Hi

I have a 3 year old laptop which has suddenly developed the infamous
spontaneous abrupt shutdown syndrome. This is usually attributed to
overheating caused by dirty fans, but I am not convinced.

- I have tested various combinations of CPU speed settings, workloads,
power-management modes, AC vs batery, ambient temperature, etc.
- Shutdowns happen in the BIOS too, so it cannot be a software issue.
- I have enabled self-tests in the BIOS and used memtest86+.
- I have fully disassembled the laptop and reseated every connector.
- I have removed all non-essential modules and peripherals: bluetooth,
wifi, modem, keyboard, LCD panel (running on VGA with a USB keyboard).
- I have cleaned the copper air ducts and operated the laptop with
external fans blowing on all large chips.

Before I call it quits I want to try one more thing: reflow the PCBs.

The idea is not to melt everything, but only to reconnect any shaky pad
and destroy possible tin whiskers (this is an early lead-free design).

Recommendations regarding prerequisites, choice of air gun vs oven,
and safe temperature cycles for this particular purpose would be
appreciated (I have zero experience with reflow soldering).

AC

"A. Caspis" wrote in message
...
Hi

I have a 3 year old laptop which has suddenly developed the infamous
spontaneous abrupt shutdown syndrome. This is usually attributed to
overheating caused by dirty fans, but I am not convinced.

- I have tested various combinations of CPU speed settings, workloads,
power-management modes, AC vs batery, ambient temperature, etc.
- Shutdowns happen in the BIOS too, so it cannot be a software issue.
- I have enabled self-tests in the BIOS and used memtest86+.
- I have fully disassembled the laptop and reseated every connector.
- I have removed all non-essential modules and peripherals: bluetooth,
wifi, modem, keyboard, LCD panel (running on VGA with a USB keyboard).
- I have cleaned the copper air ducts and operated the laptop with
external fans blowing on all large chips.

Before I call it quits I want to try one more thing: reflow the PCBs.

The idea is not to melt everything, but only to reconnect any shaky pad
and destroy possible tin whiskers (this is an early lead-free design).

Recommendations regarding prerequisites, choice of air gun vs oven,
and safe temperature cycles for this particular purpose would be
appreciated




(I have zero experience with reflow soldering).

AC


IMHO that means you will need to practice and destroy a lot of motherboards
before you start getting things right. This really isn't the kind of thing
you should be trying at home unless you have a lot of money to spend on the
right equipment..


Try googling "reflowing BGA's". You will find some claims of success here,
though probably not many admissions that the fix only lasted a few days.
Mostly you will get total failures, I did. I would guess that the ratio of
reported sucesses to reported failures is quite large



Good luck.



Gareth.

The idea is not to melt everything, but only to reconnect any shaky pad
and destroy possible tin whiskers (this is an early lead-free design).
Recommendations regarding prerequisites, choice of air gun vs oven,
and safe temperature cycles for this particular purpose would be
appreciated

If there is any hope of the board working again, this would mostlikely
exterminate all hope. There are many factors of the sudden death syndrome.
some could be solder and tin whisker related.. However, with the heating
and cooling cycles of laptops, delaminated contacts and broken traces on the
inner layers of the board would seem morelikely. Reflow won't help those
situations. Heating can cause solder to move away from the contact points.
If you've ever seen some of the early 90 model TVs, a lot of the picture
problems I found on those sets was witht the CRT socket solder connections.
Crt heats up, melts the solder, solder moves away from the pad and creates
gaps between the pin and pad.





--
--------------------------------- --- -- -
Posted with NewsLeecher v3.8 Final
Web @ http://www.newsleecher.com/?usenet
------------------- ----- ---- -- -

A. Caspis wrote:

Before I call it quits I want to try one more thing: reflow the PCBs.

What make & model laptop? HP by any chance?

Find a company with the equipment to remove and replace BGA packages.
They will have to be replaced with new chips though.

Here is a description of a rework station ...
http://www.digifixltd.co.uk/articles.html#anchor

and a UK company who offers a service (but not in the laptop field)
http://www.sers.me.uk/ic_fitting_service.htm

--
Adrian C

I would first look at the CPU heat sink. Make sure the fan is clean and
working properly, and there is heatsink compound under it. The might be a
"pad" which is heatsink material also
Frank


"A. Caspis" wrote in message
...
Hi

I have a 3 year old laptop which has suddenly developed the infamous
spontaneous abrupt shutdown syndrome. This is usually attributed to
overheating caused by dirty fans, but I am not convinced.

- I have tested various combinations of CPU speed settings, workloads,
power-management modes, AC vs batery, ambient temperature, etc.
- Shutdowns happen in the BIOS too, so it cannot be a software issue.
- I have enabled self-tests in the BIOS and used memtest86+.
- I have fully disassembled the laptop and reseated every connector.
- I have removed all non-essential modules and peripherals: bluetooth,
wifi, modem, keyboard, LCD panel (running on VGA with a USB keyboard).
- I have cleaned the copper air ducts and operated the laptop with
external fans blowing on all large chips.

Before I call it quits I want to try one more thing: reflow the PCBs.

The idea is not to melt everything, but only to reconnect any shaky pad
and destroy possible tin whiskers (this is an early lead-free design).

Recommendations regarding prerequisites, choice of air gun vs oven,
and safe temperature cycles for this particular purpose would be
appreciated (I have zero experience with reflow soldering).

AC

Frank S wrote:
I would first look at the CPU heat sink. Make sure the fan is clean and
working properly, and there is heatsink compound under it. The might be a
"pad" which is heatsink material also

Also look for 'bad caps' as well, come to think of it.

--
Adrian C

On 14 Jan, 10:20, "A. Caspis" wrote:
Hi

I have a 3 year old laptop which has suddenly developed the infamous
spontaneous abrupt shutdown syndrome. This is usually attributed to
overheating caused by dirty fans, but I am not convinced.

- I have tested various combinations of CPU speed settings, workloads,
* power-management modes, AC vs batery, ambient temperature, etc.
- Shutdowns happen in the BIOS too, so it cannot be a software issue.
- I have enabled self-tests in the BIOS and used memtest86+.
- I have fully disassembled the laptop and reseated every connector.
- I have removed all non-essential modules and peripherals: bluetooth,
* wifi, modem, keyboard, LCD panel (running on VGA with a USB keyboard)..
- I have cleaned the copper air ducts and operated the laptop with
* external fans blowing on all large chips.

Before I call it quits I want to try one more thing: reflow the PCBs.

The idea is not to melt everything, but only to reconnect any shaky pad
and destroy possible tin whiskers (this is an early lead-free design).

Recommendations regarding prerequisites, choice of air gun vs oven,
and safe temperature cycles for this particular purpose would be
appreciated (I have zero experience with reflow soldering).

AC

If it's one of the early model Thinkpads, they're notorious for
inductors dropping off, sometimes completely. Many people report
effecting a cure by resoldering just these and anything else which has
a large enough pad to do safely with your smallest soldering iron tip.

Google for a newsgroup or forum specialising in your particular model
and see what people there advise.

Chris

A. Caspis wrote:

I have a 3 year old laptop which has suddenly developed the infamous
spontaneous abrupt shutdown syndrome. This is usually attributed to
overheating caused by dirty fans, but I am not convinced.

- I have tested various combinations of CPU speed settings, workloads,
power-management modes, AC vs batery, ambient temperature, etc.
- Shutdowns happen in the BIOS too, so it cannot be a software issue.
- I have enabled self-tests in the BIOS and used memtest86+.
- I have fully disassembled the laptop and reseated every connector.
- I have removed all non-essential modules and peripherals: bluetooth,
wifi, modem, keyboard, LCD panel (running on VGA with a USB keyboard).
- I have cleaned the copper air ducts and operated the laptop with
external fans blowing on all large chips.

Before I call it quits I want to try one more thing: reflow the PCBs.

Check web sites with information about laptop computer repair.
Notebookreview.com is a reasonable starting place.

The Oct. Elektor Electronics (www.elektor.com) has a project for a
reflow soldering oven, but reflowing your main board is as unlikely to
fix your problem as pouring gasoline on it and setting it on fire
will.

Forget about resoldering the entire board. Instead examine it much
more thoroughly, using a 3x - 10x magnifier and strong light, and look
for hairline cracks and solder that looks rough, discolored, or balled-
up. A Google image search for "bad solder" will return many
examples. Do look at the usual suspects - areas of high mechanical
stress (under the keyboard, around external connectors and heavy
objects) and high concentrations of heat (not the CPU but voltage
regulator power MOSFETs). Measure voltages, including AC with an
oscilloscope. Excessive AC voltage is often a sign of bad
capacitors. First practice your soldering skills on worthless surface
mount boards, and don't quit until you confirm that they're restored
to normal operation.

"Meat Plow" wrote in message
...
On Wed, 14 Jan 2009 10:20:44 +0100, "A. Caspis"
wrote:

Recommendations regarding prerequisites, choice of air gun vs oven,
and safe temperature cycles for this particular purpose would be
appreciated (I have zero experience with reflow soldering).

689 degrees Fahrenheit to bake surface mounts in convection oven.
Duration depends on the size of the SMD.

495 degrees Fahrenheit for solder flow on through the board
connections.

These examples are used in the construction of CPU chips.


Doesn't a motherboard contain various plastic items? Connectors, SIMM
sockets etc? I shudder to think what would happen to my laptop should I do
the above.



Gareth.

"Meat Plow" wrote in message
...
On Thu, 15 Jan 2009 20:21:49 -0000, "Gareth Magennis"
wrote:


"Meat Plow" wrote in message
...
On Wed, 14 Jan 2009 10:20:44 +0100, "A. Caspis"
wrote:

Recommendations regarding prerequisites, choice of air gun vs oven,
and safe temperature cycles for this particular purpose would be
appreciated (I have zero experience with reflow soldering).

689 degrees Fahrenheit to bake surface mounts in convection oven.
Duration depends on the size of the SMD.

495 degrees Fahrenheit for solder flow on through the board
connections.

These examples are used in the construction of CPU chips.


Doesn't a motherboard contain various plastic items? Connectors, SIMM
sockets etc? I shudder to think what would happen to my laptop should I
do
the above.



Gareth.


I didn't suggest doing anything and gave an explicit example.



Yes you did, indirectly. The question was recommendations for reflowing the
PCB, airgun or oven. Your reply was 689F in an oven.
You did not clarify that you were answering a different question.



Gareth.

"Meat Plow" wrote in message
...
On Fri, 16 Jan 2009 20:26:26 -0000, "Gareth Magennis"
wrote:


"Meat Plow" wrote in message
...
On Thu, 15 Jan 2009 20:21:49 -0000, "Gareth Magennis"
wrote:


"Meat Plow" wrote in message
...
On Wed, 14 Jan 2009 10:20:44 +0100, "A. Caspis"
wrote:

Recommendations regarding prerequisites, choice of air gun vs oven,
and safe temperature cycles for this particular purpose would be
appreciated (I have zero experience with reflow soldering).

689 degrees Fahrenheit to bake surface mounts in convection oven.
Duration depends on the size of the SMD.

495 degrees Fahrenheit for solder flow on through the board
connections.

These examples are used in the construction of CPU chips.


Doesn't a motherboard contain various plastic items? Connectors, SIMM
sockets etc? I shudder to think what would happen to my laptop should I
do
the above.



Gareth.


I didn't suggest doing anything and gave an explicit example.



Yes you did, indirectly. The question was recommendations for reflowing
the
PCB, airgun or oven. Your reply was 689F in an oven.
You did not clarify that you were answering a different question.

Well looks like you're the only one whining. Next time I'll dumb it
down to be 100% sure whiney idiots like you understand my reply and
you wont roast your plastic parts in a ****ing 700F oven.



No, next time you be sure to answer the question asked, not quote some other
information with no reference to what you are actually referring to.

Calling people silly names and swearing doesn't do much to strengthen your
credibility here.



Gareth.

"A. Caspis" wrote:

The idea is not to melt everything, but only to reconnect any shaky pad
and destroy possible tin whiskers (this is an early lead-free design).

And your opinion of the EU Commission is ?

Graham

Thanks for all the advice. You have convinced me not to
rush the whole thing into the kitchen oven

To answer some of your questions:

Yes, the PCB has plenty of plastic parts. But I assume
it was assembled in a reflow oven, and therefore there
must be a safe temperature range (hence my posting here).

The laptop is a subnotebook from a major manufacturer.
No need to put a black mark on a specific one - I have
read reports of spontaneous shutdowns affecting HP, Dell,
Toshiba, etc. It is a Pentium M with i801/i855 chipset.

The failure scenario probably involves mechanical stress
and vibrations. As long as I do not type on the internal
keyboard or tap on the PCB, the laptop is quite reliable;
unfortunately I still cannot locate any specific faulty
component. On the other hand, I would expect a cracked
pad to cause the laptop to hang rather than shutdown.
Or maybe these chipsets are designed to shutdown when
ECC-like errors occur ?

I have seen my share of leaking electrolytic caps, but
this board has only SMT capacitors. Is it possible to
visually recognize a blown one ?

AC