Web NoticeBoard

The web server for the NoticeBoard application has been running for the last few days and seems to be working well. It allows people to check the status of the notice board from a web browser by going to this site.

Currently I am redesigning the interface which is very inflexible at the moment. The look and feel will also be changed slightly as a result of this, making it easier on eye.

I have stopped trying to get the old DS2405 switches used in the current pins to work with the HA5 controller. It is definately doable but I am not sure it is worth spending the time on it, as they will not get used when the new ones arrive. The new DS2406 switches should be here soon and they will be much easier to control and read their state, as well as providing a built-in 1kb of memory.

A Broken Pin

A second pin has broken when trying to pin it into the surface. The place where the outer (gold) and inner (silver) sections meet (separated by a plastic insulator) is definately the weak point of the pin. The inner pin becomes very thin at this point as it travels through the outer core. Usually it is not a problem - the pins bend a little with use but they can be carefully straightned. If they bend too much though they are liable to break off like this. They can't really be reinforced, as this would entail making the pin thicker which in turn would make it harder to push into the surface. The solution is probably to experiment with different materials to provide the insulation between the active layers in the surface itself so that the pins don't require so much force to be inserted...

Testing LEDs

Finding the right LED to use in the new pin design is important. How much power the LED consumes will determine how many pins can be powered at any one time. It is also essential to make sure that the LED will be bright enough when powered by the irregular source that the surface provides. I tried a number of surface-mounted LEDs from the Farnell catalogue. The tests were carried out using 4.5V DC power supply and an 8kOhm resistor to simulate the Pin&Play surface.

In the end I chose a pair of blue and a green LEDs made by a company called MARL. Not only were they far brighter than the other LEDs, but the power consumption was really low (about 0.3mA). And the blue ones add several points to the coolness factor of the design.

The Why of One Wire

In an effort to get my head round the lower-level functions of the HA5 adapter, I have been reading up on the workings of the 1-Wire network in general. A comprehensive overview of working with the 1-Wire network explains clearly why and how the protocol does the things it does.

New Pin Design

After two days of learning how to use Eagle CAD properly, I managed to get the program to do my bidding and produce the board layout for the new pin design I had in mind.

The new design features:

- Small size: 1.6mm in diameter, about the size of a 5p coin
- Flatter: It uses all surface-mounted components, so it should be only a few mm thick
- Two LEDs, controllable through a single DS2406 addressable switch which also provides 1kbit of memory
- Better efficiency as the LEDs have much lower power consumption than the standard 3mm ones we were using before.

The LEDs (blue!) arrived today from Farnell, but it will be at least another week before the PCBs arrive and maybe even longer for the switches to arrive from Dallas/Maxim.

The EAGLE 4.1 files are here.