ASUS EeePc fan replacement

My 2009 Asus EeePc (model 1005ha) started to make loud noise. I figured it was the fan. With an aid of this YouTube movie (see below) I was able to replace the fan and repair it. Thanks luisfilipe1983!

Assembling the tools

Opening the back

Removing the keboard

Going inside

Getting closer...

Deep inside

Fixed!

The YouTube Movie:

EQMet TSA-SMA setup for SeisComp3

The following post will describe how to install EQMet INTERNET-READY STRONG MOTION ACCELEROGRAPH (TSA-SMA) and set it up for SeisComp3.


Following the instruction in the manual:
The initial steps are to connect the TSA-SMA to electricity, GPS and a computer via the USB device connection.
Then, connect to the device via serial connection.
On windows: read the manual and follow the orders using putty. Don’t forget to download the driver for the device from EQMet site (http://www.eqmet.com/eng/Linux-cdc-acm.zip)

On Linux:
install cu:
# sudo yum install uucp
get the tty of the connection:
# dmesg | grep 'tty'
on my system it was: ttyACM0 but it can be ttyS[1234] or serial.
next I connect to the device:
# cu -l /dev/ttyACM0 -s 115200
it takes a while to prompt for user and password (root,kmi) press enter if does not prompt.
I need to assign an IP based on the mac address so I check the mac:
# ifconfig
after I got the IP, it is easy to connect via http sftp and ssh.

for static IP, edit /etc/network/interfaces:
replace:
iface eth0 inet dhcp
with:
iface eth0 inet static
address <IP>
netmask <MASK>
gateway <GATEWAY>
dns-nameservers <DNS-SERVER>

connect to the device via http.

Next go to Parameters:hardware menu option.

update to your preferable network-station-location-channel ID in:

- Network ID

- Unit ID and Site ID

- dig1 ChX ID

- dig1 ChX Location code

where X in channel 1, 2 or 3.

press OK.

Next, following the response mail I got for EQmet support regarding seedlink connection:

Sir,

The limited version of Rockhound in the TSA-SMA does not support continuous recording. However, you can approximate continuous recording by adding the SEEDLink mirror module and configuring it to accept data from the RockToSlink module at address 127.0.0.1, port 18000 (or as configured). You should add this module after installing and adding in the RockToSlink module.

This module writes the received MiniSEED packets out to storage and breaks the files at approximately the specified duration (depending on MiniSEED packet contents). If you add this module, you should also consider adding an auto file delete module to avoid filling the storage. A file sender module may also be useful (SSH, FTP, or email).

The RockToSlink module for Rockhound is provided by ISTI, New York. The module and documentation is available from ISTI at http://www.isti2.com/RockToSLink/

Technical support for RockToSLink is available from ISTI at isti-info@isti.com. SEEDlink is a public domain program and protocols. Support for these programs including detailed protocol descriptions or installation and setup of the programs is not provided by Kinemetrics.

Regards,

EQMet, Technical Support
support@eqmet.com

installe the RockToSlink module and add it to the workflow:

1. Download the rocktoslik module from http://www.isti2.com/RockToSLink/dist_pub/:
> wget -v http://www.isti2.com/RockToSLink/dist_pub/RockToSLink_v1.0_dist.zip
2. Unzip it:
> unzip RockToSLink_v1.0_dist.zip
following the instractions in http://www.isti2.com/RockToSLink/current_dist/docs/RockToSLink_ovr.html:
3. sftp to the device:
> sftp root@<IP>
4. Change directory to module location:
> cd  /usr/rock/SMARTSDist/injar
5 Upload the module-update jar file:
> put RockToSLink_update.jar
6. connect via ssh:
> ssh root@<IP>
7. Change directory to module location:
> cd  /usr/rock/SMARTSDist/injar
8. Rename "RockToSLink_update.jar" to "update.jar"
> mv RockToSLink_update.jar update.jar
9. Reboot the instrument:
> reboot
Now add the RockToLink module to device layout:
10. In the device http menue Parameters: Advanced features, add options for Add/Remove Modules, Replace Modules and Advanced Modules.
11. In Module Add, add the RockToLink module.
12. Press apply changes now.

Now go to system operation and make sure all settings are correct.

enjoy.

LSM303 accelerometer on RaspberryPi

After getting my new LSM303 accelerometer, I got my great electronics tec. Daviv Shtibelman to wire it up for RPI.
Connecting LSM303 pins 1 (SCL), 2 (SDA), 6 (GND) and 8 (V3V) to pins 5 (SCL), 3 (SDA), 6 (GND) and 1 (V3V) on the Pi board.
On my ArchlinuxARM OS I had to install i2c-tools:
> pacman -S i2c-tools
And add i2c modules:
> echo i2c-dev > /etc/modules-load.d/ i2c.conf
> echo i2c-bcm2708 >> /etc/modules-load.d/ i2c.conf
> reboot

now to check it works:
> i2cdetect -y 1
got a message like this:
     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
00:          -- -- -- -- -- -- -- -- -- -- -- -- --
10: -- -- -- -- -- -- -- -- -- 19 -- -- -- -- 1e --
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
50: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
70: -- -- -- -- -- -- -- --

Now, following this, I test it and it works. Now I'm up for creating my own code.

Made some changes to the code:

main.cpp

#include"LSM303DLHC.h"
#include<stdio.h>
#include <unistd.h>


/*
   getHeading() calculates a tilt-compensated heading.
   A float between 0 and 360 degrees is returned. You need
   to pass this function both a magneto and acceleration array.
 
   Headings are calculated as specified in AN3192:
   http://www.sparkfun.com/datasheets/Sensors/Magneto/Tilt%20Compensated%20Compass.pdf
 
*/

float getHeading(LSM303& lsm303dlhc);

#define PI 3.141592654


int main(void)
{
    uint8_t bajt;
    const char *fileN = "/dev/i2c-1";
    LSM303 lsm303dlhc(fileN);

    lsm303dlhc.enalbe();

    while(1)
    {
        lsm303dlhc.readAccelerationRaw();
        lsm303dlhc.readMagnetometerRaw();
        printf("acc [m/s^2]: \e[27;1;31m %f \e[m \e[27;1;32m %f \e[m \e[27;1;34m %f \e[m mag:  \e[27;1;31m %d \e[m  \e[27;1;32m %d \e[m \e[27;1;34m %d\e[m %fdeg\n",
               (int16_t)lsm303dlhc.acc_x_raw*0.00957,(int16_t)lsm303dlhc.acc_y_raw*0.00957,-(int16_t)lsm303dlhc.acc_z_raw*0.00957,
               (int16_t)lsm303dlhc.mag_x_raw, (int16_t)lsm303dlhc.mag_y_raw, (int16_t)lsm303dlhc.mag_z_raw,getHeading(lsm303dlhc));
        usleep(10);
       }

    }


float getHeading(LSM303& lsm303dlhc)
{
  float heading,pitch,roll,xh,yh,zh;
  // see section 1.2 in app note AN3192
  int magValue[3];
  float accelValue[3];
  magValue[0] = (int16_t)lsm303dlhc.mag_x_raw;
  magValue[1] = (int16_t)lsm303dlhc.mag_y_raw;
  magValue[2] = (int16_t)lsm303dlhc.mag_z_raw;
  accelValue[0] = (int16_t)lsm303dlhc.acc_x_raw*0.000976531;
  accelValue[1] = (int16_t)lsm303dlhc.acc_y_raw*0.000976531;
  accelValue[2] = -(int16_t)lsm303dlhc.acc_z_raw*0.000976531;

    // see appendix A in app note AN3192
  pitch = asin(-accelValue[0]);
  roll = asin(accelValue[1]/cos(pitch));

  xh = magValue[0] * cos(pitch) + magValue[2] * sin(pitch);
  yh = magValue[0] * sin(roll) * sin(pitch) + magValue[1] * cos(roll) - magValue[2] * sin(roll) * cos(pitch);
  zh = -magValue[0] * cos(roll) * sin(pitch) + magValue[1] * sin(roll) + magValue[2] * cos(roll) * cos(pitch);

  heading = 180*atan2(yh,xh)/PI;

  if (heading <0)
    heading += 360;

  return heading;
}

LSM303DLHC.cpp:

#include"LSM303DLHC.h"
#include<math.h>
#include<stdio.h>

/*Conection to Raspberry PI:
 LSM303     Raspberry PI
 VDD    ->  3V3(PIN 1)
 SDA    ->  SDA(PIN 3)
 SCL    ->  SCL(PIN 5)
 GND    ->  GND(PIN 6)
*/

#define LSM303DLHC_MAG_ADDRESS            (0x3C >> 1)
#define LSM303DLHC_ACC_ADDRESS            (0x32 >> 1)

LSM303::LSM303(const char * i2cDeviceName) : i2c_lsm303(i2cDeviceName)
{

}

uint8_t LSM303::readAccRegister(uint8_t regAddr)
{
    i2c_lsm303.addrSet(LSM303DLHC_ACC_ADDRESS);
    return i2c_lsm303.readByte(regAddr);
}

uint8_t LSM303::readMagRegister(uint8_t regAddr)
{
    i2c_lsm303.addrSet(LSM303DLHC_MAG_ADDRESS);
    return i2c_lsm303.readByte(regAddr);
}

void LSM303::writeAccRegister(uint8_t regAddr,uint8_t byte)
{
    i2c_lsm303.addrSet(LSM303DLHC_ACC_ADDRESS);
    i2c_lsm303.writeByte(regAddr, byte);

}

void LSM303::writeMagRegister(uint8_t regAddr, uint8_t byte)
{
    i2c_lsm303.addrSet(LSM303DLHC_MAG_ADDRESS);
    i2c_lsm303.writeByte(regAddr, byte);

}

void LSM303::enalbe(void)
{
   writeAccRegister(LSM303_CTRL_REG1, 0b10010111);
   writeAccRegister(LSM303_CTRL_REG4, 0b00001000);

   writeMagRegister(LSM303_MR_REG, 0x00);
}

void LSM303::readAccelerationRaw(void)
{
    uint8_t block[6];

    i2c_lsm303.addrSet(LSM303DLHC_ACC_ADDRESS);

    i2c_lsm303.readBlock(0x80 | LSM303_OUT_X_L_A, sizeof(block), block);
    acc_x_raw = (int16_t)(block[0] | (block[1] << 8)) >> 4;
    acc_y_raw = (int16_t)(block[2] | block[3] << 8) >> 4;
    acc_z_raw = (int16_t)(block[4] | block[5] << 8) >> 4;

}

void LSM303::readMagnetometerRaw(void)
{
    uint8_t block[6];

    i2c_lsm303.addrSet(LSM303DLHC_MAG_ADDRESS);
    i2c_lsm303.readBlock(0x80 | LSM303_OUT_X_H_M, sizeof(block), block);

    mag_x_raw = (int16_t)(block[1] | block[0] << 8);
    mag_y_raw = (int16_t)(block[5] | block[4] << 8);
    mag_z_raw = (int16_t)(block[3] | block[2] << 8);

}

void LSM303::readAcceleration(void)
{
    readAccelerationRaw();
}

RaspberryPi initial steps

I initiated a new project involving low-cost micro-computer RaspberryPi.
In order to begin I ordered some basic parts (costs are in USD for Israeli market, should be cheaper elsewhere):
a RaspberryPi (55$)
5v charger with microusb adapter (10$)
8GB SD card (12$)
HDMI cable (8$)

and some extra parts:
3.5" LCD screen (45$)
micro-keyboard (62$)
micro-sd low profile adapter (11.9$)
wifi dongle (18.6$)

Not so cheap but might be worth it.

after I got a couple of RaspberryPi (model B rev2 512MB RAM) devices for a project I'm working on (details about the project will come later if it will work).
I bought a SD 8GB card and hooked it to the my desktop computer.
I downloaded the ARCLinux image unzipped it and followed the instructions:

Replacing sdX with the location of the SD card, run:
dd bs=1M if=/path/to/archlinux-hf-2013-06-15.img of=/dev/sdX

In order to increase the SD space to it's full 8GB I used gparted.

upgrade system:
> pacman -Sy pacman
> pacman-key --init
> pacman -Syu
sync (3 times):
# sync
reboot:
# reboot
Now I can add packages using pacman. to look for a package use:
> pacman -Ss <package name>
then select the one you need and install by:
> pacman -S <package name>

I am currently testing the Pi with the screen and 8 AA batteries to see how long it can work. I plan to build a case with a built-in 3.5" screen and a miniature keyboard so I can have a control unit for other RaspberryPi devices. My initial thoughts of using an old Gamboy case are now switched (with the encouragement of my electronic tec. guy David Shtibelman) to build a costume made case. Of course, David is the one to build it, since I have no skills to do that.

Creating a software raid

1. format the hard drives to raid using fdisk (gdisk in fedora 17)
or with parted (for 3TB disks):

> parted /dev/sdx
        (parted) mklabel gpt
        (parted) unit TB
        (parted) mkpart primary 0 3001G
        (parted) print
        (parted) quit

2. Create the raid device (change level and devices as needed):
mdadm --create --level=1 --raid-devices=2 /dev/md0 /dev/sd[bc]1
3. format the file system of the new device:
mkfs.ext4 /dev/md0
4. create the mdadm.conf file:
mdadm --detail --scan > /etc/mdadm.conf
5. add mail alert:
echo "MAILADDR your_mail@somthing.com" >> /etc/mdadm.conf
echo "MAILFROM your_server" >> /etc/mdadm.conf
6. test mail alert:
mdadm --monitor --scan --test --oneshot
7. start mail alert deamon:
systemctl enable mdmonitor.service
systemctl start mdmonitor.service
8. edit the mail aliases file to forward root messages to your email:
vi /etc/aliases
add your mail under the line:  # Person who should get root's mail

run newaliases command

see also at:
http://www.iceteks.com/articles.php/linuxmdadmraid/

* after name change, create the /etc/mdadm.conf file and run:
sudo update-initramfs -u
or
sudo dracut -f /boot/initramfs-[current].img

HowTo use a computer without a screen attached

The Problem:
I have a Linux machine (Fedora 14)  with no screen attached. I view the graphics via VNC server. I had to adjust the system to use the Nvidia driver and tweak it for no screen.

The solution:
First we need to install the Nvidia driver. Then, we need to adjust the X configuration for no screen and our preferred resolution.

for more details see: http://forums.fedoraforum.org/showthread.php?t=204752

1. Update the kernel:

su
yum update kernel*
reboot

2. Add rpmfusion to your repository.

su
rpm -Uvh  http://download1.rpmfusion.org/free/fedora/rpmfusion-free-release-stable.noarch.rpm   http://download1.rpmfusion.org/nonfree/fedora/rpmfusion-nonfree-release-stable.noarch.rpm

3. Depending on your card add nvidia akmod to the kernel and the proper driver. for a Gforce 9500GT on my 64bit machine:

yum install akmod-nvidia-260.19.36-1.fc14.x86_64 xorg-x11-drv-nvidia-libs-260.19.36-1.fc14.x86_64

4. Edit the xorg.conf file. don't forget to  adjust the Nvidia card for no screen attached by adding:
 Option "ConnectedMonitor" "CRT"
to the device section and adjust the screen resolution under the Screen section:

Section "Device"
    Identifier     "Card0"
    Driver         "nvidia"
	Option "ConnectedMonitor" "CRT"
EndSection

Section "Screen"
    Identifier     "Screen0"
    Device         "Card0"
    Monitor        "Monitor0"
    SubSection     "Display"
        Viewport    0 0
        Modes      "1280x1024"
    EndSubSection
EndSection

How to format a hard drive on Linux

The Problem:
How to format a hard drive on Linux.

The Solution:
in order to format a HD on Linux you'll need super user privileges.
you can use sudo if you have sudo privilages or in the terminal
switch to super user:

>su
(give super user pwd)
and ignore the sudo in the following commands. if you're system can't find the commands while in su mode, try to add /sbin/ before the command.
follow the steps below: marked with •
run commnads marked with >
commands output looks like this.
Warning: formatting your HD will erase all data on it!

• Connect the HD to the computer, make sure it's on. 
• get the hard drive's (device) name on your system:

>df -h

Filesystem            Size  Used Avail Use% Mounted on

/dev/mapper/vg_insarlablin-lv_root

                      908G   77G  786G   9% /

/dev/sda1             194M   31M  153M  17% /boot

/dev/sdb1             1.4T  1.1T  193G  86% /home/novitsky

/dev/sdc1             1.4T  1.1T  193G  86% /home/novitsky/data

Now lets say our HD is sdb1.

• lets build it's partition:

>sudo fdisk /dev/sdb

The number of cylinders for this disk is set to 182401.

There is nothing wrong with that, but this is larger than 1024,

and could in certain setups cause problems with:

1) software that runs at boot time (e.g., old versions of LILO)

2) booting and partitioning software from other OSs

   (e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): 

• pressing m will get this menu:

Command action

   a   toggle a bootable flag

   b   edit bsd disklabel

   c   toggle the dos compatibility flag

   d   delete a partition

   l   list known partition types

   m   print this menu

   n   add a new partition

   o   create a new empty DOS partition table

   p   print the partition table

   q   quit without saving changes

   s   create a new empty Sun disklabel

   t   change a partition's system id

   u   change display/entry units

   v   verify the partition table

   w   write table to disk and exit

   x   extra functionality (experts only)

• pressing p will print the current partition:

Disk /dev/sdb: 1500.3 GB, 1500301910016 bytes

255 heads, 63 sectors/track, 182401 cylinders

Units = cylinders of 16065 * 512 = 8225280 bytes

Disk identifier: 0x000923e6

   Device Boot      Start         End      Blocks   Id  System

/dev/sdb1               1      182401  1465136001   83  Linux

• pressing d will delete the current partition. depending on the partition you want to delete (this case 1) press the number of the partition to be deleted. if all the disk should be formatted and there are several partitions repeat pressing d and the number of the partitions until you delete them all.
• pressing n will create a new partition.
• pressing p will create a primary partition. select 1 as the number of the partition. use the defaults (just press Enter) for the start and end cylinders.
• pressing t will set the file system type.
• pressing l will list the available file systems.
• select 83 if you want a Linux file system
• pressing w will write and exit to shell.

Now we will create a new file system on the new partition:
> sudo mkfs -t ext2 /dev/hdb1
remember to change the number at the end if your HD device number is different.
if your using different fs than ext2 (newer Linux systems have ext3 and fc11 has ext4) replace the ext2 in the command with the correct fc name.
the mkfs may take some time to run (depending on the computer and HD size).

To add a label to the partition use:
e2label /dev/sdb1 NEWLABEL
where sdb should be changed to the correct device and NEWLABEL changed to the new label.

Now add the device to the fstab file in order to mount it:
> sudo vi /etc/fstab
and add a line like:
LABEL=NEWLABEL MOUNTPOINT ext3 defaults 1 2
where NEWLABEL is the disk labe and MOUNTPOINT is where you want to mount the disk.
mount it using:
> sudo mount -a
Note you might need to change permissions and owner of the mount point:
> sudo chown username:usergroop MOUNTPOINT
when finnished - reboot.

Hard Drive Resurrection

July 22, 2009: One of my hard drives turned out to be dead.
When I disconnected it from the computer - all is well, the computer is working fine without the missing HD, only I had very important data on it and must resurrect it.
when I connect it - the whole system is freezing So I belive it is something to do with the logic board. I googled it up and came upon this.
Basically, I need another HD with the same properties (manufacturer, model,firmware) and just switch the logic board which should be easy and hopefully work.
I now need to find a similar HD for the replacement.

July 23, 2009: Well, no luck in finding a similar HD and sending the HD to data recovery is more than 150$. Before I spend so much money I would like to check the cheapest option:
Freezing the HD. lets see what happens...

July 23, 2009, supplemental: it worked!
I was able to recover the data after few hours in the freezer. I hardly believed it, but it actually worked. I'll now spend my money buying a larger HD.
wooohooo!

Working on two machines at once

The Problem:
On my work I use two computers, one is Linux based which carry out all the heavy processing and another is windows based for viewing results, office related jobs and other Microsoft compatibility issues.

My problem was I had two mice, two keyboard and two screens. it is very confusing and tiring to switch from one to another.

The Solution:
The Linux distro comes with a wonderful component - VINO SERVER a VNC server which connects to your display 0 (the main display).
so I installed REALVNC viewer on my windows computer, connected one screen to the DVI exit and one to the D-sub (VGA) exit on the graphic card.
now, all need to do is to enter the Linux machine via VNC, drag the window to the second screen and use "Full Screen" view. This allows me to use one set of keyboard and mouse and move from one machine to another just by moving from one screen to another.

Tips:

• autologin:
  Since you can't "see" the Linux machine without connecting to it, you can tweak the bootloader to load you user automatically:
  edit /etc/gdm/custom.conf (as superuser if needed) and change:
  [daemon]
  TimedLoginEnable=true
  TimedLogin=yourusername
  TimedLoginDelay=0
• ssh first:
  if you don't want to use autologin for security or other reason, you can login via ssh and start a gnome session and then use the VNC viewer. just enable ssh daemon and vino server.
• othere VNC servers:
  you can always use different VNC server and connect to other displays then 0.
• graphic card:
  I tried to use this method with an onboard G. card and only got low resolution screen size.
  it might be my mother board (intel DQ965GF) or bad X settings but using a GeForce card I managed to use my desired resolution with no problems.