[ltp] T20 won't start

[Michael Karcher]

--=-arHc8Jfdwt0p9HofHmJ4
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

On Wednesday, 11.04.2007, 15:20 +1200 wrote Nick Rout:
> My T20 (with ubuntu, see this is on topic) won't start. When it is plugge=
d
> in the battery light glows green, but pushing the on button results in th=
e
> hard disk light flashing for a brief moment, then nothing else. The hard
> drive doesn't spin up, the fan doesn't start, there is nothing on the
> screen, nada.
Same problem here. This behaviour started last wednesday after some
hours of suspend-to-ram. I was frustrated and started to take it apart.
It took some time, as I tried to do it without the HMM, remembering I
managed it some time ago. This time I did not find out how to remove the
keyboard (and especially I didn't get the clue to remove it before the
plastic part containing the power, sleep and bay LED symbol) until
looking it up. After taking out the ultrabay battery, main battery, CMOS
battery, RAM and the hard drive and having the whole thing open (I also
unplugged main power before disassembling), I just gave it a try - and
the machine started (beeping because of missing RAM). I assembled
everything - it kept working. Worked the day on it, put it to
suspend-to-ram. Half-an-hour later at home: Tried to resume, one flash
of the hard drive LED, machine off (including sleep light). Unplugged
batteries, DIMMs other things, some time later it worked again. Same on
thursday morning (again after suspend to ram). So I started to use
suspend-to-disk. The machine always started fine, even after
suspend-to-ram cycles on saturday. After a suspend-to-ram, I couldn't
turn it on this morning. Unplugging batteries and AC for some hours
(CMOS battery stayed in), the machine again works like a charm.

Clues:
1. I have ACPI problems since 2.6.21-rcsomething (running git kernels).
Sometimes after resume, I get no ACPI interrupts any more, also timeout
error messages concerning EC access, but works mostly. Turn-off on
resume has been observed sometimes, but an immediate restart using the
power button was possible. I currently (on AC, batteries charged) can't
reproduce it just now, but I also had it with the current git kernel
(2.6.22-rc1 + git patches).

2. This machine works on two mainly dead batteries (an original IBM
ultrabay battery, probably around 5 years old, bought used, ACPI
capacity around 5 to 8 watt-hours, a china main battery clone called
"THY" that worked well one year after buying (around 42 instead of the
promised 48 Wh, but never mind), and suddenly broke down to also just
around 5 to 8 watt hours on a discharge down to 5%). Especially the
china battery (in the main battery slot) seems to have a high internal
resistance (backlight flickers significantly with CPU load, also the
ACPI read voltage drops up to 1 volt on high CPU load)

3. I have a defective DIMM installed, but the defective areas (around
25M and 108M) are excluded by memmap kernel parameters.

4. I use debian-mostly-etch, hibernate as suspend script, for suspend to
ram, I use
 UseSysfsPowerState mem
 PowerdownMethod platform

Michael Karcher
--=20
Michael Karcher
Freie Universit=E4t Berlin
Institut f=FCr Experimentalphysik
Arbeitsgruppe Fumagalli
Arnimallee 14
14195 Berlin

Tel: 030-838-55591
Fax: 030-838-56299



--=-arHc8Jfdwt0p9HofHmJ4
Content-Type: application/pgp-signature; name=signature.asc
Content-Description: Dies ist ein digital signierter Nachrichtenteil

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.6 (GNU/Linux)

iD8DBQBGWggNzhek2R7EicoRAnqFAJ46odNQoCHb2WRHLc2eppC1kJG9cQCfe77P
xhPU+jlSRJGdgwREgjr0/aw=
=sHuM
-----END PGP SIGNATURE-----

--=-arHc8Jfdwt0p9HofHmJ4--