Re: Help! Another chirping IIgs PS...

[rubywand@swbell.net]

pubpc1@library.ucla.edu writes ...
> 
> eddy@cs.ucdavis.edu (arekusu) wrote:
> >To my horror, this morning when I flipping my GS on it emmited a
> >rapid chirping noise, power cycling with each chirp. (maybe 8-10
> >times a second...)
> >
> >(system description at http://wwwcsif.cs.ucdavis.edu/~eddy/gs.html)
> >
> >Pulled all cards, plugged just the GS into a different wall outlet,
> >tried again. Same thing. Reading the Hardware Ref, it mentions the
> >power supply will exhibit this sort of behavior if it detects a
> >short, or "no load"...
> >
> >any ideas? Common causes/cures?
> 
> I believe that your power supply is failing.  What you seem to
> be observing is the power supply's safety feature, which prevents
> the dying PS from destroying the GS logic board.
> 
> The cure will most likely involve getting a new power supply,
> preferably with higher capacity (an overloaded PS may fail in
> the above manner).
>

     The chirping usually means that the PS load-detect circuit thinks that 
the load is too great (i.e. that it draws too much current). This can mean 
that the power supply is bombing and, so, almost any load is too great. It 
may, however, mean that a short has developed on the motherboard.

     One test is to unplug the power supply, remove it, and do some load 
tests. If the PS can deliver +5V with a 3 Ohm power resistor connected across 
the +5V and Ground lines, +12V with a 15 Ohm power resistor connected across 
the +12V and Ground lines, and correct Voltage for -12V and -5V lines into, 
say, 100 Ohm resistors, then the PS is likely to be okay.

     If the PS chirps when small loads are applied (e.g. 10 Ohms across the 
+5V output), then it has one or more bad components. Replacing it with a new 
A2 power supply, a module from Jameco, etc. or a Buggie external PS may be the 
easiest cure.


 
and arekusu writes ...
> 
> Leif Strand (lstrand@netcom.com) wrote:
> : My IIgs died a few months ago in a similar manner.  It didn't emit a
> : "rapid chirping noise", however.  The power light blinked, and after
> : Is there any chance that my system, too, just needs a new power supply, or
> : do I need a whole new CPU?
> 
> Well, in my case it was the power supply. I've gone through oh,
> about 4 now. Replaced it with an old Apple original, and it works fine
> again, though whiney.
>

     A system which has chewed 3-4 power supplies is likely to chew another of 
the same type. Unfortunately, many GS users with higher performance systems 
may be in for a similar failure of the 'stock' power supply.

     The problem may relate to the skimpy leads on most Apple power supplies. 
These leads act like resistors. As boards are added and current load 
increases, the +5V lead's resistance becomes especially noticable. Noise 
voltage developed across the lead will propagate throughout the system and 
may cause system glitches resulting in mysterious crashes.

     Our GS is a typical, moderately 'stuffed' ROM-01 system with 4MB RAM 
expansion, Zip-GSX, RamFAST, internal hard disk, Mockingboard, stereo 
card, etc. (but, no SS card). It is using the same stock power supply it came 
with more than 10 years ago except for one mod: the wimpy +5V, GND, and +12V 
leads were replaced with heavy wire leads about the time we began stuffing the 
Slots.

     It is easy to see that the standard GS power supply should have the 
capacity to handle just about any setup. You just compare the published 
current ratings for each output to the load on each output. However, there is 
no way to draw anything like the rated current at +5V through the skimpy leads 
connecting the power supply to the motherboard.

     One might expect that, at worst, a 'stuffed' system connected to the 
standard power supply via the standard leads would simply crash due to 
glitching. This is bad enough. A wish to avoid glitching was the motivation 
for swapping in heavy leads on our PS. There was no thought that the heavy 
leads would protect the power supply; yet, this does seem to be an additional 
benefit.

     I don't know what produces persistent PS failure on so many GS systems. 
If the PS detects a simple "too much current" overload, it will protect itself 
and shut down. Possibly, voltage transients developed across the original +5V 
and GND leads damage the output voltage sensing circuitry. Or, maybe, the 
switched pulses developed across the leads trick the PS circuitry into trying 
to output higher voltages than the PS can safely handle.

     For now, even if a 'stuffed' system is not crashing often, the 
bother of installing heavier leads may be worthwhile. It _looks_ like getting 
rid of the wimpy leads is a good PS protection measure.


Rubywand