One of the difficult parts when prototyping is to find reliable power sources. Today is still hard to find the battery size we want to use because country exporting frontiers stops these chemical packages. Here I’ll show how to refurbish dead batteries by combining cells and protection circuits to preserve battery life.

An (almost) dead Apple MacBook Pro (17″) battery fell in my hands so I decided to tear it down to see if there was something profitable. Inside I found that the battery pack was composed with 6 individual cells, paired in 3 groups.

Apple Mac Book Pro 17" Battery Disassembly

Seems that the third group had a small voltage difference between cells so they began discharge between them. This leaded both cells to die, condemning the entire battery pack.

Here can be seen the individual cells:

Batteries separated from the main board.

As can be seen the cells doesn’t have individual protection circuits witch are important to avoid discharge below 2.7v to preserve battery life.

I also came across with a small photo-frame that I bough just for hacking fun as is really a bad piece with almost no memory, no SD card expansion, 128×128 display and bad electronics. It cost me 1.99€ and the battery was drain dead from the first moment I opened the package.

Small and dead LiPo battery with the protection circuit.

It was inflated and reading 0.26v. Obviously battery couldn’t be rescued, but I salvaged the power cutting circuit. Here is the reverse side of the protection circuit board. Notice the polarity when soldering the LiPo cell and wires:

LiPo Battery protection circuit downside.

Once I soldered the protection circuit board to the Apple battery cell I ran some charging and discharging tests to ensure the assembly works fine and that power is cut at 2.7v:

Working Refurbished Battery

This DSLR came to me in a little market in Barcelona called Encantes, wich is open 3 days a week. At first I though it was a good thing to buy as the camera hadn’t any scratches or visual defects. It cost me 15€ so it was worth of buying it even if was broken.

Actually I tried to power up the camera by soldering a pair of cables to the battery attachment and feeding them with the 7.4v that its original batteries supply. After a few tries an multimeter measures I ended with a diagnose of broken power supply board. It costs about $40 to replace it but I’m not totally sure about if the rests of components will be properly working, plus the camera is 10Mpx, made in 2007 and has smaller 4:3 image sensor format, wich is poorly on high ISOs, so I decided to tear it down.

Olympus E510 MOS Image Sensor

The MOS image sensor (big center colored square) is mounted over 3 frames. The first, image sensor frame moves in vertical axis meanwhile the second frame moves in horizontal axis. The third and outside frame is fixed in the camera. This conforms the IS (image stabilization) system of this camera that works regardless the lens you use on it. The bad thing about image sensor stabilization is that stabilizes image only for the final photo and live view, leaving the image unstabilized when looking in viewfinder and light/autofocus sensors.

Overall, is a really nice technology piece as the frames are moved with ultrasonic motors.

This tiny sensor board was solely designed by Mike Rankin, a friend that help me to bring life to a real custom GPS board. This time he was looking for a multi purpose board to develop anything.

Sensor Board Main display Side Showing Ponglock

Feature list:

• I2C Bus to save pins for future use.
• 128×64 OLED I2C Display: let’s draw anything
• Momentary switch on left and cross plus center buttons on right (total 6 input buttons)
• Arduino Leonardo based: USB for anything.
• Temperature sensor (I2C)
• Barometer sensor (I2C)
• Accelerometer sensor (I2C)
• Real Time Clock (I2C) plus backup battery.
• Switched power supply.
• LiPo charger.
• Tiny design: let’s make portable projects.

Sensor Board Main Component Side

I couldn’t resist temptation.. I ported the pongclock code from Rob Parrett to this tiny board. Here is the result:

Hardware is open hardware, you can obtain anything on Mike’s web or you can buy it directly from Tindie. Custom pongclock code can be downloaded from here.

About 9 months ago one Hackaday reader dropped some comments about compiling the source I published here a year ago and some troubles. After that, Mike and me kept talking and things turn out that he was PCB developer, so we though we could make together GPS cube version for real.

Together we developed a schematic and Mike built all himself several prototypes until we finally came with a (almost!) final version. So, here we are, one year after first release:

GPS Cube version 2

This version has several features:

• Everything is packed together in a single board.
• Better GPS: we switched to MT3339 PA6H instead old SirfStar III EM-411
• We added TMP75 I2C temperature sensor.
• The display PCB is also custom to run 3.3v
• Everything runs 3.3v: GPS, display, FTDI, TMP75 and atmega328p is now running 8MHz, Battery lasts longer.
• Expansion header with SPI and I2C: the board ca be used as development board for any other projects with displays or GPS.

GPS Toy Hardware Rev1.0

New GPS Board back

All the hardware was designed and developed by Mike Rankin (probably will go to Tindie) meanwhile I developed the software.

.. and couldn’t resist to tear it down >:3

Ouya Video Game System

Ouya Teardown

Ouya main board

It’s body is metal built. It gives the console a robust appearance but the WiFi antenna is damn small, so the console is pretty much signal-deaf. Also the case air flow is bad: the console may heat up. I think basing it’s main OS on top of android is a bad idea, as android can (and is) heavy, leaving less resources to games. Also case is almost empty, not well used space.

BUT, is open source and hackable :)