ATX power supply capacitor identification.

[David Farber]

There was a power surge at the customer's house. PG&E said their neutral
line failed. The computer now has an ATX power supply that smells quite
crispy. The 6.3 amp fuse was blown. I replaced it, then shorted pins 13 and
14 together on the main connector to test it outside of the box. I brought
the voltage up slowly with a variac and 100W bulb in series. The fan came on
but the bulb was a bit bright. I removed the bulb and tried it with just the
variac. The fan came on, then smoke came pouring out of one component which
was shorted. It's located between the largest two capacitors on the board
which are 470uF @ 200V. I was able to peel off the heat shrink and it
appears to be a ceramic disk capacitor. The part number starts with 241k. I
cannot read the other numbers that follow. I'm guessing it's 240pF but at
what voltage? It's wired in parallel to the previously mentioned largest
capacitor on the board. Any idea what the voltage rating should be on this
cap? Is it even necessary to replace it? The power supply works fine now
that the short has been removed.

Thanks for your reply.
--
David Farber
Los Osos, CA

[Ian Malcolm]

"David Farber" <farberbe...@aol.com> wrote in
news:ldoue5$o9d$1...@dont-email.me:

> There was a power surge at the customer's house. PG&E said their
> neutral line failed. The computer now has an ATX power supply that
> smells quite crispy. The 6.3 amp fuse was blown. I replaced it, then

> then smoke came pouring out of one component which was shorted. It's
> located between the largest two capacitors on the board which are
> 470uF @ 200V. I was able to peel off the heat shrink and it appears to
> be a ceramic disk capacitor. The part number starts with 241k. I
> cannot read the other numbers that follow. I'm guessing it's 240pF but
> at what voltage? It's wired in parallel to the previously mentioned
> largest capacitor on the board. Any idea what the voltage rating
> should be on this cap? Is it even necessary to replace it? The power
> supply works fine now that the short has been removed.
>

Its *NOT* a cap, its a varistor and is supposed to clip any spikes on the
AC in. Replace it with one rated for continuous operation at your local
supply voltage or the next time there's a surge the PSU will puke its
guts and is quite likely to take out the motherboard and/or hard drive.
There may be one across the other big cap as well and if so they should
be replaced as a pair. Its essential to replace the heatshrink (and make
sure its non-flamable) or the next failure will deposit a conductive
coating over the rest of the PSU.


--
Ian Malcolm. London, ENGLAND. (NEWSGROUP REPLY PREFERRED)
ianm[at]the[dash]malcolms[dot]freeserve[dot]co[dot]uk
[at]=@, [dash]=- & [dot]=. *Warning* HTML & >32K emails --> NUL

[David Farber]

Ian Malcolm wrote:
> "David Farber" <farberbe...@aol.com> wrote in
> news:ldoue5$o9d$1...@dont-email.me:
>
>> There was a power surge at the customer's house. PG&E said their
>> neutral line failed. The computer now has an ATX power supply that
>> smells quite crispy. The 6.3 amp fuse was blown. I replaced it, then
>
>> then smoke came pouring out of one component which was shorted. It's
>> located between the largest two capacitors on the board which are
>> 470uF @ 200V. I was able to peel off the heat shrink and it appears
>> to be a ceramic disk capacitor. The part number starts with 241k. I
>> cannot read the other numbers that follow. I'm guessing it's 240pF
>> but at what voltage? It's wired in parallel to the previously
>> mentioned largest capacitor on the board. Any idea what the voltage
>> rating should be on this cap? Is it even necessary to replace it?
>> The power supply works fine now that the short has been removed.
>>
>
> Its *NOT* a cap, its a varistor and is supposed to clip any spikes on
> the AC in. Replace it with one rated for continuous operation at your
> local supply voltage or the next time there's a surge the PSU will
> puke its guts and is quite likely to take out the motherboard and/or
> hard drive. There may be one across the other big cap as well and if
> so they should be replaced as a pair. Its essential to replace the
> heatshrink (and make sure its non-flamable) or the next failure will
> deposit a conductive coating over the rest of the PSU.

Thanks for that info. I'll remove it from the pc until I get the correct
parts. You are correct, there is another one of these across the other big
cap.

[Phil Allison]

"Ian Malcolm"

>
>
> Its *NOT* a cap, its a varistor and is supposed to clip any spikes on the
> AC in.

** Nonsense.

Large electros kill spikes just like sledgehammers kill flies.

A bridge rectifier followed by a filter electro is the best spike suppressor
you will ever see.

> Replace it with one rated for continuous operation at your local
> supply voltage or the next time there's a surge the PSU will puke its
> guts and is quite likely to take out the motherboard and/or hard drive.
> There may be one across the other big cap as well and if so they should
> be replaced as a pair.

** All they do is make the fuse blow if the electros are overvoltaged by
connection to a too high AC supply.


.... Phil

[jurb...@gmail.com]

Our buddy Phill composed, in an unusually non-berating way today thusly :

>"Large electros kill spikes just like sledgehammers kill flies. "

And they got better aim. As long as you have full wave rectification, and the rectifiers can handle alot of peak current, they are quite effective. However they are not effective so much against having simple overvoltage applied.

You are not in the US, if the power there is like the UK or whatever it is different. While you have 240 coming in referenced to neutral, in the US it is half that. The line is balanced with two 120 volt legs out of phase. Almost all smaller appliiances run from one leg to neutral. If the neutral loses continuity it can apply as much as double the effective, peak to peak, RMS and whatever other way you want to measure it - voltage to your appliance.

This cannot happen in the good countries but I am not here to bitch. At the moment...

At any rate, I have seen a bunch of equipment saved by these MOVs over the years and in fact at one time I did install them as an option on repair jobs. When I worked where they sold service contracts, I should have installed them as a matter of standard operating procedure.

Of course this does not necessarily apply in other countries.

I have also noticed a few tings that actually switch themself. What most of them do is to simply rectify without doubling and then kick it up with a chopper circuit, the usual boost thing. When the incoming is well above the 160 rectified the booster simply shuts down and passes the juice.

A long time ago I worked on a bunch of little TVs for a limo service that actually ran on 24 volts but had an upconvertor for use on 12 volts. The upside to these little Sonys (who else ?) was that they could be simply connected into 24 volt systems (tractor trailers/lorries) with zero modifications. You could fuse it a little lower but that wasn't all that bad.

I think as soon as Edison's body was cold they should have made it 240, yes center tapped like it is, but made all the outlets and appliances 240 all the time. they also should have changed the frequency to 400 Hz. (CPS LOL) Or better yet, maybe 360 or something to make it easier to make electric clocks. Synchronous motors would be a little more complex but think of all the speeds you could have.

Enough, I am not drunk enough to write more right now... lol.

[Phil Allison]

<jurb...@gmail.com>

Our buddy Phill composed, in an unusually non-berating way today thusly :

>"Large electros kill spikes just like sledgehammers kill flies. "

And they got better aim. As long as you have full wave rectification, and
the rectifiers can handle alot of peak current, they are quite effective.
However they are not effective so much against having simple overvoltage
applied.

You are not in the US, if the power there is like the UK or whatever it is
different. While you have 240 coming in referenced to neutral, in the US it
is half that. The line is balanced with two 120 volt legs out of phase.
Almost all smaller appliiances run from one leg to neutral. If the neutral
loses continuity it can apply as much as double the effective, peak to peak,
RMS and whatever other way you want to measure it - voltage to your
appliance.

** The 120V leg with the greater load will drop voltage when the neutral is
lost - so the other leg goes up until something goes bang.

This cannot happen in the good countries but I am not here to bitch.

** It happens now and then in live entertainment venues when 3 phase power
is used.

Same reason, no neutral exists or loss of the of neutral conductor.


.... Phil

[Michael A. Terrell]

jurb...@gmail.com wrote:
>
> I think as soon as Edison's body was cold they should have made it 240, yes center tapped like it is, but made all the outlets and appliances 240 all the time. they also should have changed the frequency to 400 Hz. (CPS LOL) Or better yet, maybe 360 or something to make it easier to make electric clocks. Synchronous motors would be a little more complex but think of all the speeds you could have.

You are already too drunk to make sense. There were a lot of 25 Hz
power plants for rialroads, mines and other places that required low
speeds and very high starting torque. Also, 400 hz would reduce the size
of a power grid segments, before the losses become excessive. This would
require a lot more HV AC to DC and DC to AC conversion for distribution
between grid segments. That also lowers the overall grid efficiency,
which will raise your electric bill. This is why no one asked your
opinion of our power distribution systems.

<http://en.wikipedia.org/wiki/Amtrak%27s_25_Hz_traction_power_system>


--
Anyone wanting to run for any political office in the US should have to
have a DD214, and a honorable discharge.

[Ian Malcolm]

"Phil Allison" <phi...@tpg.com.au> wrote in
news:bmb8ik...@mid.individual.net:

>
> <jurb...@gmail.com>
> Our buddy Phill composed, in an unusually non-berating way today
> thusly :
>
>>"Large electros kill spikes just like sledgehammers kill flies. "
>
> And they got better aim. As long as you have full wave rectification,
> and the rectifiers can handle alot of peak current, they are quite
> effective. However they are not effective so much against having
> simple overvoltage applied.
>

Thanks for the correction. I mistyped, omitting "and surges" as you may
have guessed by my reference to surges later on. However I do wonder
whether the self-inductance of the typical wound foil construction high
voltage electrolytic is enough to cause problems with very short
impulses if one omits the varistors.

Littelfuse's Varistor page (which contains more than yoever wanted to
know about them):
<http://www.littelfuse.com/products/varistors.aspx>
states:


> *** Speed of Response and Rate Effects ***
>
> The varistor action depends on a conduction mechanism similar to that
> of other semiconductor devices. For this reason, conduction occurs
> very rapidly, with no apparent time lag – even into the nanosecond
> (ns) range. Figure 18, shows a composite photograph of two voltage
> traces with and without a varistor inserted in a very low inductance
> impulse generator. The second trace (which is not synchronized with
> the first, but merely superimposed on the oscilloscope screen) shows
> that the voltage clamping effect of the varistor occurs in less than
> 1.0 ns.

It does qualify that by stating this speed is not achivable in a leaded
package, but I'd bet its still far far better than the electrolytics
capability to deal with fast risetime transients.

[Phil Allison]

"Ian Malcolm"
> "Phil Allison" >

>>
>>>"Large electros kill spikes just like sledgehammers kill flies. "
>>
>> And they got better aim. As long as you have full wave rectification,
>> and the rectifiers can handle alot of peak current, they are quite
>> effective. However they are not effective so much against having
>> simple overvoltage applied.
>>
> Thanks for the correction. I mistyped, omitting "and surges" as you may
> have guessed by my reference to surges later on. However I do wonder
> whether the self-inductance of the typical wound foil construction high
> voltage electrolytic is enough to cause problems with very short
> impulses if one omits the varistors.

** There is no such " self inductance" with electros or other wound foil
capacitors.

FYI:

The inductance of the PCB traces is far more than that of typical
capacitors.

You are clutching at straws - pal.


.... Phil