My son's 3-month old desktop computer failed and upon troubleshooting it, I
found that the thermistor (TH1) that is in series with a 6.3A glass fuse and
located in the 120VAC input power feed inside the power supply module had
self-destructed. I tested the soldered in fuse and it was still good, so
assume that it did not fail due to an over current situation. To determine
conclusively that the failure did not result from a PSU secondary problem, I
temporarily bypassed (shorted) the blown device with a jumper. When I did
this the PSU and computer came up normally so I conclude that this is an
isolated failure, possibly due to an under-designed component in a cheap
imported PSU that came in the tower unit he built. (I believe it is a 600W
PSU and the computer is pretty lightly loaded without many peripherals, so
that is why I suspect it just being infant mortality, or possibly an
under-designed part in this design).

I know that the purpose of the device is to limit the start up inrush
current to the PSU and protect down stream devices, so I definitely ant to
replace it. Unfortunately, I'm not sure yet what the local Radio Shack will
stock and wanted to know in advance what generic substitutes I might be able
to use for this purpose.

Basically, what I am asking is how do I read the 5D-9 value that was stamped
on the old part? Is there a current rating and if so was the old part
consistent with the 6.3A fuse or was it undersized and that is why it failed
after just 3 months of use? Does the "5" in the first part of the rating
indicate the resistance in ohms and if so, is this the steady state value or
at the start up/inrush?

Knowledge on the part nomenclature, reliability, and use in a SMPS is
desired. Suggestions on suitable replacement parts that will prevent this
from occurring again are welcome. Thanks in advance.

Bob S.

Bob Shuman wrote:
Basically, what I am asking is how do I read the 5D-9 value that was
stamped on the old part? Is there a current rating and if so was the
old part consistent with the 6.3A fuse or was it undersized and that
is why it failed after just 3 months of use? Does the "5" in the
first part of the rating indicate the resistance in ohms and if so,
is this the steady state value or at the start up/inrush?

http://www.skceramic.com/pro-ntc_power_kr.htm

Thanks for the spec sheet, but I'm not sure I can find everything I was
looking for here on that sheet or another I also found via google search.
The sheet clearly answers one question: the "5" in the 5D-9 part identifier
is the resistance in ohms at 25 degrees Centigrade (I assume this is the
startup resistance and that it decreases as the temperature increases due to
current passing through it). If I read the table at the end, it appears
that the steady state "hot" resistance drops to 0.18 ohms, but at what
current, steady state temperature, and how quickly?

More importantly, if I wanted to beef up this part to prevent this from
happening again, what generic parameters should I be looking for in a
thermistor at the local RS store? The 5 ohms and .18 ohm are a good place
to start, but I did not see any type of current limit or power rating
(assuming the .18 ohms) for the device.

Additional answers/advice would be appreciated.

Bob

"Travis Jordan" wrote in message
...
Bob Shuman wrote:
Basically, what I am asking is how do I read the 5D-9 value that was
stamped on the old part? Is there a current rating and if so was the
old part consistent with the 6.3A fuse or was it undersized and that
is why it failed after just 3 months of use? Does the "5" in the
first part of the rating indicate the resistance in ohms and if so,
is this the steady state value or at the start up/inrush?

http://www.skceramic.com/pro-ntc_power_kr.htm

Bob Shuman wrote:

More importantly, if I wanted to beef up this part to prevent this
from happening again, what generic parameters should I be looking for
in a thermistor at the local RS store? The 5 ohms and .18 ohm are a
good place to start, but I did not see any type of current limit or
power rating (assuming the .18 ohms) for the device.

If it is a 600W non-PFC power supply we might assume at 80% efficiency
the input power to be somewhere around 750W.

750/120 = 6.25A (sound familiar) - full power operation.

Per Ohm's Law: 5 ohm starting resistance = 31W if starting inrush is
100% of full power load (not likely).
..18 ohm ending resistance = 1.125W at full power.

Note that the inrush current limit for the 5D-9 is only 3A, so either
you don't have a 600W supply or (most probably) the max starting surge
is calculated at 50% of the full power load or perhaps as you stated the
device is underrated. You decide.

Travis,

Thanks again. Yes, I can do the math and understand ohm's law... and your
75% efficiency number seems about right (I assumed 80%) and consistent with
the fuse sizing, but the piece I missed in the spec sheet is the maximum 3A
in rush current. Where was this specified?

I think you've hit upon why the part exploded and burned up. It appears it
was indeed undersized for the application. So from their listing, which
part has the higher in rush current limit? Again, I did not see this
specified in the part list table. Thanks in advance.

Bob S

"Travis Jordan" wrote in message
m...
If it is a 600W non-PFC power supply we might assume at 80% efficiency
the input power to be somewhere around 750W.

750/120 = 6.25A (sound familiar) - full power operation.

Per Ohm's Law: 5 ohm starting resistance = 31W if starting inrush is
100% of full power load (not likely).
.18 ohm ending resistance = 1.125W at full power.

Note that the inrush current limit for the 5D-9 is only 3A, so either
you don't have a 600W supply or (most probably) the max starting surge
is calculated at 50% of the full power load or perhaps as you stated the
device is underrated. You decide.

Bob Shuman wrote:
...... the piece I missed in the spec
sheet is the maximum 3A in rush current. Where was this specified?

http://www.skceramic.com/pro-ntc_power_kr.htm
Column labeled 'at 25ºC (A)' - 3A
25ºC is the starting temperature and thus the 3A must be the inrush
current, or at least this is the way I read it.

I think you've hit upon why the part exploded and burned up. It
appears it was indeed undersized for the application. So from their
listing, which part has the higher in rush current limit? Again, I
did not see this specified in the part list table. Thanks in advance.

I would replace it with a NTC 5D-18 which has a 7A current rating.

Travis Jordan wrote:
I would replace it with a NTC 5D-18 which has a 7A current rating.

Oops - my error - you want the same terminal resistance so use a NTC
8D-18 (8A).

Travis Jordan wrote:
Travis Jordan wrote:
I would replace it with a NTC 5D-18 which has a 7A current rating.

Oops - my error - you want the same terminal resistance so use a NTC
8D-18 (8A).

Or a Thermometrics CL40 should work too.

http://www.thermometrics.com/assets/images/cl.pdf

http://www.mouser.com/index.cfm?&han...*&N=0&crc=true

Thanks again for taking the time to help educate me. From your explanation
of the table (which makes perfect sense), I think the 5D-15 (6A in rush)
seems to double the original current limit, but retains the 5 ohm startup
resistance and 0.18 ohm steady state resistance. The 5D-18 would probably
also do the trick ... it just has a slightly lower resistance, but I do not
think this is an issue at all since ideally it would be zero anyway.

Now to find this beast.... The local retail RS had nothing. They also did
not have anything listed in their catalog either. in fact,. the only
thermistor they carry was a 10K ohm initial value. I'll have to try some of
the local electrical supply houses. Thanks also for the alternative CL40.
I'll have to see what I can find.

Bob

"Travis Jordan" wrote in message
. ..
Travis Jordan wrote:
I would replace it with a NTC 5D-18 which has a 7A current rating.

Oops - my error - you want the same terminal resistance so use a NTC
8D-18 (8A).

Bob Shuman wrote:
Now to find this beast.... The local retail RS had nothing. They
also did not have anything listed in their catalog either. in fact,.
the only thermistor they carry was a 10K ohm initial value. I'll
have to try some of the local electrical supply houses. Thanks also
for the alternative CL40. I'll have to see what I can find.

Unless you have a distributor nearby that will sell one to you you'll
probably spend more time and effort trying to find it than it is worth -
I'd simply order it online.

Try Allied:
http://www.alliedelec.com/Search/Pro...=&DESC=CL%2D40

or Mouser:
http://www.mouser.com/index.cfm?hand..._pcodeid=52703

Yes, exactly. We did have one local electronics house here, Grainger, but
they did not stock this item. I took your advice and ordered the CL40 from
Mouser. My son is home from school for a couple weeks so hopefully we can
get this in before he leaves to return to school.

Thank you again for the help here in understanding the specifications.

Bob

"Travis Jordan" wrote in message
.. .
Unless you have a distributor nearby that will sell one to you you'll
probably spend more time and effort trying to find it than it is worth -
I'd simply order it online.

Bob Shuman wrote:
Thank you again for the help here in understanding the specifications.

No problem - and I hope you'll get his PC back up and running.

I just saw that CompUSA had an 500W Antec supply for $20 after rebates,
sale good through tomorrow. You might want to get one of these for a
spare or in case the thermistor fails again due to some undetermined
problem. Order it online and you won't even have to drive to a store.
Cheap insurance (but not as much fun as fixing a broken supply!).

http://www.compusa.com/adproducts/pr...0089&Ne=200000

OTOH if you need a good source for a replacement PS in the future I'd
recommend these guys.

http://www.newegg.com/ProductSort/Ca...sp?Category=32

Travis Jordan wrote:
I just saw that CompUSA had an 500W Antec supply for $20 after
rebates, sale good through tomorrow. You might want to get one of
these for a spare or in case the thermistor fails again due to some
undetermined problem. Order it online and you won't even have to
drive to a store. Cheap insurance (but not as much fun as fixing a
broken supply!).


http://www.compusa.com/adproducts/pr...0089&Ne=200000

Oops, just noticed that this is a CompUSA branded supply, not an Antec
supply! So buyer beware on that one (although I've installed a few 350W
CompUSA supplies and they worked fine and are still in service after a
few years).

Travis,

I'll go for the replacement part thanks since I am 99+% certain that this
will fix the supply given my previous experiment jumping around the blown
part coupled with the fact the protection fuse was not damaged. Hopefully
by putting in a replacement part that has twice the previous' current limit
this should not happen again.

I am a firm believer in minimizing waste and really hate to see anything at
all go to a landfill that can be repaired. And besides the ecological
aspect, it is so much more rewarding to fix something that is considered
dead by most people. By the way, the part (1 piece) from Mouser was $1.39,
and although I'm sure that the shipping and handling will increase that
figure, it still should be much less than a replacement supply.

Bob

"Travis Jordan" wrote in message
news
Travis Jordan wrote:
I just saw that CompUSA had an 500W Antec supply for $20 after
rebates, sale good through tomorrow. You might want to get one of
these for a spare or in case the thermistor fails again due to some
undetermined problem. Order it online and you won't even have to
drive to a store. Cheap insurance (but not as much fun as fixing a
broken supply!).

Bob Shuman wrote:
I am a firm believer in minimizing waste and really hate to see
anything at all go to a landfill that can be repaired. And besides
the ecological aspect, it is so much more rewarding to fix something
that is considered dead by most people. By the way, the part (1
piece) from Mouser was $1.39, and although I'm sure that the shipping
and handling will increase that figure, it still should be much less
than a replacement supply.

I agree, but unfortunately more and more often now I am finding that it
simply is too hard (time consuming and frustrating) to repair some of
today's electronics.

One of my Enermax high-end switching PC power supplies crapped out last
week and while I quickly found the problem (a shorted Mosfet switching
transistor and a couple of associated passive components) it was simply
impractical to repair the supply without hours and hours of labor. The
Mosfet was riveted to a heat sink that was in turn buried under several
layers of other components. Removing all the layers would have taken a
lot of time and extensive desoldering. While I would have preferred to
save the supply I decided instead to replace it at a cost of $90. But
now I've got some nice quiet 8mm and 9mm fans that I can use elsewhere.

OTOH I recently repaired a 10 year old 27" Sony TV that had a failed
power supply IC - it was easy to get to and while I was at it I improved
the heat sink so it won't likely fail again. So you win some and you
lose some.

I hope you win the PS battle and that it only costs you $1.39 plus S&H.
Post back and let us know if you succeed.

Travis Jordan wrote:

I just saw that CompUSA had an 500W Antec supply for $20 after rebates,
sale good through tomorrow. You might want to get one of these for a
spare or in case the thermistor fails again due to some undetermined
problem. Order it online and you won't even have to drive to a store.
Cheap insurance (but not as much fun as fixing a broken supply!).

http://www.compusa.com/adproducts/pr...0089&Ne=200000

Oops, just noticed that this is a CompUSA branded supply, not an Antec
supply! So buyer beware on that one (although I've installed a few 350W
CompUSA supplies and they worked fine and are still in service after a
few years).

CompUSA PSUs are made by several different companies, some great, like
Win-tact, some awful, like Powmax/Leadman, and their supplier told me
that the brand can change from week to week.

OTOH if you need a good source for a replacement PS in the future I'd
recommend these guys.

http://www.newegg.com/ProductSort/Ca...sp?Category=32

Raidmax is now supposedly OK and is made by Topower, the same company
that makes Aspire, OCZ, and Tagen. Topower's quality varies, but I
believe that the Raidmaxes are made from the better models. Still, I'd
probably rather buy a Fortron-Source, which can be as little as $40-50
for a 400-450W model.

I have a very well-built OEM PC PSU that has two thermistors, one for
each AC line. One thermistor is fairly large, the other only half its
diameter. What's the purpose of using two different sizes like this?

larry moe 'n curly wrote:
Raidmax is now supposedly OK and is made by Topower, the same company
that makes Aspire, OCZ, and Tagen. Topower's quality varies, but I
believe that the Raidmaxes are made from the better models. Still,
I'd probably rather buy a Fortron-Source, which can be as little as
$40-50 for a 400-450W model.

I use Raidmax for the low end and Seasonic for high end applications.
Enermax used to be my preferred high end PS but I've had a few problems
recently with their supplies produced in 2002/03 so I'm waiting to see
if they have addressed whatever supplier or design issue caused that.

I have a very well-built OEM PC PSU that has two thermistors, one for
each AC line. One thermistor is fairly large, the other only half its
diameter. What's the purpose of using two different sizes like this?

Good question. Are you sure that they both are thermistors and that one
isn't a fuse link or RF filter?

Travis Jordan wrote:
larry moe 'n curly wrote:

I have a very well-built OEM PC PSU that has two thermistors, one for
each AC line. One thermistor is fairly large, the other only half its
diameter. What's the purpose of using two different sizes like this?

Good question. Are you sure that they both are thermistors and that one
isn't a fuse link or RF filter?

Here's a photo:

http://static.flickr.com/42/75028029_3a711a4666_o.jpg

On the circuit board they're labelled NTC1 and NTC2, indicating that
they're thermistors. The RF filter is behind them in the photo.

Travis Jordan wrote:

I have a very well-built OEM PC PSU that has two thermistors, one for
each AC line. One thermistor is fairly large, the other only half its
diameter. What's the purpose of using two different sizes like this?

Good question. Are you sure that they both are thermistors and that one
isn't a fuse link or RF filter?

Here's a photo:

http://static.flickr.com/42/75028029_3a711a4666.jpg

I'm pretty sure that the two black things are thermistors because
they're labelled NTC1 and NTC2 The RF filter is in back of them in the
photo.

larry moe 'n curly wrote:
Here's a photo:

http://static.flickr.com/42/75028029_3a711a4666.jpg

I'm pretty sure that the two black things are thermistors because
they're labelled NTC1 and NTC2 The RF filter is in back of them in
the photo.

I'm pretty sure you are right. It is likely they are to provide two
different NTC inrush curves - I suppose we'd have to ask the designer
what s/he had in mind. Sounds like overkill to moi, or maybe they got a
'two for one' deal on the thermistors.

Have you traced the circuit and know for certain they are in the line and
neutral sides of the primary? Could they be in series or parallel in the
same line side feed? What are the values of the two components? Obviously
from the photo one is much larger so I would assume will be for higher
current/power, but possibly the thermal constants could be different and the
combination in parallel may give a better in rush current curve or possibly
protect against the problem I saw with the single under-designed device. I
guess I am just wondering if the second part wasn't a "fix" to a known
design problem

Bob

"Travis Jordan" wrote in message
.. .
larry moe 'n curly wrote:
Here's a photo:

http://static.flickr.com/42/75028029_3a711a4666.jpg

I'm pretty sure that the two black things are thermistors because
they're labelled NTC1 and NTC2 The RF filter is in back of them in
the photo.

I'm pretty sure you are right. It is likely they are to provide two
different NTC inrush curves - I suppose we'd have to ask the designer
what s/he had in mind. Sounds like overkill to moi, or maybe they got a
'two for one' deal on the thermistors.

Travis,

Per your request to post a follow up, the CL40 part arrived via UPS today
and although about 4X the physical size, installed without any problem and
the unit is working once again. Thanks again for the advice. Let's hope
that doubling the current capacity will prevent a reoccurrence.

Bob


"Travis Jordan" wrote in message
m...

I hope you win the PS battle and that it only costs you $1.39 plus S&H.
Post back and let us know if you succeed.

Bob Shuman wrote:
Per your request to post a follow up, the CL40 part arrived via UPS
today and although about 4X the physical size, installed without any
problem and the unit is working once again. Thanks again for the
advice. Let's hope that doubling the current capacity will prevent a
reoccurrence.

Congratulations on a good repair, and you are very welcome.

Enjoy the holiday weekend.