capacitor handling and mistakes and details of manufacturing

1 reply [Last post]
speedyG's picture
Offline
Joined: Nov 16 2011
Posts: 1442

In another thread a discussion started up to reasons why electrolytic capacitors can fail or become damaged....
here just an overview of post within that thread

speedyG wrote:

indeed a bad power supply can not only cause that mistake - it can ruin - if the use is continued - more and more chips with continued usage ! If the PSU smells funny ( like burned chemical ) it must be repaired before any other test can proceed..... trace where the smell is coming from... normaly you can follow that smell even if the PSU is switched off and the smell willl bring you to a burned or at least damaged component that must be replaced....
thereafter the PSU shall be tested at the outputs for the correct voltages and connecting a old harddisk with the 0 Volt, + 5 Volt and + 12 Volt is obligatory because the correct voltages can only be measured reliable with a connected load.
The measured voltages should then remain within a tolerance of +/- 0,5 Volt .......
you should not continue with tests at the mainboard unless you have confirmed the PSU to act reliable within specifications - otherwise the damages at the mainboard might be expanded !

BE CAREFULL WHEN WORKING INSIDE THE PSU ! THERE IS DANGER OF LETHAL SHOCK WITH 220 Volt or 110 Volt !

mjnurney wrote:

opened up the PSU this afternoon and the following look very ill Smile
Astec power supply AA 11040c
230vac 50mhz 0.5a
C23 capacitor 0,01uf (a) X
C1 capacitor 0,1uf (a) x
capacitors here look very old too and i'll change them..
c11 0,22uf +-10%
c17 0,022uf M
c10 0,22uf +-10%
C23 has exploded in half and smells like a burnt fire type smell Smile
can anyone guide me on the voltage of these ? i haven't taken the pcb out yet to see if its on the 230v line.

Neon Forests wrote:

The likely culprits are the electrolytics.
While other capacitors can go bad, generally most will remain fine.
Also be sure and check/touch up the soldering on the trace side.
It wouldn't hurt to check the diodes either.
If one shorted or opened, it would take out the electrolytics that much quicker.

speedyG wrote:

first the data in your inquiry, then a summary...
C1 = 250 Volt ( min. 250 Volt )
C10 = 0,01µ 1000 Volt Ceramic ( min. 400 Volt )
C11 = 100 Volt( min. 60 Volt )
C17 = 100 Volt ( min. 60 Volt )
C23 = 47µF 250 Volt electrolytic radial
There is a bridge-reticifier at DB1 ( entry of the 230 Volt line that must be able to handle at least 250 Volt ( better 400 Volt ) and at least 1 Ampere ( better 1,5 Ampere ).

The damages indicate that a shortcut has occured ( probably a lightning stroke close to the house )
and that the wall plug had no safty circuit protection against overload.... ( old house and old slow fuses )
a lightningstroke is that fast that sometimes even the internal fuse is left intact due to the fact that its too slow....

the residy oily dirt outside of the exploded capacitor is bor-acid.... this must be cleaned carefully away
- it etches copperlayers away.

In such cases the bridgereticifer MUST be replaced too as well as also the Foilcapacitor
parallel to the powerinputplug
0,47 µ 275 Volt ( carefull this one can only be replaced with EXACTLY the SAME one ! )
and this one is really difficult to get.....

If Powersupply is O.K. and starts working again you must examine very carefull the mainboard too for shortcuts !
In regelar cases a part of the lightningstroke enters the mainboard and damages at least one buslinetranciever
and one RAMchip !

I have tried several times to repair such a damage and these Powersupplies failed very short later again...
My advice is to better immediatly try to replace the powersupply as entire unit.....

by the way up to the time where the PSU is replaced you might use instead a old powersupply from an IBM Computer with some 150 Watt or so..... You just need to solder an adapter cable....
if the IBM Powersupply is ATX typ you must bridge 2 lines for the supply to start ... the instruction for that is availiable in the internet or in the mentioned page above....
sincerely
speedyG

Neon Forests wrote:

With all due respect (and 37 years of experience as a component-level tech), it's much more likely that the electrolytics blew from age than a local lightning strike.
Switching power supplies place a heavy burden on their main filter caps.
High temperatures, high altitude and age can degrade these "wet" caps even more quickly.
I've repaired hundreds of power supplies and several Apple II supplies and in most cases it was only the caps and/or diode(s).
It's true that anything can fail, but it's still more likely to suspect aged components themselves instead of lightning.
Please note also that some diodes can be fast-recovery or zeners and should be replaced with exactly the same part type.

speedyG wrote:

Hello Neon Forests,
i´d agree with most of the post - except one fact ( with respect ) - the one cap ( C23 ) exploded completly in the middle ( see explanation of mjnurney above ) and burned off..... i´ve seen this kind of damage 4 or 5 times in the past 35 years ( and once 30 years ago ) in one of my own computers , that occured during a thunderstorm and before i installed a ignition overload protection in my home powerlines.....
capacitors that die from age normally just start leaking the bor-acid - but they normally don´t blast just right in the middle....
sincerely speedyG

mjnurney wrote:

hello all,
I have changed the damaged caps and every foil cap in the psu and checked all the solder joints.
I now have a working PSU supplying all the voltages. i have wrote them down but don't have them with me. I would expect caps to fail with age - this Apple II is 33 years old now..
Anyway the graphic fault is exactly the same although the PSU no longer smells Smile
i have taken a HD pic of the ram in it original location so i can put them back exactly as they came out and placed them in line next to the Apple anyway.
CAP C23 is a metal oxide paper cap (X) it won't leak , it would just blow i would of thought?
mike.

Neon Forests wrote:

Not to elongate this or create a pissing match, but I've seen many electrolytic caps explode.
Not always from old age, but also not from lightning.
My own Imac had caps that exploded and it was a 2006 model.
Another Imac G5 did the same.
Neither saw lightning.

The pulses these caps see is far beyond what a linear supply's caps would have to deal with.
Some of these older caps are only rated at 85C too.
As they age, they become more like resistors. Their ESR changes.
So they get hotter and hotter and don't always vent from the top or push off their cans when they die.

It should also be noted that the current trend is to use multiple caps (like the Imacs.)
This is ok at first, but can create an avalanche effect of failure.
Some start going, then others follow.
But keep in mind that while capacitors have gotten smaller vs their voltage ratings, current still needs size.
I used to repair video monitors and despite a cap being the correct mfd and voltage, too small a size would explode.

mjnurney wrote:

Well we had a power cut last night as the Apple II was on and then the power came back on ...now the Apple II doesn't , blank screen that flickers when power is applied...
Sad
bad to worse....

just one remark at the moment first... i already mentioned that a repair by just changing the capacitors often leads up to a point where the PSU only does for very short time its duty and soon will fail again as it did here tooo

and surely< there is another chance to get sapacitors exploding too....
the easiest way is just to change polarity of the powersource....
this gets them exploding like firecrackers within seconds.... and with the explosion bor acid is splashed in the entire earea aound.... so don´t do this trick without security precautions... bor-acid may damage your eyes if it gets into there....

anyhow there are big differences also at the electrolytic capasitors during the ages of manufacturing ....
for example those made in the years from the 50´s up the mid 60´s have been very robust for heavy duty with tubes and they usually don´t explode at all, because they are locked in thick aluminium cups and you have to give them a lot of wrong power to explode....
those manufactured from the mid 60´s to the mid 70´s had also rather good covers and did not explode that easy....
those manufactured from the mid 70´s to the mid 80´s became much easier to fail and they also aged much faster with leaking because the containers heven´t been that good as in the ages before and in those days it was a common trend to get the size of the capacitors to shrink down as small as possible ( with the result the foils inside and the electrolytic fluid was reduced and the interspace between the foils became smaller ( i.e. closer together )- this was a development together with the shrinking of the used space on the PCB´s to and placing the IC´s closer together ...
since the mid 80´s to the mid 90´s the capacitors shrinked down in size nearly 30 %,....
at the end of the 90´s and beginning of the 2004 the capasitors again made a giant leap in being sized down another 50 % ( the so called "Super Caps" ) but they are even much more dependent to exactly defined powersupplies and they have nearly a zero tolerance to higher voltages.... the old capacitors usually had a tolerance to carry up tp 20% to 30 % of higher voltage for quite a time before failing .... with the modern "Super Caps" this new capacitors get killed by overvoltage within few minutes....

speedyG

__________________

In early days I had a lot of money but no time - now I have no money but a lot of time....
the second part includes less friends but a lot more joy on life....

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
speedyG's picture
Offline
Joined: Nov 16 2011
Posts: 1442
Re: capacitor handling and mistakes and details of manufacturing

In other words additional to the explanations above... the capacitors used in the end of the 70´s had other quality than those used later in the mac in the late 80´s.... those from the late 80´s died faster and with less "forwarning" .... those from the late 70´s died by starting to issue leakage and explosions were rather rare... those from the late 80´s ( as you explained that were used in the mac´s ) just exploded alone from heat and dry-out....

in a case like that with an exploded capacitor in the powersupply of an Apple II the oder of incidents is first that the bridgereticifier gets shot and one of the internal diodes gets shortcut and second with the next powerwave the capacitor gets the wrong polarity from the 110 / 220 Volt source and explodes because of turned polarity of the current.....

__________________

In early days I had a lot of money but no time - now I have no money but a lot of time....
the second part includes less friends but a lot more joy on life....