It’s been almost as long since I last posted. The house is new in a new neighbourhood & came with a lovely back yard. The yard backs onto a small belt of “green” land about 30 feet wide that separates us from the houses behind. When we moved in the yard was completely open to the surrounding area including the green belt and the neighbours’ yards.

Since we wanted to get a dog before the end of the year & also desired additional privacy and security, we decided it was time to plan a fence build. Being handy I couldn’t imagine paying someone else to do it, and seeing as we are not rolling in cash & especially weren’t at the time, DIY was the smart economical choice as well.

The first problem was there were several very tall and very dead trees in the back of the yard. Two of them who shared some roots were right in the corner of our property on one corner – blocking the corner.

If we left them they would have come down eventually in a wind storm. And they could’ve reached our house & destroyed part of whatever fence we put up. Some enthusiastic family friends who enjoy that sort of thing made short work of the lumber.

In retrospect I probably should have gotten rid of the stumps in the lot corner, but at the time I felt the job was too big so we planned around them instead. Perhaps one day I’ll change that.

With that out of the way we had our canvas for the artistic work that is joining bits of wood together to produce something practical and visually appealing.

After some measuring and sketching and standing around while the brain gears slowly turned, we had our plan and ordered some items.

Then came the digging. This ended up taking the longest time for one simple, uncooperative reason: rocks. Our land is full of rocks of all shapes and sizes.

We started with one hole, as you do, and had friends and family come over on the weekend to help. We ended up with about 5 or 6 started holes that we would jump back and forth on as we got frustrated or made progress with the one we were working on.
Sometimes the rocks were bigger than the holes we had to remove them from. Eventually we started getting posts set in concrete – the first time it started to hint that it was on its way to completion!

Still: more holes, more digging, more rocks. Some far too large for one person!

There was simply no escaping them. After almost four weeks of digging and excavating all by hand, all by between 1 and 4 people depending on the day of the week, we got the digging done and set all the posts.
Before we knew it we had rails on and the boards were going up super fast. And a good thing too, because due to unforeseen circumstances the puppy we had chosen had to be picked up two weeks earlier than planned.


In a flurry of activity I got all the boards up & made temporary 3-foot high barricades to enclose the yard for the dog. One Friday night I put up boards until it was too dark to see and then got up before Saturday’s dawn broke to finish.

Yes, the boards are all screwed in. Not a nail in sight. I got done, changed my clothes, got in the car and drove two hours to pick up the new puppy. No rest for the wicked!

It didn’t help that the schedule change also meant we were dog-sitting an adult dog while family was away. We had a dog to mind, a puppy to nurture, two cats to visit & feed and the fence to work on. It’s amazing that we got it as far as it did in that time, which was about a week. After that, however, we had a severe life distraction:

After we’d had puppy dearest for a month, I finally got around to finishing the holes for one side gate, setting those posts and trimming the tops of the other posts.


You can see the wonky nature of the board placement – something I later adjusted. Two weeks after that I managed to find some time to build the gate itself.

I’ll add an “it’s complete” picture this evening.

Removing the socket from the StarTAC mainboard was relatively easy – the pins are not completely flat against the board so I was able to get one side of some tweezers up and underneath them and gently pry them off the board. In the image below you can see some of the traces came off with the pins, but that’s fine.

A while ago when wondering how I was going to be able to prototype the LCD, I realised that the pitch on the socket pins was 0.5mm – the same as the socket pitch for the Telit GC864 cellular module which I failed to post about and will mention further down. I bought a SchmartBoard|ez SOIC .5mm Pitch breakout board from Curious Inventor which as you can see in the picture will give me easy access to the LCD / keypad pins.

Many moons ago in a previous post I declared that I was going to use the Telit GE863 cellular module for this project. My primary reason was that it was the only module Telit offered that was small enough to physically fit inside the StarTAC housing. The big problem with that module is that it’s pins are BGA – very hard for a hobbyist of my level to use and offers minimal flexibility in terms of avoiding screwups.

When I found the cash and the will, Telit had released several new modules including the GC864 which uses a nice 80-pin connector, with a separate antenna socket! Perfect all-round for hobby prototyping. I bought it, two circuit-board-antennas with the right type of connector plug and two mating 80-pin connectors.

Before the burglary last autumn, I was hard at work for a few weeks hooking up and testing the GC864. I found and bought the same SchmartBoard 0.5mm pitch breakout board linked above and soldered one of my connectors to it to get access to the GC864′s pins. With a lot of jumper wiring and a few hacks to compensate for my lack of female-to-female jumpers I got it working, controlled by serial over USB to my desktop PC. This mess was the result:

Power was supplied by a StarTAC battery pack which is in bad condition and required constant charging to be useful. Still! I managed to send myself an SMS message and place and receive phonecalls, though I did not have a speaker or microphone connected so they were quite calls.

Exciting times are ahead! Continued work on the project is planned so I hope to post further results before 2011

Here’s one last picture, a StarTAC housing and Telit cellular module and an SD card for size comparisons:

Arduino is also more mature now as a platform – there are many hardware options to suit different designing requirements. Sparkfun’s Arduino Pro series appealed to me and got me thinking that maybe I could switch (or rather, upgrade) to an Arduino-based environment for my RetroTAC project. Since the hardware is ATmega and the coding is similar to C, it’s not a big leap or a start-from-scratch kinda thing – more of an evolution.

So last week I bought an Arduino Pro 3.3v/8MHz unit from Sparkfun and on the weekend I set about getting it going.

I soldered male headers to the port pins (later realising that Arduinos normally have female headers, but oh well!) and jumpered up my USB-to-TTL serial adapter to the programming port pins. For power I will be running the RetroTAC from a 3.7v Li-Ion battery so I soldered jumper wires to the empty shell of a StarTAC where the flat power jack comes out, and hooked it right into the BATT of the Arduino.

At first the Arduino software gave me an error:
avrdude: stk500_getsync(): not in sync: resp=0x00
avrdude: stk500_disable(): protocol error, expect=0x14, resp=0x51

Which I found was more than likely related to the programmer being unable to reset the Arduino hardware before attempting to upload my sketch.

There’s a break in the trace (intentionally, but who’s intention and for what purpose I don’t know) which doesn’t link the RST pin on the PROGRAM header. I wired the USB adapter’s reset line to RESET on the Arduino and had the Blink ‘Hello World’ sketch running within minutes.

What I discovered was that the keypad, LCD and LCD controller are all on one board. It looks as if I could basically just ‘replace’ the mainboard and keep the existing keypad board if I can work out how to interface to the LCD’s controller IC.

Handily, the IC is labelled clearly with its model number: SED1530DAA.

Some googling reveals several versions of the datasheet for the SED1530, some more useful than others. Basically it is a “single-chip LCD driver for dot-matrix LCDs” by Epson, supporting parallel or serial (seems to be an input-only SPI bus) communication.

I’m in for a bit of work to interface to it, both physicaly and electronically, but if I succeed I’ll be able to write a driver to use the original screen hardware in my RetroTAC!

Any tips, theories, ideas or comments are welcome.

Today’s result: my little Atmel ATmega8 is telling the 4D Systems uOLED-160-GMD1 where to go! Or rather, what to say.


Using the ATmega’s USART hardware serial port, controlled in software by some USART routines written by Mihai Galos, I wrote a subroutine based on the OLED’s user manual entry for writing strings of ASCII text to the screen.

UOLED_fstring('s', 0x00, 0x00, 0x02, 0xFF, 0xFF, "Hello Interwebs");

The above is the line in the main code that sends the options to the subroutine. I’ve just realised while writing this that the first variable ‘s’ is actually not necessary as the subroutine is specific for that command.

void UOLED_fstring( unsigned char cmd, unsigned char col, unsigned char row, unsigned char font, unsigned char msb, unsigned char lsb, unsigned char *c )
{
USART_Transmit(cmd);
USART_Transmit(col);
USART_Transmit(row);
USART_Transmit(font);
USART_Transmit(msb);
USART_Transmit(lsb);
sendString_USART(c);
USART_Transmit(0x00);
}

The above is the subroutine. As mentioned, the first line ‘cmd’ is actually redundant and I will change it to always send ASCII ‘s’ / 0×73 which is the OLED’s command to print a string of formatted text.

Next up, I’ll expand this to display a short demo menu of options and code the keypad input to control the display – giving the basics of a User Interface.

The USB converter came last week (after the wrong part was sent in May. Sparkfun were quick to correct this and I appreciate their excellent service) so I flashed the uOLED to its latest firmware and set about trying to communicate with it with my ATmega8.

Once again no success after two nights of poking about and checking & double-checking code, so I did some reading and got some ideas about why other people have had issues. Most of them relate to something wrong with the serial connection causing the auto-baud feature to not correctly ‘lock on’ to the serial signal. You see, the uOLED serial interface can run at serial speeds between 300 to 256K baud. To configure it to a certain baud your micro must send it the ASCII character “U” (56 hex) after a minimum of 500ms (allowing the uOLED to ‘boot’).

Problems arise if the serial lines are not stable when the display begins checking for “U”.
This was my problem – my serial lines were floating in the time before the microcontroller got controll of them, and the display was misinterpreting the garbage as a signal. I noted that a fellow who had wired a Telit GSM module to an ATmega8 had tied his serial lines high (connected them to the + power rail) with 10kohm resistors. I tried the same thing, cleaned up my messy “it’s not working and I’m getting frustrated” code and immediately had solid, reliable control over my OLED display.

Success! Now some tedious work to write various lower-level routines for the screen. My motivation for that will come – I’m just really glad I managed to get the comms working.

Special thanks, if you happen to read this one day, to Alexander at Tinkerlog!

Telit GE863

Telit have recently updated their website and it is a LOT easier to navigate now – well done!

Also they have a few more datasheets and application notes up, including one for battery charging, audio and antenna notes.

For the module I am planning to use, the info is available on their website.

It is based around the Telit module GE863 which supports quad band GSM, serial interfacing and even has a SiRFIII GPS chip inside. It will be hooked to the other primary component, an AVR ATmega microcontroller.

Coupled to the GE863 will be the original StarTAC’s microphone and earpiece, a GSM antenna and sim card holder. Coupled to the AVR will be the original keypad and a 4D Systems 1.7″ OLED display.

Power will be provided by an aftermarket StarTAC Li-Ion 3.6v battery and either an original StarTAC charger or a modified one.

I’ve so far put up the basic project page for my retro-style mobile phone, the RetroTAC. I’ll do more work on that soon and blog about my progress on the project so far.

To come there’s my gruntily-named “MaxxMonitorr” telemetry system for my Traxxas E-Maxx RC truck, as well as my digital timer photo capture circuit to link to my digital SLR camera. It has possibly been named the “Capture XTi”. Better ideas are welcome!

Here we go! DIY Gadgets!

I made this movie with my brother and our friends in 2002. It is plot-thin, and was primarily made for fun, and to test my skills at video editing and post-production special effects. I cringe at several things when I watch it, but I think overall it’s a good effort.

17Mb, 8min 40sec, Credits feature outtakes so watch to the end!