Hello all...

Let me start by saying that these are no ordinary computer power supplies.
They are Delta 335 watt supplies that come from IBM PS/2 Servers. They are
rather complicated inside. This is a long post, but please bear with me.

The mode of failure is that the supply just plain quits...mine have either
died silently while running or have not come back on after power failures.
There are no signs of distress in these supplies--they just quit running out
of the blue.

As a PS/2 collector and hobbyist who keeps his PS/2s in regular use, I find
this failure very troubling. These supplies are solidly built...it is hard
to imagine them just silently dying, but that is what happens. Conversation
with others who have these systems have revealed that this is not something
happening only to me.

I am desperately seeking assistance in troubleshooting these supplies. I
don't have the skills or comfort level to diagnose them. What I'm looking
for is someone who would be willing to have a look at these supplies and at
least provide a reason for their failure. If they can be fixed--that's
great--but it is not the primary goal. I want to know *why* they died.

Sam Goldwasser (whom I sincerely hope does not mind my mentioning his name)
looked at one of the supplies and was able to find out some information--if
my understanding is correct, he found that the standby supply was working
but the TL494 controller was being told to shut down with the "dead time
control" pin 4.

I'd really appreciate any help with these supplies. I'm desperate for help
and running out of places to look. I am more than willing to pay for your
time and shipping of the supply both ways.

If you can help, please post back here to the group or send an e-mail to:
wct atsign walshcomptech dot com. Thank you.

William

William R. Walsh wrote:
Hello all...

Let me start by saying that these are no ordinary computer power supplies.
They are Delta 335 watt supplies that come from IBM PS/2 Servers. They are
rather complicated inside. This is a long post, but please bear with me.

The mode of failure is that the supply just plain quits...mine have either
died silently while running or have not come back on after power failures.
There are no signs of distress in these supplies--they just quit running out
of the blue.

As a PS/2 collector and hobbyist who keeps his PS/2s in regular use, I find
this failure very troubling. These supplies are solidly built...it is hard
to imagine them just silently dying, but that is what happens. Conversation
with others who have these systems have revealed that this is not something
happening only to me.

I am desperately seeking assistance in troubleshooting these supplies. I
don't have the skills or comfort level to diagnose them. What I'm looking
for is someone who would be willing to have a look at these supplies and at
least provide a reason for their failure. If they can be fixed--that's
great--but it is not the primary goal. I want to know *why* they died.

Sam Goldwasser (whom I sincerely hope does not mind my mentioning his name)
looked at one of the supplies and was able to find out some information--if
my understanding is correct, he found that the standby supply was working
but the TL494 controller was being told to shut down with the "dead time
control" pin 4.

I'd really appreciate any help with these supplies. I'm desperate for help
and running out of places to look. I am more than willing to pay for your
time and shipping of the supply both ways.

If you can help, please post back here to the group or send an e-mail to:
wct atsign walshcomptech dot com. Thank you.

William


Well, it would take a fair amount of time to do it right.
If i were to guess, bad solder joints and capacitors are
the primary suspects. The bad solder joints are pretty
straight forward if you know what to look for.
The caps would take a ESR meter at minimum to check for
series resistance. This might involve taking them out of
circuit in some cases.

If one was going deep into this, you would have to reverse
engineer a schematic along with data sheets to figure out
what's really going on inside. And that's highly time
consuming. You got the funds to to that?

There is always the possibility that there is some latent failure
of the power transformers or such. Those parts will NOT be
readily available if they can be found at all. A 494 is a pretty
common switcher control IC. Used commonly in car stereo amps.

Bob

----== Posted via Newsfeeds.Com - Unlimited-Uncensored-Secure Usenet News==----
http://www.newsfeeds.com The #1 Newsgroup Service in the World! 120,000+ Newsgroups
----= East and West-Coast Server Farms - Total Privacy via Encryption =----

On Sun, 23 Oct 2005 05:38:34 GMT, "William R. Walsh"
wrote:

Hello all...


494 IC is found in some generic LCD for lamp inverter.

Cheers, Wizard

Hi!

If one was going deep into this, you would have to reverse
engineer a schematic along with data sheets to figure out
what's really going on inside. And that's highly time
consuming. You got the funds to to that?

Please define "the funds to do that". Hundreds? Thousands?

The way these supplies seem to be dropping out, I may spend the money one of
the two ways.

There is always the possibility that there is some latent failure
of the power transformers or such. Those parts will NOT be
readily available if they can be found at all. A 494 is a pretty
common switcher control IC. Used commonly in car stereo amps.

I just don't know where the trouble lies. These supplies die silently and so
far there is only one documented case of one actually blowing a
transformer...which was a rather smoky, fuse-blowing event as I'm told.

More than anything else, I really want to know what is killing them. I have
many other PS/2 systems operating with other supplies that don't run in
anywhere near as pleasant conditions. Many don't cool themselves as well as
these Delta 335/400 watt units do, nor do they have as much output power. As
an example, I have some 94 watt supplies used in the PS/2 Model 50(z) that
have traces which have delaminated from the circuit board and even some sign
of having been *very* hot at one time or another. These also operate under
rather heavy load with fancy video cards, 386 upgrades, memory cards and
network cards. Yet not a one of these has ever given up and quit working,
even under the odd bad set of line conditions.

William

More than anything else, I really want to know what is killing them. I have
many other PS/2 systems operating with other supplies that don't run in
anywhere near as pleasant conditions. Many don't cool themselves as well as
these Delta 335/400 watt units do, nor do they have as much output power. As
an example, I have some 94 watt supplies used in the PS/2 Model 50(z) that
have traces which have delaminated from the circuit board and even some sign
of having been *very* hot at one time or another. These also operate under
rather heavy load with fancy video cards, 386 upgrades, memory cards and
network cards. Yet not a one of these has ever given up and quit working,
even under the odd bad set of line conditions.

William




The usual killers of switchmode supplies aside from the fan failing and
overheating it are cracked solder joints and dried up electrolytic
capacitors. Probably 90% of failures are due to one or more of those
initial problems.

"William R. Walsh" ) writes:
Hi!

If one was going deep into this, you would have to reverse
engineer a schematic along with data sheets to figure out
what's really going on inside. And that's highly time
consuming. You got the funds to to that?

Please define "the funds to do that". Hundreds? Thousands?

Read time=money. IN other words, can you afford the time to trace
out the schematic?

The compounding issue is that there is nothing that requires the power
supplies to use the same design. So you trace one, and when the
matter comes up again, you may have to trace again. This is not
the same as tracing a piece of consumer equipment that you will come
across multiple times but the circuitry will remain the same, or you
can trade your finished schematic to others to get a higher return
on the work.

Michael

The way these supplies seem to be dropping out, I may spend the money one of
the two ways.

There is always the possibility that there is some latent failure
of the power transformers or such. Those parts will NOT be
readily available if they can be found at all. A 494 is a pretty
common switcher control IC. Used commonly in car stereo amps.

I just don't know where the trouble lies. These supplies die silently and so
far there is only one documented case of one actually blowing a
transformer...which was a rather smoky, fuse-blowing event as I'm told.

More than anything else, I really want to know what is killing them. I have
many other PS/2 systems operating with other supplies that don't run in
anywhere near as pleasant conditions. Many don't cool themselves as well as
these Delta 335/400 watt units do, nor do they have as much output power. As
an example, I have some 94 watt supplies used in the PS/2 Model 50(z) that
have traces which have delaminated from the circuit board and even some sign
of having been *very* hot at one time or another. These also operate under
rather heavy load with fancy video cards, 386 upgrades, memory cards and
network cards. Yet not a one of these has ever given up and quit working,
even under the odd bad set of line conditions.

William

Hi!

Read time=money. IN other words, can you afford the time to trace
out the schematic?

I don't know...depends upon what people are asking for their time and how
badly I really want to know why these things died. (And I want to know that
pretty badly. Obviously any proposal to look at it has to be
reasonable...but beyond that...I just want to know why they died.)

The compounding issue is that there is nothing that requires the power
supplies to use the same design.

For most power supplies I'd agree with you, at least to an extent. However,
the PS/2 Model 95 line used this same Delta 335 watt supply across almost
the entire model line. I've seen a lot of them, and all are the same inside.
Even the 400 watt unit is largely the same inside...key differences I see?
The fan speed on those varies in operation, circuit boards are a different
color and the heatsinks might be a bit thicker, but that was all I noticed
right away. The point is that these were a custom-built, standardized
design. As far as I can tell they did not change the design much--if at
all--during the lifetime of the PS/2 Model 95.

I've only ever seen one that was *not* made by Delta Electronics.

I might disagree with you on the frequency that this scenario will occur. I
know others who have power supplies just like this that have died. This is
what is so alarming to me...most of these machines just run and run. The
Model 95 and its 335 watt supply seem to be the exception to this rule. I've
got at least two or three here now that have all failed with the same
symptoms--they're just dead.

William

On Wed, 26 Oct 2005 04:57:27 GMT, "William R. Walsh"
put finger to keyboard and
composed:

Read time=money. IN other words, can you afford the time to trace
out the schematic?

I don't know...depends upon what people are asking for their time and how
badly I really want to know why these things died. (And I want to know that
pretty badly. Obviously any proposal to look at it has to be
reasonable...but beyond that...I just want to know why they died.)

I'd be keen to have a look at one, but I'm on the other side of the
world. Judging by Sam G's diagnosis, I expect that the PSU is either
shutting down for reasons of overvoltage or overcurrent. IME the
overvoltage protection is designed to latch on so that it can only be
reset by removing power from the PSU. Being a latch, the OV condition
should be visible on a DMM after the event has occurred. I don't know
if an overcurrent condition would be similarly trapped though.

If the fault is due to an OV, then I'd be looking around pins 1 and 2
of the TL494. Perhaps one of the feedback resistors has gone open???
Could there be a dirty trimpot???

If it's an overcurrent fault, then I'd be looking at shorted or leaky
diodes on the load side or, in the worst case, a shorted turn in the
switchmode transformer. Maybe a leaky chopper transistor could behave
this way, too???

-- Franc Zabkar

Please remove one 'i' from my address when replying by email.

Hi!

I'd be keen to have a look at one, but I'm on the other side of the
world.

Well, I could try to do that. These PSUs are fairly heavy (probably 15,
20 pounds) so shipping wouldn't be cheap...but given some time to get
the $ I could likely send it your way.

I can be contacted off the group using wct atsign walshcomptech com
if you'd like to let me know where in the world you are so I can find
out how bad the shipping would be.

William

"Franc Zabkar" bravely wrote to "All" (26 Oct 05 17:29:05)
--- on the heady topic of " Help Needed With PC Switchmode Power Supplies"

FZ From: Franc Zabkar
FZ Xref: core-easynews sci.electronics.repair:346288

FZ On Wed, 26 Oct 2005 04:57:27 GMT, "William R. Walsh"
FZ put finger to keyboard and
FZ composed:

Read time=money. IN other words, can you afford the time to trace
out the schematic?

I don't know...depends upon what people are asking for their time and how
badly I really want to know why these things died. (And I want to know that
pretty badly. Obviously any proposal to look at it has to be
reasonable...but beyond that...I just want to know why they died.)

FZ I'd be keen to have a look at one, but I'm on the other side of the
FZ world. Judging by Sam G's diagnosis, I expect that the PSU is either
FZ shutting down for reasons of overvoltage or overcurrent. IME the
FZ overvoltage protection is designed to latch on so that it can only be
FZ reset by removing power from the PSU. Being a latch, the OV condition
FZ should be visible on a DMM after the event has occurred. I don't know
FZ if an overcurrent condition would be similarly trapped though.

FZ If the fault is due to an OV, then I'd be looking around pins 1 and 2
FZ of the TL494. Perhaps one of the feedback resistors has gone open???
FZ Could there be a dirty trimpot???

FZ If it's an overcurrent fault, then I'd be looking at shorted or leaky
FZ diodes on the load side or, in the worst case, a shorted turn in the
FZ switchmode transformer. Maybe a leaky chopper transistor could behave
FZ this way, too???


Franc,

The high value base bias resistors on the output transistors have a
habit of going open circuit. I suspect it is due to line transients.
These are now typically thin film types with a lower voltage tolerance
than the old carbon type. Replace these resistors by 2 of equivalent
series value to double the voltage tolerance. Be sure to mount them
standing off the circuit board by at least 1 diameter. Perhaps the
better remedy for transients is a power line filter between the line
socket and the fuse. Some mfgrs skimp on the line filter while others
include the fancy shielded modular type with the line socket built in.

A*s*i*m*o*v

.... Wasted power is current squared times the resistance.

On Wed, 26 Oct 2005 14:02:33 GMT, "Asimov"
put finger to
keyboard and composed:

"Franc Zabkar" bravely wrote to "All" (26 Oct 05 17:29:05)
--- on the heady topic of " Help Needed With PC Switchmode Power Supplies"

FZ From: Franc Zabkar
FZ Xref: core-easynews sci.electronics.repair:346288

FZ On Wed, 26 Oct 2005 04:57:27 GMT, "William R. Walsh"
FZ put finger to keyboard and
FZ composed:

Read time=money. IN other words, can you afford the time to trace
out the schematic?

I don't know...depends upon what people are asking for their time and how
badly I really want to know why these things died. (And I want to know that
pretty badly. Obviously any proposal to look at it has to be
reasonable...but beyond that...I just want to know why they died.)

FZ I'd be keen to have a look at one, but I'm on the other side of the
FZ world. Judging by Sam G's diagnosis, I expect that the PSU is either
FZ shutting down for reasons of overvoltage or overcurrent. IME the
FZ overvoltage protection is designed to latch on so that it can only be
FZ reset by removing power from the PSU. Being a latch, the OV condition
FZ should be visible on a DMM after the event has occurred. I don't know
FZ if an overcurrent condition would be similarly trapped though.

FZ If the fault is due to an OV, then I'd be looking around pins 1 and 2
FZ of the TL494. Perhaps one of the feedback resistors has gone open???
FZ Could there be a dirty trimpot???

FZ If it's an overcurrent fault, then I'd be looking at shorted or leaky
FZ diodes on the load side or, in the worst case, a shorted turn in the
FZ switchmode transformer. Maybe a leaky chopper transistor could behave
FZ this way, too???


Franc,

The high value base bias resistors on the output transistors have a
habit of going open circuit. I suspect it is due to line transients.
These are now typically thin film types with a lower voltage tolerance
than the old carbon type. Replace these resistors by 2 of equivalent
series value to double the voltage tolerance. Be sure to mount them
standing off the circuit board by at least 1 diameter. Perhaps the
better remedy for transients is a power line filter between the line
socket and the fuse. Some mfgrs skimp on the line filter while others
include the fancy shielded modular type with the line socket built in.

Thanks for the info. I think the OP should now be able to find a local
person to follow up on your advice. Perhaps Sam could have another
look?

A*s*i*m*o*v

... Wasted power is current squared times the resistance.

-- Franc Zabkar

Please remove one 'i' from my address when replying by email.

Hi!

Thanks for the info. I think the OP should now be able to find a local
person to follow up on your advice.

I found some other repair shops that might be willing to look at it.
For now I have to see about contacting them.

Local repair shops...let's not even go there. I'm more than a little
disheartened with the seeming sad state of repair shops these days...

Perhaps Sam could have another look?

Sam offered to have another look but said that he wasn't sure when he
might be able to get to it, if ever. I'm very glad he was willing to
look at it to begin with and feel that bothering him further about it
might not be the most appropriate thing to do. I'd reckon that Sam has
a busy enough life without my bothering him. :-)

William

In message .com,
writes
Sam offered to have another look but said that he wasn't sure when he
might be able to get to it, if ever. I'm very glad he was willing to
look at it to begin with and feel that bothering him further about it
might not be the most appropriate thing to do. I'd reckon that Sam has
a busy enough life without my bothering him. :-)

William

Found the 92F0051 for $55, is this more than you would pay someone to
repair one?
http://www.acsparts.com/92F0051-Desktop.html
--
Clint Sharp

Hi!

Found the 92F0051 for $55, is this more than you would pay someone to
repair one?
http://www.acsparts.com/92F0051-Desktop.html

Buying them from resellers has been a hit and miss for me. I bought a few of
them recently and three of four were dead in the box! (Well, okay, only one
was *truly* dead. Another was intermittent under load and the third one had
been very hot at the power plugs. That might be fixable.)

Even if they do run when I get them, the chance seems very good that they
will be dead soon. So I see three things here--one, it is getting very
expensive to buy them, even at $15 or $20 per. Every one I have to buy
pushes the "acceptable margin" of a repair cost that much higher. Second,
there's no sign of catastrophic failure--so *I* think the odds of whatever
is going wrong being fixable must be pretty good. Finally, it's
environmentally irresponsible to just throw them away and move on to a new
one.

At the rate I'm going, $200 or $300 for someone to fix one wouldn't be
entirely unreasonable. So the answer to your question is "no"...I'd
certainly pay more to know why they died or to have one repaired.

William

On Sun, 23 Oct 2005 05:38:34 GMT "William R. Walsh"
wrote:

I am desperately seeking assistance in troubleshooting these supplies. I
don't have the skills or comfort level to diagnose them. What I'm looking
for is someone who would be willing to have a look at these supplies and at
least provide a reason for their failure. If they can be fixed--that's
great--but it is not the primary goal. I want to know *why* they died.

The two most common problems that I have seen are bad capacitors (high
ESR) or open startup resistors. That should not be too hard to track
down.

Where are you located? If you're not in a big rush, I might be willing
to take a look at one. How much do these weigh?

-
-----------------------------------------------
Jim Adney
Madison, WI 53711 USA
-----------------------------------------------

Hi!

Where are you located? If you're not in a big rush, I might be willing
to take a look at one. How much do these weigh?

I'm not in any big hurry. I'm located in eastern Central Illinois. I'm not
sure of an exact weight for these power supplies--would guess about 10~15
pounds.

If you'd like to, please send me an e-mail:
wct atsign walshcomptech dot com

William