Wednesday, January 20, 2016

Fixing a Kobo Original Wifi

I had a kobo Wifi (n289) that started:


  1. Not showing any files suddenly (No books found)
  2. Constantly showing "updating" forever
  3. Bootlooping (to taste)
I did not like this new behavior.

Here's how I fixed it:


  1. Fiddle around with hard resets or whatever  (This will not help, but you can try it if you want).
  2. Crack open the kobo.
  3. Take the SD card off the back and put it into a reader.
  4. Put reader into ubuntu PC. If you don't have linux, oh well.
  5. Open Gparted.
  6. Gparted will show three partitions on the sd: a 270mb one, a 530 mb one, and a third huge one.
  7. run sudo dd if=/dev/(sdcardpartitionone) of=image.img and  sudo dd if=/dev/(sdcardpartitiontwo) of=image2.img to make a backup (optional)
  8. Delete the 270mb partition in gparted.
  9. Run fsck on the 530mb partition and accept everything it wants.
  10. ???
  11. Profit!

Thursday, November 19, 2015

Agents Of Physics

I was binge-netflixing Agents Of Shield (great show) the other day on vacation. In one episode, a character with the power to cool stuff freezes all the water around a battleship in just a few seconds, preventing it from moving. This man wasn't an alien, just subjected to a machine and a bit of "lightning bolt hit him" handwave, so he's prolly guided by physics, even in-universe.

When I watched this, I thought: where's all this energy going? Seriously: he dropped the temperature of the water to 0 degrees from ambient, then froze it. That has to be a substantial amount of energy. How much is it?

The answer actually depends on what season it is. According to the Agents of Shield wikia (of course it exists), this took place in the Port of Casablanca. The coastal temperature swings between 17-23 degrees celcius each year. It looked reasonably chilly during the scene, and everyone was wearing a jacket so I'll assume that this took place in one of the cooler seasons with 18.5 degree water.



The enthalpy of fusion of water (energy to freeze) is about 6 Kilojoules per Kilogram, and the energy to drop water by 18.5 degrees is about 77.5 kj. That's a pretty substantial amount of energy; added together, that's about 20 grams of TNT per kg of water frozen. Speaking of which, my new favorite unit for measuring specific heat is the kilogram of TNT per degree kelvin, which I propose to name the "shockwave enthalpy":

Unit cancellation is weird.

Now we need the volume of the water frozen. Looking at a still of the scene,

You can see that the ship in question looks like a sort of battleship. I'm not well-versed enough in shippery to know exactly what kind of ship that is, but it looks like it has a gun, and the first battleship I could find was 260 meters long, and perhaps 30m wide (a sharp eyed redditor found a width of 21m, so the calculation is under 10% off.). Good enough for a first approximation. Now, the ice penetrates quite deep. It looks like it's half as deep as the ship is wide, so we can say that it's perhaps 10m deep, and a perfect half-annulus, which gives us a volume of about 163,363 m^3, which is (breaking out scientific notation) 1.6×10^8 kg of water. Given that the displacement of $battleship is about 35 million kg, that seems about right. To freeze it from ambient, we get 1.36×10^10 joules of energy. That is a crapton. Actually, it's 3.2 metric craptons of tnt equivalent (a cube 13 meters per side):

To put that in perspective, you know that huge explosion on Mythbusters? With the cement truck? That was only 1/8th the energy this dude just sucked out of the water.

Also, that amount of energy is the same as that contained in 73,000 strips of standard 44 kCalorie bacon.

So the area was as cold as a bonfire of a hundred thousand strips of bacon is hot.

That's a nice unit.


This was fun; I think I'll be doing this more often here; perhaps make it into a weekly thing.


(Note: I hope I didn't mess up anywhere. Please do let me know if you spot anything wrong.)

Thursday, August 20, 2015

Plasma speaker

One 10v-4200v transformer + audio amp = plasma speaker. Yes, it's that simple, though good luck finding the transformer. In reality, it seems to put out about 40,000v; takes 10000v to ionize one cm, and the longest arc I've gotten is about 4cm. If I touch a dry 2x2 to the positive, you can actually feel a tingle, which is interesting. For some reason, I seem to be getting a yellow color from the arc at the highest power levels along with the typical purple; I'm not quite sure why. Ionizing CO2, perhaps?




Should try and see what this does to the sonoluminesence.

Monday, July 6, 2015

Igniting sonoluminescence

After a few months of trying to get all the right equipment in the right place and not on fire at the same time, I've finally managed to ignite sonoluminescence. Obligatory proof photo:



Yes, that is a tablet case in the background. It's the only black thing I could find at the time [1], OK? Sheesh. This was taken in the darkest conditions I could manage in the room; the room looks all glowy-like because this was a 4 second exposure. It's really quite surprising to see how much more sensitive our eyes are than cameras at low-light conditions: when everything is tuned up correctly, you can see the glowing bubble quite readily with the naked eye. But even through a magnifying glass, the point-and-shoot camera I was using saw nil on video mode and normal picture mode.

Credit where credit is due; one of the best resources for this phenomenon is techmind.org/sl. I drew a lot of information from this site, and the chart of transducer levels to bubble regimes was particularly useful. I actually used a rather different setup than he did, but it's possible that his system would be easier to setup for some.



The components I used:

Transducer:

I used one single 25khz 60w piezo ultrasonic cleaning transducer bought off ebay, which ran me about $20. I don't seem to be able to find the original seller. I originally used a small piezo disk (glued somewhere on the flask) as a sense transducer, which worked very well until I destroyed it.

Amplifier:

I tried a 100w 28khz ultrasonic cleaner driver board, with the intent of hacking into the oscillator and de-tuning as necessary, but this didn't work out as I imagined, even though the very helpful ebay supplier supplied his schematic. Also, as a result of me being rather reckless with 115v systems, I lost a few pieces of equipment trying to use this. Also I blew two of the driver boards up quite spectacularly, but we won't talk about that.

I ended up using a regular audio amp (I used a 7 watt kit from sayal, but other ones turned out to work just as well), connected to a 115v-6.3v line transformer run backwards, with the 115v side connected to the transducer, and the 6.3v side to the amp.

Signal generator:

 I purchased a velleman HPG1 signal generator, which was 150 dollars very well spent. I did manage to blow the output op-amp nary an hour after I got it, but that was a relatively easy fix. As were the next two times I blew it up.

Flask:

This ended up being quite important. I initially tried an off-round flask from Sayal, but this didn't have any sort of reasonable resonance. Don't get anything that isn't spherical, because they won't even.  I found a completely spherical distillation flask at amazon, which worked quite nicely indeed. I tried the 50ml version first, but the resonance was at too high a frequency, outside my amplifier's range. The 150ml version worked perfectly, and is what you see in the photo above.

The setup was very easy; connect some 12v power supply, amp, transformer, and transducer, glue the sense transducer and the power transducer onto the flask, and it's done.

The device only consumes about 1.1 watts while operating, which includes a large loss from the amplifier.

Now I just need to experiment around with it. Suggestions welcome for experiments I should do!


From Putterman et al, we see that about 10^6 photons are emitted in each acoustic cycle (taken from the techmind page).


[1] The black baffle was required to stop light from the signal generator's 10e42 watt apparent brightness LED backlight, which cannot be disabled or dimmed at all. All it would take is five lines of code and one single transistor.....