Common SMPS failure mode?

[Daniel Haude]

Hey folks,

first off, this post is somewhat OT because it isn't a design
but rather a repair question.

I've got this old Sun SPARC computer that's probably worth no more than
EUR50, and that only because it's in a nice 19" case. Its power supply
broke, and a new one costs EUR500. Unfortunately, this computer is part of
a EUR50000,- laboratory setup, so it's indispensable, and the power supply
will have to be fixed.

Now while it's financially no problem to just buy a new PS (which is what
I'm going to do anyway because there's a time factor involved), the
thought of trashing a EUR500,- unit in which probably only a cheapo part
needs to be replaced bugs me. However, I think that SMPSes are generally
not worth repairing -- or downright unrepairable. I'm posting here in the
hope that someone will pipe up and say: Oh, if this-and-this happened it's
always this-and-this which'll have to be replaced.

OK, the this-and-this that happened is that the P/S failed sometime at
night while the computer was running. In the morning we found the breaker
had tripped. We turned the breaker back on, turned on the computer, and
the breaker tripped again. After four more attempts the internal fuse of
the SMPS blew.

I inspected the SMPS and found no exploded parts, and no exploded-parts
smell. Is this thing worth looking at, and what should I look at first?

Thanks,
--Daniel

 

[Barry Lennox]

Hey folks,

snip

I inspected the SMPS and found no exploded parts, and no exploded-parts
smell. Is this thing worth looking at, and what should I look at first?

Thanks,
--Daniel

Based on my limited experience of about 4-5 SMPS and a tech I know who
fixes them for a living, about 95% of the faults are diodes or the
main switching transistor (or FET) Grab a meter and check them one by
one.
Barry Lennox

 

[James Meyer]

OK, the this-and-this that happened is that the P/S failed sometime at
night while the computer was running. In the morning we found the breaker
had tripped. We turned the breaker back on, turned on the computer, and
the breaker tripped again. After four more attempts the internal fuse of
the SMPS blew.

I inspected the SMPS and found no exploded parts, and no exploded-parts
smell. Is this thing worth looking at, and what should I look at first?

Thanks,
--Daniel

The first thing you should look for is someone with a little more
experience than you have. I'm sure there is someone available locally who can
take a look at the supply and fix it.

It took you four times tripping the main breaker and then the PS fuse
before you figured out that something was wrong?

Jim

 

[Jim Yanik]

Hey folks,

first off, this post is somewhat OT because it isn't a design
but rather a repair question.

I've got this old Sun SPARC computer that's probably worth no more
than EUR50, and that only because it's in a nice 19" case. Its power
supply broke, and a new one costs EUR500. Unfortunately, this computer
is part of a EUR50000,- laboratory setup, so it's indispensable, and
the power supply will have to be fixed.

Now while it's financially no problem to just buy a new PS (which is
what I'm going to do anyway because there's a time factor involved),
the thought of trashing a EUR500,- unit in which probably only a
cheapo part needs to be replaced bugs me. However, I think that SMPSes
are generally not worth repairing -- or downright unrepairable. I'm
posting here in the hope that someone will pipe up and say: Oh, if
this-and-this happened it's always this-and-this which'll have to be
replaced.

OK, the this-and-this that happened is that the P/S failed sometime at
night while the computer was running. In the morning we found the
breaker had tripped. We turned the breaker back on, turned on the
computer, and the breaker tripped again. After four more attempts the
internal fuse of the SMPS blew.

I inspected the SMPS and found no exploded parts, and no
exploded-parts smell. Is this thing worth looking at, and what should
I look at first?

Thanks,
--Daniel

Electrolytic caps with too much ESR,shorted diodes,open or shorted
switching transistors.Many switchers have current-limit circuits,so if a
diode shorted,the supply would not start,or try to start,then shut down.
Blowing fuses would possibly indicate a primary fault such as open and/or
shorted switcher xstrs or diodes.
Bad electrolytics on the primary might mean not enough input line DC for
the switcher,on the secondary would mean too much of a load causing
shutdown.
Ohm out the primary switcher xstrs and diodes,use an ESR meter to check
electrolytics.

 

[Spehro Pefhany]

I inspected the SMPS and found no exploded parts, and no exploded-parts
smell. Is this thing worth looking at, and what should I look at first?

Google for the SMPS repair FAQ, it's got a lot of good information.

You might also want to look for a used on on Ebay. I see a Sparc
Station 20 PSU with a BIN of $9.99. and a 440W SPARC 1000 Power
Supply, also starting at $10 (+ a lot of shipping).

Best regards,
Spehro Pefhany

 

[R.Legg]

OK, the this-and-this that happened is that the P/S failed sometime at

night while the computer was running. In the morning we found the breaker
had tripped. We turned the breaker back on, turned on the computer, and
the breaker tripped again. After four more attempts the internal fuse of
the SMPS blew.

I inspected the SMPS and found no exploded parts, and no exploded-parts
smell. Is this thing worth looking at, and what should I look at first?

The responsible diagnostic procedure, when dealing with a tripped
breaker or a blown fuse, in any device, is to reset or replace the
fault limiting device ONCE(!).

Repeated reseting or replacement, after the first confirmation, merely
adds unnecessarily to any damage that was present.

If there had been one typical failure to check for initially, you have
succeeded in developing a completely new train of damaged parts that
may be unrelated to the first symptom. The lack of exploded parts and
fire is an indication of a well-designed and safe product that was
originally intended to be repairable under repeated single-fault
abnormal conditions.

A tripping breaker is the last symptom of a chain of events, that
could have started anywhere in the product, and could be the
culmination of a number of discrete, non-critical failure mechanisms
accumulated over the entire life of the product. Now that fuses are
additionally popping, this is a sign that a hard semiconductor failure
has been achieved in the input rectifier or converter section, as
well, through your efforts.

If and when you send it out to qualified service personnel for repair,
you might save yourself time and money by describing fully the latter
events. This will alter their estimate of any costs involved, and may
affect your decision re repair or replacement.

RL

 

[Eric Y. Chang]

www.repairfaq.org for some excellent repair tips
www.tinaja.com for the correct way to make ground referenced measurements

 

[Frithiof Andreas Jensen]

On 12 Aug 2003 10:13:24 GMT, the renowned Daniel Haude

You might also want to look for a used on on Ebay. I see a Sparc
Station 20 PSU with a BIN of $9.99. and a 440W SPARC 1000 Power
Supply, also starting at $10 (+ a lot of shipping).

....Which you can then sell to the desparate customer at the BARGAIN PRICE of
USD 375.99 ;-)

BTW: Auctioning off the *defect* one on Ebay will probably pay for the
shipping since Ebay does not give a rat's arse about complaints!

 

[Daniel Haude]

On 12 Aug 2003 09:55:25 -0700,

The responsible diagnostic procedure, when dealing with a tripped
breaker or a blown fuse, in any device, is to reset or replace the
fault limiting device ONCE(!).

Repeated reseting or replacement, after the first confirmation, merely
adds unnecessarily to any damage that was present.

Are you suggesting I did something wrong?

--Daniel

 

[Watson A.Name - 'Watt Sun']

Hey folks,

first off, this post is somewhat OT because it isn't a design
but rather a repair question.

I've got this old Sun SPARC computer that's probably worth no more than
EUR50, and that only because it's in a nice 19" case. Its power supply
broke, and a new one costs EUR500. Unfortunately, this computer is part of
a EUR50000,- laboratory setup, so it's indispensable, and the power supply
will have to be fixed.

Now while it's financially no problem to just buy a new PS (which is what
I'm going to do anyway because there's a time factor involved), the
thought of trashing a EUR500,- unit in which probably only a cheapo part
needs to be replaced bugs me. However, I think that SMPSes are generally
not worth repairing -- or downright unrepairable. I'm posting here in the
hope that someone will pipe up and say: Oh, if this-and-this happened it's
always this-and-this which'll have to be replaced.

OK, the this-and-this that happened is that the P/S failed sometime at
night while the computer was running. In the morning we found the breaker
had tripped. We turned the breaker back on, turned on the computer, and
the breaker tripped again. After four more attempts the internal fuse of
the SMPS blew.

I inspected the SMPS and found no exploded parts, and no exploded-parts
smell. Is this thing worth looking at, and what should I look at first?

Often the fan freezes up, then things overheat and the heat destroys
something. But if you have a hard short, the rectifier or other
components very close to the incoming AC line are usually where the
problem is. Could be a bad FW bridge, or even a bad capacitor or
inductor. Your ohmmeter can tell you what's shorted.

Thanks,
--Daniel

--
@@F@r@o@m@@O@r@a@n@g@e@@C@o@u@n@t@y@,@@C@a@l@,@@w@h@e@r@e@@
###Got a Question about ELECTRONICS? Check HERE First:###
http://users.pandora.be/educypedia/electronics/databank.htm
My email address is whitelisted. *All* email sent to it
goes directly to the trash unless you add NOSPAM in the
Subject: line with other stuff. alondra101 <at> hotmail.com
Don't be ripped off by the big book dealers. Go to the URL
that will give you a choice and save you money(up to half).
http://www.everybookstore.com You'll be glad you did!
Just when you thought you had all this figured out, the gov't
changed it: http://physics.nist.gov/cuu/Units/binary.html
@@t@h@e@@a@f@f@l@u@e@n@t@@m@e@e@t@@t@h@e@@E@f@f@l@u@e@n@t@@

 

[Daniel Haude]

On Tue, 12 Aug 2003 05:03:28 -0700,

Perhaps this:

http://focus.ti.com/lit/an/slva085/slva085.pdf

Might be of interest and relevant to you.

It is. Very nice paper!

BTW, thanks to everybody for the suggestions. I've indeed found a pair of
dead IRF840 switching transistors in the main 5V branch. The +/-12V
branch, which runs independently off the same primary DV voltage, works
fine. Now I'm just waiting for the new transistors. It looks as if I could
turn a EUR750 purchase (the 500 turned out to be bogus) into a EUR5
repair.

BTW, when testing the supply I looked at the voltage across the main
storage cap while slowly turning up the AC input voltage. At first the cap
voltage followed the input, but upon reaching about 50V AC input, the cap
voltage suddenly jumped to 400V (accompanied by a short 2A input current
surge) and stayed there throughout the input voltage range. Is this a PFC
circuit at work?

--Daniel

 

[Daniel Haude]

On Wed, 13 Aug 2003 16:44:05 +0000 (UTC),

So, there's no shame in using the light bulb, even though some older
designers will sneer "chicken".

Quite true. I omitted the lightbulb, relying on the 2A breaker and the
ammeter in the isolated variac. Of course that doesn't replace a real
current limiter like a resitor or lightbulb, but it looks as if I got away
with it. Do you think you could keep me out of your "serious idiot" box,
considering that I used an isolation variac instead of plugging the supply
straight into mains?

Thanks,
--Daniel

 

[Frithiof Andreas Jensen]

Daniel Haude ([email protected]) wrote:

So, there's no shame in using the light bulb, even though some older
designers will sneer "chicken".

Of Course We Do - Think of the sheer entertainment value, the tense moments
before the switch is thrown, the thrill of the sight and smell of a new
innoncent engineers finest work creating a neat fireball on the workshop
desk acompanied by a muffled "BOOM" from the sub-basement .......
Ahhh.......makes it all worthwhile, does it not!!

 

[R.Legg]

Of Course We Do - Think of the sheer entertainment value, the tense moments
before the switch is thrown, the thrill of the sight and smell of a new
innoncent engineers finest work creating a neat fireball on the workshop
desk acompanied by a muffled "BOOM" from the sub-basement .......
Ahhh.......makes it all worthwhile, does it not!!

Safety Glasses, gentlemen, please.

Tinted lenses might also avoid Zoro-esque permanent markings on the
backs of your retinas.

RL

 

[Da Man]

Based on my limited experience of about 4-5 SMPS and a tech I know who
fixes them for a living, about 95% of the faults are diodes or the
main switching transistor (or FET) Grab a meter and check them one by
one.

Don't forget to look for dried out capacitors - that's another big one, but
they don't short usually.

 

[N. Thornton]

Do you think you could keep me out of your "serious idiot" box,

considering that I used an isolation variac instead of plugging the supply
straight into mains?

A variac is not an isolating transformer. Try not to fry yourself.


Regards, NT

 

[Daniel Haude]

On Sun, 17 Aug 2003 14:16:04 +0200,

He said "isolation variac". I have one on my desk, isolation transformer and
variac in one.
Hopefully that is what he really has...

Don't worry. It is. Still charges the storage cap to 400V, isolated or
not.

--Daniel