How Safe Is That Ultrasonic Bath For Flux Removal?

How do you clean the residual flux off your boards? There are plenty of ways to go about the job, ranging from “why bother?” to the careful application of isopropyl alcohol to every joint with a cotton swab. It seems like more and more people are turning to ultrasonic cleaners to get the job done, though, and for good reason: just dunk your board and walk away while cavitation does the work for you.

But just how safe is it to sonically blast the flux off your boards? [SDG Electronics] wanted to know, so he ran some cleaning tests to get to the bottom of things. On the face of it, dunking a PCB in an aqueous cleaning solution seems ill-advised; after all, water and electricity famously don’t mix. But assuming all the nooks and crannies of a board can be dried out before power is applied, the cleaning solution itself should be of little concern. The main beef with ultrasonic cleaning seems to be with the acoustic energy coupling with mechanical systems on boards, such as crystal oscillators or micro-electrical-mechanical systems (MEMS) components, such as accelerometers or microphones. Such components could resonate with the ultrasonic waves and be blasted to bits internally.

To test this, [SDG Electronics] built a board with various potentially vulnerable components, including the popular 32.768-kHz crystal, cut for a frequency quite close to the cleaner’s fundamental. The video below goes into some detail on the before-and-after tests, but the short story is that nothing untoward happened to any of the test circuits. Granted, no components with openings as you might find on some MEMS microphones were tested, so be careful. After all, we know that ultrasound can deal damage, and if it can levitate tiny styrofoam balls, it might just do your circuit in.

29 thoughts on “How Safe Is That Ultrasonic Bath For Flux Removal?

  1. Speaking of ‘Danger’:

    The head of the local Makerspace has two different sized (small) sonic cleaners from ‘Chairman Mao’s Dollarama’, (as one of our irreverent YouTubers puts it)…

    *He’s* concerned about potential hearing loss, which leads him to retreating from the room to another part of the building when it is on, and advising others about the potential hazard of very high frequency audio.

    So is *that* a thing? For health and safety purposes, should we be taking these little Harbor Freight units out onto the front porch and watching them from inside with shooter’s muffs on?

    1. I wonder about this too. Even though you cant hear it, it’s still blasting you little tympanic coupled hairs :-).
      For Die Bonders as well as these cool ultrasonic cleaner kits you can get on Ali-express. I have a couple of transducers to play with for cutting and bonding using a hand-piece.

    2. 1st of all, there are special solutions for cleaning circuitry with an ultrasonic, made by Elma I believe. Theres actually a lot of specialty solutions for ultrasonics if you know where to look (try proper jewelers and watchmakers supply houses like Jules Borel and Otto Frei), and I do believe there are commercial operations cleaning electrical equipment with these.

      2nd- normal ultrasonics are nowhere near as high frequency on the transducers as stuff like my Elma S40H, which has 3 36k transducers. commercial units are waaay more powerful- and I have never noticed any hearing loss after years of using them.

      I think that fear is imaginary.

        1. I used to sit near the ultrasonic cleaner at one of my college summer jobs, and I can still hear the gentle chirping of summer construction jackhammers. Also those stealth airliners they use nowadays don’t fool me, I can hear them coming hundreds of feet off. ;-)

    3. I don’t think there is a mortal danger because of ultrasonic waves.
      But then people including me can hear frequencies that others can’t:
      There are cars with marten repellent systems that cause me a headache because of the sound they emit.

      So your guy might experience something like this.

    4. From memory the recommended limit is 105dB at 30kHz, mainly to avoid damage from sub-harmonics. Ultrasound rapidly attenuates in air, a small cleaner’s transducer is probably under 50W, and the tank/air interface is hardly ideal for transmission so I would be astonished if they posed a hearing risk unless you stuck your head in the tank, but you could get a sound meter and check. I suspect your colleague is erroneously conflating high frequency with high volume.

  2. I wouldn’t worry about blowing out the 32.xxx Khz crystal… I would worry about managing to mechanically excite it such that it socked out voltages that could fry sensitive CMOS chips in circuit. I’m not sure I see anything on the board that would be that fragile to confirm or deny such in this test.

    1. We had those watch crystals drift, I believe that it was caused by physical stress but now that you mention it, it could have very well been excitation playing hell on the chips input, loading it down and causing the drift

      1. The cheapy ones aren’t meant to be all that thermally stable, go a bit faster and slower over “extended” temperature ranges. As a desk clock in an air conditioned room they’d stay spot on, but sitting on your wrist in summer sun and winter cold seeing -10 to +40C in the case they’ll gain and lose.

  3. My old job we had a dishwasher to clean the boards. We used organic flux. And it would grow stuff if not cleaned. Always thought it strange we used a dishwasher to clean the boards

    1. We used dish washers to clean Vaseline off boards used in chyro environments
      I know that water and electronics mix fine as long as you dry them off before power up but it still felt wrong

    2. We used dishwasher, but 7t was kind of warranted, for we used sunflower (cooking) oil as flux in wave soldering line.

      Ah, the aroma of soldering a batch of cards in the morning, smells like popcorn!

    3. At work I also stuff boards into a dishwasher, then dry them off with compressed air….

      Seems like the worst combo ever, water + electronics for 2-3 hours, then followed by fast moving dry air.
      (Fast moving dry air creates charge buildup on a lot of surfaces, a big reason why pneumatic stuff should be well grounded if working with explosive gases.)

      But a finished board isn’t really that high resistance, so the little charge the manages to build is dissipating fast enough to not be a danger.

      And the warm water bath washing over the boards for 2-3 hours isn’t long enough for it to be corrosive.

      Once forgot the boards in the dishwasher over the weekend though…. But they turned out fine without any signs of corrosion.

      In the end, ultrasonic cleaners aren’t really that needed to clean boards, compared to a good old dish washer.

  4. I used to use some nasty flux cleaner in school, but I don’t think you can buy the same formulation anymore. I also used lead solder which has rosin that was easier to dissolve than then new stuff, so maybe my old chemicals weren’t really super strong, just super dangerous. ;)

  5. First hand experience: Piezo buzzers (in this case small SMD buzzers) generated voltages from the ultrasonic excitation. The micro on the PCB had some pins fried because of it. This was during a first production run, so in this case the solution was to use a different cleaning method, but the point is it can certainly cause problems. And the reason can be _very_ difficult to find ;)

  6. You should not be cleaning anything that contains any type of silicon with these ultrasonic cleaners. You will weaken the bond between bond wire and bond pad on the die within seconds for many semiconductor processes. I’ve seen both gold and copper bond wires completely separated after less than 5 minutes on lowest setting on a good quality “spread spectrum” type cleaner. Don’t do it. Learn to solder without making a mess and use the least agressive cleaner you can get away with, and make sure it is compatible with your substrate and components. You shouldn’t be using any cleaner on things like plated contacts, those should be soldered after cleaning messes on a PCB and you should avoid getting anything on these contacts at all costs!

  7. Learned this the hard way back in 2015 when a supplier’s “board cleaning service” drifted standard crystal “watch” oscillators so bad they were unusable
    That was far from the only problem too, the bath left a very thin film behind, loosened one 1206 cap and left alcohol inside of a screw terminal.

    They refunded the run cost, changed methods and we never had a problem since but we were very confused on how the watch crystals drifted until they announced the change and indicated it was ultrasonic, we assume the transducer operated at around 32khz and that’s what caused it

  8. When I was at university, I worked summers at a local firm assembling PCBs and rack instruments for the North Sea Oil boom. To clean off the boards, we were sent out onto the fire escape, with a gallon can of Carbon Tet and a brush, and were not allowed back in until the board was spotless and had dried. Ultrasound? pah.

  9. The “spread spectrum” cleaners that are supposed to solve this issue are just as bad. Ultrasonic is the wrong way to clean PCB’s. I assume someone in the marketing department at an ultrasonic cleaner manufacturer started this nonsense with a well-meaning sales pitch and chaos has ensued.

    1. I have to agree that ultrasonic cleaners aren’t really all that good for cleaning PCBs.

      Cleaning rusty bolts and such, it is fine at.

      But a board full of electronics has all sorts of components that simply aren’t built to survive such a treatment.

      And in the industrial scene, ultrasonic cleaning isn’t used for populated PCBs. The good old dishwasher is actually for more common. (And the even larger machines with conveyor belts are practically dishwashers too…)

      Though, regular dish soap isn’t all that good of a fit, sometimes warm water is good enough on its own. Sometimes adding a bit of Isopropanol is a good solution, and other times one can need something else entirely.

  10. Cleaning keyboards in the dishwasher has been a common practice for a long time (and as already mentioned, they should be given enough time to dry up completely before being used again).

    1. I used to do it in the shower when I was a broke student. Bought pallets of old video terminals, cleaned them up, mated them with a cheap modem and re-sold. It was a good gig while it lasted.

  11. I have had problems with large ceramic caps cracking in my Horrible Freight sonic cleaner. The small ones seem OK, but I have never done electrical testing on them after cleaning. If you look at the data sheet for ceramic caps, some of them vaguely indicate there may be a problem with sonic cleaners.

    At this point, Isoproply Alcohol and a toothbrush are my goto method for flux cleaning.

  12. Dishwasher or no-clean flux. I know people who clean reloading brass in ultrasonic baths, and jewelers, and equipment with tiny orifices places that can’t be easily reached. Everything I see on PCBs these days can be washed. I would consider it for boards made of special materials to be cleaned for high vacuum, optics, medical, or ?

Leave a Reply to Ostracus Cancel reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.