Author Options:

Ultrasonic soldering bath Answered

Making a working ultrasonic soldering iron is not as easy as I though it would be.
Finding tanrsducer of suitable design and size is even harder.

So I thought I start with something easier and share the thoughts here.

If you need to solder impossible to solder things then quite often you could get away by wetting the entire area.
For example the end of a wire or a lug where it won't matter that you can solder on the bottom as well as the top.
Back in my days flux core solder was a rare and very expensive thing to find.
So we had a little soldering pot and flux pot instead for working with lots of wires.
Dip, dip, done....
The pre-soldered wires where then easy to work with and the ramaining flux on then was enough.
Doing this for metals like aluminium, stainless steel or even ceramics seems impossible at first sight.

China offers cheap ultrasonic transducers including the required driver electronics for very littel money these days, despite the trade wars.
The most obvious solution would then be to get a cheap and big enough soldering bath and to attach the transducer to it....
Won't work though and if it does then not for long.
Problem is firstly the heat transfered to the ceramic parts of the trandsucer and secondly the fact that most of these soldering baths use quite thick steel for the container.
Add the that you deal with quite some grams of molten metal and you know where I am going.
Building your own ultrasonic soldering bath to solder the impossible with ease!

Project costs:
40kHz transducer with driver board : about 50 bucks.
Thin walled stainless steel bowl ( about 50 to 100ml but go bigger if you like) : about 2 bucks.
Leftovers for an enclosure can be wood, plasic or your favourite 3D printer.
Ultrasonic horn: About 500 bucks from your favourite engennering company or you need to make it yourself - I prefer the later.

Main design considerations for the horn:
We need something to keep the heat away from the transducer that also amplifies the power coming from it.
That is why we can use a bowl or container that has a small bottom daimeter as the transducer if need be ;)
There is a good reason a commercial horn costs a lot of money.
They are preferably made from titanium and they need to perform as advertised right from the start.
We substitude by using some aluminium round stock and a lathe.
It is advisable to leave the transducer as it is!
Do not take it apart to mount your horn directly onto the ceramics!
Use a long enough set screw or include the required thread on your horn to mount it onto the transducer.
If you prefer to use stainless steel doe to the lower heat conductivity then be my guest.
The horn should have the same diameter as the mating part of the transducer for a quarter of the wavelength of the transducers frequency in the given material.
Please look up how fast sound travels in your choosen material and calculate it properly.
Having the lenght of the thick part right is quite cruicial.
The thinner part that amplifies our movements should be about a quarter of the diameter of the transducer.
For example: if the mating face of the tansducer is 40mm in diameter then the thin part of the horn should be 10mm.
The length again is a quarter of the wavelength or the same as the thick part.
Where thick meets thin please allow for a 3 to 5mm radius and make sure this area is nice and smothly finnished.
Now, length is quite critical here....
As we will mount our finnsihed actuator free hanging under the bath we need a feasable way to comapensate for our tolerances by creating our horn without a simulating software.
I found that welding a short stub onto the container works best but with aluminum it is harder.
I assume most will opt for welding a 6mm soft steel threaded rod onto the container.
Either way the container surface must be kept flat for the mating surface of our actuator rod.
So it is best to make the stud yourself or to use a suitable replacement - like using some flux and your stick welder for create a makeshift spot welder ;)
If you decided on using steel for the horn then of course you can just mill a 10mm piece with a suitable thread and flat mating surface...
What you want to end up with is a screw connection that has a flat mating surface and no empty spaces, fine thread prefered.

Tuning the horn....
The ensclosure is easy to make as a box, so the only thing to worry about is insulation but nothing to affect performance.
So I just assume you have it all ready ;)
With the horn at one quarter wavelength either end our thin end will be too long unless a short stud is used for a direct fit.
So whatever you had to add for the part on your container or bowl need to be removed from he horns thin end.
Try to keep the gad for the threaded part as small as possible as it affects the resonace.
As things never turn out perfect the first try I prepare some thin steel washers - 100mm outer diameter in case you wonder and stick with the above example.
I use a strong neodymium magnet and belt sander to create washers from very thin to slightly thinner ;)
Taking off slightly more from the horns end will then allow toadd these washers if required - but please do a try as it is first when you think you got the measurements all right!
For an aluminium horn you will of course use aluminium washers here.
To do so fill the container with some water and place a sheet of thin alumiium foil on top of the water.
Turn it on and within a few seconds you should see holes appearing in the fiol or even small fractures.
If nothing but noise happens it is quite certain your rod will be a bit too long.
Unscrew and take about one tenth of a mm off the thin end of the horns mating surface to shorten it.
Try again with the foil and if no better remove some more material.
Once you see some action try adding a layer of aluminium foil between the mating surfaces - screw it tight!
The foil won't last long but if the action on the water is far better until it fails you know you took off too much.
The washers come into place if the tuning won't work at all.
Sometimes you can cut off a little bit again and again but the piece will remain too short ;)
Especially if you have an aluminium horn and needed to use a steel screw on the bowl...
So once the shortening of the horn fials you add a washer to get slightly above the original length and start replacing the differently thick washer until you find a sweet spot.

The tricky part is over, now to solve the heating poblem...
Using some glass seal as used on wood fire ovens not olnyl provides good insulation to our enclosure but also prevents the vibrations from spreading too far.
As our hardware store won't just give use the little bit we need the rest can be used to insulate our container.
Dending on the size and shape of your container I hope you decided to buy a container tha fits your heating element...
I found that replacement coils for lab heaters work fine but some small fan heaters also use round heating elements instead if wire spirals.
For a custom shape it is quite easy to use a coil of heatin wire rated for your mains voltage and a glass fibre sleeve for insulation.
To keep it all in shape just wrap some steel wire over it - over the insulated coils of course.
The temperature control can be as fancy as with a microcontroller or as simple as using a dimmer like I did.
Most heating elements will go glowing red hot if the mains voltage is not reduced.
It makes sense to limit the dimmer's movements accordingly by testing it.
Just do it in the dark afeter exposing a small bit of the heating wire from the insulating sleeve.
Once you see a faint glow coming dial it back a bit until you can see any glow - that should be the max setting.
For a big bath or to save time you can of course crank it up to what the glass insulation can tolerate but be aware that solder can boil over!
I do a temperature check either with a touch free IR thermometer of by checking how quickly some rosin boils off.
If you need to dip bigger parts you need a higher temperature, so I think a digital or sensor temp control is not really required.
Once you found a sweet spot to hold the solder temp long enough without getting too hot or cold just mark it for reference ;)

Using the ultrasonic soldering bath correctly.
Cavitation is what the work for us, so we only need to activate the ultrasonic part when we need it with a push button or food pedal switch.
We do not use any flux or resin!
That means if you used the bath for normal soldering and or resin then clean the remains off the surface first.
A shiny and clean surface is best but the oxidisation will happen quickly so don't be too disappointed ;)
Start by dipping in a clean copper wire.
Some solder might stick but it won't look proper.
Now dip it in again and while it is in push the button for about 3 seconds.
Like magic, if tuned properly your wire is soldered and properly covered to where it was dripped in.
Try the same with some slightly sanded or at least clean aluminium wire, but use the button right away for about 5 seconds.
The wire should be coated with solder once more.
You can try a glass rod or some stainless wire next but I guess the working principle is clear now ;)
Not everything will bond with solder, especially not if it is not clean.
A piece of glass with your fingerprint on it might just fail and some ceramics will only let the solder stick without actually bonding.
You should always check the mechanical strength of your soldered connection before having to rely on it ;)

And why would you need such a machine?
Well, most people won't have any use for it.
Those who do might not be able to afford a commercial model.
And there is always those who just want it all...
If you know why you need such a thing than you have an alternative now at a fraction of the cost.
You only need a lathe or someone who can machine the horn for you.
Another benefit is that for smaller containers it is possible to weld a small "bridge" over the top.
Should be placed so the bottom is in the solder while top is above it.
In many cases you will then be able to use this plate to heat up whatever you need to solder on.
Like a glass plate where you would like to solder a wire to.
Once up to temp turn the ultrasonic part on and use a normal soldering iron and flux flree solder.
Works quite well for these small solar panel kits...

Ok, and how far away is our cheap ultrasonic soldering iron?
Not that far :)
I already have a topic for this though....


The forums are retiring in 2021 and are now closed for new topics and comments.

1 year ago

Finally a cheap soldering pot arrived from China.
First order of business was to take it apart - isn't that what everyone does with goods from China? ;)
Screws missing, parts loose, bad design throughout, but what to expect for less than 30 bucks...
To my total disappointment these guys changed the design from what I had a few years ago.
The solder pot is no longer removable but instead its bottom is screwed to a cross bar in the housing.
The heating coil is no longer a nice wire inside some glass insulation but a clamped sleeve.
By clamped I mean the type you need special pliers for and that can' be undone without breaking things...

I saw the srew mount at the bottom as an oppertunity at first to mount the transducer with some crappy horn.
An initial test however showed that no matter how I try there is no good vibrations as in the famous song...
I guess the mounting bracket is one culprit here but the main problem seems to be the pot itself.
In the past these were made from a quite thin alloy that had a very high nickel content.
Solder wouldn't stick to it and with some pliers it was easy to filly a small mold with pewter or similar.
The new pots are made from thick stainless steel that was annealed to get a thicker and colored oxide layer to prevent getting wet by the solder.
The small pot is actually heavier than the entire rest of the scrap of the "machine".
In hind sight it would have been better and easier to buy one of these soldering pots on a handle....

But there is (somehow) good news too.
With the pot disaster and my initial plan falling apart I got word that there is another option to deal with the heat and vibrations.
Instead of a feed horn, sonotorode or whatever you call it in your region a tube can be used.
Not nearly as affective as an amplifier as a conical horn but with 50W I think still enough.
Downsides of using a tube or pipe on the transducer:
# The length is still critical as otherwise the pipe just "wobbles" or contracts and expands in diameter instead of the length.
# Mounting is a true pain in the behind!
You can't just add a thick enough plate to screw onto the transducer as it would mess up the contractions and tuning.
You can't really create a fitting just on the outside ot the transducer either as it would fail too quickly.
# It will be very loud, so even though we can't "hear" the noise, hearing prtotection would be highly recommended.
Some positives about using a pipe:
# If I could find something with the right diameter AND fine pitched threads on it then tuning would be way easier.
Instead of cutting and sanding in the hope to find a match I would only have to deal with the thread glue and turning the pipes to adjust the length.
# A pipe puts the load in a concentrated form onto the attached pot.
With thin materials for the pot it would creat something similar to a speaker cone in the area of connection - more vibrations will be transfered but at the cost of material fatigue, something to remember and check every now and then.

With all the things starting to get more compicated than what I remembered from my last experiments in this region I am wondering....
I checked some videos on what is all possible with ultrasonic soldering, benefits, problems and so on...
What started to stand out is that in many cases you could do without all the complicated stuff.
Take glass as a prime example.
Be it small solar panels or fancy looking circuits on a drinking glass - the glass needs to be heated to the required soldering temperature.
And the same way I already cheated some years ago it still happens in the professional area like this.
Either the glass is heated on a plate and an ultrasonic soldering iron is used once all is hot enough, or the glass is heated over and ultrasonic soldering bath and a normal soldering iron is used.
So why not try it the dirt simple way first? ;)

No solder soldering pot?!!?
I could imagine use a modified 3D printing hotbed to get things to solder temperature.
Bit thinner though and small size.
If a stianless steel set screw is used to mount the horn plus something like a mica sheet to block the heat it should be possible to do single solder runs before the transducer gets too hot.
Some 3D printed enclosure with a fan should keep the horn cool - imagine it sitting in a tube with constant airflow...
Of course some suitable clamps are required to keep the glass or whatever needs to be soldered firmly in place and to allow a good heat transfer - maybe with a drop of oil...
Only downside would be that you need a drop of solder on your iron and that you might have to scratch a little bit.
Not 100% sure how this will affect a decent soldering iron but will start my test with a cheap one ;)
Any clue how long a neodyminum magnet would survive vibrations and soldering temperatures?


Reply 1 year ago

you have dipped an aluminum wire into this bowl of molten solder, and succeeded in getting a layer of solder to actually adhere to the wire


9 months ago

Thanks for posting your notes - saves me a lot of legwork. I'm thinking now a quick and dirty solution may be to have a stack of stainless steal washers fixed to the transducer as a heat break and mount everything below the pot but not contacting so that a mechanical leaver is used to press the two in to contact for the brief time necessary while tinning the substrate. Depending on what sort of duty cycle was required a fan could be attached to ensure the heat break was back down to safe limits before the next time the transducer was required and the fan along with the transducer powered down after a grace period.

Jack A Lopez
Jack A Lopez

1 year ago

Have you actually done this?

I mean, have you successfully coupled a bowl of molten solder to an ultrasonic generator?

Also you have dipped an aluminum wire into this bowl of molten solder, and succeeded in getting a layer of solder to actually adhere to the wire?

And you have repeated this trick with other materials, including a "glass rod" and "stainless [steel] wire"?


Reply 1 year ago

I did and I have...
Although only with a cheap 40kHz actuator and driver board.
Back then I though it is smart to use a smal soldering pot and just attach the actuator directly.
You know these $20 pots with crappy heat control and no real safety, made in china...
Worked quite well for about a year until it totally failed with smoke from the driver.
Who would have thought that mounting the aluminium part of the actuator directly to the solder bath would cause a heat trasfer into the transducer??? LOL
Let's just say the ceramic disks don't do well in heat...
When I comapered my crappy design with a commercial unit I realised they used a stainless steel horn to keep the heat away from the disks and to increase the amplitude of the vibrations.

For the rest of your questions:
Soldering aluminium wire was my main purpose for building it.
But with just 40kHz I had to rely only on the about 50W it produced.
Unlike water you don't really see much movement in the solder, so finding the sweet spot for the fill level was crucial.
For my tiny $20 pot that was somewhere between half and 3/4 full.
I moved the wire around in the bath and up and down until it got a good coating.
It had a sweet spot for this too, about half way down in the center but was not always able to use it due to wire insulation and such.
And for some reason lead free solder works better here than lead containing solder.
You can somewhat feel it happening when you hold the wire.
A spot where you get a slight tingle means it is vibrating badly and will have solder on it.
Stainless steel wire worked the same way but required a higher temperature.
For example if normal soldering worked at around 220°C then stainless would require about 25° more.
I guess it has to do with the bad heat transfer stainless has.
Trying with some old stainless however is almost certain to fail.
The oxide layer needs to be removed by chlorides or fine sanding first.
I think with a frequency of around 55kHz like we had for our pro models at work would be better but I could not find any transducers for these frequencies.
Glass and ceramics however can be a pain!
Small stuff like a stirring rod can be diped in after a nice acetone cleaning.
The wetting can only happen in areas where you get strong cavitation effects on the surface of the glass.
To actually solder on glass I found it much easier to use spring clamps (like on a microscope) to hold the glass over the almost empty bowl.
Or on a bigger one to screw a steel plate over the bowl.
Once heated up you can use a normal solderin iron in the node areas to get the solder to wet and stick to the glass.

I am considering to buy another cheap soldering pot, mainly because it is still cheaper than buying all components seperately.
But I need to desing a suitable horn to limit the heat transfer and increase the amplitude of the vibrations.
Will be a rather high soldering pot as a result LOL
Sadly I can't machine stainless on my tiny lathe without ruining tools and motor, so I will have to figure something out first.
I am contemplating about using an electromagnet instead of the transducer.
Small transformer cores out of ceramic can be salvaged from old switchmode power supplies.
Would allow for a wide frequency range but comes at the price of getting it to work reliable.
If I find an easy way for either solution I will post something for other tinkers to replicate.

Jack A Lopez
Jack A Lopez

Reply 1 year ago

Well, I am impressed by this.