All posts by leviathan

Sunxi U-Boot SD card, done right

First get the most recent u-boot code with:
git clone https://github.com/jwrdegoede/u-boot-sunxi.git
cd u-boot-sunxi
git checkout sunxi-next

compile the code:
CROSS_COMPILE=arm-linux-gnueabi- make Cubieboard2
and now flash it, where /dev/sdc is your SD card device:
sudo dd if=spl/sunxi-spl.bin of=/dev/sdc bs=1024 seek=8
sudo dd if=u-boot.img of=/dev/sdc bs=1024 seek=40

MendelMax (times two)

The following picture shows the recent status of my “MendelMax times two” project.
I scaled all the extrusions by two, the bottle of water is there in order to give you an impression of how big the printer will be in the end.
50cm x 50cm x 50cm is the area I will be able to print at least for sure.

20131201_013837

Update: R2C2 firmware (cooler PWM)

Hi
I’ve done some tinkering at the R2C2 firmware in order to parametrize the maximum duty for the heating unit and also optimizing the power of the cooling fan of the extruder in order to provide a more stable temperature.
I haven’t yet tested my code but you can already take a look at: http://git.o2s.ch/?p=r2c2_firmware.git

Psychopath probability index

If you meet someone and you aren’t sure if the person in front of you might be a psychopath, there is a pretty simple formula which appears to be correct most of the time.
Let’s say we take $$p$$ as the number for the ranking someone has in a certain hierarchical structure and $$m$$ for the number of possible available rankings within the said structure (for instance top management), then we norm with $$n$$ which is the number of subjects involved within the leadership of said hierarchical structure layer (for example chair holder, or board member), then we can formulate the psychopath index with:

$$\psi(n,p,m)=\frac{p}{n \cdot m}$$

Examples 1:
You meet a top manager of a multibillion pharmaceutical company.
The company has around 20 different rankings you can get, the woman you meet is under the manager which answers to the board, so we have $$p=18$$ and $$n=20$$ as well as $$m=1$$ since she has no coworkers in her management layer.
The probability that the person in front of you is crazy is around 0.9 or 90% which tells you that’s a pretty bad idea to turn your back to this person.
If she had a coworker the probability for craziness would have been less then 50%, but since she is working alone, maybe she has blown up the car of her co worker or has her burned alive in car wheels as the Mexican drug cartels are used to do. You’ll never know.

Example 2:
You go to a quite place in Elsace where you have at least 3 hectares for yourself, but there is a neighbor with his own house some mile away.
You meet, so there are only two people, of which one lives self sustaining.
With $$m=p=1$$ and $$n=2$$ we get a fifty fifty chance that one of these two is a psychopath, which turns to be correct.
Bazinga!

Lemma:
“American psycho” proofs to be quite realistic.

Fish and chips

First the chip issue:
I’m developing the A1X SD/MMC host driver for linux-next right now. As soon as Emilio has found a solution for the accessing of the clock delay registers of the mod0 clock I will have a functioning driver. Then I can clean the code base up, rebase it and move it from the plsdsif-branch (Please don’t shout it Finnish) into the for-next from which it should be merged into Torvalds’ branch soon.

Now the fish:

20130907_195618

This son of a b** ate away my bate three times before I finally was able to get him onto the hook and drag him out.It was the biggest what was left after these three Russians who were there before me had fished out every big fish in the area >_>

RepRap

I’ve now finally fixed some minor issue with the x-y-ratio and the printer finally prints virtually perfect circles, when it is required by the CAD.Also the dimension appear to be correct now.
Only a small issue with the bonding of small areas, as they exist within the x-carriage.stl of the PrusaMendel plate, still persists.
Maybe I have to raise temperature and slow down the movement of the carriage at such small perimeters a little bit further.
I’ll try it next Wednesday, when I’m in the HackerSpace again.