The flux I would consider essential. I use silver solder frequently with stainless items, and the flux makes the solder flow and stick much better than without.

A good idea when brazing a stud like this onto thin sheet is to put a thick backing plate against the back side of the sheet. It'll soak up any excess heat and help stop any warping.

Every aspect of DIY is thought over meticulously after confronting the problem. The description alas is fantastic

Props to the seller, lots of AliExpress sellers have basically zero support.

Now that is dedication, I simply wanted to trouble shoot a possible dud transducer, started with your first video and was instantly grateful for your efforts. Both these videos were brilliant and went through some of the practical solutions I was about to attempt. Genuinely impressed with your scientific and engineering rigor. I'll be watching your videos in the future. Much respect.

Great work, lots of discovery there, really interesting for me (my company makes transducers). I couldn’t see if you prepared the tank surface. You increase the adhesive strength a lot if you sandblast or roughen the surfaces. This gives the adhesive a larger area and surface features to key into. Pulling the transducers when the one surface can flex with result in a weaker bond, more like a peel test. Every transducer has its own frequency, not always exactly the same but close. Because they are so narrow-band, you could be closer to the resonance on some transducers than others, it will work harder and get hotter if it’s closer to resonance. As for the polarity, it does not matter, transducers are insensitive to polarity but if you put 2 near each other it would be preferable to match polarity so that they drive the same way. If you parallel two transducers they definitely need their polarity matched. Also, the front and back metal blocks of the transducer are connected to one of the wires so you may get “live” voltages on the tank. Hopefully the electronics has its output electrically isolated so this won’t be an issue. How well the output is isolated is the real question, is it safe? I seriously doubt they have safety approval ratings on the electronics and considering the high voltages involved, this can become dangerous.

As many have mentioned, flux will make it much easier to flow the solder and get a reliable bond. If it is soft solder, the plumbing torch is a bit overkill and will make heat management a bit tricky for the thin wall tank. One of those pocket butane torch lighters would work perfectly. Stainless heat soaks very easily due to very low thermal conductivity, so less heat than you think and a bit of patience is best to let it come up to temperature without creating hot spots. If it starts to glow, it is definitely way too hot for soft solder. The small pencil flame of a butane torch lighter would help make it easier to focus the heat primarily on the stud and just flick the flame off to the tank just enough to bring up the temperature to where the flux starts cleaning. Getting solder flowing on the stud will help conduct heat into the tank and flow it to the tank without too much heat deformation from too much direct heat. Just before glowing is about good brazing temperature for a Harris stay silv brazing filler if you are doing it that way.

Ultra-sonic cleaning improves almost everything I use it for. Even in plating processes, it causes the build to be more diverse and compacted. You can also add a sonic horn and other techniques to the ultra-sonic without causing harmful effects to each process. I was going to name all of the functions it improves, but there are just too many. Whether its medical equipment, automotive, industrial cleaning, high-tech production parts, surface polishing, and plating, etc........ Wiring , welding, width of wall thickness are all super important. Most of the liquids used in the US, are wetting agents to help wet the solution to an altered state of slipperiness. This makes the dislodging of the smallest particles very easy.

Amazing patience with trial and error. Thank you 🙏

KZfaq - would you like to watch a guy make a ultrasonic cleaner Me - sure why not Not a single shred of interest in making one but watched the whole thing. Very well put together video

I love how your wiring diagram is the equivelent of "just plug the things into the things, don't think about it" I don't think it's really a project that needs a wiring diagram

Use a capacitor discharge stud welder they are specifically designed for those studs

That solder test was hecka punckrock!♡ luv it

You could also try friction welding - put the screw in a mill or drill press. High RPM & lots of pressure. The heat generated is very localised. I've used it for welding studs in place (not on a cleaner though)

Thanks for this follow-up video! I'm definitely going to build my own knowing how to fix the issues you had earlier.

Thank you. Now that the bugs are worked out, it's time to watch the video series, make a parts list and let the fun begin.

Thanks so much for the follow up ^_^ really appreciate you sharing what you've learnt with everyone. Good to know you solved overheating problems

I've not seen a reasonably priced u/s cleaner that has cooling fans .. and I've had a few blow up on me ! So I think you have built a better machine than the off-the-shelf units. I may now rebuild on of my old dead "shells" using your specifications. Many thanks

TIG welding (FUSE) or even silicone bronze filler is probably the best way to mount the transducers. If you are going to use silver solder make sure that you use silver solder flux!

Bruce always fun and always informative, if I ever build my own tank I will know what to do thanks to your journey of discovery!

Thanks for doing the work Bruce, this is a great diy resource!

Very useful series. Thanks for all your trouble.

Great work, thanks!

That was just 10/10 YouTubing. Amazing stuff. Thanks man

Very helpful testing! I was convinced that brazing would be good and I was sure that welding would fare better than silver solder. It's a good thing you did this testing. Thank you.

So much of this just highlights how using the cheapest parts available on reseller sites like Aliexpress is a recipe for hours of frustration. Using known good parts from known good sources might be significantly more expensive, but it's worth it to avoid all the troubleshooting.

Thanks for sharing your experiences. I like that t-shirt. English for Spanish speakers.😂 Greetings from Spain

when brazing, the stainless steel has time to anneal, therefor it has become softer stainless steel then it was new, hence causing it allmost imposible to tear or crack, a good point indeed, the ultrasonic vibrations when in use will workharden it but that takes forever with the water dampening it. Whe welding, the heat is short and intence, depending on the type of stainless, the short heatcycle even can harden the stainless steel localy, giving the steel a higher change to crack on load and even from the ultrasonic vibrations.

Omg i LOVE this soo much! Thank you for making these videos!

Fantastic! A must watch video! Thanks again for sharing!

Another GREAT video!

Super helpful video!

Dang! Well done Bruce

Great Job Bruce Good Follow Up

Enjoyed this. The information you learned from troubleshooting the problems alone is worth its weight in gold. It does leave me wondering though if a tanks thickness makes any difference in cleaning efficiency. I am guessing it shouldn't but I also see that most the commercial tanks are pretty thin as well. Thanks for this.

Just started watching this channel to learn how to repair a rare peavey guitar amp. The amp model is quite state of the art for being as old as it is but suffered from a factory flaw. When manufacturing they used a battery to hold memory and they bust and leak acid on the board eating all kinds of stuff up. I'm being told by the guy who sells a mod to fix a non leaked units that my amp is toast and that he's basically never been able to bring one back to life. But I going to attempt it anyways.

This are great videos. Thank you.

Brilliant Bruce, well done. Thoroughly enjoyed your video and glad you now have a 'product' that suits your needs. I'm almost certain somebody would want to buy one😁

Brilliant!

Hi Bruce, I think the transducer falls from fatigue not from overload, so testing strength by applying a load is probably irrelevant. After seeing the first video I assumed maybe the bond had a small gap between them, it could cause: 1. The vibration of the transducer makes the trapped air compress and contract at a very high speed which causes micro fracture in the adhesive -> causes the transducer to fall 2. The adhesive that has been half failed causes the transducer to hit the tray repeatedly -> causes the transducer to heated up Keep it up! I'll be waiting for your next follow up video

Very informative video, I like it. I know your idea for the tank was easy & really a good starting place. You should look into Hammer forming your own tank. This will allow you to get the exact size you are looking for.

Great video Bruce. I'd like to see you build a dining table sized tank with about 100 transducers (give or take) so you can clean about 18 (6 x 3) Mac II logic boards at once! Also, people should not get upset with you not providing diagrams of your work. After all, it's your build and you can choose if you want to share or not.

A bolt welding available to amateurs is friction welding: Clamp the bolt in a drill press' chuck and run it on the surface you want it to weld to until the contact surfaces get red hot, then stop the drill and keep pressure until cold. It goes without saying that you best try this on a sample piece first, to establish the best parameters. It still probably weakens the base material more than the silver solder method. You never mentioned this, but I think it is obvious that using the drivers with the epoxy cracked greatly reduces the sonic transducivity of the joint. I see the advantage of the bolts mostly in helping with installation of the tranducers, after that, the adhesive should really do the whole job.

If you could find a suitable tray, a tall, thin cleaner would be useful for pcbs, like a toaster. you could attach the transducers to the sides instead of the bottom. this might reduce the liquid volume and hence the power requirements.

Sheet metal spot welder would do this also, like the kind a body shop would use to replace panels that looks like a large pinch clamp, one side on the bottom of the pan the other side on the bottom of the bolt, get it lined up and hit the foot pedal and instant bolt attachment. Silver soldering seems like the best alternative, just use flux and a little less heat to prevent distortion.

I’m surprised you said flux for the silver solder was recommended, but not required. Your advice about using generous amounts of flux in electronics soldering has made my simple excursions into electronics soldering far easier than before I took your advice to heart. If you can elaborate on it sometime that would be appreciated. Thanks for the tutorials!

I wonder how effective spot welding would be, for the bolt to the tank? Significantly more heat applied to a significantly smaller area. Less bowing of the tank, but you most likely would need to tack a couple of times to get enough of a secure bond. Often used to weld lipo batteries together into battery packs securely.

Nice content, Bruce👍

Wow, you can now go into business as an industrial scale denture cleaning service! You'll be rich!! 👍😃

The KZfaqr "Project Farm" tests a lot of "stuff" WITH measurements. Talks fast, but well worth a look... This is a great project!!! Arduino is awesome ..

Fantastic build and an amazing work there! Very well done! 5*

If I was going to tackle this project I'd drill through the tank, drop in a few stainless steel blind-hole nutserts to suit and use some allthread to mount the transducers. No welding needed... But you'd need the nut insert tool. A dab of silicone at the joints should keep it watertight.

I have the best success with epoxy metal bonds by sand blast etching both surfaces to provide tooth

I think the main reason the drilling and bolting isn't used is because at any kind of medium to large production scale the production cost matters a lot more than in a one off project. Welding is just cheaper at scale.

Acrylic structural adhesive may suit your needs if do build again. The most popular brand of thread locker has a line these types of adhesives I have used in manufacturing. Bonded metals typically have 6k psi tinsel strength. I don't know what the vibration will do to it but I am about to find out. These cleaners fail quickly for me. I purchased 4 since you made this video. I need one that doesn't break. It'll be fun to build one even if it costs more than a new cheap one. Thanks for inspiration.

I honestly don't know why so many people are in an unreasonable love with JBWeld. It is okay-ish for a lot but - like WD40 - not really great at anything. It is - in all honesty - just an epoxy with metal powder stirred in. Can be good, can be bad.

Great follow up video! Soldering really seems to be a good alternative to welding. Though it'd be interesting so see their strenght compared over time under actual load. Because I do know that stainless steel welds have a tendency to easily crack under constant stress/vibration and soldering might have similar problems. I personally think drilling a hole and just straight up bolting the thing on would work just fine as well. You can get it watertight by just using a copper washer between screwhead and tank. If the screw is glued in it won't loosen from the vibrations either and screw connections are generally more resistant to stress and vibration compared to stainless steel welds. I'm certain the only reason the screw method isn't done commercially is because of the "unclean look" and the fact that a smooth container with no screw heads at the bottom is easier to clean.

For lower thermals, make sure they use low Rds-on driver FETs.

Hi Bruce, what "I think" might be a problem with epoxy is that it needs air to cure. With two non breathable surfaces the epoxy in between could posssibly not cure properly, or take a really, really long time, leaving only the epoxy in direct contact with air to cure normally. A thick layer in between might help. Maybe a thin layer is always a problem ? I don't know for sure. I would be interested to see more epoxy tests to prove or disprove this (like project farm). Good luck with your projects.

9:50 I didn't know that what I was looking for will be in your follow up video . Kind off about ungluing the transducer but with bolt and glue it will be impossible to separate I think .

what a lovely video! thank you for this, but also want to ask something extra. as still feel somewhat unanswered or unknown question. which is nothing to do with your stuff in the video here. but since you have just become our collective industry expert in the field, it seems like you are now the right person to ask it to; my question is in regards to tank configurations / dimensions and effectiveness.... see when we are shopping around for ultrasonic cleaners the big question always is what size to get. and it seems like the optimum size is nearly always long in 1 dimension. then either narrow and taller. or wider and then more shallow. this is either because we want to save on desk space / the device footprint. since these things are pretty bulky. or it is instead because we do not want to have to fill up and use too much liquid into the tank. not to have the water level higher than necessary for convenience it filling it up or moving it. or for a given chosen tank volume and hence purchase price. however the vast majority of ultrasonic tanks being sold appear to be long in 1 dimension. and then approx the same in both the width and height. which is generally less ideal for the typical shape of pcbs. which are of course flat rectangles. and we don't want to angle our pcbs diagonally because that is then a huge volumetric waste of cleaning liquid. so my question is clearly to try to find out why the ultrasonic manufacturers do that. because the reason is not known. but it could be for one of several reasons. and this then informs both our purchasing choices. and also the tank design for a self built device. it could be (for example) that these are being sold mostly for general purpose parts cleaning. with pcbs being a minority usage. then those tank dimensions would make more sense for other random parts which are normally not flat pcbs. car engine parts for example. or it could be because the ultrasonic waves need a certain volume of water to be more effective. wheras a too shallow depth of water would cause the waves to reflect back down from the water surface. and just not work as well for the shallow water scenario or conversely it could be because the 'deep and narrow' configuration is too narrow, for example for a single row of standard sized transducers. again to work most effectively. without problems with the waves being reflected back off the side walls or it could be because deep and narrow stainless food trays are not common. or more expensive and difficult to manufacture or get good cheap supply of? wheras shallow food trays also dont offer the same band for buck perhaps when they can sell for more dollars a higher volumetric capacity unit with the same number of transducers and parts cost in it etc. in other words market forces and customers comparing units by volume capacity class... and then always the more regular shaped tubs wins with the lower bom cost to manufacture i suppose there could be other reasons too. but you get the picture. clearly it would be helpful to uncover some answers here. to be better informed and not end up with making the wrong decision. whilst also not being lumbered with a machine that is more bulky and takes up too much more space than is actually necessary. would be pretty helpful anyhow thanks for this follow up video. learned a lot from it. very grateful & well done for being so clear about all of it. very good troubleshooting and test methodology. kind regards

Silver solder it is then. That’s something I can do.

I think the cause of aluminum foil perforation is using 40 kHz transducers. Perhaps you could use 100 kHz or 132 kHz transducers for your next cleaner to reduce the damage. I imagine that some capacitors on the motherboards wouldn't like to be "penetration tested"

I would think a local muffler shop, turbo installer, or tuner shop would be more than happy to have one of their guys tig-weld those bolts for a nominal fee.

Thank you for this. I had assumed it was the transformer (inductor) on the board that was causing my problems (noise fades after several 10's seconds, just buzzes rather than "sonicing"). Not only can I not find out how to determine what L a given inductor is supposed to be, I don't know how to measure that, what that failure mode might look like nor how to order a replacement... or upgraded item... so I'd more or less given up. I hadn't considered the transducer could fail - it's such a simple physical device... In terms of bolts, I've been looking into spot welding and come to the conclusion that it would work fine but that it's a ton of work. Then I was thinking about bolting through the tub... and ... leaks... What I am keen to try doing is taking a chunk of steel bar , drilling and tapping holes through it, attaching the transducers to that and then gluing that to the bottom of the tank and using bolts to reach around to the top with clips to hold it snug against the tank bottom.

Thanks for this Ultrasound cleaner series. I have had a very similar fight for a long period of time. My transducers also fell off one by one, and I also didn't weld or solder the threaded rods to the tank. I will now silver solder them. I also fried the generator a couple of times. One question: You diagnosed the defective transducers by their working temperature. Now that you KNOW which ones fail, are you able to maybe measure their impedance or resistance to see if they also differ in these values depending on if they are defective or not?

Good video again but was surprised your circuit diagram didn’t show any earthing after your comments regarding the need for earthing. Those of us with medical equipment backgrounds are very into earthing and leakage currents. 😊

i know very little about ultrasonic cleaners and even less about transducers but i do quite a bit of welding my perferd method is gas welding also know as oxy acetylene welding a regular oxy acet torch is what i weld most anything together with unless its really thick and i got a big machine that does heliarc or smaw but most of my welding these days is done with a harris torch with a standard mixer and whatever size tip. its alot cheaper than using tig and i have welded the crazyist things known to mankind such as welding a cast iron pipe fitting to a stainless banjo fitting i once welded a motorcycle carburetor float post. stainless will require a flux to unless you want to use mild steel filler then not necessary. cast aluminum is the biggest pain in the ass to weld cus of all the contaminants in the cast a tig would be best for this but is still a serious pain to weld ive welded whole vehicles this way and the only 2 accepted ways of welding aircraft frames is gas welding and tig welding so if you got a few dollars to spare on a torch and tanks you could probably get a set up goin for less than 400 usd but i dont know what the availability is like down under either. but like you found out silver soldering / brazing is over kill for something like this in the good ole days most body shops brazed vehicles together and it will hold even to the point of ripping 18 ga steel the bond will be unfazed im surprised you didnt use a pid for the water temp in the beginning with the analog timer if i build one it would use a timer like that just for reliability sake that and i dont know much about electronics

Nice job you done there 👍🇮🇪💚🙏🏻

Did you think about "Spot welding" it, with just electricity and a microwave transformer? I know it is creating a new machine in between but it is easy, economic and you can use it to weld battery packs in the future. I hope you get to create a great product. Thank you for being so thoughtful and detailed with your project!

First thing I thought of when I watched your original video - buy a manufactured ultrasonic that meets your wattage needs and modify to suit. Far less headache and probably wouldn't have cost much if any more.

Thanks for the video, didn’t had a clue before that I could diy a unltrasonik cleaner. I will start today my project to clean car parts in a tank 2-5 litres. 150 was is enough though?

You may have done this and/or already suggested but just in case, before epoxying surfaces should be scuffed up with course grit paper, rougher the better.

Now if you can program into the Arduino some pwm flex you will have a really awesome diy project that competes with pro-grade. Sweep and degassing would be nice too.😊

Very interesting. Apropos the attachment of the transducers to the tank. Considering the tank manufacturers actually just use an adhesive to attach the their transducers, can I suggest you talk to an industrial adhesive manufacturer, like Loctite or Scotch, explain your problem and allow them to suggest a suitable adhesive and preparation regime? I always feel it's best to ask the experts, they've usually done it before? Additionally, spot welding the studs would work well also?

Buena polera, señor Bruce :-)

You could try a plug weld (drill a hole, align the bolt on outside, weld a large bead on the inside over the hole and the material). That should let you get a decent weld without blowing through the thin stainless. You'll be welding onto the thickness of the bolt.

Great video! One observation: why print a plastic box when a couple of bits of wood will work just as well? Planet, life, .... you know.

You could drill a minuscule hole and add a point of solder to breach the hole and hold the bolt in position from the inside, then solder the outside without it moving

I have seen videos on DIY spot welder from old microwave oven transformer. I wonder if such could be reimagined to use for your studs. Also noted that the welded driver that tore the pan when you pulled it off, was the one you mentioned you'd burned a hole and patched same, when initially doing the weld. Probably weakened the pan at that time, otherwise it may have bent same as with the silver soldered join. Great video BTW. I've thought about making one myself, but don't have as much need as I am focused more on building electronic prototypes from scratch. 👍🤙

Great video. One thing I didn't see you address is the epoxy cure times. As epoxy full cure time can be up to a week long.

should have put a foam to catch a falling transducer. they are likely piezo and impact might crack the ceramic

My transducer seems not to have a real polarity as the signal is feed from the transformer and a coupling capacitor so no real connection with the live. I just make sure the parallel set is paired with the the same polarity scheme. Not to the driver itself. Still on the way to build a box but I think will build two boxes apart, one for the driver and another for the pair. Simple ultrasonic cleaner anyway with just one pair of transducers also silver brazed. Will also add a timer from Aliexpress which allows programming on off periods and a fan. If you have oscilloscope you can tune the oscillation frequency to the driver by changing which output one of the transformers is in use and maximizes the voltage output wave (changing the inductance changes the resonance frequency). These chinese drivers have a short which can be changed in between the different pins of the transformer thus changing the inductance of the output circuit.

I am trying to build my own ultrasonic cleaning, along very similar lines. However there are a few difference I would like to get your (and your viewers) advice on. 1) My cleaner is to be mounted to the wall (It is also a stainless steel kitchen sink). What is the best way to minimize the ultrasonic losses from this connection. If I use continuous metal - then this will transmit the sound from where I want it, to where I don't. If I use some rubber or silicone at the boundary this will absorb the sound which will reduce the efficiency and will generate heat. If I have an air gap, then I still need some way of holding the sink up, while creating a place that water can spill! 2) I would like to try to combine a vacuum chamber with an ultrasonic cleaner. Both are good ways of degassing epoxy or silicone rubber before putting them in a mold. Any advice on how to do the two together. A vacuum chamber needs to be strong. This means much thicker stainless steel and means any welds could be weak points. Thicker steel requires more power, Any advice on mitigating this. 3) Any advice on the drainage? I have been thinking that detachable drainage might be the way to go. That removes one source of energy loss. 4) Is it possible to combine 1-3 in one build?

Hello I have better suggestion for you to test this idea of Manny transducers cancelling each other which mainly caused by phase difference between many drivers .... For testing try this.... Make differential transformer by winding 3 different coil on same torroid and connect two coils with two Drivers in opposite polarity ( depend on winding directions ) with current limiting resistor and connect 3 coil to led with resistor or any measuring device so if both have same phase ( if powered simultaneously ) it should cancel each other otherwise if phase isn't the same than led connected with 3rd coil will glow because un cancelled wave induce current in the 3rd coil so it will help to check the phase difference of board but for transduser , you could use other transduser as receiver and check if signal amplitude increase whenever you add transmitter transduser in parallel or not ! All this test can be done with low power for better control .

thank you for the infomation. im a car builder and i want a way to build a tot tank that i can out axles housings in to get the super clean and with help of another video and yours i think i have come up with the plan. now i have to get it on paper.

When you said you had no means of turning the mechanical timer off once you started it, you never tried just turning the knob back counter clockwise? Literally every mechanical timer I've ever encountered either at home or hotels for heat lamps can be overridden in this manner. That said, the Arduino method along with your other control is a nice addition.

If you build another I'd suggest 132khz transducers, I have a commercial unit which was way over $1500 but compared to my old 40khz machine it cleans way way better in fact it's like chalk n cheese , it's that much better on pcbs ..

5:08 can you explain this specific wealding for same purpose I seen on marketing video that it's nail wealding ..! So what is nail wealding ?

Awesome 😎😎😎✌🏻✌🏻✌🏻

It might be a good idea to passivate the area of the tank that was discolored on the inside. Acid is I believe how you do that. Pickling gel or something like will passivate the surface so you don't end up with rust.

i think you need to have equal length wire to your transducers 265 mm to 275 mm find the sweet spot. to sound/heat

These are great videos. The transducers are already threaded. Why not bolt them onto the pan???

Hi, I'm looking to build an US cleaner to clean up some engine parts and in particular aluminium crankcases. I will work on your rating of 100W per gallon, however , one of the aluminium cases is larger than the rest of the parts. If I size the tank for this larger part it will be larger than need be for the rest of the parts. If I have a large tank but only partially fill it for the majority of the parts I will only need about 8 off x 50W transducers. If I then fill the tank to cover the large part will the 8 transducers eventually clean the larger part if I leave it in there for a longer time OR will I need more transducers to work effectively with the larger volume? ( I want to use US cleaning as it will retain the existing die cast surface finish whereas if I use vapor blasting the surface finish will be slightly rougher and give a "keyed" surface which will take on dirt more quickly ). I have been told that using a caustic free aluminium cleaner in US cleaner will give the best result.

The epoxy they use is probably formulated specifically for vibrations and stress. High strength may not mean much stability in the long term

7:04 yes it's right choice because .... I have seen USC manufacturer KZfaqr answering someone's comment about glue , and it was indeed araldite with temperature rating of 120C M

You are still putting a whole lot of heat in with the silver soldering. Are you familiar with stud welders? This will only create a short pulse and should damage the steel less. Of course these aren’t cheap, but if you have one, you’ll be using it a lot. Colin Furze actually has one, can be seen in his tunneling videos. Or maybe you can build your own stud welder 😜

If you decide to build another one, why not bring the tank and studs to either a weld shop or maybe a tech college. The students would weld it up for you at no cost. I'm sure the instructors would want to see their students practice on things like they would find on the job, and this fits the bill. I've trained a lot of welders, and I always told them if they have something in need of repair or to build to bring it in. Best things to learn on instead of the metal the facility has to buy.

The biggest issue I see with this design is the transitor, pmw. If you had a constant power supply and hbridge controlled with a FPGA it would save space and money

Nice trouble shooting, looks like a solid build now.