"Randy" wrote in om:
"Vanguard °°°" wrote in message
news:qr_Ta.139523$H17.48792@sccrnsc02...
"V W Wall" wrote in message
...
He also mentioned that he did not have a grounded line power outlet!

He must've been communicating to you separately via e-mail. I read
both
of Randy's posts in this newsgroup and neither mentioned the use of
an
ungrounded outlet.

I did not communicate separately... my outlet is grounded.


The motherboard also removes the "turn-on" signal when the momentary
front panel switch is held closed for ~5 seconds. This is not
needed
when testing a PS, but most do require a load of ~1A to turn on.

I'll have to check this on my next system build to see if a load is
really needed or if just shorting the PS-ON signal pin to ground will
work. If it doesn't then I'll know that I need to generate a load
(but
would that be 1A across all taps or just 1A on one of the taps)?

The ones I've seen only use LEDs to check for proper voltages. They
do have a "power on" switch and a load to insure proper start up.
You
really need a digital VM and a variable load to properly check out
a supply.

The Antec PS tester does let you use a voltmeter but does provide the
green LED "idiot light" to show the PSU is okay. It has a 25W
5.4-ohm
load to create a load (but they don't mention if they load all
outputs
or maybe, say, just +5V), but it's a fixed load so you cannot
increase
the load (current draw) to check for degradation in regulation (which
would require an oscilloscope to see the change in the ripple,
anyway).

The input line filter in the PS, which tom_w is making such a fuss
about, can only be effective if a proper ground is supplied.
Otherwise
the conducted noise has no place to go but onto the input power
lines,
which then act like a big antenna to radiate the interference. The
basic frequency is ~40-50hz, but having fast transitions, it is rich
in harmonics, hence the broad band interference.

Again, I didn't see Randy mention using an ungrounded wall outlet. I
did see him mention about trying to use a surge protector that also
included some line conditioning (although it looks like simply a
passive
setup using a capacitor) but then that wouldn't protect the line from
getting noise from the output side of the surge protector. Putting
both
the computer and radio on the line conditioning surge protector
wouldn't
protect the radio since both devices are on the output side of the
filter. Maybe if he had plugged the computer into the wall outlet
instead of the surge protector and then plugged the only radio into
the
surge protector he might've seen some relief if the line
conditioning in
the surge protector was of any value (but it sounds like the problem
might be with RF instead of induced EMF).

first, thanks to everyone who has responded... it has definitely made
me think about a few things.

To set the record straight I just finished many tests.

My outlet is grounded. With the computer off but the ps switch in back
set to "on" I get am radio interference (bad) in all that circuit's ac
outlets in the room. I also get interference on a battery operated
radio in the same room... maybe a smidgen better but still
unlistenable to. I had plugged the computer into the new surge
protector and it didn't help. I also plugged the clock/radio into the
surge protector and it didn't help. My case is metal. The power supply
is 300w adt-300 (came with my DTK WT-PT074W ATX Tooless case).

I disconnected everything and shorted the power supply... the fans
came on and it still wrecked the AM radios. I turned off the power
switch in back of the case, left it connected to the motherboard but
tried disconnecting each drive in turn... still AM radio problems. I
then left the drives connected and disconnected the motherboard...
there was a quick (maybe 1 second) AM radio disturbance and then it
faded away to become pretty good. Plugging the motherboard back in
made it bad again. I disconnected all the cards from the motherboard
but it was still bad if the power supply was plugged into it.

Now, could it be the motherboard? I think not because this is the
second motherboard I had in this case and both were AM radio
destroyers. First a ECS-K7S5A and now an Asus-A7N8X. I'm thinking that
maybe the motherboard connection clicks something special on the power
supply that shorting it does not.

Anyway there it is in a nutshell. I think I might try a better more
powerful (430w) ps. With a dvd, dvdrw, 2 hd, floppy (useless), canopus
analog video converter, radeon 9600pro, modem and three case fans I
might be pushing it with the original 300w cheapo ps. But... I've
never had a single problem with it other than the AM noise.

Randy

No, plugging it back into the motherboard just allowed the +5V turned on
(even with the unit powered off). As long as AC power is supplied to
the power supply unit (because the rear switch is on), the +5V power is
active in an ATX-style power supply. Someone else already re-mentioned
this to me since I had missed or forgot it. When you shorted the
power supply (presumably meaning PS-ON signal to ground), this would
have enabled the +5V tap to go active again, and when it did you got the
AM radio noise again. You've had 2 motherboards in this machine but the
same power supply and the problem continued. With no power to the
motherboard or drives, the problem lingered. Sure looks like a power
supply problem to me.

How old is the power supply (or the case setup that included it)? Is it
still under warranty? Maybe you could get a warranty replacement power
supply. It's possible that this power supply was defective from the
start and that a replacement of the same model would eliminate the
problem. However, the cost of shipping it to them might be too
expensive and could be better used in purchasing a new one.

I didn't find any computer cases at http://www.dtk-computer.com.tw/, and
none of the cases listed at
http://www.dtk.at/products/products_cases.htm were for your case model
(and none of the cases that they list show a model of the power supply
that matches yours). So I can't lookup what power supply was included
with the case. The model number you mentioned might've been a part
number from the reseller from whom you got the case.