Dead computer?

Patty wrote:

with the Antec Power Supply Tester, it should make life
much simpler. From what I understand from reading about
it at Tom's Hardware site, you plug it into the power
supply (main power plug that goes to the motherboard),
it puts a load on the power supply, and then has
places that allow you to check the various voltages.
Sounds simple enough, hopefully I can get to Circuit
City tonite to pick it up and give it a try.

If the mobo didn't work even with a new PSU, maybe the mobo has been
zapped or it has bad capacitors (see www.badcaps.net) in its voltage
regulator (long row of capacitors between that long tan slot in back
and the rear connectors -- if any are bulging or leaking on top, they
may all be bad).

PSU testers are so bad that they'll say that a PSU is OK even if it's
voltages are all way off. I had one do that even though the +12V rail
was too low to make the HD spin.

It makes no sense for you to buy a PSU tester because you already have
a digital multimeter that's a lot, lot more accurate.. All you need is
a paperclip to turn on the PSU by shorting the green wire (pin 13) to
any black wire (pin 14 or 15 will do) plus maybe something to load down
the PSU, like an HD or a couple of resistors. The +5V can be loaded
with a 5-10 ohm resistor rated for at least 10 watts and connected
across red and black wires (some PSUs won't even start without a load
on the +5V). The +12V can use 10-20 ohms, rated at least 20W and
connected across yellow and black wires. A working mobo can be the
load instead, and it's OK to short the green wire to ground even when
the mobo is attached, or you can blip the two mobo pins that normally
connect to the power button.

Patty wrote:
On Sun, 30 Jan 2005 08:28:02 -0500, w_tom wrote:

I
can't get any reading with the power supply connected to the motherboard, I
think my probes are too big to fit in the backside of the connector.

Use some slender conductors such as paperclips to reach inside the molex
connnector, then touch your test leads to those conductors. Radio
Shack's #270-354 Alligator Clip Adapters
http://www.radioshack.com/product.as...t%5Fid=270-354
are useful in this context.

Don't accidently short across the contacts whose voltage you are testing
or you may have some unwelcome smoke or a blown fuse.

On Sun, 30 Jan 2005 17:09:47 GMT, Matt wrote:

Patty wrote:
On Sun, 30 Jan 2005 08:28:02 -0500, w_tom wrote:

I
can't get any reading with the power supply connected to the motherboard, I
think my probes are too big to fit in the backside of the connector.

Use some slender conductors such as paperclips to reach inside the molex
connnector, then touch your test leads to those conductors. Radio
Shack's #270-354 Alligator Clip Adapters
http://www.radioshack.com/product.as...t%5Fid=270-354
are useful in this context.

Don't accidently short across the contacts whose voltage you are testing
or you may have some unwelcome smoke or a blown fuse.

Well, I went ahead and used the power supply tester. I had already gotten
+4.99V on the purple with just using the meter. With the tester, I got
+5.16V on the +5V, +11.92V on the +12V, and +3.4V on the +3.3V. (Now that
I think of it, I remember these types of numbers when I used to have the
power supply in my system and checked it using MBM5.) Can I assume now
that the power supply is ok? Or should I try to test more wires by
following your instructions? I really don't want to short out my power
supply, I may need it in the future. It's an Enlight 300w that originally
came with my case (was replaced with an Enermax 350w.) I'm still thinking
it's the motherboard that has a problem since it didn't work with its
original power supply (which I don't have since the person they took the
computer to in the first place said it was fried and kept it... yeah right)
and this one is my own. I still think two different power supplies, same
result... dead motherboard or CPU.

Being an owner of two Abit KT7 motherboards, I know all about bad
capacitors and how to recognize them. ;o)

Thanks much for your help.

Patty

Patty wrote:
On Sun, 30 Jan 2005 03:40:49 GMT, sbb78247 wrote:

Kinda
funny though the Dell's have a poor track record and the machines
that I have built keep on running without incident.
Hmmmmm............

While for the most part, this is true, most computers (even
homebuilt) are only as good as the parts that comprise them. I once
built the computer from hell. Yes, that thing gave me nothing but
problems. After much fooling around with it, upgrading BIOS,
tweaking, I finally got it mostly stable. The mainboard was an ABIT
KT7 w/RAID (now known as the bad capacitor boards) and I finally
removed it when the capacitors began to look iffy and another system
with the same board that I had built had already died, and I now run
a much more dependable system. Sometimes you never know, you try to
buy what you think is good and it doesn't always turn out that way.
However, when you've built it yourself you only have yourself and
maybe these knowledgeable folks in these newsgroups as your tech
support (which can be MUCH better than Dell or other tech support).

You know DISCUSSION includes personal experiences so others can draw
on it and make their own conclusions, or does your brain not work
that way?

I was just pointing out that we were considering a different
situation. I do remember when Dell out of the country tech support
did handle business computers. Where I work, one of the ladies
called Dell support and that's exactly what she was told, so she did
reinstall Windows. What a mess. Everything on that system needed to
be reconfigured (network logins, etc.) and she lost all her programs
as well which needed to be reinstalled. She did not know this would
happen. Since it's a small company, we do not have an IT department
(only 4 employees total, real small). I told her next time she had a
problem to call me and I'd help her work on it.

Now off you go to sift through yours and everyone you know's junk
pile to hopefully come up with something that works.

Yeah, I'm sure we'll come up with something. This lady doesn't need
much. Since she was running a 266Mhz system with only 32MB RAM (and
it was sufficient for her needs) almost any old computer (better than
a 486) will more than likely do well for her. Besides, I have lots
of old junk I'd like to get rid of... don't you? vbg Btw, I have
an old modem in this system now, that I got from who knows where,
(56K modem) and after searching around, found drivers for it that
work and it's used for faxing. So, old "junk" can be useful at times
too. ;o)

Good Luck, HTH, GFIA

Thanks. Much appreciated.

Patty

56K internal??? My fax modem is a 28k external - you must really be up with
the times.

I just sold the majority of my old junk around Christmas to a guy that
wanted a system for his kid, P4/1.7, 80g hd, 845 based mobo, antec case,
52x cdrw, 512mb mem, basically a hotrod from 4 years ago. They were happy
and I got some coin out of the deal.

S

No reason to apologize. You are doing things that even
computer assemblers fear to learn. That battery measurement
was so simple. You did something that many computer
assemblers fear to do and then say it was all unnecessary -
which is why they also purchase insufficient power supplies
costing only $25 or $40. In short, I am pleased you instead
want to be smarter.

The purple wire was 5.0 volts. But when you press the power
switch, no other voltages appear. No problem. As those
previous newsgroup posts noted, the motherboard controller
turns on power supply. So we still do not disconnect any
wires. We measure voltage on the Green wire. This is a
signal - called Power On# - that tells PSU to power up. Put
that meter on green wire to measure more than 0.8 volts -
probably 1.8 or more. When power switch is pressed and
released, that green wire voltage drops to 0.8 volts. We have
just tested other two components of power supply system.

Let's assume that **green wire voltage does not drop**. Now
we suspect the other two power supply 'system' components.
First verify the power supply switch. Is it even making a
connection? Does it really switch?

Numerous ways to answer that question. One is to use finer
wires to connect meter probes into switch wire connector.
Then monitor that voltage as switch is pressed and released.
Obviously that voltage will change in response to switch
pressed and released - if switch works.

Other way is to disconnect the switch cable from
motherboard. Then use meter's continuity tester (or put it in
low ohms) to see resistance drop to zero and go to maximum as
switch is pressed and released. Some meters have a continuity
setting so that it beeps when leads are shorted together;
makes it easier to detect continuity.

Third way is to disconnect switch from motherboard, monitor
voltage on that green wire, and then short the two motherboard
pins that connect to switch. If motherboard controller is
working, then that green wire voltage should change in
response to those shorted 'switch connector' pins.

Notice how quickly we isolate the problem to specific
components without doing all this removing and swapping power
supplies. A power supply tester cannot test power switch or
motherboard's power supply controller - two other components
of the power supply 'system'. Also an intermittent connection
or wire might not be detected by moving things too much.
Multimeter accomplishes a complete test without breaking any
wire connections and without accidentally 'fixing' an
intermittent connection.

Now let's assume the **green wire voltage does change**.
Monitor red, yellow, and orange wire voltages as power cord is
disconnected and reconnected, and then power switch is
pressed. You are looking for voltages that rise up to their
3.3, 5, or 12 for 0.4 to 2.0 seconds.

Again I reference those previous posts that noted this
test. Power supply turns on, then monitors those voltages.
If any one voltage does not appear, then all other voltages
are turned off internal power supply functions (an essential
function that is sometimes missing in those discount $25 and
$40 power supplies). If motherboard controller does tell
power supply to turn on (green wire voltage changes when power
switch is pressed), then which voltage (red, yellow, and
orange wire) does not turn on? More important information.

Procedure is stepping us right to the 'reason for failure' -
and still we avoided changes inside the computer and never
needed the power supply tester.

Patty wrote:
I got +4.99 on the purple. But I can't seem to get the others to
work. I can't get any reading with the power supply connected to
the motherboard, I think my probes are too big to fit in the
backside of the connector. I got the +4.99 with the motherboard
disconnected from the power supply and by putting the red probe in
the purple and the black probe in the closest black wire in the
end of the connector where it would attach to the motherboard.
Power supply plugged in and turned on, computer switched off,
of course.

Please forgive my questions as I am a novice at checking power
on a computer. Give me a wall socket and I'm ok. ;o)

Patty

On Sun, 30 Jan 2005 19:12:37 GMT, sbb78247 wrote:

56K internal??? My fax modem is a 28k external - you must really be up with
the times.

I just sold the majority of my old junk around Christmas to a guy that
wanted a system for his kid, P4/1.7, 80g hd, 845 based mobo, antec case,
52x cdrw, 512mb mem, basically a hotrod from 4 years ago. They were happy
and I got some coin out of the deal.

S

Oh, I wish you could see the boxes of junk I have. Would you believe an
old 120MB hard drive? Yep, it was top of the line in its day. I got a
386SX board in a box if anyone wants it. ;o) This computer I'm looking at
is turning out to be quite a dinosaur too. Just checked the hard drive.
Seagate Medalist Pro 2520 (2.5Gb) so, I think the ole 200Mhz might just
work out fine for her. I'm gonna stick a Diamond 4MB video card in it and
floppy drive and see if it will boot up to a floppy. Then we'll go from
there.

Darn, I gotta get rid of this old junk! ;o)

Patty

Previously noted was that a meter cannot harm power supply.
Why? Intel specs only repeat what was industry standard even
30 years ago. Take all outputs from the power supply. Short
them together. Power on supply. Power supply still cannot be
damaged. If you perform that test and suffer damage, then you
have identified a power supply that was defective when it left
the factory and was probably also missing other components
necessary for computer hardware safety.

That's right. Even a shorted power supply must never be
damaged as was defacto standard so many decades previous.

Don't worry about shorting PSU because shorts should cause
no damage to supply. Meter selected to DC volts also will not
be damaged. Just don't touch anything inside the AC power
cord or inside the case that says "Danger".

Worry more about your own body static electricity. Static
electricity is far more dangerous to computer.

Patty wrote:
Well, I went ahead and used the power supply tester. I had already
gotten +4.99V on the purple with just using the meter. With the
tester, I got +5.16V on the +5V, +11.92V on the +12V, and +3.4V on
the +3.3V. (Now that I think of it, I remember these types of
numbers when I used to have the power supply in my system and
checked it using MBM5.) Can I assume now that the power supply is
ok? Or should I try to test more wires by following your
instructions? I really don't want to short out my power supply,
I may need it in the future. It's an Enlight 300w that originally
came with my case (was replaced with an Enermax 350w.) I'm still
thinking it's the motherboard that has a problem since it didn't
work with its original power supply (which I don't have since the
person they took the computer to in the first place said it was
fried and kept it... yeah right) and this one is my own. I still
think two different power supplies, same result... dead
motherboard or CPU.

Being an owner of two Abit KT7 motherboards, I know all about bad
capacitors and how to recognize them. ;o)

Thanks much for your help.

Patty

Windows. Not something I'd want to leave up to a computer illiterate old
lady.

DELL swears they have free on-site tech support :-)
Seriously, you are going to regret this one !!!

johns

On Sun, 30 Jan 2005 14:36:29 -0500, w_tom wrote:

Let's assume that **green wire voltage does not drop**. Now
we suspect the other two power supply 'system' components.
First verify the power supply switch. Is it even making a
connection? Does it really switch?

Numerous ways to answer that question. One is to use finer
wires to connect meter probes into switch wire connector.
Then monitor that voltage as switch is pressed and released.
Obviously that voltage will change in response to switch
pressed and released - if switch works.


Switch turns on the cpu fan, case fan, and power led light. The power
supply fan also comes on when the computer is switched on. I think I'm
done fooling with this thing. I think the power supply is fine (it was
when I last used it) and everything else has checked out fine on it. I'm
still thinking it's a motherboard or cpu problem.

Had a similar problem once when I built a new system, nothing on power up
except fans. Turns out motherboard was shorting to case, a quick fix
solved that problem. Since this motherboard has been in the case for a few
years, I'm guessing something else is shorting out on it perhaps. For what
the equipment is (at least 5-6 years old), it's just not worth pursuing any
further.

Thanks for all your help. I'll print out the info you've given me and put
it in my toolbox for future reference. I'm happy I've learned something
about using a meter for testing. )

Patty

Patty wrote:
On Sun, 30 Jan 2005 19:12:37 GMT, sbb78247 wrote:

56K internal??? My fax modem is a 28k external - you must really be
up with the times.

I just sold the majority of my old junk around Christmas to a guy
that wanted a system for his kid, P4/1.7, 80g hd, 845 based mobo,
antec case, 52x cdrw, 512mb mem, basically a hotrod from 4 years
ago. They were happy and I got some coin out of the deal.

S

Oh, I wish you could see the boxes of junk I have. Would you believe
an old 120MB hard drive? Yep, it was top of the line in its day. I
got a 386SX board in a box if anyone wants it. ;o) This computer
I'm looking at is turning out to be quite a dinosaur too. Just
checked the hard drive. Seagate Medalist Pro 2520 (2.5Gb) so, I think
the ole 200Mhz might just work out fine for her. I'm gonna stick a
Diamond 4MB video card in it and floppy drive and see if it will boot
up to a floppy. Then we'll go from there.

Darn, I gotta get rid of this old junk! ;o)

Patty

Oh yea? Well top this! I have a couple of 286s up in the attic! one is a
blazing 12 mhz and the other is a 20! I think I still have isa video for
them and add in controller cards for the monsterous 40 mb hd! and if i look
around I think I still have some old 5-1/4 and possibly some 8" drives
around. Also, I think in the storage room at my parents house, they still
have an old IBM PC jr.! 8086 cpu and no ram to speak of!!!!!

Oh wait, I found an 8088 in my attic.

S