Ultrasonic cleaner size: 10L vs 6L

[SolderOcelot]

So I bought

this cheap ultrasonic cleaner from AliExpress as my first ultrasonic cleaner ever to undo liquid damage on laptop motherboards. Yes, I cheaped out, but since these are being sold at big chains in US and UK I thought it might not be too bad as an entry model.

I'm looking for advice from experienced users, specifically on the size. I'm very obsessive about getting the right size and I doubted long and hard and got the 10L instead of the 6L variant. Only the width is different and thus in the 10L variant you can lay the motherboard flat and in the 6L often times you have to lay it at an angle so it fits.

• Basket size 6L: 30cm x 12cm x 15cm (200W heating unit / 180W transducer)
• Basket size 10L: 30cm x 24cm x 15cm (200W heating unit / 240W transducer)

Round 1
So I took a laptop motherboard with some water corosion as a test sample. Here is the process of the first run:

• Fill it with Distilled water / IPA @ 6:1 ratio
• Heat the tank to 60°C
• Let it run for 5 minutes without board to degass because it has no dedicated "sweep mode" for degassing
• Place the board in (damage facing down for optimal agitation) and let it run for 5 minutes
• Result: corosion only ~25% removed

Round 2
I talked to my neighbour and he said that I should have gotten the smaller tank because the agitation / vibration would then reach the board easier and you would need less runtime. So I did a second run with longer agitation time:

• I use the same liquid (lol)
• Heat the tank to 60°C
• Let it run for 15 minutes
• Place the board in (damage facing down for optimal agitation) and let it run for 5 minutes
• Result: the gunk is gone and you can even see factory flux getting removed / displaced as well so effectiveness increased. However, you can see ultrasonic cleaner damage to the paper sticker, see photo. So a 15 minute runtime might be risky.

I have 2 straightforward questions:
1) Do you think the 6L basket size is enough for most laptop motherboards?
2) Is it true that less runtime is needed for the smaller tank and thus is safer for laptop motherboards (or any other populated pcb for that matter?

[/list]

[SteveThackery]

I'm afraid I cannot contribute anything useful regarding the suitability for a motherboard.  However, I have been using ultrasonic cleaners for 35 years or so.

The ultrasonic power is important, and in particular in relation to the volume of the fluid. Some low cost cleaners use the same circuit board and transducers over two or even three different sized models.  50W of ultrasonic power in a small tank can punch holes in metal foils, but in a larger tank can be insufficient to do a decent clean.

I've never heard of mixing water and alcohol for use in an ultrasonic cleaner, but I must emphasise again that I haven't used an ultrasonic on electronic boards.  In the early days I used trichloroethane 1.1.1, which worked brilliantly, but of course every time you uncapped the bottle you wiped out another cubic mile of ozone (this is an exaggeration, but you know what I mean!).  Nowadays most ultrasonic cleaning fluids seem to be aqueous: surfactants plus water, although those intended for use on metal parts usually have a hydrocarbon base.

In terms of cleaning time, I often let the machine run for up to 30 minutes.  There is a reason: if your cleaner does not have a heating element then the longer you run the cleaner the hotter the fluid gets, and that always (in my experience) helps the cleaning action.  A heating element is definitely worth paying a bit extra for, in my view.

I realise you know this, but I want to emphasise it: the transducers are affixed to the bottom of the tank by a special cement which is glass-hard, in order that it doesn't absorb any energy from the transducers.  It is therefore very important NOT to pour hot water into the tank, because the sudden differential expansion CAN cause the cement to crack.  Not always, but yes, it has happened to me.

Nearly all cleaners use a swept frequency, but usually swept at the mains electricity frequency which I suspect is too fast.  I modified one of my cleaners to sweep much more slowly, and I think it is better because it allows the vibration patterns to build up in the fluid.  Some cleaners offer a manual control over the frequency (you should set it for maximum agitation), and others have a closed loop system where they automatically find the frequency at which maximum energy transfer from the transducers to the fluid takes place.

[SteveThackery]

About your question 2 specifically: it isn't a straightforward answer because it depends on the power density in the fluid.  If the same wattage is being concentrated in a smaller tank then yes, the necessary runtime is less.  Otherwise you'll need to compare the watts-per-litre of each machine to compare.

About degassing: perhaps my experience is insufficient, but I've not come across a "dedicated sweep mode" for degassing. All of mine pulse the power at about 0.5Hz and 50% duty cycle to do the degass.  Maybe the more professional ones use special sweep modes?

Edit: I'm too embarrassed to put up yet another post, so I'm going to cheat by adding some stuff to this one!  Another point I want to add: some cleaners have a plastic lid, others a metal one.  In general I've found that the plastic ones tend to be a better fit, and importantly, they mute the horrible noise from the cleaner more thoroughly.  I cannot guarantee this is true for all cleaners, but perhaps you will arrive at your own conclusion.  I can't stay in the room for one of my cleaners because the noise is unbearable - a really harsh, high frequency, teeth-tickling racket.

Be aware, though, that if you do use one with a plastic lid then you probably won't be able to use anything other than water (and some alcohols, maybe?) in the tank. In my experience most solvents eventually end up causing the plastic to fog, craze and - if you wait long enough - crack.  Some make the plastic sag instead.

If you want to put lots of small, and possibly delicate, parts in the cleaner you ought to put them on a platform so they are off the base of the tank.  Plastic trays absorb ultrasonic energy, so something stainless steel would be better. Fine meshes (like a tea strainer) greatly attenuate the ultrasonic vibrations.

One last thing. The design of the cleaner in terms of keeping the cleaning fluid away from the electronics inside is obviously very important, but you'd be amazed by some of the crap designs.  They all have a seal around the top of the tank, but the good ones also have a kind of double-walled arrangement whereby any fluid that does get past the seals is caught and directed to a drain hole underneath the cleaner.

Sorry - this has become a bit of a brain dump.

[SolderOcelot]

First of all, thank you for all your input. Although it doesn't directly address my individual issues, still an interesting read.

Some low cost cleaners use the same circuit board and transducers over two or even three different sized models.  50W of ultrasonic power in a small tank can punch holes in metal foils, but in a larger tank can be insufficient to do a decent clean.

The 6L variant has 200W heating power and 180W transducer power. The 10L has 200W heating power and 240W transducer power. Seems like they took this into consideration.

I've never heard of mixing water and alcohol for use in an ultrasonic cleaner, but I must emphasise again that I haven't used an ultrasonic on electronic boards.

Finding the right recipe is the next project! Here in NL there aren't many ready made stuff to be found.

In terms of cleaning time, I often let the machine run for up to 30 minutes.  There is a reason: if your cleaner does not have a heating element ...

True, but mine does! :-) it takes 45 minutes to heat from 24C to 60C.

Edit: I'm too embarrassed to put up yet another post, so I'm going to cheat by adding some stuff to this one!  Another point I want to add: some cleaners have a plastic lid, others a metal one.

Mine is metal and I can not sit in the same room..

[SteveThackery]

The 6L variant has 200W heating power and 180W transducer power. The 10L has 200W heating power and 240W transducer power. Seems like they took this into consideration.

Looks like the 6L one should be a bit more punchy, then.  By the way, I'm sure some boffin will tell you that "watts per litre" is not the whole story, but it seems about right to me.

[metrologist]

I got the 10L version and am happy with it, although I think the heating power is 300W on mine. I believe the 6L tank has 3 transducers and the 10L has 4.

Boffin: Yes, you have to take into account power losses from the surface area of the tank relative to the volume 

[SolderOcelot]

Boffin: Yes, you have to take into account power losses from the surface area of the tank relative to the volume 

Surface area meaning also the walls right? I'm asking this because my theory is that the waves are stronger in the bottom of the tank and weaker in the top part by diffusion. In that line of thinking a 10L with lower wattage per liter would clean better than a 6L because the PCB's can lay flat and have most of their surface closer to the floor of the tank.

Or am I overanalysing it?

[tszaboo]

Get the bigger one. The items in it are not supposed to touch the wall. You are loosing space here and there, and the first time something doesn't fit you regret it.
I certainly did when only half my H0 model train tracks fit into it for cleaning, because the others were too big. Also you want to run it in another room or wear ear protection (or preferably both).

[SolderOcelot]

Get the bigger one. The items in it are not supposed to touch the wall. You are loosing space here and there, and the first time something doesn't fit you regret it.
I certainly did when only half my H0 model train tracks fit into it for cleaning, because the others were too big. Also you want to run it in another room or wear ear protection (or preferably both).

This comment is more or less the type of confirmations I was looking for. I already had made the decision and have the 10L sitting here. I was just slightly tempted to swap it for a 6L because I thought it would be more effective for my use purposes because of
1) my neighbour's statements
2) seeing Louis rossman using a 6L which barely fits his logicboard



For the record by the way: I always use a basket and never let anything touch the tank walls.

[metrologist]

Boffin: Yes, you have to take into account power losses from the surface area of the tank relative to the volume 

Surface area meaning also the walls right? I'm asking this because my theory is that the waves are stronger in the bottom of the tank and weaker in the top part by diffusion. In that line of thinking a 10L with lower wattage per liter would clean better than a 6L because the PCB's can lay flat and have most of their surface closer to the floor of the tank.

Or am I overanalysing it?

Yes, somewhat overanalyzed because these tanks are really on the small scale, and yes the walls of the tank are where more losses occur. Wave propagation through the liquid is not much an issue, more so when reflecting off the sides. You can also do tinfoil tests to see any hotspots, but I could not really find any. I would have opted for a 15L if I had planned on regular use, but glad I did not get a smaller one.

A tip is I've seen folks use jars to contain smaller pieces and more expensive cleaning solvents, so you don't have to fill a 10L tank with expensive cleaning agents, and then they're easier to reuse. It works great.

[CaptainBucko]

A tip is I've seen folks use jars to contain smaller pieces and more expensive cleaning solvents, so you don't have to fill a 10L tank with expensive cleaning agents, and then they're easier to reuse. It works great.

I don't use a jar myself, but I do use smaller zip locks bags so I can just use plain tap water as the main liquid in the tank, both with ISO and PCB Cleaner solution where required. It works really well.

[SolderOcelot]

I don't use a jar myself, but I do use smaller zip locks bags so I can just use plain tap water as the main liquid in the tank, both with ISO and PCB Cleaner solution where required. It works really well.

I thought plastic is too good at dampening the soundwaves..

[CaptainBucko]

I have not noticed any issues myself, but I don't doubt it is (slightly) less efficient in sound wave transmission.
The benefit of using the barest minimum amount of cleaning product is significant however.

[MathWizard]

So what happens if you don't degas ? Do bubbles just stick to the PCB and not get cleaned as good ?

I've watched video's of UC, but don't remember the noise, certainly not a loud one anyways.

Do people dip big things like whole computer PSU's in them ??  I have no plans to get an UC. Just that much distilled water and whatever else must get expensive.

[CaptainBucko]

In the early days I used trichloroethane 1.1.1, which worked brilliantly, but of course every time you uncapped the bottle you wiped out another cubic mile of ozone (this is an exaggeration, but you know what I mean!). 

1,1,1Tricho was awesome stuff - I used to joke it was the main flavoring ingredient in Terry's Chocolate Orange due to its subtle sickly sweet citrus odor.

[CaptainBucko]

• Fill it with Distilled water / IPA @ 6:1 ratio

Try using a proper PCB cleaner, not IPA + Water Mix. Depending on your country there will be various suppliers. I find the PCB cleaning solutions much more effective and removal of contamination, so you don't need to spend so much time with the item in the ultrasonic cleaner.

[SteveThackery]

So what happens if you don't degas ? Do bubbles just stick to the PCB and not get cleaned as good ?

Truth be told, I haven't noticed any difference. Also, I've observed that no matter how long you degas, you still get bubbles forming and rising to the top.  Or at least, that's what I've found with mine.

Perhaps its importance depends on the particular fluid you are using.

[SolderOcelot]

So what happens if you don't degas ? Do bubbles just stick to the PCB and not get cleaned as good ?

More bubbles = more risk of damage. Look up "aluminum foil in ultrasonic cleaner test".

Try using a proper PCB cleaner, not IPA + Water Mix. Depending on your country there will be various suppliers. I find the PCB cleaning solutions much more effective and removal of contamination, so you don't need to spend so much time with the item in the ultrasonic cleaner.

I wonder what it is about these solutions and why I wouldn't be able to make it myself. So far I've understood these are the essentials:

• Carrier (distilled water)
• Solvent (IPA or something else)
• Surfectant (dish soap)

[SteveThackery]

More bubbles = more risk of damage. Look up "aluminum foil in ultrasonic cleaner test".

I'm not sure that follows. There is a mechanical cleaning action with an ultrasonic cleaner, and any mechanical action can lead to damage. You manage this by keeping the cleaning time as short as possible commensurate with getting a successful clean. Also, and perhaps obviously, the higher the ultrasonic power density in the fluid, the greater the risk of damage.

I don't see how either of these relate to the visible bubbles.  In fact, the proposed relationship seems to be the other way round.  That is, degassing to reduce the bubbles increases the cleaning action.  So it would seem that the bubbles themselves actually reduce the mechanical cleaning action (by absorbing the shockwaves because the bubbles are "springy") and thus the risk of damage.

Only if the bubbles are used as an indicator of the power density could you say that more bubbles = more risk of damage.  That's why I think the opposite to SolderOcelot: insufficient degassing reduces the cleaning action and the risk of damage.

(PS:) You can't have a mechanical cleaning action with no risk of damage. The fun you can have perforating aluminium foil is simply a demonstration of that fact.  You can also use it as a crude indicator for how powerful the shockwaves are.  That is all.  Degassing the fluid increases the shockwave power, thus also the cleaning action and the risk of damage.

[tooki]

I wonder what it is about these solutions and why I wouldn't be able to make it myself. So far I've understood these are the essentials:

• Carrier (distilled water)
• Solvent (IPA or something else)
• Surfectant (dish soap)

Commercial PCB cleaners/flux removers for ultrasonic are aqueous, i.e. not solvent-based. The key ingredient is a saponifier (lye or similar), plus surfactants (probably something closer to what’s in all-purpose spray cleaner like Cilit Bang), plus corrosion inhibitors and probably redeposition inhibitors. These cleaners don’t rely on solvents to dissolve the flux, they use the saponifier to make the flux water-soluble.

(Solvent-based cleaners that can be used in ultrasonic do exist, but they don’t contain water. At least I’ve never seen one that does.)

Edit: I’m not so sure now. One aqueous cleaner (Safewash Super, SWAS) does contain various alcohols, but they aren’t normal alcohols (isopropyl, ethanol, etc) and I don’t know enough about chemistry to understand whether the alcohols used are actually “solvents” in the usual sense or not.

[SolderOcelot]

Commercial PCB cleaners/flux removers for ultrasonic are aqueous, i.e. not solvent-based. The key ingredient is a saponifier (lye or similar), plus surfactants (probably something closer to what’s in all-purpose spray cleaner like Cilit Bang), plus corrosion inhibitors and probably redeposition inhibitors. These cleaners don’t rely on solvents to dissolve the flux, they use the saponifier to make the flux water-soluble.

(Solvent-based cleaners that can be used in ultrasonic do exist, but they don’t contain water. At least I’ve never seen one that does.)

I'm not a chemist, but the fact that they rely so much on saponifiers is maybe the reason why so many people swear by just distilled water & dish soap?

Also, as a Swiss I'm going to assume you use ultrasonic cleaners daily on Rolex parts 

[jpanhalt]

@tooki
I wonder whether Kester has spies on EEVBlog.

Kester's SDS for aqueous flux removers (e.g., #5768) used to say what was in it (2-butoxyethanol and ethylamine).   The same ingredients are in acrylic floor stippers.  Now Kester is more generic:  https://www.macdermidalpha.com/sites/default/files/sds_pdfs/NA/4060034_KESTER%20SDS%20GHS%20AMERICAS_ENGLISH%20%28US%29.PDF

Look for a household or commercial product with those two ingredients.  They are extremely common, and the cellosolve is generally considered non-toxic.  The recommended final concentration of butoxyethanol ("butyl cellosolve" and other names) is about 10%.  The amine is the "saponifier."  Add a little non-ionic surfactant, if that makes you feel better.  I use Zep brand heavy duty acrylic floor stripper from my local big box store.  I believe it is distributed in Europe, but not in Australia yet.  It's about 30%, so I dilute 1 part to 3 parts total with water.  You could probably use even less of the butoxyethanol.  Then rinse well with water (I give two minutes in the ultrasonic with the final rinse and same time for initial cleaning.)

[SteveThackery]

I use non-aqueous fluid for small steel components like watch parts, in order to avoid any risk of rusting.

[tooki]

Commercial PCB cleaners/flux removers for ultrasonic are aqueous, i.e. not solvent-based. The key ingredient is a saponifier (lye or similar), plus surfactants (probably something closer to what’s in all-purpose spray cleaner like Cilit Bang), plus corrosion inhibitors and probably redeposition inhibitors. These cleaners don’t rely on solvents to dissolve the flux, they use the saponifier to make the flux water-soluble.

…

I'm not a chemist, but the fact that they rely so much on saponifiers is maybe the reason why so many people swear by just distilled water & dish soap?

I don’t think I am getting your point — dish soap isn’t a saponifier. (Saponifiers turn something else into soap. Like using lye to turn vegetable fats into soap.) So you’re not using a soap to remove rosin, you’re turning the rosin into a soap.

And who are these “many people” who rely on dish soap and water?

Also, as a Swiss I'm going to assume you use ultrasonic cleaners daily on Rolex parts 

Well, I live in Switzerland and have citizenship, but I’m not actually Swiss.

I think the last time I wore a wristwatch was in the 90s. As soon as I had a cellphone (1998 IIRC) I stopped. And it was just a Swatch. :p I’ve seen Rolex watches in shop windows though!

[SteveThackery]

I’ve seen Rolex watches in shop windows though!

I think Rolex are seriously over-rated.  People in the know buy Zenith.