Here it is :) My not-so-secret project was to build a custom handheld. See below (previous news) for the beast, assembled :

You can check all the previous news I wrote about this project, that were offline until today. All their titles start with "MEX", that stands for "Micro ElectorluX" :)

Thanks to Devpsx for his help on this project, I will put all the related files for download in a few (Eagle files, C source...).

I spent some time today building a package for the system. It's basically made of balsa wood, as it is easy to work with. But even with that, it's still looking quite bad, as I don't like "standard" do-it-yourself tasks :)
Anyway, I'm satisfied with the results, here is how it looks like :

Here is the contents : main board, SparkFun header board holding the LCD screen, keypad board, and some last-minute parts (I forgot to tie P0.14 high, and the power switch).

Next step, maybe paint it and, more important, make a game for it :)

As you can see, my Lab is now a mess, dedicated to PCB building for the moment.

I'm working on a "secret" project, for which I had to build a PCB. See the FPGA section to see my progress in PCB building techniques :)

It is now close to the end, and once I've made the first "official" public demonstration, I will put all the news I wrote about it online.

I just connected my keypad to the board, on the places left for this purpose, and wrote a simple program to show up button status on the LCD, and everything's fine !

It means that the hard work is now done, I will now try to make a not-too-ugly packaging, and see if I can add sound support (I had to remove the dedicated pin when I fixed the design, but I can still use the UART0 pins that can do PWM).

Ok, the board is now working (see previous "public" news). I connected pins P0.0, P0.1 and P0.14 to the FPGA, and re-used a simple design I used for programming the Olimex LPC2106 board.
I've been able to flash the LPC2106 using Philips ISP software, and run the well-known blinking LED program. Good!

Next thing to do, put the Sparkfun header board carrying the LCD on the place dedicated to it, and check if everything's working as it used to do on the breadboard, but I will do that later.

Meanwhile, Sparkfun has restocked some Hirose DF23 compatible connectors, so I could do without the header board but well, as having this first board up and running was quite painful, I think I will choose the easy path for now :) However, I've ordered some of those connectors, just in case of...

After another set of attempts, I finally got my board working !!!

I think I've probably made all the mistakes one can make : stupid design errors like inverting pins on the voltage regulators, placing the main component the wrong way (and finding it out only once all pins are soldered, of course). But well, this is history now, as this frigging board now works !

After some other attempts, I managed to get a quite clean PCB. So I went into soldering, which was painful especially at the beginning, before I had the idea of using solder flux, which makes things way easier :)

It was also the opportunity to find some (obvious) mistakes in the board design too. Anyway, I managed to make some workarounds, and things should have gone OK... but they didn't. My first soldering (and desoldering) attempts might have destroyed the central component (I started with it, thinking that if I wasn't able to solder it properly, it would be useless to solder the remaining parts).

Anyway, things are getting better and better, so let's continue :)

Well, I thought I could go on with the board I've built, but it seems that I were wrong.
I hardly managed to repair some bad tracks, but my attempts of soldering the center component failed (and soon led to some other broken tracks).
I probably need to change some things in my method. I learned some tricks on the web about insulation, that I will apply for the next board, and I will probably try something like using solder paste and an oven.
Meanwhile, I'll probably work on a new board using some components a bit larger (0.8mm pitch instead of 0.5mm) and I think I'll be able to succeed this time.

As you can see on this picture, I finally managed to have a not-too-bad board. The printings made by Devpsx were better than mine (thanks again) and after another failed attempt, here is the board, etched and stained. Some tracks didn't show up properly, but none of them is critical, and I think I would be able to correct them by hand. Good!

I've put a 10 euro-cent coin next to the board, as it helps to understand how tiny the component in the center is...

Following the advice of Devpsx, I changed the design of my board, to make most of the tracks larger. I also changed the paper, which is less transparent as the previous one, but retains ink better.

It now looks much better, as you can see on this picture : on the left, the old design, on the right, the new one. Unfortunately, the LQFP48 pins are not as clearly separated as they were before, and are unusable. So I probably need to take the best things out of my two attempts : use the new design printed on the transparent paper I used yesterday, and everything should go fine.

These last days, I haven't worked on sound, but on my "secret project" instead. I won't tell much about it for the moment, just that it is now quite advanced, at least in my mind. I've designed a board, then unpacked most of the tools I bought some time ago to build PCBs on my own. Tonight, I made some PCB building attempts, and even if they failed, I'm quite satisfied with what I've already done.

This board is designed to use SMT components, and especially one packaged in LQFP48. This kind of board is probably not something that you can do with the usual amateur PCB building tools : the different sites I found on the web show examples of boards that use through-hole components. So I will probably have this board manufactured by some PCB prototyping company. But at least, I'm now a bit more familiar with the tools, and I'm now confident that they will be great for many other projects.

Well, if I want to have this board manufactured, I will have to change many things in its current design, that means many hours of work, then (usually) some weeks to wait for the board to come. So, once the design is done, I'll probably be back to my previous work (sound experiments) meanwhile.

EDIT : I will make some extra attempts with a different kind of paper. It may give better results on thin tracks.