Can ultrasonic cleaners kill PCBs or components?

[exxos]

It does seem apparent that by my own experience that oscillators can be killed in my ultrasonic cleaner. But after constant drama with the ST536 project where I seem to get progressively worse for no apparent reason, I started to question I actually causing the problem myself (probably) and overdoing things in the ultrasonic cleaner..

I've been looking around the old Internet and as per usual there is conflicting information. I asked all the regular AI models, and similar conflicting answers. Some say yes definitely and others saying no not really.

After some AI discussion.

Your 30-minute cycle at 70°C exceeds every single one of these thresholds. The case for ultrasonic damage to the XC95288XL on the ST536 is now very well supported by multiple independent sources across five decades of research.

I use 60-70c as thats what the cleaning solutions says use. But I never considered anything breaking before. I just crank up 30mins and leave it to finish. Some boards have been in the cleaner multiple times. So I am wondering if this is why I am getting all sorts of unexplained issues all the time with all sorts.

The alleged evidence seems to be more towards the bond wires inside integrated circuits.

Q-Tech Corporation White Paper (2005)
URL: https://q-tech.com/wp-content/uploads/2 ... eaning.pdf

Ramsey stated that very long immersion times (30+ minutes) in laboratory-type ultrasonic cleaners were necessary to damage the bonds

Hackaday / practitioner evidence
https://hackaday.com/2019/12/05/how-saf ... x-removal/
I have not watched a video but basically they are saying no.. But then someone in the comments says..

Bond wire and bond pad on the die can be weakened within seconds for many semiconductor processes. Gold and copper bond wires have been seen completely separated after less than 5 minutes on lowest setting on a good quality spread-spectrum type cleaner.

Is Your Ultrasonic Cleaner Damaging Your PCBs?
https://www.allpcb.com/allelectrohub/is ... c-cleaning

Bond Wire Breakage: In integrated circuits (ICs), the tiny wires connecting the chip to its package can break under excessive vibration.

The Effects of Ultrasonic Cleaning on Device Degradation
https://www.emerald.com/cw/article-abst ... m=fulltext

Damage can only be induced by use of anomalously longer times or higher power densities. In all cases in which damage has been induced, it is of a purely mechanical nature due to fatigue, and is located on the device bond‐wires and/or the package legs. Cleaning using CFC‐based solvents under standard ultrasonic conditions of power density and time etc. is readily achieved within 2 minutes, even with a minimum stand‐off height.

The Effects of Ultrasonic Cleaning on Device Degradation
https://www.researchgate.net/publicatio ... bond_wires

Cracks or fractures were found at the heel of bond wire after decapsulation and subsequent ultrasonic cleaning. No such cracking was found prior to cleaning. The effect of ultrasonic cleaning on the bond wire was simulated and analysed using finite element analysis (FEA). The results show that the heel crack is likely induced by resonant vibration of the bond wire

Ultrasonic Cleaning on PCBA: A Comprehensive Guide to Benefits and Best Practices
https://wellerpcb.com/ultrasonic-cleaning-pcba-guide/

The intense energy from imploding bubbles can crack ceramic capacitors, damage wire bonds in ICs, and erode vulnerable surfaces. Limit cleaning cycles to short bursts of 2–5 minutes. Longer is not better

Is Your Ultrasonic Cleaner Damaging Your PCBs? A Guide to Safe Practices
https://www.aivon.com/blog/pcb-knowledg ... practices/

Bond wire breakage within integrated circuits is a documented risk. It is advisable to limit cleaning cycles to 5–10 minutes per session.

[rubber_jonnie]

I have read that crystals and oscillators and to a degree some capacitors can be affected by a quick dip in the hot tub.

I only have a small U/S cleaner so tend to drop only the smaller stuff in, but I do try to remove as much as I can.

I suppose that given crystals are thin pieces of quartz it does make sense that they could be damaged by the vibrations in a U/S cleaner.

[alexh]

Interesting ideas but these components are typically designed to survive multiple times G and vibrations many times that of an ultrasonic cleaner for weeks if not months.

For example being picked up and put down. Being dropped. Being in motor vehicles. Being transported tens of thousands of miles in container ships. Being in countries where the temperature and humidity are many times that of the UK day in day out.

I don't believe, but I have no evidence to the contrary, that electrical components are as fragile as your research might otherwise indicate

[stephen_usher]

The ultrasonic bath works by causing cavitation in the liquid. To do that the accelerations involved are very high, but over a short distance.

Also, frequency is important. Stiff materials will have internal resonant frequencies far higher than day-to-day jolts. The high frequencies of ultrasound are more likely to cause microfracturing of substrates such as the resins in PCBs and the cause delaminating between the copper traces and the resin base.

[exxos]

One thing is missed.

If the solution temperature surpasses 60°C, it can induce stress or cause warping in components or the board itself

Thing is, on my cleaner the temp gets hotter when the US is running. If I set it to 60 it can hit 80. I mitigated this by having the lid not fully flush to let some heat escape.

Also spied this talking about failures of JLC.

https://manufacturingreports.com/jlcpcb ... ce-review/

Rough HASL Finish
Solder Mask Overlaying Pads
Damaged Vias
Damaged Footprints
Solder Mask Erodes Easily During Rework
Order Delayed

Indeed JLC seemed to have ghosted me about the damaged vias on one of my H5 and ST536. The solderrmask and print issues I have also seen recently. In fact I think it's what caused frack damage on one board as they so fragile. I don't recall having those issues until more recently.

That site suggests PCBWAY.. Well not for me. Not after they can't seem to solder throughhole parts.

I think I'm not going to use US cleaning anymore. Slightly annoying as I have a barley used US cleaner which is large enough to fit a motherboard in.

I'll order a small air pump and pond airstone to circulate the solution a bit. Problem being there's probably not much room under the basket for one.. So I'll have to think about that one..

[rubber_jonnie]

If the solution temperature surpasses 60°C, it can induce stress or cause warping in components or the board itself

Thing is, on my cleaner the temp gets hotter when the US is running. If I set it to 60 it can hit 80. I mitigated this by having the lid not fully flush to let some heat escape.

I wonder if under some circumstances it could get hot enough to popcorn PCBs? It seem unlikely but who knows?

[exxos]

I wonder if under some circumstances it could get hot enough to popcorn PCBs? It seem unlikely but who knows?

They supposed to be rated at 135c. I think the copper is failing as its so delicate. Vias likely don't like the heats and US cleaning either. But considering some of failed even before hitting the US..

JLC have different options.

1.PNG

2.PNG

3.PNG

[exxos]

Realised I missed off some of the options...

1.PNG

2.PNG

3.PNG

[exxos]

The problem with the PCB angle is that I did not manage to detect any broken traces (though I did not check absolutely everything) or bad vias. But I have had one board which has had a bad via so far, and have not really made up that many boards yet either. Probably only several.

There has been a issue where it seems like repeatedly plugging in and unplugging seems to damage something. Though I already changed the socket on the motherboard. Even putting them in at a angle to increase pressure on contacts does not really help. So it is like something has failed on the board but I have not yet found out what overall.

There has been a couple of boards which only work and putting pressure on a particular corner like where the ROM socket is. or near the regulator etc. I was not able to ascertain much but one board appeared to work after re-flowing the ROM socket.

I think possibly HASL is not a good choice either and the flatter gold plating would be more reliable. As I believe the regulator and one board was not soldered down correctly causing it to not work at all. Indeed some of the HASL is really rough on some of the blank TF536 boards.

But for any future readers, I want to make it clear, that I have not been able to conclusively prove the PCBa have failed, other than one board with a bad via. I think a lot of boards I have because of the repeated plugging in and unplugging could ultimately bend them and damage them over time.

So I do think I may well go for a more solid material with gold finish in future runs and not take the risk. Plus also pay extra for proper testing.. I don't want to be fighting PCB quality issues during development,that has probably contributed to the hassle I have had greatly.


Interestingly PCB have moved away from King board (standard JLC option)

https://www.pcbway.com/blog/News/PCBWay ... a6d1c.html

Dated May 29,2024.

PCBWay announced they were upgrading from Kingboard material to Shengyi material across their 2-layer and 4-layer+ boards, absorbing a 10–30% increase in production costs to do so and providing the upgrade free of charge to customers. Specifically, KB-6165F TG150 was fully replaced with Shengyi S1000H, and KB-6167F TG170 was replaced with Shengyi S1000-2M

But doesn't seem to explain why the change was done.. But obviously they have moved away from Kingboard for a reason...

PCBWay's blog post is frustratingly vague on the specific technical reasons. It says "to improve product quality and meet market demands" but doesn't spell out what was wrong with Kingboard.

Capture.PNG

Capture.PNG

https://www.pcbgogo.com/current-events/ ... rates.html

Interestingly dated Apr 24.2024 , so pretty recent really as well...

Why Choose Shengyi

Selecting Shengyi Technology's materials reflects our dedication to quality and our responsibility to our customers. Shengyi's global ranking and brand advantage, coupled with the widespread application of its products across various industries, attest to the superior performance and reliability of its materials. We believe that by utilizing Shengyi's high-performance materials

Outstanding Advantages of Shengyi Material

- Superior CAF Resistance: Shengyi substrates offer excellent resistance to CAF, significantly enhancing the long-term reliability of circuit boards.
- Lead-Free Compatibility: Adapts to lead-free soldering processes, meeting environmental standards while maintaining high-quality welding.
- High Thermal Resistance: Maintains stability under high temperatures, ensuring circuit board performance under extreme conditions.
- Through-Hole Reliability: Superior through-hole design strengthens the overall performance and durability of the circuit board.
- Mechanical Processing Capabilities: Easy to process, improving production efficiency while maintaining high-quality standards.
- Low Water Absorption Rate: Maintains performance stability in humid environments, extending the service life of the circuit board.

The pattern that emerges
The fabs that made a clean, documented, public switch — PCBWay and PCBGOGO — are the volume prototype fabs most comparable to JLCPCB. The timing of both announcements (mid-2024) and the identical cost figure (10–30% increase absorbed) strongly suggests this was an industry-wide reassessment rather than individual decisions. JLCPCB notably hasn't followed, which means if you order standard FR4 TG135 from JLCPCB today, you are on Kingboard while two of their main competitors have explicitly moved on from it.

TL;DR

Yeah switching to Shengyi boards seems to be happening for some reason. While I am reluctant to use JLC after the trash job they did on my H5 board, they do offer that material. Main problem is I have a lot of parts in my personal inventory which I want to use up first anyway.

I will move to goldplating as well because I think that may well be a problem. My UK PCB guy goes on a bit of a rant about all that stuff.

So while the PCB should be physically better quality and stronger. I'm still reluctant to use them in a US cleaner, as I cannot determine if the ultrasonic cleaning is damaging the chips and / or the PCB. These things are not cheap to keep binning because of failures. So all I can do is reduce all the possible failure points as much as possible.