Hey all, regarding this here is something to read...

http://www.chemi-con.com/u7002/life.php

This article shows that operating an aluminum electrolytic capacitor (which is 
the type I assume everybody is referring to) below it's rated voltage has only 
a minimal impact on it's life.  A more important factor in determining lifespan 
is the operating temperature.  Additionally, while not discussed in the 
article, it is very important to understand that for an aluminum electrolytic 
the rated capacitance is AT the rated voltage...  As you decrease the voltage, 
the capacitance drops off rather dramatically.  I no longer recall the general 
decay rate, but that 100uF 16V capacitor when run at 5V is likely to be closer 
to 50uF (or less).  The odd thing is that these aluminum electrolytic 
capacitors are a bit of a throw back...  They work well at line frequencies and 
are suitable for linear regulators, but really not much else.  Unless you get 
high frequency capacitors they are just about worthless in modern electronics 
when used by themselves.  In most of the systems I have designed, the 
electrolytic capacitors were closely coupled to a solid tantalum and a small 
ceramic to deal with any instantaneous current requirements (perhaps 100uF, 
2.2uF and .1uF).  When it comes to solid tantalum capacitors I would always 
recommend running them well below their rated voltage which will certainly 
improve the lifespan without impacting the capacitance.

As for replacing the aluminum capacitors on the motherboards...  Here are my 
suggestions based upon a great many years of soldering and countless board 
repairs:

Generally I do not advice destructive removal of a faulty component, but in the 
case of Apple's circuit boards I do.  Some boards are capable of being reworked 
without great fear of lifting a pad, unfortunately the Apple boards i have 
worked on are not this type of board.  While far from the worst boards I have 
worked on (where the pad lifts as soon as heat is applied) they are rather 
delicate.  Carefully cut the capacitor in half and gently remove the can and 
any additional pieces leaving just the two leads on the pads which should come 
off without great difficulty.  To assist you in the removal of the leads a 
small drop of liquid rosin will aid in the wetting process...  Alternatively 
you can use a bit of fresh solder.

When replacing the aluminum electrolytic capacitors use solid tantalum 
capacitors as a replacement.  Ideally you should use surface mount and not 
through-hole devices with the leads bent over to match the pads.  The reason 
for this is the force placed on the pads by having an elevated mass is not to 
be ignored...  An unfortunate bump can cause the capacitor and pads to be torn 
from the board.  Once the capacitor has been replaced and the board has been 
cleaned, I would suggest using a small dab of epoxy resin to ensure the 
capacitor stays put.

Cleaning the board is quite important!  People have hemmed and hawed about 
using a dishwasher or immersion in a water bath, but do not fret.  With rare 
exception most electronics can be totally immersed in water without harm...  
Providing the board is thoroughly dried prior to applying power.  Generally 
anything not sealed would be a non-submersible part including (but not limited 
to) some switches and audio enunciators (the little beepers/buzzers you 
sometimes see in electronics).  One important note on water is that ideally you 
should use distilled (or better distilled & deionized) water for a final rinse. 
 Most of the water we drink has dissolved minerals in it which, while not 
likely, could cause a problem once it has been deposited on the components.  
Isopropyl alcohol is a great solvent to clean the circuit boards, but keep in 
mind that when the alcohol evaporates it does not take the impurity with it.  
To clean properly with alcohol you must clean the wet alcohol off the board 
(Kimwipes are ideal for this).  If you do not wipe up the alcohol the residual 
flux or whatever you are trying to clean off the board will simply be thinly 
spread of a larger area (the area covered by the wet alcohol).  If you are 
really crazy, a thorough water flush followed by a distilled-deionized wash and 
then an immersion in trichlorotrifluoroethane will do the trick (after drying 
of course)!  In my younger Navy days I once had a radio room flood with salt 
water and we had a couple dozen transmitters totally immersed.  We used the 
triple wash method and only had 2 or 3 transmitters not come back to life.

If there is interest, I can take some pictures of the process (I have a great 
many boards to recap) and put them...  Somewhere.  This is not a difficult 
process but it does take some finesse and it helps to have good equipment.

Without re-reading this long mess I think it is all correct and coherent.

Derek
ex-Navy 2-M, C Level micro tech (for those who know about this sort of thing)

On Jan 7, 2013, at 2:59 PM, Jeff Walther wrote:

> 
> 
> On Dec 30 2012, 12:26 am, James Fraser <wheresthatistanbul-
> [email protected]> wrote:
>> Hello,
>> 
>> --- On Sat, 12/29/12, Doug McNutt <[email protected]> wrote:
>> 
>>> And while I'm at it, Apple's choice of aluminum
>>> electrolytics seems to have used a lot of caps rated at 16
>>> volts where a 6 volt rating would be good enough. It's
>>> likely a perceived saving in parts where variety makes
>>> automated assembly more difficult.
>> 
>>> Six volt tantalums are cheaper and smaller if you can
>>> separate the 12 volt bus from the 5 volt bus.
>> 
>> I'll bite: when you say "cheaper," are you talking on an industrial scale
>> (like Apple would use) or on a (typical) hobbyist's scale?
> 
> One of the members of 68kmla.org dug up an old white paper on
> capacitors and expected life times and such and the paper concluded
> that capacitor's have much longer lives if they are never operated
> close to their actual voltage rating.  The paper or a link is posted
> somewhere in the forums over there.  I didn't keep a link, so to find
> it you'd have to search.  I think JDW may have posted it, but I'm not
> certain.
> 
> If you actually look at the expected hours rating on some of the
> electrolytics, it is as low as 1000 hours.   Even in a home
> environment, that time would be consumed quickly and in an office
> environment the capacitors would not last out the year.
> 
> The paper recommended voltage ratings approximately three times the
> expected operating voltages.  So, in a 5V circuit, 16V is just about
> perfect.   The 1uF 50V capacitors Apple was so fond of are a little
> odd, given that they're probably in, at most, a 12V circuit, so for
> those I have a 1uF, 35V cap. as the 50V variety was difficult to find
> in the small package size (physical envelop) needed.
> 
> So, while it would seem that one would use 6V capacitors in a 5V
> circuit, according to some evidence some of us read once, 16V is a
> much better choice.  :-)
> 
>> I ask because, just out of curiosity, I pulled up an old post that listed the
>> caps that Jeff Walther was offering a couple of years back (NB: circa 2010;
>> he may or may not be offering the same selection in 2012) and didn't see
>> any caps with less than a 16V rating:
> 
> The selection has changed slightly and some new stuff has been added,
> but it's still basically the same.  My posting is sticky'd in the
> "Trading Post" forum on 68kmla.
> 
> I wish I had time to do capacitor installations too, but I just
> don't.  I have a job, a house, a 10-year-old son, fish, a vegetable
> garden, I coach LL baseball two seasons a year, and I have an anemic
> thing resembling a social life as well...
> 
> Jeff Walther

-- 
-----
You received this message because you are a member of the Vintage Macs group.
The list FAQ is at http://lowendmac.com/lists/vintagemacs.shtml and our 
netiquette guide is at http://www.lowendmac.com/lists/netiquette.shtml
To post to this group, send email to [email protected]
To leave this group, send email to [email protected]
For more options, visit this group at 
http://groups.google.com/group/vintage-macs

Support for older Macs: http://lowendmac.com/services/

Reply via email to