Post by Stu M
I'll lay out my concern plainly...
I've dealt with hundreds, if not thousands of modern desktop computers
with destroyed motherboards due to power supply failures. Either by short
circuit, or because the power supply's regulation was too far out of
spec, damaging the VRM (fine voltage regulation section) of the
motherboard by trying to make up too large of a variance.
Of course, the power supplies involved were total garbage, and this
includes systems from all of the major manufacturers.
Unfortunately, too many times when a power supply fails in a "modern"
computer, it takes the rest of the system out with it.
Simply put, I don't want to see this happen to any of my vintage computers.
Yes, back when the Apple II's were built, Apple was building things to a
much higher standard of quality than many of todays electronics.
I feel that with every year that passes, that the risk of catastrophic
power supply failure increases.
With next year being planned as the last year of the garage giveaway, and
with eBay prices reaching astronomic levels, I'd like to do what I can to
preserve what I have. I've smelled the magic smoke many times in my life,
and the last place I want that coming from is one of my Apple II's.
Let me ask this; What can I do to ensure that a power supply is not on
the verge of failure, putting the connected system at risk?
"The art of prophecy is very difficult, especially with respect to the
future." (Mark Twain)
And certainty is even harder! It's worth considering likely failure modes.
Short answer: it is highly improbable that an Apple II power supply will
put a main board at risk even when it fails.
I agree that electrolytics are the most likely to fail, and the vast
majority fail in the direction of "open"--usually by increased equivalent
series resistance. Such failures are extremely unlikely to result in damage
to the load (main board).
In fact, loss of regulation leading to significant overvoltage is the only
"risky" failure I can think of, and that would require the failure of one
of a couple of resistors or a transistor in the feedback loop (I'd have to
review the schematic to be more precise.) *And* the failure of the supply's
The best predictor of a supply's robustness is its performance under full
(or 10% greater) dummy load. If it works for 30 minutes under load, it's
very likely to be quite trustworthy. (Shaking or banging it a bit is also
Post by Stu M
Do you agree with the general statement that most aluminum electrolytic
capacitors at this point (~30 years) have either dried up, leaked, or
otherwise not performing correctly?
I think experience shows that the vast majority of those original
capacitors are still doing fine, and can probably be expected to continue
to operate satisfactorily. If one should fail, it will very likely fail
open, resulting in no damage, just incorrect operation.
What do you suppose is the probability of a new capacitor experiencing
infant mortality? Certainly greater than zero.
The Apple II supplies were designed with a crowbar circuit that shuts down
the supply in the event of overload or overvoltage, which provides
protection against the fault you're concerned about. (That's why Apple II
supplies "click" when shorted.). Of course, it's possible for the crowbar
to fail, leaving the load exposed if a second improbable failure occurs,
but now we're getting into the weeds. ;-)
Just consider all the Apple II's that have experienced power supply
failures--what fraction of those have led to consequent main board
failures? I'm sure there are a few, but consider the much larger fraction
that only required a replacement power supply. Apparently, Apple supplies
seldom fail catastrophically.
Post by Stu M
What other components do you feel should be checked or replaced at this point?
I would not replace any properly functioning component unless it showed
clear signs of degradation, say by overheating.
And I would never desolder a component in a functioning supply just to
measure it--that would only be justified by strong suspicion of its
My first job was as a radio/TV repair technician, and that exposed me to a
lot of malfunctioning analog electronics. About 10% of the problems I saw
were the unintended consequence of an earlier repair! No intervention on a
circuit board is without some risk!
Rather than rattle you with non-zero probabilities of catastrophic failure,
I'd like to encourage you to have confidence in a well-designed system
(unlike some Chinese PC power supplies). ;-)
Post by Stu M
By the way, you were greatly missed at KFest this year!
Thank you, Stu--I really missed KFest and all the KFesters, too! ;-(
-michael - NadaNet 3.1 and AppleCrate II: http://michaeljmahon.com