Can bad caps on a motherboard kill a PSU?


S

spodosaurus

Hi all,

In a recent thread here ("") I was testing out a new tool to quickly
diagnose whether a PCs issues were PSU related. Turns out the power
supply tester and the PSU were defective (behold the power and of the
multimeter). Anyway, The Antec TruePower 380 PSU is only about 26 months
old and seems a tad young to be dying seeing as my other Antec TruePower
380 is over four years old and running perfectly. The motherboard in the
system with the bad PSU is one of several Gigabyte boards I've owned
that have been affected by the bad capacitor issue. I'm wondering if the
buldging caps on the GA-7VTXE+ board could have caused the early demise
of the 5V rail on the PSU?

PSU readings on the 20 pin ATX connector:

Pins 1, 2, and 11 (3.3V): 3.28V (good)
Pins 4, 6, 19, and 20 (5V): 4.68V (bad)
Pin 9 (5VSB): 5.05V steady (good)
Pin 10 (12V): 11.66V (good)
Pin 12 (-12V): -12.58V (good)

Ari

--
spammage trappage: remove the underscores to reply
Many people around the world are waiting for a marrow transplant. Please
volunteer to be a marrow donor and literally save someone's life:
http://www.abmdr.org.au/
http://www.marrow.org/
 
Ad

Advertisements

S

SteveH

spodosaurus said:
Hi all,

In a recent thread here ("") I was testing out a new tool to quickly
diagnose whether a PCs issues were PSU related. Turns out the power
supply tester and the PSU were defective (behold the power and of the
multimeter). Anyway, The Antec TruePower 380 PSU is only about 26
months old and seems a tad young to be dying seeing as my other Antec
TruePower 380 is over four years old and running perfectly.

26 months isn't that old for an Antec PSU to be dying. I've had 4 pack up
that were less than 18 months old (all under guarantee of course). I gather
Antec did have problems a while back and posted on one of htese groups to
that effect.

SteveH
 
P

Paul

spodosaurus said:
Hi all,

In a recent thread here ("") I was testing out a new tool to quickly
diagnose whether a PCs issues were PSU related. Turns out the power
supply tester and the PSU were defective (behold the power and of the
multimeter). Anyway, The Antec TruePower 380 PSU is only about 26 months
old and seems a tad young to be dying seeing as my other Antec TruePower
380 is over four years old and running perfectly. The motherboard in the
system with the bad PSU is one of several Gigabyte boards I've owned
that have been affected by the bad capacitor issue. I'm wondering if the
buldging caps on the GA-7VTXE+ board could have caused the early demise
of the 5V rail on the PSU?

PSU readings on the 20 pin ATX connector:

Pins 1, 2, and 11 (3.3V): 3.28V (good)
Pins 4, 6, 19, and 20 (5V): 4.68V (bad)
Pin 9 (5VSB): 5.05V steady (good)
Pin 10 (12V): 11.66V (good)
Pin 12 (-12V): -12.58V (good)

Ari

(Your motherboard ?)
http://www.fmc.com.tw/product/gigamb/pic/7vtxe+.jpg

Does the voltage on the 5V rail, pop up to normal levels
if the supply is not powering that particular motherboard ?

It could be that the 4.68V level is there, because an extreme
amount of amps are being drawn. That is one reason I like to
have a clamp-on DC ammeter, to quickly check the current draw,
and detect an overload in progress. What you do with the clamp-on
ammeter, is open the jaws, and place the jaws around all the
5V wires at the same time. The clamp-on ammeter can sum the
current flow in all the wires, and give a total 5V ampere reading
flowing into the motherboard. And because the meter is non-contact,
you don't need to cut wires or anything. The meter uses Hall Probe
technology, and senses the magnetic field. The only caveat, is
limited bandwidth, and the inability to handle complex AC
waveforms properly (does a lousy job of measuring standby power
flowing in the AC line cord).

A 60W Athlon would be 5V @ 12A. The Vcore converter is 90% efficient
or less, so the amps need to be bumped to 12/0.90 = 13.3 amps. So,
if measuring the 5V into the board, I'd expect to see a bit more
current than that, when the board is running Prime95 or CPUBurn.
How much more than 13.3 amps, depends on what other stuff runs
off +5V.

This is my meter. The 40 amps DC range is good for motherboards.
I use the 400 amp DC range, to work on my car :) Normal multimeter
current ranges, seldom exceed about 10 amps. These are very handy
to have, and if your business has a budget for test equipment,
add one to the bench.

http://exphil.com/images/products/Extech/380947.jpg

The lifetime of a power supply, could be affected by the load
placed on it. Some of the Antec supplies are known to use a
poor brand of caps. All capacitors are sensitive to temperature,
and the higher the temp, the faster they dry out. Which is
one reason I am not happy to see the fan spin so slow on some
power supplies. Higher fan speed means lower internal PSU
temp and longer life.

Paul
 
N

nemo

Hi all,

In a recent thread here ("") I was testing out a new tool to quickly
diagnose whether a PCs issues were PSU related. Turns out the power
supply tester and the PSU were defective (behold the power and of the
multimeter). Anyway, The Antec TruePower 380 PSU is only about 26 months
old and seems a tad young to be dying seeing as my other Antec TruePower
380 is over four years old and running perfectly. The motherboard in the
system with the bad PSU is one of several Gigabyte boards I've owned
that have been affected by the bad capacitor issue. I'm wondering if the
buldging caps on the GA-7VTXE+ board could have caused the early demise
of the 5V rail on the PSU?

PSU readings on the 20 pin ATX connector:

Pins 1, 2, and 11 (3.3V): 3.28V (good)
Pins 4, 6, 19, and 20 (5V): 4.68V (bad)
Pin 9 (5VSB): 5.05V steady (good)
Pin 10 (12V): 11.66V (good)
Pin 12 (-12V): -12.58V (good)

It is entirely possible that the bad caps could have shorted. I
remember the capacitor problem since I had a mobo repaired under
warranty from that. In fact, they replaced all of the bulging caps
and none of the others, which I thought was odd. I figured it was
just a matter of time before the others went, but so far, so good. On
the other hand, my PSU also died after the machine started to have
trouble powering up. The symptom was that I could only get it to
power up by disconnecting other parts of the computer from the PSU.
Once the mobo would power up, I could plug them back in. I blew up a
CD driver (not the more expensive burner) when I couldn't see the
connector I was plugging into. Seems the polarization will prevent
the connectors from securely mating, but won't stop the contacts from
touching... OUCH!

I have wondered if my real problem is the mobo caps going bad and
drawing too much current from the PSU on startup. It is a 6 year old
machine, so I don't know if I am going to worry too much about it.

In your case, I don't know that excessive current draw could easily
pull the 5 volt rail out of spec. I guess it could be seeing
excessive voltage drop across the PSU wires and connectors, but that
would be a lot of current and the heat would have to go somewhere. Do
you have any hot components other than the CPU?

These PSUs don't start up without some load on them. But if you have
a way to hang a load on the 5 or 3.3 volt rails (I don't remember
which one I had to load to make it work) you can power it up and
measure the voltage open circuit. I seem to recall using the heating
element out of a 110 volt space heater. You don't even have to take
it out of the case, just use an alligator clip to give you a variable
tap point. It would draw a couple of amps from the 5 volt rail I
think and that would let me test all of the voltages open circuit.
See if your 5 volts comes back up under light/no load.
 
S

spodosaurus

Paul said:
(Your motherboard ?)
http://www.fmc.com.tw/product/gigamb/pic/7vtxe+.jpg

Yes


Does the voltage on the 5V rail, pop up to normal levels
if the supply is not powering that particular motherboard ?

No. Same if it's powering the power supply tester and IDE drives, or
just the paperclip jumper method + IDE drives.
A 60W Athlon would be 5V @ 12A.

Athlon XP 2400+
The lifetime of a power supply, could be affected by the load
placed on it. Some of the Antec supplies are known to use a
poor brand of caps. All capacitors are sensitive to temperature,
and the higher the temp, the faster they dry out. Which is
one reason I am not happy to see the fan spin so slow on some
power supplies. Higher fan speed means lower internal PSU
temp and longer life.

The fan in this PSU spins at a constant speed (I don't even think this
PSU has the fan throttling circuitry and the motherboard connector that
the TruePower II series has). The case also has a small rear exhaust fan
(60mm Vantec Stealth) so ventilation could be better.

Ari


--
spammage trappage: remove the underscores to reply
Many people around the world are waiting for a marrow transplant. Please
volunteer to be a marrow donor and literally save someone's life:
http://www.abmdr.org.au/
http://www.marrow.org/
 
P

Paul

spodosaurus said:
No. Same if it's powering the power supply tester and IDE drives, or
just the paperclip jumper method + IDE drives.


Athlon XP 2400+


The fan in this PSU spins at a constant speed (I don't even think this
PSU has the fan throttling circuitry and the motherboard connector that
the TruePower II series has). The case also has a small rear exhaust fan
(60mm Vantec Stealth) so ventilation could be better.

Ari

If the voltage had popped up, under no load, that would suggest that
the computer was loading the thing heavily. If the voltage is relatively
the same, with a load connected, or with no load, that tends to suggest
the supply for some reason, is always outputting on the low side.

4.68V is out of spec, so I'd replace it. Ideally, the supply should
have some loading on it, in case the design is such that it goes out
of regulation when the minimum current loading is not provided. But
since you've measure the voltage while in-circuit and got 4.68V,
then I'd change it out.

I do occasionally replace supplies, before they have a chance to fail.
On my first home build, I swapped out the supply when the fans started
to run erratically. A sign that the 12V was starting to "wobble".
That supply still works, but I won't be using it again. Using the
multimeter, I think it is still in spec - I just don't like to see
a supply that changes output voltage for no reason. (I.e. Fan speed
would change while sitting in the BIOS.)

Paul
 
Ad

Advertisements

L

larry moe 'n curly

spodosaurus said:
In a recent thread here ("") I was testing out a new tool to quickly
diagnose whether a PCs issues were PSU related. Turns out the power
supply tester and the PSU were defective (behold the power and of the
multimeter). Anyway, The Antec TruePower 380 PSU is only about 26 months
old and seems a tad young to be dying seeing as my other Antec TruePower
380 is over four years old and running perfectly. The motherboard in the
system with the bad PSU is one of several Gigabyte boards I've owned
that have been affected by the bad capacitor issue. I'm wondering if the
buldging caps on the GA-7VTXE+ board could have caused the early demise
of the 5V rail on the PSU?

More likely, the bad caps in the PSU caused the demise of its +5V
rail.

www.badcaps.net has lots of info about caps, and one of the brands
they dislike is Fuhjyyu, which are common in TruePowers and
SmartPowers. If you find one cap bulging, change all the others that
are of the same brand, color, and markings (except for voltage and
capacitor values) because they'll likely fail soon, too. With my 300W
Antec, it took just two weeks for a cap that tested and looked good to
fail after I replaced the bulging one next to it.
 
J

John Doe

....
www.badcaps.net has lots of info about caps, and one of the brands
they dislike is Fuhjyyu, which are common in TruePowers and
SmartPowers.

I don't believe that any major power supply manufacturer would
continue using bad capacitors, especially not when the manufacturer
caters to power users. It makes no sense, but maybe it makes for
good conspiracy theory.
If you find one cap bulging, change all the others that are of the
same brand, color, and markings (except for voltage and capacitor
values) because they'll likely fail soon, too. With my 300W Antec,
it took just two weeks for a cap that tested and looked good to fail
after I replaced the bulging one next to it.

All I've used are Antec SmartPowers and currently a TruePower II, for
gaming and with a relatively low wattage rating (230 to 380), and none
of them have failed.

When a product is very popular, there are going to be a larger number
of negative experiences. You have to work with percentages.
 
Ad

Advertisements

L

larry moe 'n curly

John said:
I don't believe that any major power supply manufacturer would
continue using bad capacitors, especially not when the manufacturer
caters to power users. It makes no sense, but maybe it makes for
good conspiracy theory.

But Channel Well Technology is still putting Fuhjyyus in those Antecs
and still running the cooling fans really slow for low noise and hot
capacitors.
All I've used are Antec SmartPowers and currently a TruePower II, for
gaming and with a relatively low wattage rating (230 to 380), and none
of them have failed.

My 300W developed bulging caps even at 60W (measured with a clamp-on
ammeter), and those caps were on the +12V rail, which in this
particular computer powered only the 5400 RPM HD and a CD writer.
When a product is very popular, there are going to be a larger number
of negative experiences. You have to work with percentages.

At Overclockers.com, HardOCP.com, or JonnyGuru.com, an Antec employee
(goes by AntecRep) said that there are no high failure rates for those
PSUs but there were defects, including a high voltage capacitor in the
standby +5V and bad solder on the newer, lead-free versions (labelled
"RoHS"). But many users there, including JonnyGuru himself, report
high failure rates due to Fuhjyyu capacitors.
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top