Jul 27


The newest addition to my swarm of never ending pile of interesting electronics is the Arduino Yún.

As I was getting some parts from Newark Canada, I just thought, well i’ll get myself an Yún to play with as well.

Yún means Cloud in Mandarin, so it’s quite obvious it is an Arduino meant to be connected to the cloud (Internet of Things).


As I usually do, I hack into the deeper workings of devices, and in the case of the Yún, there will be no difference.


I will dig deeper into it’s internals – have You ever seen the “Black magic” hiding under the shield? I haven’t.

Also, later I will look into how much power this unit takes, and to see if there is any way to tweak the power usage.



Setting up the Yún was a breeeze – I followed the instructions over at http://arduino.cc/en/Guide/ArduinoYun

Quickly I had it connected to my own WiFi-network, and programming it over Wifi with Arduino 1.5.7 (Beta at the moment) worked quite nicely.


Well, let’s get to the point, what exactly is hiding under that EMI-shield?

A quick desolder job with the Solder-wick (Chemtronics 10-100L, the best money can buy!) and the cover was off.

Yun uncovered - overview

Yun uncovered – overview


Aha! – three chips – that ain’t much.

On the left we have a Winbond W9751G6KB-25 – 16 Bit DDR2-RAM – this is the main memory for the DD-WRT linux-environment that the Yún runs.

In the center there is the Atheros  AR9331-AL1A Wireless SOC – This is where all the magic happens.

And on the Right, once again from Winbond, a 25Q128FVSG – a 128 MegaBit Boot-flash, using SPI-interface.


So, now you know what hides under the shield, and you do not need to void the Warranty on your Yún to find out 😀

– Due to time constraints, I will leave the power-measurements to the next blog-post, as I need some time to gather readings, and present them in a usable manner.

Apr 23

I got my sticky hands on some very nice vintage LED-displays, and some days ago i got a pack of nice prototype-PCB’s from eBay, so of course i had to make two small test-boards for these displays 😀

This one is for 4 pcs. of TIL311, intelligent Hexadecimal display (shows 0-9, A-F)

TIL311 LED Display

TIL311 LED Display


The next board is for a DL1416 display. It can display all numbers and some Alphanumeric digits. It has a 7-bit Data-bus.

DL1416 LED display

DL1416 LED display


The plans for them? Well i have enough of both types to warrant the time needed to layout a design for a very limited-edition wristwatch :)

Jul 13

Finally got time to shoot a vide of the cube working, what do You think ?

Feb 08

First initial testing, after fixing some minor errors i made in the layout (Swapped RX/TX, not first time) and swapped D & S on the Fets. Grrr…

Dec 17


I am in the process of installing a Nexus 7 Tablet permanently in my Car – Using it for Navigation software and playing music.

I need to be able to tell the Nexus 7 to pause playback of music, and go to sleep. The easiest and most simple way is to touch a NFC tag to it, that tells it to do just this (Using a combination of NFC Task Launcher & Tasker)

But, how to enable the tag at will ?!

Solution: Put a CMOS switch in series with the antenna of the tag – this way it will only be readable when i tell it to!

A good solution is the 74HC4066 or even the DY411DY, but those are just too big an complicated. Why not resort to a single CMOS switch – Enter the 74LVC1G66.

A small SOT353 device that does it all, albeit a bit small for most soldering irons, but nothing my Weller WD1000M can’t handle.

I drew up a board in Eagle cad and milled a PCB. It didn’t work in the first try, it seems I damaged the RFID chip when removing it from the blob of glue inside the tag, so next chip i mounted sits on a big glob of glue, no need to risk damaging another chip!

I the second try, it works perfectly! – The chip needs a supply voltage of 3,3 or 5V and a logic high signal to trigger the Tag “ON”


Here is a Video of it in action!:


Attached is the Eagle files for your perusal.




Feb 12

Recently, i purchased a big lot of Philips ZM1000 Nixies, and needed a test PCB to test out if all was good – i got sockets for a number of the tubes, so i settled on making a PCB with the correct socket, so I easily can test them.

Being fond of the Arduino, i just thought, why not make it an Arduino shield – this way it’s a nice compact solution, powered by the USB-port, and no extra cables needed.

I earlier on aquired one of the real nice high voltage supplies, perfect for nixies from Taylor Electronics

So i made a single PCB that fits the small power supply module, a socket for a 74141 Nixie driver, and a socket for the ZM1000 tube.

I milled the PCB on our PCB milling machine in the local hackerspace where i frequent, Labitat

I did not mount all the headers on the finished PCB, since they are not in used. The use of stacking headers are a little waste, since there is no way I can mount anything on top of the Nixie PCB, but they make the PCB headers more mechanically stable, since they are soldered on on the bottom side of the PCB.

A small video of the circuit in operation, running a simple counting program, testing all digits: Youtube



You can get the Eagle PCB files for the project here: ZM1000_Shield

Jun 10

A member of Labitat, [Kulitorum] donated around 50 telephones to labitat since his work place scrapped them.

I took one of the phones home to reverse engineer the LCD. After about 10 minutes with the logic analyser, it was clear that it’s based on the Hitachi HD44780 or compatible controller.

Just connect the contrast pin to GND, then contrast is perfect for most projects – unfortunately the LCD is meant to be viewed from an angle, and is not good for a front panel etc. that you will view front-on.

I drew up the connections here:


Some pctures of the LCD and the adaptor i soldered up.














The phones contain some other nice parts, a DC-DC converter module some Xtal’s, Electret microphone, speaker etc.


Jun 01

Was cleaning out in the shed today, and found an old piece of kit, some kind of controller box from Evans & Sutherland with 8 encoders and 8 pcs. 8-digit LED readout

The Display units are intelligent Alphanumeric LED readouts, have to do something with them some day… Anyway, the encoders is HP QEDS-7596, enclosed optical encoders, 512 pulses per revolution. Very nice encoders, must have cost a small fortune!

Wired it to my Arduino (yep, still playing with it, hehe) and opened a serial terminal to see the position of the encoder wheel, works nicely, the resolution of these encoders are just insane.

Googled around, i wanted to show the result on my computer (i’m using a Mac) and have read about processing, but never tried it. Now i had a reason to try it out.

Downloaded it, and got it running in a few minutes, it’s just as plug and play as the arduino, very nice!

I’ll upload some pictures of the result soon.


May 29

I have these displays from an old Ship navigator i scrapped for parts.

It was a Philips Navigator, model F 505-2

I kept the displays as the three PCF8576 controller chips sat nicely behind the display glass, so i cut the PCB along the smart mechanism that holds the display glass and elastomers tight to the PCB :-)

It’s a quite curious design, the first 9 digits are standard 7-segment letters, and a bunch of symbols.

The last 18 digits are 11-segments, nor 12, 14 or 16 which seems to be the “standard” around…

Each of the last 18 characters needs 12 data bits, so writing them becomes quite a challenge, as one can only send 8 bits at a time with the I2C-bus. So writing the first character is 8 bits and you can control 8 segments, and the next byte, the first nibble is the remaining segments for the first digit etc.

So addressing the display is like this

  1. DIGIT1(8 Bits)
  2. DIGIT1(4 Bits)/DIGIT2 (4 Bits)
  3. DIGIT2(8 Bits)
  4. Etc….

Before scrapping the navigator completely, i hooked up my logic analyser to see how the chips were adressed via the I2C-bus.

A shot with all segments lit:

I wrote a simple and _ugly_ program for the Arduino to test the display, and the results can be seen here:

You can get the Arduino-program here.

May 22

Waiting for my RepRap to be finished i could just as well get to know the Arduino hardware, so i installed the programming environment from Arduino on my Mac. That went real easy and it worked in the first try. Success!

What do you start out on with a new board/compiler ? Blink an LED of course! Looking in the /File/Examples/Basics/Blink there was a test program to blink the LED on the Arduino board. Worked just fine :-)

Next i needed to try something a bit more exciting, so i dug up my trusty ‘ol PLED display (Organic LED) with a KS0066 controller (it’s HD44780 Compatible) and tried to get it work – no go, it simply wouldn’t work, so maybe it got damaged during moving my stuff – who knows.

I whipped up some wires and a bit more standard LCD display, the 162COG-BA-BC and it worked fine.

But what is fun using a standard LCD – that has to be the oldest hack in the shed – apart from the Blinking LED example.

I knew i had some NOS VFD displays kicking around somewhere, that i got from a former employer. They was controlled by SPI bus and were 1×16 character standard ASCII.

I dug them up from my Display box (yep, i actually have two boxes ONLY with displays and display-related technology, among Nixie tubes etc.)

Tried getting them to work by adjusting the example suppled with the arduino found in /File/Examples/SPI/DigitalPotControl

It simply wouldn’t work reliably, i got i to show some very dim characters and it gave me a audible whine when i changed the display contents, but almost nothing was showing on the display.

Then i caught the smell of fish! – it turns out that two capacitors on these NOS displays was leaking Electrolyte all over the PCB and main switcher, i just hadn’t seen it. Incredible for NOS electronics that has never been used (i hooked these up for a short moment when i got them years ago, but otherwise unused)

Capacitors C4 (10μF/50V) and C9 (100μF/10V) seemed to be bad – especially the C4 (SMD) it was leaking all over the switcher. I replaced both and the display sprang to life immediately.

A datasheet for the display can be found here: CU165ECPB-T2J and here: CU165ECPB_T2J_specification

You can get my test routine here – it’s very ugly but it works.

UPDATE: After looking a bit around on the net, i actually found someone else that has been playing with the same display, just on a PCB from HP. You can find documentation and a library here