Need To Repair Battery Leak Damage

[Tincanalley]

My Supermicro P5STE board has suffered some damage from a leaking battery. It was in storage for a bit of time and seemed to work just fine when I fired it up, other than loss of settings. It ran fine for an hour or so, then the mouse went crazy. It shot all over the screen and at the same time it was clicking on stuff. The only way to exit Windows was to power it off. When I restarted I tried it again and after a few mouse clicks in Windows, it went crazy. I blamed Windows of course, but decided to try a serial mouse. That worked fine. I then reconnected the PS/2 mouse and enabled mouse support in BIOS that enables it for BIOS setup. When I went back into BIOS setup, I got one mouse move and one click before it went crazy. So I concluded it's the PS/2 port on the board. I put it away for a few months to work on other things and when I finally got back to it, I opened it up and found the battery had leaked. I removed it and clean neutralized it with distilled vinegar. I in no way removed all the corrosion as I was noticing traces that might be damaged and I stopped, unsure what to do. Now that I have a little more soldering under my belt, I want to attempt to fix it.

All other functions of the board work perfectly. I could just clean it better, use a serial mouse and call it a day, but I want to repair it. I'm sure the mouse is acting up because of a damaged trace. There are a couple on a bank of ferrite beads as well as a few spots on traces connected to the PS/2 header. I've attached a couple of pics and can post more if needed. In looking at them, do you think this is repairable, and if so, worth the effort to get the mouse back?

 

[Hugo Holden]

Its pretty obvious FB1 and FB2 need to be removed and likely replaced, you can tell from the green color of copper corrosion and salts there, they will be damaged. Once they are removed, the conformal coating should be removed in the darkened areas over the copper tracks, if found to be defective. then thin link wires (such as those extracted from wire wrap wire) can be added on the top of the tracks.

The pcb needs a water leaching treatment to remove the remaining ions, that I have described on other threads. Don't add any chemicals at all to the pcb in an attempt to "neutralise anything" it will only make matters worse. Then fit the new parts, and all should be ok.

 

[norwestrzh]

You might benefit from watching some of the "Adrian's Digital Basement" videos on YouTube? He has often covered repairs of the effects of battery leakage -- some of it pretty severe.

 

[Phil_G]

Q-Tips dipped in white vinegar have served me well over the decades on worse damage than this

 

[daver2]

It probably is repairable.

You need to use a multimeter (and your eyes) to chase, and check for continuity, from every pin of the PS/2 connector - paying particular attention to the damaged tracks on the outer side of the PCB and any vias between the outer tracks and the inner (hidden) tracks.

Dave

 

[Tincanalley]

Thanks for all the replies. If I grab the FBs with tweezers and use the tip of the iron on each end, will it release? Most of what I've watched involved a hot air tool, but that's not in the cards at this time.

Will a fiberglass brush remove the top of the traces to see below?

Lastly, are all FBs the same, just different sizes? If not, where do I get the information on the ones installed to find replacements?

 

[Tincanalley]

Okay, feel like an idiot. I posted about this years ago and completely forgot. Guess I'm getting old. Anyway, I'm now at the point I feel I can fix it, so I'm going to take from earlier thread and this one and give it a try. I'm going to clean off the leads of the FBs and just do some basic resistance tests. Also go some leads with very sharp tips so I can pierce the traces, find the bad spots and repair from there.

 

[Hugo Holden]

There is all sorts of advice about cleaning circuit boards which have been contaminated by either leaking electrolyte from electrolytic capacitors, or from leaking solutions from batteries, all over the internet and it is all wrong.

While the notion of trying to neutralise either acidic or basic solutions seems appealing, you will never get a perfect acid-base balance. Ionic and electrically conductive species will remain, on pcb's especially in the porous conformal coating, or in the case of phenolic pcb's in the pcb material itself. Never add any ionic solution, such as vinegar, to a pcb, it is a very bad idea.

I'm not sure why this has not been figured out by all the youtube posters. Perhaps they did not have degrees in material science and basic inorganic chemistry combined with a knowledge of electronics. You need all this to understand what to do.

Of course some leakage currents between pcb tracks won't significantly affect low impedance circuitry, especially TTL and, you would never know that ionic species conducting currents even existed on the pcb. However in other circuits, especially Cmos, where the resistances get above about 20k to the 1Meg Ohm region, pcb leakage from these chemical spills can be devastating and completely disable a circuit.

But, what to do about it ? Cleaning a pcb with contact cleaner will not help I'm afraid. Some people in desperation remove the pcb's conformal coating, that can work but it makes a horrible mess of the pcb.

The only way I have found that works properly, after decades of dealing with this problem, is the process of ionic leaching. I have explained this in a number of articles I have written on the topic of dealing with leaked electrolytic capacitors contaminating pcb's, but the same applies to leaked battery electrolyte too. I could explain how to do it again but the stuck record is getting worn out. Maybe I should post a video on youtube, but there are so many idiots posting things there, I don't want to become part of their crowd.

 

[Tincanalley]

Okay, did some tracing. I find continuity between the data pin on the header to pin 92 on the IO chip. However, the CLK pin on the header has no continuity with pin 91 on the IO Chip.

I also found that the data pin on the header runs through FB2. I'm not sure if the clock also runs through one of the FBs due to the bad traces around the FBs. If the clock is supposed to run through one of the FBs, that might be where it is losing connection. Any opinions on jumping clock pin from header directly to the IO chip for testing? Figure I could use a wire, hook up the PS/2 port, connect mouse and fire up into BIOS and see if the mouse works correctly.

 

[Tincanalley]

There is all sorts of advice about cleaning circuit boards which have been contaminated by either leaking electrolyte from electrolytic capacitors, or from leaking solutions from batteries, all over the internet and it is all wrong.

While the notion of trying to neutralise either acidic or basic solutions seems appealing, you will never get a perfect acid-base balance. Ionic and electrically conductive species will remain, on pcb's especially in the porous conformal coating, or in the case of phenolic pcb's in the pcb material itself. Never add any ionic solution, such as vinegar, to a pcb, it is a very bad idea.

I'm not sure why this has not been figured out by all the youtube posters. Perhaps they did not have degrees in material science and basic inorganic chemistry combined with a knowledge of electronics. You need all this to understand what to do.

Of course some leakage currents between pcb tracks won't significantly affect low impedance circuitry, especially TTL and, you would never know that ionic species conducting currents even existed on the pcb. However in other circuits, especially Cmos, where the resistances get above about 20k to the 1Meg Ohm region, pcb leakage from these chemical spills can be devastating and completely disable a circuit.

But, what to do about it ? Cleaning a pcb with contact cleaner will not help I'm afraid. Some people in desperation remove the pcb's conformal coating, that can work but it makes a horrible mess of the pcb.

The only way I have found that works properly, after decades of dealing with this problem, is the process of ionic leaching. I have explained this in a number of articles I have written on the topic of dealing with leaked electrolytic capacitors contaminating pcb's, but the same applies to leaked battery electrolyte too. I could explain how to do it again but the stuck record is getting worn out. Maybe I should post a video on youtube, but there are so many idiots posting things there, I don't want to become part of their crowd.

I would like to see a video. Visuals are always a plus.

 

[Hugo Holden]

I would like to see a video. Visuals are always a plus.

There are so many idiots with no knowledge of mathematics, physics, chemistry (organic & inorganic), electronics, biology, medicine, mechanical engineering, posting on youtube and making videos, that I'm afraid I will not be a part of it. Forget youtube videos, get some textbooks and do some reading.

However, if you look at this article page 18 or thereabouts you will see how the water leaching method was used to help repair a 2465B scope which had pcb contamination from leaked electrolyte:

https://worldphaco.com/uploads/2465BPSU.pdf

 

[Phil_G]

Perhaps they did not have degrees in material science and basic inorganic chemistry combined with a knowledge of electronics.

Such insulting arrogance. You're right to point out my degree isnt in material science. I do have a lifetime in electronics and taught electronics and microprocessor based systems at Sheffield Polytechnic. What I said was that "Q-Tips dipped in white vinegar have served me well over the decades on worse damage than this" which was a suggestion based on my own personal experience, offered as an electronics professional not as an idiot.

 

[Hugo Holden]

Such insulting arrogance. You're right to point out my degree isnt in material science. I do have a lifetime in electronics and taught electronics and microprocessor based systems at Sheffield Polytechnic. What I said was that "Q-Tips dipped in white vinegar have served me well over the decades on worse damage than this" which was a suggestion based on my own personal experience, offered as an electronics professional not as an idiot.

Not sure where you got the idea that I was referring to you, personally in any way at all. I was referring to the nonsence I see on youtube .And I was replying to @Tincanalley's post above too.

I think you got the wrong end of the stick.

I don't know you, your background or your capabilities (how could I point out what degrees you did or did not have, I'm not Psychic), yours may well exceed my own in many areas for all I know, I don't know, how on earth could I insult you ? I have no idea.

I think you somehow assumed an insult for yourself. Sometimes people do this when their methods are questioned, but it pays to keep an open mind. Getting angry never helps science, or electronic repairs either.

In any case if you feel insulted by my assertions on how to deal with specific technical problems, I apologise.

We could throw a lighthearted note on it. There was a very funny scene on the movie Starship Troopers, where the idea that the Bugs on the Planet Klandathu were smarter than Humans thought they were , and that they might be able to think and make plans. They interviewed an Entomologist on TV and asked him about it, he replied something to the effect: "Frankly I find the idea that a Bug can think is offensive"

 

[andy]

The first thing you need to do is scrape away the green solder resist from all the blackened areas since electrolyte has gotten inside these areas. This will better allow the contamination to be removed, and allow you to see if the copper traces have been eaten away. Next you should clean the area using the method of your choice. I personally use an ultrasonic cleaner, but other methods are also valid. The important thing is that every trace of electrolyte is removed from the board.

If you find a gap in a trace, repair it by soldering some fine wire across the gap (after thoroughly cleaning the trace so it will accept solder). The most complicated problem you may run into are the vias. The vias on this board appear to be filled which makes repairing them more difficult. Internal layers may also be involved, which makes things even harder. Hopefully none of the vias are open, but you won't know until you scrape away the black areas.

Only after repairing all the obvious damage do I start more detailed trouble shooting. If you still have problems, then you can try to visually trace the signals from the mouse port, and hopefully you'll run into the break before it gets to a hidden internal layer.

 

[daver2]

I don't think anyone has stated it yet, but I think the FB component is a surface mount ferrite bead.

This is a piece of wire (so should appear as a very low DC resistance between the two 'pins' surrounded by a ferrite block of material).

The component acts to block the propagation of Radio Frequency Interference from being received (or transmitted) down the mouse cable.

Dave

 

[Tincanalley]

I have small kit of dental tools, don't know how, though, and never tossed it. In it is one of the angled scrapers with a very sharp point. Is this suitable to remove the resin on the darkened traces? Just some light pressure and a scraping motion?

I believe the FBs will be fine with better cleaning. I orders some of the dental flossers that are basically mini pipe cleaners. I feel they should be able to clean away the corrosion under them. I don't know what runs through FB1, 3 and 4 yet. FB2 is definitely Vdd from the controller chip and it is getting to the port header. Ground and data are good as well. Maybe the keyboard is using some of the other FBs.

Is it possible for the mouse to "work" without a clock signal? Would the lack of one cause it to be erratic or not work at all? The last time I tried the port, it work until the first click, then went crazy.

 

[HaHa I]

Q-Tips dipped in white vinegar have served me well over the decades on worse damage than this

I second this, wile its not a end all be all its a wonderful first step in fixing a problem like this. I would ignore anyone that says vinegar will not fix or help it as it clearly does and there is a mountain of evidence to back it up as a good way to clean a board like this. I my self used to buy and fix boards like this all day long back when running a eBay store was still worth wile for a small fry.

This board is not to bad off and some cleaning and bridging a few traces might be all you need to do if you just want get it back up and going. Now if you want to do it right there would be more steps involved. I would also like to say that andy posted a lot of good info on fixing the board too.

Is it possible for the mouse to "work" without a clock signal? Would the lack of one cause it to be erratic or not work at all? The last time I tried the port, it work until the first click, then went crazy.

You likely would need it for it to work.

 

[Tincanalley]

I second this, wile its not a end all be all its a wonderful first step in fixing a problem like this. I would ignore anyone that says vinegar will not fix or help it as it clearly does and there is a mountain of evidence to back it up as a good way to clean a board like this. I my self used to buy and fix boards like this all day long back when running a eBay store was still worth wile for a small fry.

This board is not to bad off and some cleaning and bridging a few traces might be all you need to do if you just want get it back up and going. Now if you want to do it right there would be more steps involved. I would also like to say that andy posted a lot of good info on fixing the board too.


You likely would need it for it to work.

I'm going to clean up the FBs, check the trace for the clock signal and hopefully fix it. If it doesn't work, it's not the end of the world as all other functions on this board work perfectly and I can deal with a serial mouse.

 

[Tincanalley]

Well this is of no help. This is the "official" Supermicro PS/2 mouse header pinout. Not quite right as ground is pin 5 in my tracing and goes to the controller chip pin 90. Hell, according to this, there are only 3 connected pins. Not sure how you get Vss, Vdd, CLK and Data with only three pins. How could the manufacturer get it so wrong?

 

[HaHa I]

I'm going to clean up the FBs, check the trace for the clock signal and hopefully fix it. If it doesn't work, it's not the end of the world as all other functions on this board work perfectly and I can deal with a serial mouse.

might be worth finding someone locally who fixes stuff like this to help. could be a good learning experience.