer. :D Thanks.

[@tajpibrallan]

Thank you for an excellent video! I've read through all the comments but there's one thing I can't wrap my head around. If the only parameters that are important for the horn are length and material, why are not all horns shaped like a simple cylinder? Most of the horns for ultrasonic welding are either stepped or hour glass shaped. For example, are the two horns in your video the exact same length? (assuming it's the same material)

[@imajeenyus42]

Thanks! It's a good point, something I never actually mentioned. The reason for a stepped horn is to amplify the vibration from the transducer. A stepped horn might increase the vibrational amplitude by a factor of 2-3x, while decreasing the force by the same amount. Think on it like an electrical transformer - if you step up voltage, it'll step down the current as well. There are also weird things called boosters, which also seem to amplify, but don't appear to be stepped - however I haven't a clue how they work!

[@tajpibrallan]

@@imajeenyus42 okay makes sense! Thank you 🙏🏼

[@igortsukrenko7465]

Thank you! Really useful information!

[@shitalkakade3828]

Hey Brother.... thanks for the Video, If we don't have scope and we give parallel connection of Function generator to Micro Ampere meter then (other parallel connection will go directly to transducer) , at the frequency where meter needle fluctuates to Max. is resonant frequency? Thank you in advance.

[@imajeenyus42]

Yes, that is correct - a series current meter would give a maximum reading at resonance. However, you would obviously need a meter that has sufficient frequency response to reach the frequency that you're operating the transducer at. Most digital multimeters only go up to maybe 10kHz so wouldn't be suitable.

[@johnfish837]

Excellent video. Very informative. Thanks, Lindsay!

[@avineshjain481]

Hello Lindsay...1) What if the tool is directly connected to the bare transducer? Then there will be need of horn? 2) How to calculate how much amplitude will be there if bare transducer works at 25 kHz?

[@imajeenyus42]

You probably could connect the tool directly to the transducer, but the vibration amplitude will be less (the horn increases the amplitude). Besides, it then gets very awkward to actually use, because the tool is so close to the (wide) transducer face. As for amplitude, I'm afraid I have to idea how to calculate it - depends a lot on the loading, efficiency of resonance etc. Very, very roughly, amplitudes are in the tens of micrometers for these sort of devices. Check out my other video where I measure the peak-to-peak displacement amplitude: kzbin.info/www/bejne/hXq0ZIePZtyejqs

[@babakzolfaqari2594]

You have done very great job. Absolutely brilliant.

[@kv4atv]

Is the Transducer being in direct contact with the table something that will alter the frequency due to it will pass some of the frequency to what it is in contact with just like the horns you put on it? Just something to consider...The horn is no different than the length and diameter of an Antenna. The horn is an antenna. Higher frequencies use shorter or "smaller" antennas. There are simple formulas that will allow you to calculate the length of an antenna depending on Frequency. In HAM and MARS radio we build out own antennas all the time.

[@imajeenyus42]

The effect is pretty insignificant, and it only really damps the resonance very slightly, it doesn't change the frequency. If you look at 17:30, where I'm showing the effect of pinching the horn at various points, you can see how the resonant frequency doesn't shift.

[@realcomsolucoes7403]

thanks for sharing all this nice work. That's really a Good Lesson on KZbin.

[@kishoregowda4455]

I have purchased a new 36 kHz 1200 watt transducer as the original stopped working. The new transducer is heating up within a couple of minutes of being connected to the generator, even when the bare transducer is connected. Would you say the problem is with the generator or the transducer.

[@imajeenyus42]

I'm afraid I really couldn't say. Does the generator have any sort of retuning process that you have to go through when changing a transducer? Maybe it's not driving it quite at the resonant frequency and that's why it's heating up?

[@i2c_jason]

Love the video. Can you operate this transducer in a translational mode with a high voltage input to use it as an actuator? Or are there not enough Piezo donuts to be practical?

[@imajeenyus42]

Thanks! I don't think so - if I have my figures right, a 5mm thick piezo disc made from PZT will change in thickness by only about 60 nanometers with 200V applied. Things are obviously different at resonance, where the amplitude is far greater - I measured about 20 microns peak-to-peak. Piezo actuators generally have many stacked elements, or additional mechanical amplification with a flexure.

[@reatoruv]

Thanks for sharing. All knowledge of ultrasonic horn I was needing I found in your channel. You won a subscrite.

[@zxz2530]

Great video, finally can understand it well, thank you !

[@danbrunermer4295]

Another question, if I might. Can you use the measured resonant frequency of a TXR + Horn or the difference between the two resonances to calculate the speed of sound thru this particular piece? You were using 5800m/s, but as you noted, that 'published number' is all over the charts. So you can't make perfect predictions about length prior to having the object you're going to attach (like your piece from the waste bin). But after you've measured the new resonance and found this frequency, can you somehow do the calculation backwards? I mean, you take the piece and then go turn it down 'a couple of millimeters'. Could you have taken a precise measure of the horn's length and the TXD length and figured out a precise number to shorten the horn by?

[@imajeenyus42]

Unfortunately not ;-) You might get an approximation, but again nothing accurate. The transducer + horn assembly doesn't vibrate exactly as a simple half-wavelength system, there are little variations, breathing modes etc. However, you could make a rough guess - e.g. if taking 1mm off raises the frequency by 1kHz, and you need to raise it by 5kHz, then PROBABLY taking off 5mm would do the trick. It'll likely be a linear response over small adjustments. But I'd still sneak up on the final value ;-) [Having said that, I did take a big chuck off my original horn by mistake (screwed up calculation), but luckily it ended up pretty much exactly where I needed it!]

[@jshaw4757]

​@@imajeenyus42hello mate 6 years late is OK too talk

[@imajeenyus42]

@@jshaw4757 What did you need to know?

[@jshaw4757]

@@imajeenyus42 Hello mate..I was just wondering are these the piezoelectric transducers and can they be used too generate power ..I'm very interested in batterys n energy n wish too make a system myself and I just wondered if these things could play a part in such a system as I plan too combine things too try n come up with something...but yh or just short info on them in lamens terms I prefer talking too people than just using Google..thanks alot

[@kred7776]

Great video! What are you using the ultrasonic drill for? You should do more videos about working with glass

[@imajeenyus42]

Hi - glad you like it! I originally wanted the drill to make holes in ceramic discs for supporting discharge-tube electrodes, but I've kinda got away from that stuff now. I've got a page on the drilling here - imajeenyus.com/workshop/20110516_ultrasonic_drilling/index.shtml. Yeah, I've got a couple videos in the pipeline about laser-cutting of glass tube, should be interesting!

[@WhatAWondWorld]

Thanks for the video.

[@matteorossi5128]

why didn't expose on paper the theory to calculate the length of the attached body, It was such a cut of time for me lol, Getting back on what you studied 20 years before just because you need, it becomes an adventure! lol, but thanks for the video You just simplified my research a lot! thanks for refreshing my memory in a simple way!

[@danbrunermer4295]

Lindsay, thank you so much for making this video! This is so much easier than the method from imajeenyus. I have a question though. I was looking at components to make my own transducer, and I guess the rings themselves have a specific frequency. Can you speak briefly about the importance of that, and could you explain what a PZT maker means when they specify 'Radial Frequency' and 'Thickness Frequency'? Which is more important for a Langevin style device? Thanks again!

[@imajeenyus42]

Thanks, glad you liked it! Regarding making your own transducer, it's probably not something I'd recommend, given (probably) critical dimensions, plus the availability of cheap transducers, but it should be possible. I'll send to a direct message with links to two papers which describe the construction of bolt-clamped Langevin transducers, they might explain things a bit better. A PZT ring will have a thickness resonance, where the thickness increases and decreases. This is generally quite high, and isn't of interest in a Langevin transducer. They also have a radial resonance, where the ring "breathes" in and out - this is the one we want. It's a little counterintuitive - although the ends of a Langevin transducer vibrate axially, the MIDDLE vibrates radially, which is why you want to excite the radial mode of the PZT ring. Unfortunately, since PZT ceramic is very weak in tension, the entire transducer has to be pre-stressed using the bolt down the center, and the amount of tension applied will vary the resonant frequency (quite a bit). The papers show the effect of different bolt tensions. Most of the mass dimensions, preload tensions etc. are calculated through finite-element simulations - because it's such a complex vibrating system, it's impossible to do from simple equations!

[@imajeenyus42]

Actually, I can't direct message you for some reason. Could you drop me an email at [e-mail] and I'll send the papers to you?

[@victorst5997]

Well done. But in my opinion, there are several pieces missing. For example, the horn is a step type. Not very reliable and primitive for my taste. Just a few more volts at resonance and it'll crack. Also, although you mention the wavelength, half wavelength and a little bit about the tuning, you left out the constants k, kl1 and kl2. There's just no way to machine a horn, be that step type, conical, exponential, catenoidal or Fourier, or Barbell for that matter, without the constants and the interpolation tables. All in all, your presentation seems fine for an introduction to power ultrasonics and sonotrode design, but even a speed course should mention the procedures available to calculate and design the horn of your choice. I hope you show us what else you've got. But good work. Keep it up.

[@clebermarcelino673]

Hi, Thank you very much for your videos. they are help me a lot with my studies. I have watch lots of your videos but I still have a doubts about transducer for cleaning, How much voltage can I put on a transducer? Does has a way to calculate a better voltage and current to operating a transducer? thank you very much and please don't ever stop making your videos they are really goods and important. Thank you.

[@gpcoelho13]

Great video! Do I always have to calibrate the assembly (transducer + horn) to the transducers' resonance frequency or can I use the frequency of the whole assembly as the input? For example, I have a 40Khz transducer, and the assembly resonates at 35Khz, can I input 35Khz and operate the machine with this frequency or do I have to use the 40Khz and calibrate the horn for well function?

[@imajeenyus42]

Thanks! To be honest, I'm not entirely sure, but my feeling is that you should optimise the horn so that the entire system operates at the same frequency as the bare transducer. The transducer has presumably been optimised to operate at 40kHz - at that frequency, the piezo elements will experience a longitudinal node, so they are only vibrating radially, which is where they are happy. At any other frequency, they would experience other sorts of vibrations, which might damage them. So better be on the safe side and get everything working at 40kHz!

[@AlejandroEspinel]

Good morning, at what voltage do the transducers used in cleaning work and if the applied voltage is sinusoidal or square?

[@imajeenyus42]

Several hundred volts. The ouptut from the inverter is squarewave, but this is filtered by the series inductors to produce a more sinusoidal signal which is applied to the transducer.

[@johncoppens849]

One observation: I suspect that in the original application of the transducer the side with the nut is screwed to the housing of the application. Wouldn't leaving that side unconnected (as in the video) make that side vibrate much more intensely than the business end? (estimate inverse proportional to the relation of the masses)? And wouldn't that way much of the power be lost? So, shouldn't all the measurements be done with the transducer 'grounded' at the nut side? (i.e. connected to a large mass) I suspect that the 'natural' resonance would change noticeably too. I've no experience at all with mechanical resonators, just with how electronics would work...

[@imajeenyus42]

Nope! The "back" of the transducer, the side with the nut, must be left free to vibrate as well. If you check out a photo of transducers used in a cleaning application (e.g. www.ultra-piezo.com/2011/0211/378.html), you can see how only the front is attached to the tank - the back is left free. In operation, both the front and back of the transducer oscillate in and out, and the back mass of the transducer is providing something for it to "push against", as it were.

[@thetatek6634]

Great video. What value probe are you using 1x, 10, 100x, 1000x?

[@imajeenyus42]

Thanks. Just 1x, the voltages are pretty low.

[@MarvinMalane]

I am wondering if I can use VNA to determine the resonant frequency, I suspect it should be possible but I wonder if the fact that the VNA is optimized for 50ohm impedance might mess with it. Either way this video is just awesome, thank you for making it :) You helped all of us. EDIT: One more question, lets say i have a horn that resonates at frequency rather far away from the original frequency of the transducer, do I lose a lot of performance by just driving it at the new resonant frequency or should i rather modify the horn to resonate at the original resonant frequency of the transducer?

[@imajeenyus42]

A VNA is basically exactly the concept you want, but like you say the impedance will probably mess with things. Also I think most are designed for much higher frequency use and they might not reach down to the tens of kHz which is needed for these sorts of transducers. As for a horn at a different frequency, you would really be best to modify the horn to suit the transducer. You could drive it at a different frequency, but the efficiency will be lower, and the piezo elements in the transducer wouldn't be happy since they would no longer be located at a node. They might experience mechanical stresses outwith what they can handle.

[@MarvinMalane]

@@imajeenyus42 Makes sense, thanks so much for such a perfect explanation. My VNA goes down to about 10kHz so I will give it a shot. If I get some interesting results I might attach it to this comment for others to see (as it could potentially make the measurement even easier). Regarding the horn, i dont have much freedom to change the horn but, you gave me new idea. I can try to get the transducer back to the node by designing a "counterbalance" on the other side, then it should be just matter of measuring where the node actually is. Which I dont know how to do yet, but I assume I should be able to just attach the entire assembly at various points and measure the attenuation, if its held at the node the attenuation should in theory be the lowest.

[@imajeenyus42]

@@MarvinMalane It would definitely be interesting to see how the VNA works with it! You get dedicated frequency analysers for use with ultrasonic transducers, but they're pretty expensive. As for holding the horn, it gets tricky. Although the piezos are located at a longitudinal node, they are at a radial antinode - they "breathe" in and out. So you can't clamp anything rigidly, it would need to be some sort of compliant mount like rubber or sponge. A counterbalance may work, but attaching it would be difficult - do NOT, incidentally, dismantle one of these transducers! Unless you have a torque wrench and means of measuring the resonant frequency ;-). The frequency actually shifts depending on the tension of the bolt. You need to keep the piezos in compression at all times, hence why there is usually a significant bolt holding the two halves together.

[@MarvinMalane]

@@imajeenyus42 You just saved my transducer, I was just about to take it apart

[@tyhuffman5447]

Outstanding! Very well done!

[@MrShwaggins]

I have a silly question regarding ultrasonic horns/transducers you see on cleaners in various wavelengths and strengths. Is there one metal that the horns prefer? What about material? All the commercial ultrasonic cleaners are made from some sort of formed stainless steel. Has there ever been a plastic drum or plastic container with decent performance? Its a silly question but I can't google this answer thats why I'm deferring to people that know more about this.

[@imajeenyus42]

The transducers themselves generally use an aluminium "front mass" (the part that's attached to the tank) and a stainless steel "back mass". Fancy transducers may use stainless steel or even titanium, but you're unlikely to see that in a cleaner. As for the tank material, metal is the only option - plastic would absorb too much of the ultrasound as it passes through. Additionally, it would heat up and possibly melt from the ultrasonic vibration - a fact used in ultrasonic plastic welding!

[@MrShwaggins]

@@imajeenyus42 Thank you for answering my question so fast. TBH, I'm thinking about making a metal utility sink in our shop into an ultrasonic cleaner. Think utility sink but made of metal. I would want to power it with a dial timer that goes up to 15 minutes so it can't be left on. I think I'm going to go with 40 khz transducers since I want to keep it simple.

[@martinchabot_FR]

You can use X-Y mode to get a better reading of resonance frequency

[@imajeenyus42]

You can, but I find it easier to just do the two sinewaves - that way you can see the frequency at the same time.

[@hikari8852]

2 Questions: 1. Do you have to wear ear protection when turning on the transducer? 2. When turning on the transducer, can you still touch its body without getting shocked, i.e. is the transducer body charged when connecting to power?

[@imajeenyus42]

It's probably advisable to wear hearing protection, especially if you have a higher power transducer. Although you can't hear it, you do get a "sensation" of the sound ;-) I would avoid touching the transducer unless you are using an isolated driver and are sure of which side is grounded.

[@hikari8852]

This is the transducer I got: www.allendale-ultrasonics.co.uk/ultrasonic-transducer-40khz does that help in answering?

[@thaitrungnguyen9640]

Dear Lindsay Wilson. Would you please enlighting my very silly question. I have an 600 watt generator connecting to 4 x 100 watt transducer. I increasing the ouput watt so the noise coming out from transducer getting louder. But if i keep increase watt to the point which make the sound suddenly getting quite. I am very sorry for troubling your time, i try to search on google but there no answer or i must miss read somewhere. Many thanks to you.

[@vincenzolauria5944]

Hello Lindsay. Very interesting. How about when using the transducer attached directly to a Cup of stainless steel to use the ultrassonic waves to clean? Do i need some specific size for the container for best performance or there is no need because it is a direct sonication?

[@imajeenyus42]

As long as the transducer is well-coupled to the cup/tank (e.g. by gluing the entire face of the transducer), there shouldn't be any issue. The water in the tank presents a large load to the transducer, which broadens its resonance significantly, so the exact driving frequency isn't quite as critical.

[@vincenzolauria5944]

@@imajeenyus42 yes the transducer Is fully glued with JBWeld under the container. Thanks a lot for your answer.

[@luisfranciscopinedavega6067]

Hi Mr. Lindsay, great video! . Im trying to design an ultrasonic horn in COMSOL for a 40 kHz transducer to generate acoustic cavitation, but i´m having issues with the correct geometry. Could you share the dimensions of your 28 kHz horn to understand better the effects of geometry? Greetings

[@drmohsintiwana1020]

Great Video. Have you tried designing a roller type horn? Any video / ref will be a great help

[@ahmadrinno]

Hello Lindsay. Great video. I'm a bit confused regarding the harmonics. When you find the third harmonic you divide the frequency by 3, shouldn't you multiply?

[@imajeenyus42]

Thanks! The resonant frequency of the transducer is being excited by the 3rd harmonic of the driving squarewave, so the _fundamental_ frequency of the squarewave is a third the resonant frequency. Hope that makes sense!

[@ahmadrinno]

@@imajeenyus42 That helps a lot, thank you. But how do you know, you have the third harmonic? Just by looking at the oscilloscope?

[@imajeenyus42]

@@ahmadrinno Well, I know what frequency the function generator is putting out, and I know what the resonant frequency of the transducer is, so it's quite easy.

[@ahmadrinno]

@@imajeenyus42 that's a good point, thank you! Again, great video, very informative!

[@gustavogarcia7442]

It will be highly appreciated if someone explain me why an industrial sonotrode has two resonant frequencies, one at 17.705 Khz and the second at 19.754 Khz. Thanks in advance

[@mikesuttenberg7763]

Excellent tuning 101 video.

[@davidemasiello8020]

Great video, thanks! When the added mass changes the resonance frequency, does it also affect the transducer displacement (and therefore the produced sound power)? Otherwise it sounds like I can just get any frequency of vibration from a Langevin transducer by attaching long enough probes, regardless what its original resonance frequency is.

[@AlexiLaiho227]

great video! i wish i was half the engineer you are

[@owmamagoto]

really interesting video sir. Is the current you are measuring is the same as the echo coming out from the transducer? because i can't still figure out about how to measure a reflected echo coming out from a single ultrasonic transducer.

[@imajeenyus42]

Thanks! No, the current is simply the drive current passing through the transducer - it is being continuously excited. I'm not sure how you'd measure the echo - usually things like rangefinders have some sort of blanking or filter to prevent the drive pulse from being coupled directly to the receiver circuit.

[@297645]

Great video! I'm working on an acoustic levitator for my physics undergraduate thesis, as I intend to study deformations of water droplets in an acoustic field, and I'm so grateful I found your page. It happens to be, that I have the exact same 28kHz transducer and now I need to make a horn for it. I was wondering whether I could get ahold of the dimensions of the "step-down" horn you made, plus the disc tip you use to "correct" the resonant frequency? Thank you very much!

[@imajeenyus42]

Hi, glad you like it! I've got a drawing of the horn and the flat tip here: imajeenyus.com/temp/horn.pdf imajeenyus.com/temp/flat_tip.pdf (Best save them, I clean out the temporary folder from time to time...) Both made from aluminium. I have to point out again that you MIGHT get lucky with the resonant frequency if you machine a horn to those exact dimensions, but there's always the chance it'll be slightly off (different alloy, temperature, whatnot), so it's best to do it a little bit longer and shorten it to tune. I really should sometime get a chunk of a specific alloy and make a horn from it, so I have something repeatable to go on! Hope that helps.

[@297645]

Thank you for your answer. You sure saved me a lot of work :)

[@Mr0rris0]

Thanks for your comment.. I've been wondering why the algorithm wants me to know about ultrasonic transducers so much... This confirms its all spawned from an acoustic levitation vid I watched recently

[@mistyfrequency7261]

Excellent 👍 thanks you make it straight forward.

[@akkaanantashesha3672]

amazing brother . precise n concise

[@MrSemperfidelis225]

Great stuff, thanks for sharing. I wonder what the resonant frequency of the piezoelectric ceramic pieces alone is? Your 'bare' transducer has about 40-50mm (?) of aluminimum already attached to it. I would expect resonant freq a bit higher, the Q to be very high, and the resonant peak very narrow band. But not completely sure why I'd expect that. What does the mass on the back do to frequency? Have you tried heavier/lighter mass?

[@imajeenyus42]

Thanks for commenting. The bare piezo disc also has the same resonant frequency. For example, here's a couple of discs which would be used in a 44kHz transducer: www.steminc.com/PZT/en/piezo-ceramic-ring-38x127x63mm-44-khz It's important to remember that , although the ends of the overall transducer move axially, the middle, where the piezo disc is, moves radially ("breathes"), so it's the radial resonance of the piezo disc that's important, not the thickness resonance (which will indeed be much higher). As for the back mass, it's basically something for the piezo to push against. I don't know all the ins and outs, but there's a (very length!) description of transducers here which is worth a look: www.ultrasonic-resonators.org/design/transducers/transducer_design.html The material of the masses, as well as the tension on the bolt, all have an effect on frequency and efficiency. As for trying different masses, that is not something which a user would typically be concerned about. The masses and bolt tension will all have been chosen to optimise the transducer, and changing anything will likely be detrimental.

[@victoryfirst2878]

Can you tell me what the efficiency of the transducer horn assembly ???? Thanks

[@imajeenyus42]

I've seen figures quoted of around 90% for transducers used in cleaning applications, but I don't know about a horn, sorry. I imagine it will depend a lot on how the horn is loaded, the driving frequency etc.

[@victoryfirst2878]

@@imajeenyus42 I should have been a little more clearer with my question. When a transducer is matched to a horn of the proper frequency would the matched pair increase the sound pressure, and if so by how much would you say Lindsay ?? Since the transducer is more or less a piezo crystal powered by an electric current would varying the frequency electrically be feasible to increase the cleaning power ?? Thanks, V

[@paulbyerlee2529]

Hi great video I was thinking about modifying a cheap ultrasonic cleaner to have a sweeping frequency. My question is does an ultrasonic transducer have more than one harmonic that they switch between or do they just alter the frequency and loose some amplitude as a result. Thanks in advance.

[@Suvatn]

Excellent video.It's make clear understand.

[@hubercats]

Fantastic video. Thank you!

[@sprintdigital9914]

Excellent video. 2 Questions 1) Is the grade of Aluminium important? 3000 series vs 7000 series? I know that 7000 series is harder and does not need heat treatment post-machining but any other difference? 2) Any difference in initial length calculations if its a bar-shaped horn as opposed to circular one?

[@imajeenyus42]

Thanks! I don't know about the aluminium grade, but I would guess that one with a higher tensile strength would be better, as it would be able to support a larger oscillation amplitude before failing. E.g. 7075. Tensile strength is one reason why "proper" horns are made from titanium. As for length, it won't vary hugely - if you start off with a cm or so longer than the theoretical half wavelength, you should be good in both cases.

[@hubercats]

Nicely done. Thank you!

[@Electromechaniac]

Hi there! I'm confidence that I'm not the first who asking it, but actually I've a little bored to read all the comments. So why are You not drilling holes (for studs) with an extra depth initially (before shortening the horn)? Is pontentially existing empties make occur a problems? P.S.: Thank You for Your great job! Your video is fascinating and so helpfull! P.P.S.: Forgive me for my poor English - I'm working to improve it.

[@imajeenyus42]

Mainly because I didn't know in advance how far I would have to shorten the horn, and I wanted to have as much solid material left as possible.

[@Electromechaniac]

​@@imajeenyus42 Thank You for your fast response and answer! I've understand your position.

[@mjolinor]

I do not understand why the resonant frequency of the bare transducer is relevant as long as you drive the assembly at its resonant frequency making sure that you are at the fundamental frequency with the anti-node at the business end you will achieve what you wanted. What am I missing?

[@ihtsarl9115]

5800m/s is illustrated in one of the links as the speed of sound through steel.

[@theoracle6005]

Very informative video, thanks

[@francisamor6276]

What’s the difference between non ultrasonic homogenizer and ultrasonic homogenizer? Thank you

[@ryanalbert1807]

Wonderful video! I am looking to use ultrasonic transducers in an art project. Can I use any ultrasonic transducer to pump any ultrasonic frequency? Basically I'm looking to amplify any frequency all the way up to 500 kHz. How would I be able to do this? Thank you!

[@NoahSpurrier]

Why not have a transducer driver that automatically tunes itself? Periodically the driver could sweep a frequency range and detect the frequency that causes the highest power load on the driver and then drive the transducer at that frequency. Or just have a manual tuning of the driver frequency and set it to the highest power load. It seems like it's far easier to tune the driver circuit electrically than it is to tune a physical object. ... adjust within some reasonable narrow frequency window, of course.

[@imajeenyus42]

Most transducer drivers do automatically tune themselves to the resonant frequency, e.g. by tracking the phase of the current and using a PLL. However, the horn must still be tuned (machined) to match the resonant frequency of the transducer, since that's where it was designed to operate most efficiently.

[@karlharvymarx2650]

@@imajeenyus42 I was going to ask the same question as Noah. How does the inefficiency manifest itself? The current looked about the same. Is there less mechanical movement? Does the piezoelectric element wear out faster?

[@imajeenyus42]

@@karlharvymarx2650 The impedance reaches a definite minimum at the resonant frequency - for example, here's a graph: imajeenyus.com/electronics/20110514_power_ultrasonic_driver/photos/Graph2.JPG. Resonance there is at 28.25kHz. This is the point at which current draw from the supply will be greatest. It's perhaps not perfectly clear in the video (in the case of the horn) because the signal generator's output actually drops slightly as you try to draw more current, so it might appear that the current stays relatively constant. However, if you watch the part with the bare transducer, you can clearly see how the current peaks. Yes, you will get less mechanical movement if operating away from the resonant frequency. At some point I want to demonstrate this with say a non-contact optical sensor to measure the vibration amplitude versus frequency and show how maximum vibration does indeed occur at the resonant frequency.

[@anullhandle]

@@imajeenyus42 check out applied science video on a laser as a transducer. Also an old phonograph cartridge might show displacement as you sweep the horn along its length or sweep frequency.

[@astraymark244]

@@imajeenyus42 Another point missed is that as the frequency changes all the nodes will move and as such so does the ideal mounting point, thereby reducing the efficiency of the entire system.

[@farezaaditya2333]

Hi Lindsay, I know this is an old project but do you mind telling me what is the white cable around the transducer for ? I'm assuming they are connected to the poles. Thanks !

[@imajeenyus42]

It goes to a high-value resistor (I think 100k, can't remember) which bleeds off any charge that forms on the transducer. If you pick up a bare transducer you'll sometimes get a small shock because the body has expanded/contracted and generated a charge on the piezo discs!

[@farezaaditya2333]

Nice! thanks for your reply and great work !