Saturday, May 16, 2015


A friend of mine ordered 50pcs of 24C02 EEPROMs in SO8 packages for really cheap and decided to just hand me 1/5 of them. Apparently he doesn't have any use for 2Kibits of memory.
Datasheet says it's in the quite usual form of 32 pages of 8 byte blocks. They are interfaced with a general two-wire serial interface. And unfortunately need to be addressed by a hard wired address of 3 bits. That allows for only 8 of them on a single bus.
So these little chips are just general 2Kbit EEPROMs.
Apparently I don't have any use for them either, they are too small in memory size and have a too big of a footprint to fit in any of my designs.
But they are cool to play around with.
So just for the learning experience I decided to add another project in my ever lengthening list of them: A RAID system of tiny EEPROMS.
I'm thinking it would better my understanding of the inter-workings of such a system. 8 memory devices in different RAID configurations.. or maybe even try implementing ZFS in an embedded system... Sounds fun.

I broke out one for testing.

And even auto-routed a mock-up PCB, still needs some resistors and capacitors but that's about how I envision it.

Friday, April 10, 2015

Dissolving copper in a chemical bath by immersing and agitating.

WARNING: The chemicals used in this process are dangerous!
Dihydrogen Monoxide is a colorless and odorless chemical compound, it is a polar solvent capable of solving hydrophilic substances. 
Iron(III) chloride is toxic, highly corrosive and acidic. The anhydrous material is a powerful dehydrating agent. Ingestion of ferric chloride can result in serious morbidity. 
These experiments were conducted in a well ventilated area while wearing all of the appropriate safety gear and having taken precautions against any spills or splashes.

Dissolving copper in a chemical bath by immersing and agitating.

The use of iron(III) chloride to dissolve away copper from a double sided 35 µm copper-clad fiberglass board.

The solution was prepared:

As the iron(III) chloride dissolves in dihydrogen monoxide at the rate of 92 g/100 mL at 20 C the acid was added to the solvent at the rate of 110 g/100 mL at 30 C.
An amount of 20 g of sodium chloride was added to further increase the chloride concentration.
The solution becomes a iron(III) chloride - chloride ion - dihydrogen monoxide system.
The container used was a clear plastic box with a detachable lid. The dimensions were as follows: 15 cm * 10 cm * 5 cm.
The amount of solution produced was 320 mL.

The subject was prepared:

The test subject was a 2 cm * 1 cm piece that was attached to a 6,5 cm long 1mm diameter wooden dowel by a narrow strip of a double sided adhesive tape.
A thick layer of oxide was cleaned by the process of abrasion with a 1000 grit Aluminium oxide/Silicon carbide sandpaper.
The copper surface was then cleaned of any oils by application of a thin coating of industrial grade distilled ethanol with a soft cotton material.

The experiment:

Immediately after the last visually distinguishable signs of solvent had vaporized, the subject was introduced into previously prepared solution.

While maintaining constant depth the subject was moved around in the solution and observed.

At 20 second mark a discoloration of the surface of copper was observed. The copper seemed to acquire a lightening pinkish hue and became dull.
At 65 second mark only approximately 10% of the copper remained.
At 90 second mark the reaction was considered complete and the subject was removed from the solution and placed on a stack of porous cellulose material to absorb the solution.

The subject was then immersed in a tank of dihydrogen monoxide and agitated for 4 seconds.

Subject was again placed between layers of a porous cellulose material.
Momentarily pressure was applied and then the subject extracted.

Subject was then carefully detached from the wooden dowel and analysed to detect any remains of metal.

As by optical analysis no traces of metal were found the experiment had succeeded.

The use of iron(III) chloride to dissolve away copper is a viable method up to at least 35 µm * 2cm *1cm * 2 of copper.
The experiment can be further optimized by using a less pure solution as presented in another study "Effects on etching rates of copper in ferric chloride solutions" by Jian Cai ; Dept. of Mater. Sci. & Eng., Tsinghua Univ., Beijing, China ; Ma Jusheng ; Wang Gangqiang ; Xiangyun, T.

I wanted to know how long it takes to etch a PCB.
So I stuck a piece of double sided FR-4 copper board in new salted FeCl3 solution and got less than 2 minutes when you clean the board properly and move it around.

Thursday, April 2, 2015

An SMD 4 digit 7 segment diy display

I was designing an electronic clock to see time easier at night. And while at it, I came up with a nice idea:

Having used a lot of perfboards(dot pcb) to prototype my projects, I thought of a way to make a 7 segment display out of smd leds. Making a segment out of 2 common grounded leds..
I stumbled upon it while trying to determine a nice size for my clock. I drew a mask on perfboard with a marker. Later I redrew it and cut it out:

I soldered on all of the 58 smd leds:

And then used a 0.25mm(30AWG) coated wire to make connections:

That gave me a nice bunch of leds:

All of the segments are connected in the usual matrix style. Every segment is a row and every digit is a column. 
The 12 pin header is configured: A, B, C, D, E, F, G, Dots, D1, D2, D3, D4

I forgot to connect the dots ground. It can be connected to whichever digit as another extra segment.

Thursday, October 30, 2014

Wireless chargign

I am tired of plugging the charger in my phone. And I'm afraid the socket is going to break soon, seeing as it's about the fourth cable that has broken on me..
So as a solution I ordered a QI wireless charging set. A Note II receiver that claims to have charging current of 500-1000mA at 5V, works up to 5mm from the reciver at the frequency of 100-200kHz with efficiency of more than 70%.

I also bought 2 transmitters. One nice round plastic disk and the other one a bare PCB with a coil.
They both take an input of 1500mA at 5V and claim a 75% conversion efficiency.

And the other one:

The coils are made of strands of coated wire that are about 0.05mm(45AWG). The whole cable is ~1mm in diameter. That's about 300 strands!
It's stranded to increase the surface area to transmit more power.
Unfortunately the receiver's coil is just a copper track on a flex PCB. And has some surface area but not as much.

The charging rate is just average. Gets 460mA through. Not as claimed but good enough, falls in the 70% of 500mA. (The phone is claimed to charge at up to 2A!)

Tuesday, October 14, 2014

Yubikey Neo

So, I have wanted to buy this nifty little device called Yubikey Neo for a long time now.
It does authentication over USB and NFC. Another layer of security for me.

But there has been a tiny little problem. The website said that they'll be releasing a new version 'soon'.
I checked the site today, as I have done irregularly for the last year or so, and finally found out that "Today, Yubico is releasing its YubiKey NEO with support for U2F and delivering it in two form-factors. This key will hold the promise of a significantly more secure online consumer experience, and a dramatic increase in enterprise security and ease-of-use." That was published October 13, 2014. Yesterday.

Well, it was about time.. Unfortunately all of my money is tied up right now. So that $50 purchase will have to wait..

Wednesday, October 8, 2014

Smartwatch update 1

So, Now all I have to do is wait for the components to arrive.
In the meantime: Pictures!
All of them done by hand. No libraries. This will be purely my personal project. At least for now.
Not finished, will tinker with them a bit more. I'll probably publish all of the project when I get a working device that does at least the minimum that I expect it to do.

Smartwatch update 0

So I ordered a few important parts:

No pictures as of yet, too soon for that. Datasheet links will have to suffice.

The mcu was chosen mainly for the availability. I couldn't get my hands on the newer one at any reasonable price. The Nordic Semiconductors seems to be the industry leader right now. And because I wanted an integrated solution this series seemed like the obvious choice. Not the most energy efficient but it'll do.

I also got a ceramic antenna that looked good enough and was also, available to me at a reasonable price.

The charger and gauge for monitoring and charging, respectively. Because I want it to be a one easy to handle unit.

All of the datasheets will be in a later post uploaded to my own hosting server and made readily available.

I have the major parts chosen for now. I will tell more about the development and my choices in a more detailed future post.

Cost so far: 30.57 USD
That's about 24€ out of my 30€ budget.
Probably could have gotten a few items cheaper... but oh, well. 
If I fabricate the pcb myself, that'll be right on the mark. If not, it'll go over by quite much.

20x Neodymium Disc 1/8" x 1/25" Super Strong Rare Earth Magnet 3x1mm
US $0.99

HTU21D Temperature & Humidity Sensor Breakout Board Module
US $3.89 + US $1.74 shipping

5V/3.3V 240x320 2.4" SPI TFT LCD Touch Panel Serial Port Module with PBC ILI9341
US $6.66

Free Shipping NRF51822 Bluetooth Module / Networking Module / Wireless Communication Module
US $6.69

Walsin RF Product 2.4GHZ Ceramic Antenna 1206 (3216) Size LTCC Multi Layer Chip Antenna - RFANT3216090A0T
US $2.00 / lot of 20

MAX1555 MAX1555E MAX1555EZK MAX1555EZK+T , TSOT-5,electronic components ,ICs, &Free Shipping
US $3.20

TDFN8 MAX17043G + T MAX17043 MAX17043G
US $5.40