Anyone cleaning the flux off using an ultrasonic cleaner?


  • Hero Member

    I saw a youtube video where the guy was cleaning the flux off his boards using an inexpensive ultrasonic cleaner and denatured alcohol. Seemed to work for him. Amazon price for an ultrasonic cleaner is around $30.

    I've seen others where they just use water in their ultrasonic cleaner and then rinse off with something else to displace the water.

    Heating the solution to 50 or 60C is alleged to help, but apparently cleaning at room temperature will also work. Not sure. Anyone here doing it have an opinion?



  • @NeverDie I just use isopropyl alcohol and a paint brush! Cheap and effective.



  • I own a regular household ultrasonic cleaner. 50W, 46 kHz, without manual heating control. I've tried cleaning a PCB once with distilled water, but it wasn't very promising. Not all of the flux came off and the board was still sticky after blowing it dry with compressed air. Then again, I'm using rather tacky homemade flux made of rosin, isopropyl and a bit of glycerin.

    Anyway, I wouldn't want to use large amounts of alcohol or similar volatile solvents in a semi-open device, that also heats up when in use. That just sounds like a fire hazard waiting to happen, since it evaporates rather quickly and the flash point is below room temperature (e.g. isopropyl is somewhere between 10 - 15°C). With enough vapours in the air, a spark (static discharge?) near the cleaner might be sufficient for a neat little explosion, in a worst case scenario. Industrial ultrasonic cleaners might have some kind of safety features to reduce the risk of explosions and / or fires, I guess. I'd rather try a less dangerous, commercial ultrasonic cleaning solution for electronics, but I have no suggestions for that.

    Since I'm not dealing with large amounts of PCBs that need to be cleaned, I stick with isopropyl and a tooth brush, like skywatch.


  • Hero Member

    @BearWithBeard said in Anyone cleaning the flux off using an ultrasonic cleaner?:

    isopropyl and a tooth brush,

    Exactly what I've been using to date, and then blowing it dry with compressed air afterward. It works, but for some reason I hate doing it, so I was looking into whether there's some easier way.



  • I get what you mean. It's dumb labour, messy, splatters everywhere.

    I guess, since isopropyl alcohol is a good flux solvent on its own, pouring some in a small sealable container, in which you can then barely submerge a PCB for a few minutes, should work aswell. Maybe even agitating it a little bit. Similar to how people etch PCBs in ferric chloride.


  • Hero Member

    @BearWithBeard Many if not most of the ultrasonic cleaners have lids on them, so that puts a limit to how much solvent evaporates (due to vapor pressure). Thinking about it now, I'd wadger the explosive potential is quite limited. For sure you'd want a lid that doesn't lock tightly closed, so if it were to blow-up, it would just blow the lid off/open, not create shrapnel. I suppose a hinged lid would be best so it doesn't go flying. For that reason, I think the risk is probably low.

    Denatured alcohol is cheaper than iso, so presumably it wouldn't cost much to submerge a pcb in it. I can't say that I've tried it though. Any reason it wouldn't work just as well as iso?



  • @BearWithBeard I do the 2 container method for cleaning pcbs with isopropyl. First container is the "dirty" one which gets the pcbs first and they soak for a couple minutes. Then each pcb gets the electric toothbrush treatment above the dirty container and then they go for a rinse in the "clean" container. Once the clean container starts to soil it becomes the dirty container and the old dirty container is cleaned and fresh isopropyl is added. Just like washing dishes. I actually worked for a place that used a standard dishwasher to clean (large) pcbs. It worked surprisingly well.


  • Hero Member

    I guess there's something wrong with denatured alcohol. Otherwise, I'm sure everyone would be using it instead.

    Buying 99% ISA in 12 bottle quantities from amazon would save over the one at a time approach, so maybe I'll go that route.



  • @NeverDie Well, I have safety concerns because of the relatively large amounts of liquid that are needed. IIRC, my ultrasonic cleaner requires to be at least half-full, which is a little more than 250ml, or one bottle of isopropanol. And after the 5 minute bath, what do you do with it? Dumping it seems wasteful, transfering it back into a bottle is inconvenient and again, potentially dangerous, depending on the environment. Leaving it in the cleaner for the next use? Nah, I'm not risking it. It might be fine 99% of the time you use it, but you also wouldn't make a narrow cut on a table saw without a push stick, would you?

    What I think could be a good compromise, is filling the ultrasonic cleaner with water and place the PCB in a zip bag filled with enough solvent to submerge it, squeezing the air out and then place that bag in the cleaner. Like sous-vide cooking. It provides isolation and the solvent, of which is much less required, can be handled separately in a save location.

    Oh, and regarding isopropanol vs ethanol: Here in Germany, I can get a liter of 99% isopropanol for about 3 EUR. That's not really much more than 96% denatured ethanol. With ethanol however, I found that it tends to leave some sort of film on the PCB, which has a kind of dull and blochy appearance. Not the uniform, shiny finish you would like to see after cleaning. Also, while making my own flux, I found that pure rosin dissolved better in isopropanol than in ethanol. And while some of the rosin in ethanol started to crystallize after a few weeks, the isopropanol-based flux is still completely liquid after over a year.



  • to reduce evaporation i saw a guy doing the following method. he fills the container with regular water up to middle mark and then gets some fridge ziplock bags fills them with cleaning solution or just IPA puts the pcbs in the bag locks them and tosses bag into the water filled ultrasonic cleaner. when it finished just takes out the pcbs and locks the bags again for later use until the solution or alcohol is not usable. sounds very practical but i dont have experience with it.



  • @NeverDie Here in Canada we can buy 99% ISA from our local pharmacy for a couple bucks a bottle so that's why I normally go that route. Question, have you tried the actually flux cleaner like this before?


  • Hero Member

    @Jon-Raymond No, I haven't tried that one. Looks pretty toxic, though I imagine it works a treat. Is that what you use? I've only used IPA with a toothbrush. I started off using 99% IPA from aerosol cans but later migrated to the pharmacy stuff because of cost.

    Here pharmacies sell 91% and 70% IPA because they have more disinfection power than the 99% (due to longer dwell time).



  • @NeverDie Before when I used SN63/PB37 Rosin Core I needed to mix a solution of IPA and Acetone to clean the flux residue off. IPA alone wouldn't do it. I have since switched to using SN63/PB37 "No Clean" solder for all through hole parts and have found using only IPA does a good job of cleaning.

    You're right about the toxicity of the commercial (and home made products). Whenever possible I do all my board cleaning under a fume hood that I also use when reflowing boards. It's a simple enclosure with an exhaust fan that vents outside.


  • Hero Member

    @Jon-Raymond Acetone has a tendency to dissolve many plastics, so I'm not sure how compatible it is with IC packaging and such. Unless maybe they're made from polypropylene? I just don't know.

    I've read that brake cleaner (the kind containing Tetrachloroethylene ) can be quite effective as a flux remover. It's also cheap, which is why some people use it. However, without a hood such as you have, it can be quite bad for you biologically. That's why I thought perhaps using an ultrasonic cleaner with less toxic (but comparatively less effective) flux cleaners is better for my environment.



  • @NeverDie IC encapsulation is almost always a thermoset polymer as they need to withstand the reflow process. The added benefit of this is that they are not dissoluble by acetone. I wish they were as it would make de-encapsulating IC so much easier. I generally have to use an acid to de-encapsulate. I haven't tried brake clean but would think it would work well.



  • I gave the method with the isopropanol-filled zip bag a try. I soldered the two PCBs in the photo below by hand and have been intentionally messy. I applied lots of rosin based flux, although it wasn't needed, except for maybe the FFC connector. The white solder mask certainly emphasizes every bit of flux.

    mys_ultrasonic_before.jpg

    I put them side by side in a small zip bag with approximately 50ml of isopropanol, squeezed the air out and let the ultrasonic cleaner (filled with water) do its thing for 2 minutes. Here's how the PCBs looked right out of the bath - just gave them a splash of clean isopropanol. No further cleaning.

    mys_ultrasonic_after.jpg

    The board on the left has a little bit of flux residue near the TQFP, in a place where the flux was quite dark and crusty, but the board on the right looks super clean. The isopropanol in the zip bag might be good for another few uses before it has to be replaced. It is already slightly discoloured from the flux.

    Anyway, besides the one spot of leftover flux, they look fantastic. Especially the connector cleaned well. And the best of all, for the first time ever my hands didn't feel like I dipped them in a big pot of sticky honey after I cleaned a board! 😁

    (Yes, I forgot to solder one side of C10 on the right PCB. Shame on me! I fixed it already.)


    A cheap ultrasonic cleaner might not be a good choice though...

    Apart from this, I also took some time to read about ultrasonic cleaning. Upfront I want to say, that I'll not continue to use my ultrasonic cleaner for this task in the future despite the pleasing results - this was just a test to see how good a simple household-grade cleaner of the 30 - 50 EUR / USD price range does the job.

    You may know that ultrasonic cleaners rely on cavitation action to clean. Basically, they create lots of microscopic bubbles which rip the contaminants off the parts when they implode. The frequency which is generated by the transducer(s) determines how big those bubbles are. Lower frequencies cause larger bubbles, which are helpful to clean gross contaminants, like grease and grime from mechanical parts (usually around 25 - 30 kHz). Higher frequencies are less harmful, due to smaller cavitation bubbles and can be used for more delicate cleaning jobs. I read that 40 - 60 kHz are generally prefered for PCBs.

    Professional ultrasonc cleaners use frequency sweeping to continuously vary the generated frequency by a few kHz to avoid standing waves in the container, where the cavitation action would be far more vigorous and therefore more harmful to the parts in specific areas. Frequency sweeping seems to be crucial for consistent results with a very low to non-existent risk of damage. Cheap cleaner, like mine, usually don't have this feature and are therefore more likely to damage components. You can test how uniform your cleaner works by placing a sheet of regular thin aliminium foil (10 - 30µm thickness) in a basket, start the cleaning procedure (1min max) and inspect it. It should ideally have a consistant pattern of tiny, roughly equally sized perforations. Here's a photo with three separate passes at 60 seconds each of my cleaner. Looks... totally uniform...?! (Not really)

    mys_ultrasonic_test.jpg

    I also read that a power density below 11 W/liter is considered to be quite save to use. For comparison: my cleaner has a nominal power rating of 50W and a small 0,6l tank. Then there are manufacturers who advice to be cautious when using ultrasonic cleaner, for example MLCCs could crack (TDK, Murata).

    Now I'm not dealing with super expensive parts, so it's not that I'm worried to replace a 50 cents component if I get very unlucky, but debugging an invisible / internal fault that is potentially caused during the cleaning process could be a very frustrating task, I reckon.

    I think it wouldn't be a bad idea to invest in a higher quality model to clean PCBs. I'll stick with a toothbrush and cotton swaps.


  • Hardware Contributor

    @Jon-Raymond said in Anyone cleaning the flux off using an ultrasonic cleaner?:

    @NeverDie Here in Canada we can buy 99% ISA from our local pharmacy for a couple bucks a bottle so that's why I normally go that route. Question, have you tried the actually flux cleaner like this before?

    Haha off-topic but the French translations on this can look like they've been outsourced in China. Smells like Google Translate !


  • Hero Member

    @BearWithBeard Wow! Thank you. That is the best review I've ever read on this topic!

    As a side note, I've lately been working on some ultra low energy circuits (ones meant to draw less than 30 nanoamps of current), and that exposes a whole new world of awareness about leakage currents, which at that level are everywhere. Part of what I've learned is that keeping moisture out of the FR4 is absolutely essential, and so the last thing I'd ever want to do now is clean a PCB meant for such a circuit in a cheap IPA that has lots of water in it, let alone soak it or do anything that might drive water into it. I mention it now because the trendline is toward lower and lower power, and even if it doesn't affect you today, it probably won't be long before it does.

    So, what am I using now? Contact cleaner: https://www.amazon.com/CRC-05103-Electronic-Cleaner-11/dp/B000BXOGNI/ref=sr_1_4?dchild=1&keywords=contact+cleaner&qid=1585329683&sr=8-4 It's relatively cheap and it dries almost instantly. It also sprays out forcefully, so you can kind of blast the flux off, not just soak it off.



  • We always "bake" PCB to get humidity out, and then afterwards those are stored in special non-humid cabinets until we take them out for SMD manufacturing

    typically baked for two hours at 105 – 120 °C
    http://www.surfacemountprocess.com/uploads/5/4/1/9/54196839/ipc-1601a.pdf


  • Hero Member

    @bjacobse That makes perfect sense, but how are you removing the flux after the SMD manufacturing? Or, are you also heat soaking them after cleaning them?

    I've also read about certain conformal coatings being used, especially to slow down or maybe (?) even prevent moisture intrusion after the flux is cleaned off. Advanced Linear Devices appears to use it on their zero threshold voltage mosfet demo boards, which as you can imagine are highly sensitive to leakage currents.



  • Those few IC that we dip in flux, is because they have solderballs on the IC from supplier, the remaining components will have solderpaste on each solderpad in PCB.
    We don't "remove" flux, we just forward the PCB on conveyor belts into heating oven (REHM reflow solder oven, with Nitrogen gas, to get better solderability as there will be less oxygen interefering during soldering in oven)

    https://en.wikipedia.org/wiki/Reflow_soldering



  • @NeverDie
    We don't use conformal coating. But we use hydrofob coating of the whole device, so w avoid water ingress through the plastic box


  • Hero Member

    @bjacobse Excellent! You're the first and only person I've ever met who knows anything about this very important topic.

    By "hydrofob", I assume you mean hydrophobic coating? If so, just how effective are they at preventing moisture intrusion, and which type or brand of hydrophobic coating is the best?



  • Thank you for the nice compliments, but I'm not an expert, I just know something, and listen to our technicians at work. I work for a hearing aid company, in our PCB manufacturing plant

    Yes hydrophobic coating, the machine is quite expensive, so a now retired technican made the machine himself, (He is also a quite bright guy), anyway a similar machine can be bought from Europlasma https://www.europlasma.net/products.html
    as I understand the gas (teflon I think is used as gas) is sprayed inside a vacuum chamber, and then the hearing aids with electronics are rotating (Like a slow tumblking washingmachine),
    and then after a while not sure but about 20-25min the process is finished.


Log in to reply
 

Suggested Topics

  • 175
  • 33
  • 25
  • 3
  • 48
  • 930
  • 1
  • 3

248
Online

9.2k
Users

9.9k
Topics

103.3k
Posts