<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
<title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
Kimmo Hämäläinen a écrit :
<blockquote cite="mid1158821859.17746.162.camel@localhost.localdomain"
type="cite">
<pre wrap="">On Wed, 2006-09-20 at 23:17, ext Frantisek Dufka wrote:
...
</pre>
<blockquote type="cite">
<pre wrap="">So it really seems related to wi-fi.
</pre>
</blockquote>
<pre wrap=""><!---->
Thank you for the information, I'm trying to keep the internal
investigation on-going (and not just assuming that HW is broken). These
kind of hints should help to nail it down.
BR, Kimmo
</pre>
<blockquote type="cite">
<pre wrap="">Frantisek
</pre>
</blockquote>
</blockquote>
Hum ... this time I spent much more time on this and i have done a lot
of tests.<br>
I try to use a scientifical approach so that I could say "yes the
problem happen 5/100 if condition x ..."<br>
The results "for my device" are that the problem seem not to be related
to<br>
- empty battery <br>
- wifi<br>
- temperature<br>
- high power requirements<br>
- bad memory region<br>
or any combination of these parameters.<br>
<br>
You can have the illusion that there is a correlation with that or that
but no, the stats say that it's just a coincidence.<br>
<br>
I have notice that when memtester find a bad address line then the
probability to find other bad address in the same run is very high.<br>
<br>
So after all that I REALLY don't know what the problem is <span
class="moz-smiley-s2"><span> :-( </span></span><br>
Need somme new ideas ... or an oscilloscope ... to go deeper.<br>
<br>
Regards,<br>
Olivier ROLAND<br>
</body>
</html>