Skip to main content
Home Forums Capacitor replacement Capacitor replacement
Thread

Capacitor replacement

Capacitor replacement Troubleshooting 28 posts Aug 2, 2007 — Nov 29, 2007
For those who have replaced capacitors on old Macs, did replacing the capacitors solve audio issues?

I have a Classic and an LC II that have audio issues. There are a lot of squealing noises and static but you can still make out the audio. I believe it is capacitors that filter the audio that have failed, probably the 50V capacitors for sound out and in. That is my theory, because I have not replaced the 50V capacitors yet on my Classic but I replaced 5 other ones with little or no audio improvement yet.

Replacing capacitors often solves audio issues. I have an LC that "squeals" from the internal speaker while on, and replacing the capacitor fixed it.

You can fix any piece of electronics if you replace every component on the board [:o)] ]'>

I've not looked at a Classic motherboard, but I figure that an audio circuit that's 20 years old is bound to have transistors or a simple 6 or 8 pin IC that you can replace. (Tip: I solder IC-sockets back onto the board so the same component can be repaired again later without have to use a soldering iron) I'd suspect almost every capacitor in the circuit would ba little tired and I'd probably just replace everything - even resistors, which are the most robust component usually.... but this is a 20 year old Mac. If it didn't work after that I may be forced to break out the multimeter [:)] ]'>

Having fixed more Macs than I can count, I can offer these (admittedly unscientific, biased-sample) statistics:

1) Bad semiconductors are almost never the cause of audio problems in these all in ones. In fact, I am unable to recall a single case.

2) Bad capacitors (and this includes the goop they can leave behind) are almost always the cause of audio problems in these all in ones.

3) The number 2 cause of audio difficulties, and it is a far-lagging second at that, is bad contacts on the headphone jack.

So, if my stats are at all representative of the larger population of bad macs as a whole, then a prudent troubleshooting strategy would start with an examination of the two common sources.

I have never needed my multimeter to fix any audio problem on these compact macs. Chances are that you won't, either.

You can fix any piece of electronics if you replace every component on the board
Take an Xacto knife and slice through a few traces on your logic board. Now replace "every component on the board." Does it now work? No. Why? Because "replacing every component on the board" does not guarantee 100% perfect "like-new" operation. In my own experience, I have found this to be true.

Why do I say this? Because I am a strong advocate of "replacing every electrolytic capacitor" on logic boards, especially the SE/30 board; but you must set out to do such work as a labor of love, keeping in mind there could be other problems that capacitor replacement will not solve. I have one logic board like that. All the caps were replaced and yet I get horizontal lines at boot, no chime, totally dead. But again, this should not faze those of you who need to fix problems with your logic boards. Replace the caps. But just keep in mind that there is a small chance it may not solve every problem on a really beat up board.

As to the question on whether replacing caps will solve audio, the answer is "yes, almost 99% of the time." And you need not replace every cap on the logic board either. Only the caps around the audio section of the board -- but I only state that for the chronic "lazy" among us. My father often told me, "when you do a job, do it right, and finish what you start." And so I give that same prudent wisdom to all of you. Despite the fact that the replacing of only a few audio caps will likely solve an audio problem, you should replace all the electrolytic caps on a given board, because there are likely other problems other than audio you are not yet seeing. And sometimes those problems only appear rarely and they seem not related to capacitors (like a crash now and then you cannot explain).

"But, JDW, I lack the knowledge, experience and tools to replace the caps!" I don't mean to offend many of you by saying this, but such remarks are a cop-out for not doing what you need to do! It's not as difficult as you think. But for those of you who have embarked on this soldering task and failed, then please know there are some on this site (yes, even in your own home country), who would likely do the work for you for a small fee (and of course, you would need to pay shipping both ways). But again, you can save more money by doing it yourself, and I can tell you from experience that it is not THAT hard. Plus, it can be a rewarding, educational experience, and may even be fun for you!

I may be an engineer, but I am by no means a "modder" or "hacker" or "lover of all things solder." I only set out to accomplish "the accomplishable." And because I know I am by no means the best solderer on the planet, I feel confident that "if I can do it, so can you."

Good luck.

Looks like I trashed the Classic II board.

I pulled off solder pads for one capacitor by mistake. Now the audio is completely gone. I had replaced every capacitor near the audio section except for one; if I had replaced the one successfully, audio would probably be working fine now.

The only effect on functionality is that audio is completely dead now, however, since audio was in effect useless before, it is not a huge loss. The Classic II was sold on eBay "as is" and I never stated the audio was working.

And...?

I am completely baffled how you could give up so quickly, in light of the work you already performed. So what if pads were ripped off? Pads come off. It's not a problem. I know because I've pulled them off my board even while taking great care NOT to do so.

You can clearly see to what locations those pads (and traces) lead to on your logic board. So run a couple wires from their origins to both sides of your cap and you're done. What wire? Any wire! It doesn't really matter at these voltages and current.

It really is that simple! Two pieces of wire and four soldered points and you're audio is back in business!

I'm going to look at it tomorrow. I'm not too sure how to attach the wire. It looks like the trace goes into a little hole on the motherboard. I can stick a wire in there, but how do I keep it there? Glue?

I have not given up. I am working on my LC II board now. I have replaced many of the capacitors, and the audio has improved greatly. The only thing that has stopped me is that I ran out of 10uf capacitors. Tomorrow I'll get some more and finish it up, and my LC II will be in perfect condition.

I just got in a Performa 550 with audio problems. It's probably going to need new caps. There's always an SE and SE/30 that need attention too.

Once you get the hang of it, it's not so hard. Getting - and + right is a little confusing, but I have discovered, that on all caps there is always a kind of darker or lighter line on one side, on the cap, and that's always the negative side. I've had no problems with that.

I will say it is a rewarding experience to be able to restore an old motherboard instead of letting it fail even more or trashing it.

Mr. Richardson, can you please explain to us how you broke these pads in the first place in light of the fact you clearly know nothing about "solder"?? And how did you replace those capacitors you say you've replaced? Glue? Indeed not. You used solder. Copper wire is made to be inserted into those "holes" of which you speak. And to affix that wire to those holes permanently, you simply use solder.

I'm quite confused by your last post, so perhaps you should explain to us more what is going on. Because I for the life of me cannot see why you are having trouble unless you don't understand solder. But again, if you don't know what solder is, then how did you replace those caps?!

Regardless, I can only add that any old DMM (multi-meter) with a continuity check feature will be useful to you in that it will "eliminate doubts" about where to wire something. If you have a doubt where one of those dislocated pads originally led, and if you cannot follow the trace with your eyes for some reason, you simply put one problem at the trace edge where the broken pad used to be and put the other probe where you think the trace originates -- in many cases, it could trace to multiple locations (any of which you can connect to, but the way). When you've found the origin, your meter will beep at you to let you know there is continuity between the probes. If your meter doesn't beep, you simply need to probe another spot. And once you've found the right spot, you just solder one end of a wire to that point. It doesn't matter if that point is a thru-hole in the PCB or a pin on an IC. Just solder one end of the wire there. Then solder the other end of the wire to the appropriate leg of your capacitor, making sure you get the polarity right. Do the same for connecting the other wire and you're done!

That's it!

Mr. Richardson, can you please explain to us how you broke these pads in the first place in light of the fact you clearly know nothing about "solder"?? And how did you replace those capacitors you say you've replaced? Glue? Indeed not. You used solder. Copper wire is made to be inserted into those "holes" of which you speak. And to affix that wire to those holes permanently, you simply use solder.
I'm quite confused by your last post, so perhaps you should explain to us more what is going on. Because I for the life of me cannot see why you are having trouble unless you don't understand solder. But again, if you don't know what solder is, then how did you replace those caps?!

Regardless, I can only add that any old DMM (multi-meter) with a continuity check feature will be useful to you in that it will "eliminate doubts" about where to wire something. If you have a doubt where one of those dislocated pads originally led, and if you cannot follow the trace with your eyes for some reason, you simply put one problem at the trace edge where the broken pad used to be and put the other probe where you think the trace originates -- in many cases, it could trace to multiple locations (any of which you can connect to, but the way). When you've found the origin, your meter will beep at you to let you know there is continuity between the probes. If your meter doesn't beep, you simply need to probe another spot. And once you've found the right spot, you just solder one end of a wire to that point. It doesn't matter if that point is a thru-hole in the PCB or a pin on an IC. Just solder one end of the wire there. Then solder the other end of the wire to the appropriate leg of your capacitor, making sure you get the polarity right. Do the same for connecting the other wire and you're done!

That's it!
I knew a few things about soldering, and I know a few more things now.

The first capacitor I replaced took about 30 minutes for me to do, in order to figure everything out and get the process down. I knew basic stuff like, heat up the metal (solder), etc. and I pretty much just figured it out. Now it takes 5 or 10 minutes to do a capacitor. I knew what iron to buy (a 15 watt one with a smaller tip, which is more suitable for these things) and some stuff like that.

Solder does not seem to affix to plastic. There is no metal left on the board for this particular capacitor except for what's inside of the little holes. Any solder I put there falls off because it does not affix to anything.

The pads broke because I removed the capacitor too aggressively. I cannot seem to remove them by simply heating up the pads because most of the pad is under this plastic stuff, so I have been rocking the capacitor back and forth gently until it breaks off, and then I remove the little metal bits it left behind from the little globs of solder.

Five minute epoxy can be used to glue down an improvised replacement pad made out of copper foil or even the flattened end of a piece of bare or tinned copper wire, cut long enough to lap slightly over the remaining trace. Use only the smallest touch of epoxy. After gluing the epoxy will really set almost immediately with the heat from the subsequent tinning/soldering operation, but best not to use tip hotter than 700F. As long as the epoxy has not formed a bead surrounding the repair pad, you can scrape off some solder mask on the trace leading up to the edge of the repair pad and use solder to easily bridge from the repair pad down to the trace surface. If there is an epoxy excess you carve it out of the way with an xacto blade so you can solder bridge the replacement pad to the original trace, and also so the replacement capacitor tab can sit flush on the replacement pad.

I appreciate the detailed explanation. The content of your earlier posts makes a bit more sense now.

At this stage, it may be prudent for you to post some photos for us (on Flickr or a similar free hosting site). That way we can more clearly see what you are seeing and offer detailed advice. I personally would rather guide you based on what I see in a clear photo rather than based on what I read in text form. This is especially true since you are just starting out in the world of soldering.

I'll try and get some good macro shots of the board tomorrow of where it's broken off.

I need to get a LOT more capacitors. I've got a Performa 550 here with bad audio, and you know what that means...

I attempted a repair; I put small bits of solder over the holes, then soldered a wire onto each of the small bits, and then put some hot glue over the whole thing to secure it. I soldered a capacitor to the wires. It did not work. I have removed all of it and now it is back to the way it was in the picture above.

I attempted a repair; I put small bits of solder over the holes, then soldered a wire onto each of the small bits, and then put some hot glue over the whole thing to secure it. I soldered a capacitor to the wires. It did not work. I have removed all of it and now it is back to the way it was in the picture above.
No offense, Mike, but your soldering skills are making me cringe. lol. I think the board is repairable, but your skills and possibly tools and solder type need some evolution before that can happen.

If I were in this boat, I'd be using some fine wire strands (take apart a multistrand wire) to put down into those holes then solder them in there since I do see metal in each one.

~ J

I think it's better to use the types of tools the cellphone repair guys use. These are the ones with minute jets of superheated air, instead of the the regular pencil types (unless it's an ultra pencil type.)

The actual operation of touching the iron to the work and applying solder should only take seconds at most. Looks like you've got a heat problem ;) Yeah the soldering could use a little work - but that only comes with experience (and starting with a dual-sided circuitboard doesn't make it any easier! [:D] ]'> )

For your remaining capacitor I'd just stick it upside down on that burnt patch of circuitboard with some glue, then I'd bend the legs 90 degrees and cut them short adding a couple of thin insulated wires and running them to the appropriate place on the circuit.

OMG, that photo looks like it came from Iraq! (which I may remind the uninitiated is widely considered one of the birthplaces of civilization, but sadly our lunatic leaders have bombed back to the stone age...)

Anyways, don't mean to poke old threads, but I can't believe NOBODY mentioned using FLUX! Hello!? Who does soldering without flux guys? You need to deoxidize the surfaces, no wonder Mr. Richardson's solder doesn't stick.

And one more thing: Tyleress is right about "replacing every component will not make something work". Other than the broken trace, or lifted pad, or damaged via [duct], what can happen which is a *major* bitch to fix is the following: a damaged part (call it A) causes a part downstream from it (call it B) to burn out. So you replace B, but before A, since you don't know it's the source of evil. You try the board and B gets burnt again, because A is still faulty, yet you DON'T know this has happened, since B is "brand new" as far as you know - you just soldered it! Then you continue and replace A. Try your board again and you still find your board not working. Basically you're screwed.

Of course one way to get around this is to replace all components at once, and then try your luck. But unfortunately you must be a God of perfection, because almost no human can achieve the same faultless run as a robotic parts machine.

And one last thing that I didn't see mentioned:

before you replace *all* the caps, see if any leaks exist (easily found by smell or if the solder around the caps isn't shiny, but very dull, or even yellow or green), and if they do you'd better clean them up with PCB cleaner or 99% pure alcohol, because they will corrode the board and traces and eventually no matter how much cap replacement you do, it will never work again!

Cheers and happy surgeries!

before you replace *all* the caps, see if any leaks exist (easily found by smell or if the solder around the caps isn't shiny, but very dull, or even yellow or green), and if they do you'd better clean them up with PCB cleaner or 99% pure alcohol, because they will corrode the board and traces and eventually no matter how much cap replacement you do, it will never work again!
What he said, except the part about it never working again.

This thread has been focused on soldering, so the following wasn't really topical, but it probably needs to be said, in case someone reads this thread without reading some of the other cap replacement threads.

A cap replacement job has three basic steps:

1) Remove the old caps

2) Clean the board. This includes use of solvent and/or detergent to remove all the residue which may have been left by leaking capacitors.

3) Replace the caps with new ones. Tantalum replacements will not leak later.

If you don't clean the board, the electrolyte which has leaked onto the board will continue to corrode any metal parts it touches.

An example of the damage leaking caps can do was my old IIci board. One day around 1995 it just wouldn't power up any more. It turns out that the caps leaked onto the board, But the caps weren't content with no longer doing their job. The leaked fluid corroded away the solder connection to some component pins and more importantly actually ate the solder and copper out of one of the vias (hole) in teh circuit board.

That copper/solder filled hole was meant to connect a trace from the front of the board with the back of the board and now that connection was gone.

So, using my DMM (as mentioned by an earlier poster) I traced the connections forward and back from the damaged point. Once I found accessible points forwards and backwards of the damage, I ran a wire between the two points and soldered it at either end. Voila, board repaired. BTW wire wrap is very good for this kind of thing and available in small spools from Radio Shack.

The damage done by capacitor electrolyte can be very difficult to detect visually. A good 5X or 10X magnifier can be handy. Corroded contacts may look fine until you poke at them with a (e..g) dental probe and find that the metal is no longer solid.

This corrosive effect is why it is soooo important to replace those old caps at the first opportunity. They can destroy your board.

All that said, I haven't replaced all the ones on my old boards yet. There's just never enough time for all the projects. Any day, I'm going to order those two reels of tantalum caps and start replacing... :-)

I fully agree with Trag's advice and can only supplement it by stating what I have said many times before: use the continuity check feature of your DMM to check trace integrity if your board isn't working right after you've changed the caps.

As for da9000's "God of perfection" remark, I can only say that I have replaced all the caps at once on two SE/30 boards without problem, and I do not have a Heavenly abode (not yet anyway). That doesn't mean one should not take care when doing a cap job. Indeed, I have mentioned that before when people say they don't have a tweezer style soldering tip (which I do have). Breaking traces is possible. But it's also not the end of the world. Again, use the continuity check feature of your DMM on the traces that attach to both sides of each cap, and trace them back to their origins. That's what I do to ensure a 100% fault proof cap job. Again, this is not "speculation." I've done two boards as have others on this site. Take care, but don't avoid doing the cap job. Most all SE/30's need to have a cap job done now as most of those SMD electrolytics have leaked (whether you can see the leakage or not).

tweezer style soldering tip
Can you provide a link to what this is? I'm not positive what it is you're using.

Thanks

We have a ESA-safe HAKKO 937 soldering station here at work, used to make prototype electronic devices and for product repairs. The tip I spoke of is the HAKKO 950 Hot Tweezer. Both sides of the tip are changeable and we have a kit of various widths that make it easy to desolder SMD components of almost any size, including the SMD caps used on the SE/30 logic board. While this may be overkill for personal or hobby use, I highly recommend it if you happen to be a bit wealthy 8-o or very serious about soldering! It's not THAT expensive though. Just compare it with the price of an oscilloscope and it will look dirt cheap!

No doubt, some will contend (as many have in the past on this site) that the desoldering job can also be done with two $9 soldering irons placed on either side of the cap. I agree that such is possible. But I myself don't go that route because, well, I have a HAKKO 937 and 950 here at work! And honestly, a professional device makes the job faster, safer and cleaner.

Thanks for the link. Looks very nice. The dual soldering iron method sounds tricky as well. Am I the only person that uses desoldering wicks? Would that be a bad idea when doing a cap replacement?

There are not a lot of "bad ideas" in terms of desoldering because there are many ways to attack the problem. But some means of attacking the problem (capacitor replacement) are better suited to some than others. Some methods require more patience and skill than others. Although I have an EE degree and have years of soldering experience, I still prefer pro tools to make my job easier and ensure the job is professional and quite nearly "perfect."

Just don't scratch your board, which could brake a trace. That could happen if you apply too much pressure to a normal soldering iron (or two, one in each hand), and your hand slips and a sharp iron tip sinks into your board and digs a 4cm trench in it! I've seen this happen!

Don't apply too much heat. With a professional tool you can adjust the heat, but with a $9 iron, you cannot. But if you buy a low wattage iron (around 25W) it may not provide enough heat so you may hold it on the board too long and over heat some parts. There is a lot that can be said about this point, but it's too much for me to write here. Suffice it to say, you need to apply "the right amount" of heat to melt the solder and remove the component.

Solder braids or wicks will help this sometimes, but you can't always pull all the solder away from SMD components this way (I know this firsthand), and sometimes you will get frustrated and yank on the SMD part and brake a pad and/or trace. With my HAKKO tweezer this never happens because I am not interested in cleaning up the solder and then pulling on the part. With the tweezer, I heat the solder to melt it on both sides of the part and then easily pull off the part. It works well, without any desoldering braids or pumps.

Anyway, you need to desolder all your caps (making sure you first have the new caps handly), then clean the board with something like 99.5% pure dehyrated Ethanol alcohol (I advise against "rubbing" alcohol), then solder in your new caps. (Sorry for the Japanese link, but I live in Japan and source my Ethanol here too.)

Now I have long contended, and still do, that you should replace all the caps, including the two axial caps, in one fell swoop. While others may say you can see or smell leaked fluid, I for one cannot. I tried easy method once (replacing only the caps around which I physically could see leaked fluid). That didn't solve my problem. I then replaced all the caps and the problem went away. And since most of these electrolytics contain fluid, and since many have leaked, and since all will leak at some point, why NOT do all of them at once! Because if you only do a few, what if you get a system crash now and then you cannot explain? There will always be the possibility that you didn't replace one or more caps that you should have. So I strongly recommended doing the ENTIRE BOARD, SMD caps and the two axials too.

Lastly, while I am by no means an "expert" on SE/30 repairs, if you wish to read much more on this subject, you can just search this site for "caps" or "capacitors" or "Ethanol" along with "JDW". You can Google for that too, as I am known as JDW on most other classic Mac sites, and I have written many of the same words over and over and over again on many sites. Some threads contain more detail the use of Ethanol as opposed to other cleaning methods, other threads talk more about types of capacitors to use a replacements, etc. Just do a little searching with "JDW" as a keyword and you're bound to find a wealth of information. And no, this is not a plug for me. I simply like to participate in these discussions to see what others have done and to share my own experience to help others avoid the same pitfalls that affected me when I first started out.

@trag:

I stand corrected about the "never again". I'm an optimist, what was I thinking!!! :)) ))

Also thank you for writing out so clearly what electrolyte does to the boards. That's what I was obviously getting at.

BTW, what recommendation would you make for a 5x+ magnifier?

@JDW:

When I mentioned replacing components all at once, I meant *all* motherboard components :p I'd wager a big bill on most people not getting it right the first time when there are 100+ components. Anyways, don't mean to be argumentative, just naming realities and probabilities.

As for your methodologies and using the DMM, I totally agree and practice the same. No sense in not checking your work before firing up the machine. Always double + triple check your work, and like mentioned earlier, it will also avoid exploding (tantalums!) capacitors.

BTW, have you found a similar Ethanol product in the US that you can recommend? If not, don't worry.

@equant:

I use desoldering wicks! Well, not all the time. One of the problems with them is picking the right granularity. If it's too coarse, it won't suck small solder joints and small gobs of solder. Perhaps a worse thing about wick is the fact that it damages the tip of your iron, by scratching the nickel (?) plating, and thus making your iron less useful over time. Lastly, it also runs out, and quick if you're replacing 16x20pin chips :)

Happy desoldering everyone!

have you found a similar Ethanol product in the US that you can recommend?
I've spent the past 14 years in Japan. So I can only say what I use, which is a Japanese product available only here in Japan from a Japanese supplier. No doubt there are places in the US where you can buy it. But I wouldn't know where. Also, I've heard from others that it's "hard to find" in the US, but I don't know what that means. I can only assume it means that some restrictions may be placed on it in the US. But it's easy to find here in Japan, perhaps in part due to the fact that the Japanese don't get their jollies off drinking what should otherwise be a cleaning agent! 8-o

Hahahaha, yeah, I'd rather drink cheap beer! :)

Oh well.

mp.ls