[Tech tips] Fixing an overheating Toshiba Portegé M200 tablet PC (2)

This is an additional note to a previous post. It is a tip on how to orient the tablet when using it upright (i.e. using it as a tablet on a book stand as depicted below, or held in your lap).

I have found out that the Portegé M200 tablet runs much cooler (and quieter) when the left side (nearest to tab key) or the back side (nearest to function keys) faces up (towards the ceiling) than if the right/bottom side faces up.

GOOD ORIENTATIONS... back side up / left side up


BAD ORIENTATIONS... front side up / right side up


The reason is easy to understand given the design of the cooling system (see pictures). Hot coolant rises up, so coolant circulates much better if the radiator is physically higher than the CPU. If you do the opposite, the heat will tend to rise towards the CPU, away from the radiator, and cooling will be less efficient.

Even if you use the tablet flat on a desk, tilting it slightly in the above orientations, with a book or something, should help to cool it too (I haven't tested this).

[Tech tips] Fixing an overheating Toshiba Portegé M200 tablet PC

[rough translation from Japanese]

I've been using my Portegé M200 tablet PC (sold in Japan as dynabook SS M200) for over a year now. It's followed me around the world, and runs through almost all of my waking hours, so it's amazing that it came this far without a hitch.

Just recently, I'd noticed that the fan was running all the time, and the exhaust was pretty hot. Then, all of a sudden, in the middle of a calculation, it just shut down without any warning. No blue screen or anything, the fan just ramped up for a second, and then everything went black, as if I had pulled the cord and removed the battery. I restarted the computer, but it would only run for a few minutes before dying in exactly the same way. I tried running it with low CPU clock, and it ran fine without this problem.

Given this, I diagnosed overheating, and proceeded to clean the cooling system. This computer does not have a fan directly on top of the CPU, but instead, leads the heat to a radiator on the very edge of the computer, and blows air through this radiator. (Photo: The X-shape bracket on the lower left is holding down the CPU. The copper heat conduit tube leads to the radiator on the lower left corner. This radiator is cooled by the main fan.)


I took out the fan and found a huge buildup of dust on the inside edge of the radiator. There was so much stuff there, completely blocking the radiator, that it's a surprise it even ran at all.(Photo: Yellow arrow, area where dust was found. Blue arrow, dust removed with a Q tip.)


After this procedure, the tablet is running just fine.

(to be continued)



+ Opening your notebook will probably void your warranty. Open at your own risk. It should be very easy to fry a chip or two with static.

+ When opening, don't forget the little screw hidden underneath the battery, and the screw in the harddisk case cover. There's another small fan on the lower plate, so carefully remove its lead from the motherboard when taking off the bottom cover.

[Cool toys] Tablet PC (2)-- Manage your reprints electronically

Most practicing scientists are inundated with countless piles of photocopies and reprints. Since my chosen field is neuroscience, the situation is particularly acute; given that neuroscience is a complex patchwork of many disciplines, the nature of the field defies any simplistic effort towards classification/filing. I also have random reprints that I've accumulated as a medical student; this only complicates the mess. Furthermore, scientific literature is multiplying steadily in accordance with de Solla Price's exponential law; the paper mess can only become worse in the future.

---

I have been addressing this problem by turning completely to PDF files on my tablet computer. Even old reprints, not available in electronic format, have been scanned. The tablet allows me to scribble, just as if I had a physical copy in front me.

I calculate that
(A) the time saved (no need to fish through piles, or re-copy lost papers) and
(B) the convenience (I have my whole library with me at all times, and it can be backed up)
more than justify the initial investment.

I use a combination of the following software packages...

• Reference Manager... this package allows me to maintain a sorted database of my references. The user interface is somewhat more to my liking than EndNote, especially because it allows fine-tuning of the display screen. I use a custom user-defined field to categorize my papers into little "piles," and by sorting with this field, I can browse through these piles in the main window.
  
  
• Ideally, I would like to switch to JabRef, which is an open-source Java database program based on BibTeX, but the lack of easy bibliography management in MS Word keeps me from switching. JabRef has a very useful categorization scheme, based on a directory-like tree structure (see upper left panel in screenshot); links from each reference can be dragged to multiple nodes on the tree, which is ideally suited for the multidisciplinary nature of neuroscience.
  
  
• PDF Annotator... Clicking on the Reference Manager entry brings up this software, with the PDF in question. The old version of Adobe Acrobat responded too slowly to pen input for scribbling hand-written notes. I haven't tried Acrobat v.8, but presently, I am quite happy with this shareware package. In combination with the tablet PC, it allows you to scribble and highlight your PDF files, just as if you had a real copy before you. Currently, it does not support text highlighting (i.e. you must carefully trace your pen over the line to highlight) or text copying, and you cannot type in comments.

[Cool toys] Tablet PC (1)-- Dissecting the tablet pen

I've been a tablet PC user (Toshiba Protege M200) for over a year now. The tablet is an amazing piece of engineering; it allows you to write directly on the laptop screen, as if it were a piece of paper. The digitizer pen writes like a smooth felt-tip. It even senses your writing pressure at 8 bit resolution, so in some graphics programs, if you press harder, you get a thicker line.

How does it work?

The digitizer pen has no power source, and is just an RLC circuit that resonates at a certain frequency-- this resonant frequency is modified by the pen pressure and by button presses.

To read the location of the pen, the tablet first sends out a pulsed radio wave through an antenna array which is embedded in the screen. This is recieved by the pen's circuitry, which resonates at a given frequency. Next, the tablet switches to antenna mode, and senses where on its antenna array the resonating pen is located. It apparently does this 50 or so times a second.

One of the few complaints that I have with the Toshiba unit is the ridiculously cheap digitizer pen. It's way too light, and feels like a cheap plastic ballpoint--which is not what you'd expect from an expensive unit.

This cheap digitizer pen broke on me recently, and I guess the least I can do is to post a photo for those who might be interested but don't want to destroy their pen for curiosity's sake. (The photo shows circuitry from the top part of the pen, which acts as an eraser. I imagine that another circuit exactly like it is embedded in the tip as well.)

Links

• How does a Graphics Tablet work?
• Wacom Graphire graphics tablet