Toshiba A200 series Proadalizer

[Ron]

A dual core dynabook was crashing etc while the psu was plugged in,
and I replaced the proadalizer capacitor and all was well, I had
Puppy Linux and Peppermint OS4 running happily with the psu.
The said Dynabook had a Japanese/English keyboard which was detected
OK by Peppermint but not Puppy, but came across a bargain replacement
keyboard so that is not an issue now.
A few months passed and then the original problem returned.
While waiting for another proadalizer to come, I tried disabling
one core of the cpu which is supposed to let you use the laptop
in this condition.
That did not help so I set it back to normal.
Puppy helped me notice the acpi=off boot setting, so I tried that
and to my surprise everything now works fine. I have also run
Crunchbang , playing a movie with VLC etc and no problems.

I think my next step is to replace the Proadalizer again,
but I find it weird how a simple thing like acpi can crash
the machine very quickly, and quite heavy cpu useage is OK
without it.
(There is no battery detection without it, but the warning light
on the front still flashes when the battery is low.)

Anyone know what is going on here?

Thanks in advance,
Ron M.

[Computer Nerd Kev]

For anyone else like me who was wondering what a Proadlizer is
and who they should use one on, it turns out it's an IC with a
lot of filter capacitors inside:

http://www.nec-
tokin.com/english/product/proadlizer/outline.html

Anyway, some aspect of the power control circuitry must be
getting confused by noise on the power input which the
Proadlizer capacitor is meant to filter out. I guess turning
ACPI off prevents this messing up more important functions.

In any case there must be something up if a new one of these
Proadlizer capacitors went in the space of a few months.
Perhaps your power supply is making voltage spikes on the DC
output? If so, it won't be doing the other components much
good either, especially now that the Proadlizer is broken
again.

If you (or someone you know) have an Oscilloscope, take a look
at the DC output from your power supply (preferably while it's
powering the laptop) and look out for said voltage spikes.
Otherwise, you might want to buy another power supply just to
be safe.

[Ron]

Post by Computer Nerd Kev
For anyone else like me who was wondering what a Proadlizer is
and who they should use one on, it turns out it's an IC with a
http://www.nec-
tokin.com/english/product/proadlizer/outline.html
Anyway, some aspect of the power control circuitry must be
getting confused by noise on the power input which the
Proadlizer capacitor is meant to filter out. I guess turning
ACPI off prevents this messing up more important functions.
In any case there must be something up if a new one of these
Proadlizer capacitors went in the space of a few months.
Perhaps your power supply is making voltage spikes on the DC
output? If so, it won't be doing the other components much
good either, especially now that the Proadlizer is broken
again.
If you (or someone you know) have an Oscilloscope, take a look
at the DC output from your power supply (preferably while it's
powering the laptop) and look out for said voltage spikes.
Otherwise, you might want to buy another power supply just to
be safe.

I have had three different 19v supplies, one a large capacity
HP supply, and didn't get any variation. I understand that it takes a
high end scope to pick up the type of spikes that can cause trouble to a
cpu anyway.

It seemed to me that it was just a case of wether the power
was coming from the 19v-in or the battery. I may have to rig up a pure
19v supply with batteries before that is conclusive, It would at least
prove it is the spikyness of a psu in combination with acpi as the
cause as opposed to dodgy acpi switched circuitry.

I have spent quite a few trouble free hours with acpi=off,
but maybe I should try some diagnostic that can give it
a real workout.

I'm also wondering if the weak battery could also be
contributing. The laptop will only run for about 20mins
on the battery alone.

Oher series dual core laptops dont seem to be prone to this
condition, Could it be possible that they have a bad or failing
design in the acpi affected circuitry? I'm not sure exactly
what this monitoring circuitry is yet.

Hopefully I'll narrow it down, and report back my findings.

Ron M.

[Ron]

Post by Ron

Post by Computer Nerd Kev
For anyone else like me who was wondering what a Proadlizer is
and who they should use one on, it turns out it's an IC with a
http://www.nec-
tokin.com/english/product/proadlizer/outline.html
Anyway, some aspect of the power control circuitry must be
getting confused by noise on the power input which the
Proadlizer capacitor is meant to filter out. I guess turning
ACPI off prevents this messing up more important functions.
In any case there must be something up if a new one of these
Proadlizer capacitors went in the space of a few months.
Perhaps your power supply is making voltage spikes on the DC
output? If so, it won't be doing the other components much
good either, especially now that the Proadlizer is broken
again.
If you (or someone you know) have an Oscilloscope, take a look
at the DC output from your power supply (preferably while it's
powering the laptop) and look out for said voltage spikes.
Otherwise, you might want to buy another power supply just to
be safe.

I have had three different 19v supplies, one a large capacity
HP supply, and didn't get any variation. I understand that it takes a
high end scope to pick up the type of spikes that can cause trouble to a
cpu anyway.
It seemed to me that it was just a case of wether the power
was coming from the 19v-in or the battery. I may have to rig up a pure
19v supply with batteries before that is conclusive, It would at least
prove it is the spikyness of a psu in combination with acpi as the
cause as opposed to dodgy acpi switched circuitry.
I have spent quite a few trouble free hours with acpi=off,
but maybe I should try some diagnostic that can give it
a real workout.
I'm also wondering if the weak battery could also be
contributing. The laptop will only run for about 20mins
on the battery alone.
Oher series dual core laptops dont seem to be prone to this
condition, Could it be possible that they have a bad or failing
design in the acpi affected circuitry? I'm not sure exactly
what this monitoring circuitry is yet.
Hopefully I'll narrow it down, and report back my findings.

I have now ruled out the psu's as being the problem.
I rigged 5 pair's of lithium batteries to get on load voltage
of about 19.4 volts. I found this voltage was maintained
throughout the test.
System crashed around the arrival of desktop or before on
several tests, pulled out the external 19v then everyhing is OK,
including the desktop battery indicator, confirming that acpi
was on (as expected) for these tests.

Ron M.

[Computer Nerd Kev]

Post by Ron
I have now ruled out the psu's as being the problem.
I rigged 5 pair's of lithium batteries to get on load voltage
of about 19.4 volts. I found this voltage was maintained
throughout the test.
System crashed around the arrival of desktop or before on
several tests, pulled out the external 19v then everyhing is OK,
including the desktop battery indicator, confirming that acpi
was on (as expected) for these tests.
Ron M.

OK, I suggested the power supply as a cause of the early failure of
the Proadlizer capacitor, not the ACPI problems that occour after it
has failed. A basic scope would at least show any low frequency spikes
occouring on the supply line which could be feeding into the Proadlizer
and causing it to fail early, which then leads to problems relating to
the noise that the Proadlizer is there to filter out. There could be
high frequency spikes that you won't see, but it's no reason to ignore
the test.

Basically if I were you, I'd replace the Proadlizer capacitor again.
Then take a look at the output of the PSU on whatever scope I had
(even my old one that's only 10MHz) to see if it was the power supply
that killed the last one prematurely, and is going to start wearing
down the new one.

[Ron]

Post by Computer Nerd Kev

Post by Ron
I have now ruled out the psu's as being the problem.
I rigged 5 pair's of lithium batteries to get on load voltage
of about 19.4 volts. I found this voltage was maintained
throughout the test.
System crashed around the arrival of desktop or before on
several tests, pulled out the external 19v then everyhing is OK,
including the desktop battery indicator, confirming that acpi
was on (as expected) for these tests.
Ron M.

OK, I suggested the power supply as a cause of the early failure of
the Proadlizer capacitor, not the ACPI problems that occour after it
has failed. A basic scope would at least show any low frequency spikes
occouring on the supply line which could be feeding into the Proadlizer
and causing it to fail early, which then leads to problems relating to
the noise that the Proadlizer is there to filter out. There could be
high frequency spikes that you won't see, but it's no reason to ignore
the test.
Basically if I were you, I'd replace the Proadlizer capacitor again.
Then take a look at the output of the PSU on whatever scope I had
(even my old one that's only 10MHz) to see if it was the power supply
that killed the last one prematurely, and is going to start wearing
down the new one.

The new proadlizer(s) have arrived, and I haven't rushed into replacing
it as I am trying to find some sample laptop psu regulator circuit
drawings, assuming I won't get the correct one. I haven't tried any
Toshiba support or Toshiba forums yet.

I have had a few crashes even at the bios level, before linux has had
the chance to disable acpi, so it does look like electronics in the
charging circuit or proadlizer or both is the cause.

It might be something a simple as regulator capacitors but I suppose
a chip could be working, but in a noisy way.

I have an esr meter, but useally I need to know the uF values to make
comparisons with known good ones.

Thanks, Ron M.