Tolerances for capacitor replacement: electronics query

[beachycove]

Thanks to all once again.

It will be a couple of weeks at least before I can tackle this, and it might even have to wait until Christmas, since I first need to find the components, and then find the time to do the deed, but I'll post the results when I have completed it.

 

[Mac128]

Ah, I had forgotten we were talking about the SE/30, one of the first compact mostly surface-mount boards. I tend to view it in relation to my old 68 Mustang (appropriate don't you think): I could repair almost anything on it without hesitation. My 90's Jeep was on the boarder, some things I could fix others I wouldn't touch. But when it comes to my late model car, I really want nothing to do with what's under that hood, save refilling the washer reservoir.

Glad someone has experience with those pesky SMDs! Those caps certainly cannot be clipped, but a brief visual inspection confirms that. Truth is, even on the early Mac boards, only the axial mounted caps can be clipped, most of the radial are mounted tight against the board (some with spacers) and additionally glued, making them even difficult to pry up enough to slip cutters under them. But again, mine was a general comment to call attention to the obvious which some of us overlook. I was de-soldering an axial cap once, when I though: "why am I doing this? I can just clip it off".

FYI, radial and axial caps are interchangeable (leads permitting). Someone more knowledgeable than I may know if this is preferred, or the rational for using one or the other. As for myself, replacing radial with axial caps would make replacing them in the future much easier, though for those caps affected by operating frequencies vibrations, harder to dampen thus risking more solder cracks and subsequent failures.

Now, at the risk of sounding like a broken record on this forum or being unduly "matronly" or harsh ... people really should do a search before posting. I was looking for some unrelated info last night and stumbled across this:

http://68kmla.org/forums/viewtopic.php?t=677

Actually I'm surprised JDW didn't link to it in this thread.

Good luck!

 

[TylerEss]

I've only skimmed the thread, but everything I saw seemed to be good advice to me.

I think the caps are safe enough to pull off with just one soldering iron, no tweezer-tip required: usually the leads are flexible enough and have enough thermal mass that you can get 'em up without harming the pads. A tweezer-tip would be preferred, but I've done at least 6 SE/30s without one.

Make sure to clean the board really well (just water and brush is probably fine) and use lots of flux on the old joints before you try to remove them. I like to wet them with some new, fresh solder, too, to help them melt a little quicker.

If one managed to pull all the old caps off without lifiting a single pad, or if they had some of those stick-on replacement pads, the tantalum chip SMD capacitors makes a very durable and pretty (stock-looking) replacement. Personally, I use radial lead caps and bend the leads into little legs with gull-wings on the ends. I let 'em stick up about 3mm from the circuit board, to give room to get dikes in and clip 'em off to ease future repairs.

I've always used aluminum electrolytics on my own SE/30s but if the tantalum caps give you more of a feeling of security, go for it. They cost about twenty times as much, but they should be more durable.

 

[trag]

Just use two soldering pencils to remove surface mount caps. I don't know why this idea doesn't penetrate better. It's easy. It's cheap. And it pretty much completely eliminates any danger of lifting traces, as long as one exercises a bit of patience when heating the component.

15 watt grounded pencils are under $10 each at Radio Shack. A pair of those will lift any SM capacitor or resister in a few seconds without all the angst about lifting traces and whether to alternate sides or use desoldering braid or whether to spend a small fortune on special tweezer tips.

I think one of the earlier posters touched on it briefly, but it really deserves more empasis, so I'm emphasizing it. 2 soldering pencils == trouble free SM (passive) component removal.

 

[TylerEss]

Just use two soldering pencils to remove surface mount caps. I don't know why this idea doesn't penetrate better.

It penetrates just fine, I think. I personally just haven't gotten around to *doing* it yet.

+1this is really easy, hot idea.

 

[JDW]

It also has the potential to literally "penetrate" your logic board if you apply a lot of pressure to both sides of a component and your hand slips, only to have an arrow (soldering iron tip) shoot into your board (or leave a scratch on it, or slice through a tiny trace). That's why I use tweezer-tip heads. No matter how much pressure I apply horizontally to the sides of the component being soldered, my hand will not slip in such a way that the board will be marred.

With that said, your idea about using 2 irons is not a bad one. It's a great idea. But whoever goes that route needs to take care not to damage their board by a slip of hand.

You know, I think it would be fun to shoot a video of the process the next time I do a board. But since I don't have a board that may be a while. Perhaps the individual in this thread who is waiting until Christmas would be willing to shoot a YouTube video of his efforts? I think that would be informative and jolly fun to watch too!

 

[tomlee59]

If the OP hasn't had much soldering experience before, it's definitely a good idea to practice on a junk board of some kind (a great use for discarded PC motherboards).

Once you get the hang of it, make sure that you work on the Mac's board with it secured mechanically somehow, or you risk the sort of tragedy that JDW cautions against. If you do this, and work carefully and deliberately, you shouldn't have any big problems.

 

[da9000]

All the advice in this thread is pretty good. I don't have much to add, except for a big WOW on equill's impressive chemistry knowledge.

@Equill:

would you mind elaborating on what happens when the electrolytic liquids spill onto the motherboard and on traces and are left there for a while?

In my experience they actually corrode or 'eat' the traces. Don't know or understand the chemistry behind it, but I know I've got to clean it and now with your previous extended post, I know the best ways to do that!

Thanks in advance!

 

[trag]

Or you can use desoldering braid on both sides to significantly reduce the solder there if not remove it completely

This is my preferred method as well. However, I was never able to get a braid to work very well.

Braid comes in different widths. If you use too wide of a braid for the power of the soldering pencil you are using, the braid will conduct and radiate the heat away so that it can never get hot enough to melt and absorb the solder.

Of course, if you're using an expensive temperature controlled soldering station, this is less likely to be a problem.

Also, even with the width of the braid well matched to the power of my pencil, I've had trouble with most brands of braid just not working well for some reason. I feel braid impaired. The stuff they sell at Radio Shack (Easy Braid?) has never worked well for me.

However, I have found that the Chemtronics brand of desolder braid works very well for me. I don't know why the brand of braid should make any difference, but it does for me. Perhaps the Chemtronics is impregnated with a different forumuation of flux? Anyway, now days I mail order the stuff from Digi-Key in the 25 ft. roles and keep an eye on my supply so that when I make a Digi-Key order I always check whether I need to refresh my supply of braid.

 

[equill]

... would you mind elaborating on what happens when the electrolytic liquids spill onto the motherboard and on traces and are left there for a while? ...

Although it is in the nature of a 'transition metal' such as copper to be oxidized slowly by atmospheric oxygen, it corrodes most rapidly in the presence of many acids and alkalis, forming characteristically-coloured copper salts, hydroxides and carbonates.

By contrast, the 'poor metal' aluminium readily forms a hard, protective, surface oxide coating in air. In an electrolytic capacitor, that oxide coating is used as a thin-but-effective dielectric between an aluminium anode and an electrolyte-plus-aluminium cathode.

The electrolyte in capacitors of 20-odd years ago contained glycols, boric and organic acids, or their salts, with amines and ammonia, in a high-dielectric solution that was barely on the acidic side of neutrality. The electrolyte was contained in a spacer between the rolled foils, and ranged from liquid to gel-like in consistency.

Maintenance of the dielectric in a working electrolytic capacitor needs suitable pH in the electrolyte, and avoidance of prolonged reversal of polarity between the electrodes. Functionality of the capacitor as a charge store is dependent on maintenance of narrow limits of pH, water content and conductivity of added ions in the electrolyte.

When age, accident or planned ventilation cause a spill from an aluminium electrolytic capacitor onto a circuit board, the pH of the spilt goo (its acidity or alkalinity) is of less immediate concern than the aggressiveness of its solutes towards copper, perhaps amplified by remnant currents transmitted through either the goo or the traces. The chemical soup that constitutes the electrolyte needs little impetus to push it from pH~7 towards alkalinity when borate and ammonium ions are present. Certainly chemical and electrolytic corrosion of board traces is possible, as well as short-circuiting between traces. However, because water is the solvent in which chemical species must ionize in order for electrolysis to happen, water is also the best solvent to remove the goo, even if it is aided by a tiny addition of a non-polar detergent.

de

 

[da9000]

@equill:

Man, you're a mad scientist! Thanks for the treatise!

Now a few questions, if you don't mind, because chemistry was not my forte:

1)

If one leaves the "characteristically-coloured copper salts, hydroxides and carbonates" on the copper, will this further oxidation? (my intuition says no - damage already done)

2)

When you say "of its solutes towards copper", you mean the stuff you mentioned earlier (glycols, boric+organic acids, salts, etc), that is inside the electrolyte, right?

And lastly,

3)

Not a question, but I *always* had a feeling (through experience) that electricity running through the affected traces made things worse.

4)

So here are my most important questions:

If water is the best solvent, won't it matter if the water is chlorinated (tap water) or if it's got lots of impurities (typical tap water) ? Should we use distilled water? What is a brand for a non-polar detergent? And will baking powder do anything at all? (I assume it's used by people to combat the acidic substances)

5)

Easy one: what's your profession? You know so much about capacitors, I'm lead to believe you make capacitors as your day job

Thanks for all the great info!

 

[equill]

... 1) If one leaves the "characteristically-coloured copper salts, hydroxides and carbonates" on the copper, will this further oxidation? (my intuition says no - damage already done)

You need to remove the coloured deposits, if only to be sure that traces and lands have not been severed or destroyed underneath the deposits. Better to know this at the outset than to rue it later.

... 2) When you say "of its solutes towards copper", you mean the stuff you mentioned earlier (glycols, boric+organic acids, salts, etc), that is inside the electrolyte, right?

Any one or more of the chemical species in the electrolyte, alone or in concert with others.

... 3) Not a question, but I *always* had a feeling (through experience) that electricity running through the affected traces made things worse.

Which is what may happen if the logic or power/sweep board is powered-up or active while the capacitors are leaking.

... 4) If water is the best solvent, won't it matter if the water is chlorinated (tap water) or if it's got lots of impurities (typical tap water) ? Should we use distilled water? What is a brand for a non-polar detergent? And will baking powder do anything at all? (I assume it's used by people to combat the acidic substances)

It's a matter of concentration as well as pH. A few ppm of something such as chlorine residues (most of the hypochlorite has been reduced as it was acting on organisms or inorganic molecules) is of no consequence, even if you can still smell it. The aim of washing is 'infinite dilution' of the nasties by washing completely one, two, three or more times as needed, and then rinsing several times to reduce the concentration of nasties (which are not being replenished because you have already removed the bung capacitors) to infinitesimal levels. It's an asymptotic approach to zero, which is never reached, but can be approximated. Ordinary kitchen detergents are non-polar. If in doubt, consult the label on the container.

... 5) Easy one: what's your profession? You know so much about capacitors, I'm lead to believe you make capacitors as your day job

Last 25-odd years: Technical writer/editor. Previous 25 years: Biologist, biochemist, microbiologist, fermentation technologist, biodegradationist(?)—in industry—with a strong dabbling in electronics to use in those disciplines.

Thanks for all the great info!

Y' wulcum.

de

 

[da9000]

1)

What you say makes most sense indeed.

3)

Indeed.

4)

How about using baking power with warm water? Any comments on this?

As for non-polar detergents, I looked at the labels, but they don't say.

5)

No wonder your writs are most elequent (loved the sentence about the asymptotic approach)! Where can one find your editorials?

I was right about the mad scientist comment! Or more appropriately a panepisteme, or the term I prefer for myself sometimes: a polymechanic (as Homer would name Odysseus for his all encompassing characteristics).

Being about half-way to your path in life, I feel good that there's so much more to learn and such great levels to reach. Thanks again for all the wisdom!

 

[equill]

Baking soda is weakly alkaline in water. Weakly alkaline materials, and any foreign materials at all, are what you are trying to wash off. Mere drops of detergent in the wash water is all that you need, and in the rinse water, nothing. Look at, say, Wiki for info. about non-polar (non-ionic) detergents.

Mad, conceivably. Scientist by choice. I leave it to you to sort out the implications.

de :-x

 

[da9000]

Hi equill,

0)

so you're saying that the alkaline nature of baking power could actually further accelerate the corrosion of copper (since copper, a transition metal, oxidizes more readily in the presence of alkaline materials). Did I understand this correctly?

Couple of more questions for you (if you're bored already, you don't have to answer them):

1) What about using 99% isopropyl alcohol? What's your diagnosis?

2) I've been trying to combat the similar corrosive effects of NiCd batteries. Any insights into that? Will warm water + non-polar detergent work best in that case as well?

Thanks again for all the great info!

 

[JDW]

da9000, I'm no chemist, but I can tell you that I've had good experiences with Dehydrated Ethanol. I reported about that in this post, and clicking on my link in that post will take you to the exact brand/bottle I use. I don't just use it for my Mac hobby's either. We use it at the office to clean circuit boards that we repair. Works great to clean up spilled cap fluid and solder resin -- and really, just about anything you could possibly spill on a circuit board. It dries fast too, so you can plug in your electronic device very quickly after cleaning.

 

[equill]

While in no way decrying what JDW has settled on, I point out that anhydrous anything, by definition, has not water enough in it to dissolve polar materials. The inorganic (and organic—nothing to do with food faddism) non-polarized portions of capacitor electrolyte (the gelling agents) and the ionized molecules and ions (duh ...) dissolve in and are carried away best by the ionic solvent 'water'.

Truly anhydrous ethanol is so avid for water that it will imbibe water from the air, so making and keeping anhydrous ethanol anhydrous is an expensive pastime. The eutectic mixture of ethanol (what results from distilling a water/alcohol mixture) is about 92% ethanol, so you can see that solutions of ethanol in water (and water in ethanol) on either side of 92% ethanol are not stable, tending up or down towards the magic 92%.

Where anhydrous ethanol (and other liquids that can dissolve water) come into their own is as the last rinse, because they can remove remnant water from nooks and crannies (eg under surface mounted ICs and other components) more quickly than can simple evaporation of water. Ethanol, or a higher (longer-chain) aliphatic alcohol that does not deplete the logic-board's lacquer, then comes into its own. As I wrote in another post, commercial board-cleaning is usually done by sonicating boards in water containing suitable detergents, followed by rigorous rinsing. That's cheap and effective. If remnant water is dissolved off a board, as opposed to wiping or blowing it off, time can be saved. However, 98% ethanol is much cheaper for the purpose than is the fiendishly expensive 100% ethanol, or the not-much-cheaper 99.5% ethanol. Acetone is (relatively) cheap, and also infinitely miscible with water. However, it is dangerously flammable and probably can play merry Hell with board lacquers. In the end, one does what works, and swallows the cost, or seeks a less costly equivalent process.

The number of boards to be dealt with at any one time will influence the choice between quick and cheap.

de

 

[JDW]

... it might even have to wait until Christmas.

Just wanted to post this reminder...

It's Christmas!

:

 

[beachycove]

Still haven't sourced all the right capacitors, but will report back when I have done the deed.