This blog has moved to https://walduk.at

Hello everybody! I will shut down this site soon as all content has been moved to

https://www.walduk.at

Liquid thermal paste in laptop

TL;DR: Don’t use it in portable things as it cracks

My laptop has become loud and slow over the years. This came from permanent 90°C CPU temperature and thermal throttling.

From my experience, the manufacturer never uses the correct of thermal paste, so I opened my laptop up and applied Coollaboratory liquid pro thermal paste.

This stuff is much better than the usual thermal paste. It’s actually metal (indium? germanium?). It also diffuses into the copper which makes a very strong bond.

CPU cooler with crystallized liquid thermal paste. Source: mine

Learning a sine wave with LSTMs

TL;DR: Copy & paste the linked code to have working Neural Network as a starting point for your application

50 consecutive prediction steps on a sine wave. Source: mine

Introduction

I tried several times to get started with Deep Learning. But pretty much all the tutorials were outdated and it was just a lot of work with no result in the end.

This time I invested my time into learning a sine wave. And it actually works! I’ve published the Jupyter Notebook on Colaboratory which you can just copy and get started within 5 minutes.

Deep learning made easy with Colaboratory

TL;DR: Get free a GPU and environment with Jupyter Notebook with this tutorial

MNIST dataset picture from the tutorial

I always struggle to set up the environment for deep learning frameworks. It seems you need the right set and versions of hardware, OS, drivers,framework. And if you want GPU support, things get even worse.

But google wants you to use their google cloud platform. They even give you the first shot for free a free server including Linux, ssh and static ip.

And google is one of the leading companies in Deep learning. They combined both with Colaboratory. You get:

• Tesla K80 GPU shared instance for up to 12h / day
• complete software environment with Jupyter Notebook
• linked to google drive so you can share with your friends other people
• you can access the base operating system easily and install other software like opencv

Someone else invested a lot of time and work to explain this. Follow fuat’s tutorial here.

Out of memory errors

Well, this is free, so you only get the ‘free’ resources. You can check how much with this snippet

# memory footprint support libraries/code
!ln -sf /opt/bin/nvidia-smi /usr/bin/nvidia-smi
!pip install gputil
!pip install psutil
!pip install humanize
import psutil
import humanize
import os
import GPUtil as GPU
GPUs = GPU.getGPUs()
# XXX: only one GPU on Colab and isn’t guaranteed
gpu = GPUs[0]
def printm():
 process = psutil.Process(os.getpid())
 print("Gen RAM Free: " + humanize.naturalsize( psutil.virtual_memory().available ), " I Proc size: " + humanize.naturalsize( process.memory_info().rss))
 print("GPU RAM Free: {0:.0f}MB | Used: {1:.0f}MB | Util {2:3.0f}% | Total {3:.0f}MB".format(gpu.memoryFree, gpu.memoryUsed, gpu.memoryUtil*100, gpu.memoryTotal))
printm()

The good output should look like this:

Requirement already satisfied: gputil in /usr/local/lib/python3.6/dist-packages (1.3.0)
Requirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from gputil) (1.14.3)
Requirement already satisfied: psutil in /usr/local/lib/python3.6/dist-packages (5.4.5)
Requirement already satisfied: humanize in /usr/local/lib/python3.6/dist-packages (0.5.1)
Gen RAM Free: 12.6 GB I Proc size: 308.3 MB
GPU RAM Free: 11438MB | Used: 1MB | Util 0% | Total 11439MB 

If it is not enough you should kill your instance and reconnect. Make sure you comment this line again after execution!

 # !kill -9 -1 

this is the source with other scripts

Connect Batteries easily

TL;DR: Use magnets to connect batteries

Most Batteries contain Iron or Nickel. Just use a magnet to  connect the battery. On the other side you can use one of those cheap copper-clad cables from china or alligator clips

Bonus: The magnets automatically disconnect if they reach ~90°C in case of a short circuit.

Fixing a chinese 3D Printer: Part 2

TL;DR: I’ve invested way too much time and money in this thing

Introduction

I’ve printed for over a month now. Various things broke and some improvements weren’t improvements after all.

 

New bed surface

I’ve tried a lot. The best is borosilicate glass. It just works and the bottom is sooo smooth.

 

This here is a ipad screen protector. It turns out that you can’t cut tempered glass with a box cutter.

But I managed to keep it in one piece as the bottom side is made of polycarbonate.

Fixing a chinese 3D Printer: Part 1

TL;DR: You have a lot of work to do until this is usable

photo of a random print at 120mm/s

Introduction

I’ve bought an Anet A8 für 143€, everything included. It’s a clone of the the Prusa i3 and available under all kinds of brandings and with every changing components. Here are the things I’ve learned on my way.

 

Common Errors which have to be fixed

Raspbian headless install over Wifi

TLDR; Configure your new PI without a monitor

Just follow this guide to install your raspberry pi 3 without a monitor. It enables wifi and ssh in the default raspbian installation.

1. install raspbian on your sdcard:

export DEV=/dev/sdX; \
unzip 2017-04-10-raspbian-jessie-lite.zip ;\
cd 2017-04-10-raspbian-jessie-lite ;\
sudo dd if=2017-04-10-raspbian-jessie-lite.img of=$DEV bs=4M && sudo sync

1. mount boot partition
2. enable ssh:

   sudo touch /mountpoint/To/bootPartition/ssh     

3. mount root partition
4. edit /etc/wpa/wpa_supplicant.conf to:

   country=GB
   ctrl_interface=DIR=/var/run/wpa_supplicant GROUP=netdev
   update_config=1
   
   
   network={
   ssid="MyWifiName"
   psk="MyWifiPassword"
   }

5. nmap for new devices: 

   nmap 192.168.178.0/24 -p 22 --open

   ...
   Nmap scan report for raspberrypi.fritz.box (192.168.178.137)
   Host is up (0.0099s latency).
   PORT   STATE SERVICE
   22/tcp open  ssh
   MAC Address: B8:27:EB:1F:25:A2 (Raspberry Pi Foundation)

6. login with password raspberry: 

   ssh pi@raspberrypi

7. do whatever you want: 

   sudo apt-get update && \
   sudo apt-get upgrade -y && \
   sudo apt-get dist-upgrade -y && \
   sudo apt-get install --reinstall raspberrypi-bootloader raspberrypi-kernel
   

And that’s all.

Lowside current measuring gone wrong

TL;DR: Always take into account the ground bounce.

Introduction

I’m developing a software controlled Switched Mode Power Supply (SMPS). For this I need to measure current at ~250kHz. This is a recollection of the false assumptions and mistakes I made.

The circuit

I’ve decided to measure the current lowside as it seemed easier when I stared. And lucky me, this problem has been solved countless times before and the ideas collected in the application note AN105 by LT.

My circuit is basically this:

Figure 26: “Classic” Precision Low Side Current Sense. Source: AN105 by LT

Earn money with your smartphone

TLDR; Let your smartphone mine cryptocurrencies

I like money. Money is even better if you don’t have to work for it. So I set up my smartphone as a zcash node. Zcash is similar to bitcoins: You exchange computing power for virtual money.

This post is twofold: First the hardware modification to keep the CPU cool, second a rough sketch of the software.

Hardware

I’m using my old phone. This is a LG G3 D855. The used CPU is a quad core with 4 x 2.5GHz Qualcomm MSM8974AC Snapdragon 801.

Unfortunately, LG messed up. Here is a picture of the CPU:

You can see that the CPU is recessed a bit. There is a corresponding recession on the phone’s display side. But there shouldn’t be a recession, it should be slightly elevated to contact the CPU!