Author Options:

Building an ultrasonic soldering station ? Answered

I have some strage projects coming up that will require me to solder things together that usually don't really like this.
In my past job this was quite easy as we had an ultrasonic soldering bath and several ultrasonic soldering stations.
Well, the access to those is gone :(

At first I thought "How hard can it be to build one?"
Not really too hard for a half decent soldering bath but a lot harder for a soldering iron...

The soldering stations we used operated in the range of 50-60kHz.
Commonly available transducers and their drivers however you find for 28 and 40kHz.
Finding small ones in the range of 10 to 20W is also not easy.
I can deal with a soldering iron that ends up in the size of a 500W wood burning iron as long as the tip is replacable.
But what are the frequencies really required to solder for example on ceramic, titanium or lab grade glass?
Does anyone have experience with doing this on lower frequencies than what the professional stations use?

Last but not least:
Is there any software available (preferably free or as an online tool) to simulate the quite long feed horn that is required with a soldering tip?
Going with 1/4 wavelenghts is logical but how could I get the info on the actual shape required without simulating it first?
I was thinking of using a heater cartridge like used for the hotend on a 3D printer to heat the tip.
However, tests on my ultrasonic actuator showed the entire thing disintegrates within a few minutes, at least the filler and "glue" used to hold it all together.
Would a free standing heating coil around the tip work?
Heat transfer would be a pain and losses high, but should be possible?

In case you wonder why: I just don't ave the thausands of dollars at hand required to buy one of these soldering stations....



12 days ago

Ran into some slight troubles...
There is only about 4cm from the end of the handpiece housing to where the soldering tip needs to be.
As it seems to be impossible to get a M3 thread centered on a 5mm stainless steel rod I try M3 stainless steel joiners to attach a soldering tip.

Problem one:
I built a tiny heating coil with about 60W and a fibreglass sleeve.
Despite best attempts to keep it small it won't leave much room between housing and end of coil.
As a result the shaft heats up quite quickly, which is no good for the transducer at all.
I tried a heating element from the spares box of my 3D printer with a tiny soldering tip welded to the end of the heating tube.
No chance with the vibrations due to the cables.

Problem two:
Heat affects the metal :(
I tried a 2mm stainless steel rod to extent the shaft and also to get better cooling.
Took a few hours to find a working lenght of about 6cm.
Vibrations are no way usable once you touchsomething but that was not my concern at this point.
Wanted to find a suitable lenght only.
Sadly, one the tip comes to temp the steel rod at the end also get quite hot.
And once it does all it totally out of tune.

Problem three:
Salvaging the parts from the handpiece to make a custom one seems to be impossible as well.
Not only is the thing ultrasonically welded, the inner parts are also glued in before the housing is closed.
If you know the thick, heavy duty 3M souble sided mounting tape (the clear stuff) the you know the consistency of the glue around the transducer.
It is a bit softer than silicone but tuff as nails.
The end is glued in with what seemed to be resin while the shaft just goes through a rubber grommet in the cap that is welded on.
These things, with proper use, fail once the shaft seal left water inside as it will turn into steam and stay in the handpiece when sterilised in the autoclave.
Also means I wasted a $60 handpiece to find out I should just leave it one piece and add another housing around it.

Any clues how to solve the heat transfer problem or how to design a tiny heating coil with no touch to the soldering tip?
Preferably not longer than a few mm LOL


24 days ago

I played round a bit more to figure out what might work to replace the tips of a dental scaler with something to cut or solder.

The scaling tips are all about 0.8 to 0.92g in weight.
They are all identical up to the point where the wate outlet holes it.
From there out the shape and thickness changes within certain limits.
The thread used for the tip connection is M3X0.5 - nice to be a standard and nothing homebrew.

Most obvious choice is to use a M3X0.5 coupler, they are usually about 8mm long, which is about the same as the original scaler tip at the point where it gets thinner.
I am still waiting for the ones I ordered to arrive from China as there was no local supplier to find for this tiny size.
So I salvaged one of the scaling tips, cut the working part off and soldered a stainless wire onto it.
Just about 15mm long.
Did not last long but it did not show any significant cavitation action when dipped into some water.
Bend it a little bit and soldered it on again, did not last long either...
But it showed nice cavitation on three spots along the wire.
At this point I started to wonder about the claims that these things are supposed to provide a circular motion instead of the normal push-pull we see on transducers for cleaning purposes.....
Took my tiny hydrogen torch to create a nice sized bubble at the end of my wire and the dremel tool to turn the blob into a flat stamp.
No surpises after the next test anymoe as suddenly there was really nice cavitation originating at the flat stamp surface.
Going into the water like a stream from a pump - a very tiny pump :(
Means the claimed circular motion is not for the handpiece but for the working end of the scaler tip!
The actual transducer provides the same push-pull action as any other transducer out there....
Making the wire longer however caused failure of the soldered joint quicker than 3 seconds.....
Will have to find slightly thicker wire to test if the handpiece can handle the extra weight without going out of tune.
The smallest soldering tip I have clocks in at over 5g so I guess a leight weight custom one needs to be made.
It would also make sense to use Titanium wire for the connection of tip and shaft but joing these would be a true nightmare :(

Not really the topic here, just a by product.
It should be easy to use a scaler as a cutter but it really is not.
Weight and thickness of the attachment is as vital as the lenght and shape of the blade.
Ideally you would want the cutting edge alinged with the shaft of the machine but hat would mean the center of gravity is way off and all starts to swing in the wrong directions.
Going half way either side and at an angle solves this but results in the wrong type of motion on the cutting edge.
Would be more like an axe instead of a knife.
Like with the scler tips the key might be to create thin and flexible mount that transfers the motion into the right directions on the cutting edge.
But I will try to figure this out once I can solder ;)


4 weeks ago


My dental scaler finally arrived and the thing is tiny!
No clue how it can produce up to 20W but we will see...
So, what's the first impression?

As it is a dirt cheap surplus unit the hand piece can't be dismatled, not even partially like getting the cable off.
But and 30kHz hand piece should work with the electronics.
No feedback, just a water connection and the power to the transducer.
First thing was to check the various attachments in regards to how they operate.
What looked really complex comes down to basic vibrations.
The rounded tips that are longer are made from thicker material.
Means the tip vibrations are "locked" to the design and shape of the tip.
A straight tip would certainly work too, question is how long can it be so it still performs...

Is it possible to add heating and a soldering tip then?
With a bit of 3D printing it is no problem to make a suitable attachment to hold the heating coil on the handle.
What could be a problem though is the heat itself.
Without water flow the handpiece operates fine but it also heats up quite quickly from medium power onwards.
The design certainly won't allow to use water cooling just for the trandsucer.
And changes to the horn, like adding a hole and collar to have the water circulate, will mess up the resonance.
I did a wasteful test by giving one of the tips a sharpening to test it as a knife.
Required quite some afterwork in terms of soldering on some extra material to make the tip swing right at the front (tested with water dripping on the tip to see if I get a fine mist in the right spot).
Works like any small ultrasonic cutter I tried so far apart from the fact that it get a warm handle quickly.
But it shows the basic principle behind all this ultrasonic stuff is always the same ;)

Next planned steps:
1. Designing a screw on soldering tip that, if possible, should be straight and at least 3cm long.
2. Designing a mount for the hand piece to cater for the added heating coil and wires.
3. Figuring out how to deal with heat and heat transfer in/into the transducer.

The first step is by far the hardest already.
A mass that is too great will mess with the power of the transduce, apart from the resonance problem.
A tip that has a too thin shank will most likely not perform, while a too thick shank again adds too much weight.
Funny side note: Isn't it a a bad joke that the handpiece of an ultrasonic scaler is ultrasonically welded together? LOL
Now, if you think I use some pirated high end software to design the soldering tip then I have to disappoint you here.
The tuning will be done old scholl, like with any other new antenna design.
Grab a SWR meter and check how much energy actually comes out of it.
Since we deal with sound and not radio waves I will substitute the SWR meter with a fine stream of water running down the shank onto the tip ;)
If it does not work as guesstimated then I remove some material and test again, and again, and again...
Once it vibrates where and how it should I move on to check how long it takes he heat to creep up if I externally heat the tip.
And well, from there I go back to making it all thinner (stainless steel does not transfer heat very well) and considering to add a small fan and a shielding plate or tunnel in front (or beter the back) of the heating coil.
Hopefully I solve both the vibration and heat transfer problem.
If so then I guess the final step would be to repeat it all with a scaler unit that offers exchangable hand pieces for a reasonable price.
Last but not least: Ripping one of those hand pieces apart without destroying it and then create a custom handle for all the parts with build in air cooling.

Is it worth is and what are the to be expected costs?
A little ultrasonic soldering station that you would use to make your own solar panels or do some work on aluminium and stainless sets you back about 1500 bucks second hand.
A cheap, lower duty model (a bit above hobby level) still the same price.
Using a dental scaler as the base of operations means in terms of actual costs we only need to add the material costs for the added parts.
I might get a cheap temp regulated soldering station as a donor for the heating part with a custom heating element.
So preferably a model without a temp display and added sensor.
Excluding the costs to have the printed parts ordered:
Dental scaler: Around 150$ for a new model, around 300$ if it has exchangable hand pieces and independent controls for water flow and power.
Soldering station: About 50-80 bucks, maybe cheaper.
All up with some scrap stainless it should be still inder the 500$ mark if you go with cheap units and even cheaper if you opt for working second hand units.
I hope to be able to do some work on the parts and initial testing within the next 3 or 4 weeks - real life work shedules might change this slightly though.

What would be the final goal in case it really all works out as planned?
I would like to find the best base unit in terms of surviving our abuse or at least easy to replace failed parts, assuming the output transistors will be the culprits and won't destroy other bits in the process of failing.
As the only real difference in hand pieces is the frequency and here we talk about iether 20 , 28 or 30kHz models I guess the cheapest avialable it is as we would only salvage the internals.
Then find some keen people to try it out and help to improve it or make it eaven cheaper.
Last, if I would be greedy, would be to make a kickstarter and desing my own from scratch ROFL
But I guess once there is a working prototype with instructions it will be only weeks until we find a chinese copy in the usual online markets :(


4 weeks ago

As it turns out here might be an option for "cheap" convertion :)
Ever been to the dentist to get your teeth cleaned?
Those ultrasonic scalers only work at 20to30kHz depending on the model but....

Assuming the water is not just required for the cleaning but also to cool the swinging crystal inside there might be the requirement to replace the hand piece frequently...
But it seems those hand pieces are consumables anyway, at least judging by the numbers of them you find online.

How could it work?
Well, I ordered a cheap one to find out :)
My first test would be to see if it actually works as advertised and go from there by designing a suitable tip.
The design of these things seems to cause the water to run through the tip, so it is fitted onto some sort of tube.
Does not make the process of converting it into a soldering iron easier.
I guess I shall order a spare hand piece right away.
The dry testing won't be too good and well, I need something for destructive reverse engineering, don't I?

Why I think it could be ideal way out as well as being a total disaster?
For starters we have the water problem, both in terms of the flow control electronics and the most likely cooling of the internals.
Secondly there is the way these things operate.
An unltrasonic cleaner has the transducer moving in an up down motion, or expanding and contracting if you like.
A scler on the other hand works more like pendulum.
A straight tip would oscillate in a more or less circular motion.
Right now I can now find any info whether or not this rotational motion would work for molten solder.
Means I have to figure it out....

So why bother then to waste money and time on something like this?
If rotational motion is as effictive as the push pusll motion then it would make ses´nse to eliminate the fragile, ceramic oscillator part.
And instead of very complicated and highly tuned designs we could opt for a dirt simple approch.
A BLDC motor is driven by a three-phase voltage on the windings.
And those controllers can work at quite insane speeds these days - in terms of the possible PWM frequencies.
It wouldn't be too hard to reverse the mechanical vinyl player with a BLDC motor controller ;)
Instead of a magnet rotating we would use 3 tiny magnets in front of coils at a 120° angle.
Said magnets are fixed to a stinaless steel round bar of 4 to 6mm and the bar is mounted to the handle dead center and quite firm.
We only need to make sure the length is somewhat within the wavelength of 30to60kHz and that our controller can provide that "speed".

Possible benefits of a pure mechanical system....
Considering that heat and ceramic oscillators don't do well together the worst problem would be eliminated from the start.
Apart from the challenge of creating a suitable mount in the node with no movement of the "axle" the benefits (if working) are quite obvious.
You can use a threaded and replaceable tip and with that different tip sizes.
An external heating coil around the tip means no problems with vibrations destroying the coil or ceramic insulation.
Did I meantion cheap already?
If you have a half decent understanding of electronics or microprocessors you can design your driver yourself.
Otherwise a cheap BLDC controller should be able to make the impossible possible.
However, the best part I think is versatility.
Using magnetic fields means we don't have to worry about things wearing out too quickly.
And because movements are usually in the range of just a few µm in this type of application we can even use a thin rubber sleeve between magnets and coils to make it "silent"...
In theory it would even be possible to ignore the self resonance within the axle and to use a fixed system instead.
Here the amplitude of motion would be created by the length of the axle outside the handle.

Can it be done even more simple and cheaper?
I thought about it....
One idea here that is relatively easy to try fro most people with some scrap electronics is the ultrasonic hammer approach.
Imagine you have a nice round rod in a smooth and easy glide sleeve.
Any motion applied would move the entire rod in the mount.
Now, when it comes to resonant motion it is best to use a sine wave input as the sqaure waves can produce quite some nasty side effects.
Imagine you would mount a round neodymium magnet on one end of the rod and said magnet would be inside a suitable electromagnetic coil.
With a 55kHz sine wave applied to the coil our magnet would try to move at 55kHz up and down inside the coil.
Of course it makes sense to utilise some caps and ZVS system to keep it all self resonant even if we apply some pressure onto the tip.
Obvious downside is mechanical mass.
even if the rod would only have 50grams in weight it still means it needs to be moved.
Unless the mechanical mounts can provide some mechanical resonance our coil and power requirements would be rather larger.
I guesstimate around 200W to get sufficient motion transfered.
Last but not least is the resonance of the rod itself.
If the lengths allows for resonance effects at the driven frequency it coul intere with the mounts and through that drive it all out of resonance.
But if working you could build your own ultrasonic soldering machine for under 50 bucks....

Jack A Lopez

6 weeks ago

Um, this topic was the first place I actually read the phrase, "ultrasonic soldering."

But apparently this is a thing, or an art,

I have never actually seen these things, so the lack of pictures in the Wiki article is disappointing.

Pictures are out there,


Also, via those pictures, I found:

James Byron Jones's 1957 US Patent (2803735)

Hunter Scott's Hackaday.io project page

The main problem I have with trying to comprehend this is, is that there is a lot I do not grok about ultrasound, like for example,

Why are voice coil type speakers (i.e. coil + magnet) typically not used as transducers for ultrasound? And why are piezo type transducers prefered?

What materials are transparent, or opaque, to ultrasound?

Also you keep mentioning something called a "feed horn", and I am guessing that is a kind of "wave guide" with impedance that changes slowly with distance, analogous to the same thing used for RF, or microwaves, or audible sound, but I am not picturing it.

In particular I am wondering what material is inside the horn. Is it air? Is it solid metal? Or is it something else?

Downunder35mJack A Lopez

Reply 6 weeks ago

Well, instead of making a long and boring answer here I created some topics for it.
Maybe you like the ideas and want to try it one day - it is quite fun!

Jack A LopezDownunder35m

Reply 6 weeks ago

I have to admit, there is some powerful magic here; i.e. the ability to solder to glasses, ceramics, aluminum, etc, and that could be handy for many things.

It is just that I have not yet got my head around how to make, and channel, ultrasound energy.

Downunder35mJack A Lopez

Reply 5 weeks ago

It is really as simple as it is complicated.
You can screw a transducer to a stainless steel pot or even aluminium or glass to make your own ultrasonic cleaner on the cheap.
It works but it does not mean it works perfect.
Same story for soldering.
Creating a working solution that is reliable enough for commercial use is costly.
This reflects on the price you would have to pay for even second hand units in bad shape.
Again, the basics are dead simple:
You have a transducer mounted in a handpiece.
On the transducer you have a more or less long horn that ends in a soldering tip.
Around the end you have a free standing heating coil to heat the tip.
The reality is what makes it so complicated.
The transducers we can get are usually quite bulky and powerful.
about 45mm in diameter and 50W seems to be the standard.
20 to 30W would be sufficient though.
With a max of 40kHz a horn would also end up much longer than for the commercial 50-65kHz models.
Last but least is the problem of finding the right spot to join handpiece and actuator/horn.
This can only happen in the nodes with no movement and even there it already affects the resonant frequency.
Biggest problem however comes with the mounting bits.
So far I always struggled with screw connections unless they included a glued surface.
Spring washers, thread locker glue and such always failed me shortly after turning the power on.
Adding set screws or pins messes with the vibrations and resonance.
Need more tinker time but will get it working one day :)

Jack A LopezDownunder35m

Reply 5 weeks ago

I guess I need to get myself one of these transducer plus driver packages, and do some experimenting myself.

I see these things on eBay, like this listing,

The transducer in the picture, seems to have stubby little bolt in the top (I guess for bolting it to things like a pot of water, etc) and it has some scribbles on it, "HESENTEC" and "RANK: C"

That manufacturer has a web site,

I am still looking for documents, like data sheets, application hints, and howto, if anyone has bothered to publish such things. There could be some clues.

Or maybe I should just buy the kit, and try bolting it onto things.

Jack A LopezDownunder35m

Reply 6 weeks ago

I think there was a very long delay (about a day) between you creating those new topics,



and those new topics being viewable to me, and maybe everyone else.

You know, the orbit of Neptune is only about about 4 light hours from the orbit of Earth. So that if you had to send a radio message to a friend at that distance, and wait for the reply, it would be a minimum of about 8 hours.

But that would still be faster than starting a new Instructables forum topic.



6 weeks ago

Ok, due to the ongoing requests in the single form of a single user I will provide some more insight... ;)

Ultrasonic soldering works the same way as your dentist cleans plaque from your teeth.
Only difference is it uses hot solder instead of water.
But since we need to work with molten metal we need a bit more energy.
Otherwise it would be cheap and easy to just add a heated tip to one of these dentist cleaners.
What gives us the energy for the vibrations is the feed horn, sonotrode or however you prefer to call it.
Resonant frequencies are fun, same for what they can do.
In our case the horn acts like a mechanical antenna or better at hydraulic hammer.
Imagine two pistons of equal volume connected to each other.
Like a thin and long hydraulic cylinder and short but large diameter one.
Apply movement to the thick one and the thin one moves a much greater distance.
For our soundwaves it is similar.
The metal of the horn stretches and contracts at the resonant frequency and if the lenght is correct.
And like when you grandfather used his hearing horn the amplitude of the signal will increase - it gets louder.
Louder means the metal in the thinner half will contract and expand much more than the thick part.
And with the mating right at 1/4 of the wavelength the little push from the big part will be added to the already longer push of the thin part.
However, if the measurements are not perfect it all just wobbles around in all directions but lengthwise.

Tricky part on a hobby level is to find the right way of mounting it all into a suitable handle.
From there how to heat the vibrating tip without actuall ahving any contact to it.
turns out even slight pressure from the sides or adding material totall ruins the resonance and nothing works anymore.
A free standing heating coil with some soft fibreglass insulation around it seems to work sometimes.
It is not reliable and not very efficient in heat transfer either.
Currently thinking of welding a stainless steel tip to a mild steel horn so I can use inductive heating for it all.
Problem is heating the horn also changes the resonance slightly.
Wouldn't it be nice to have a driver that is self tuning and adjustable in the power output for both the heating and the ultrasound? :(