January 22, 2011 § Leave a comment
I got the two BS2 that I was waiting for, that was fast. I thought of putting them both in one of my small boards and then the LEDs repesenting the domino tab in another; but they do not fit together in one board by one pin. I will need to re-design how I am going to organize my test environment; I do not want to end up with 6 boards for such a simple project.
So the new design of the test environment is finished, it looks like in the picture. So everything will fit in three boards, that is going to be the next step, building a full clock using three boards with a design similar to the one in the picture.
January 19, 2011 § Leave a comment
I have been keeping logs on a notebook, with my developments in this project, but this is not the way of doing these things in Century 21. So at the advice of Hack a Day editorial, I started this blog yesterday … and today is finished. Now I will reply to them, to see what they say. Hopefully I will get a note in their blogs!
January 18, 2011 § Leave a comment
It has been a while, more than two weeks, that I do not work in the project; busy at (real) work, finishing some real projects. Today I started to see how to add the other two digits.
If I use a BS2p40, for either the hour tab or the second tab of the minutes, since it has more ports, I can easily accommodate figures like, for example, 8|0 and 8|1 to represent 8 and 9 respectively. Now the problem is that the testing board that Parallax offers (Professional Development Board) is way too expensive and I do not need not even 20% of what it offers. So I started to look for a solution in different forums.
At the end, in order to gain simplicity and keep this project to the minimum budget possible, I decided to cancel the idea of using the BS2P40, and will be using three BS2, one for each tab. That means that my clock will never see an 8 or a 9 in any of the faces of the tab.
I placed the order today, will pay tomorrow, I should have them in my hands in one or two days (they are delivered from a local provider in Tokyo, so it is really fast!)
January 5, 2011 § Leave a comment
Today I got an answer from Hack a Day and they are interested in seeing what I have done so far, so I just need to show them this blog! (which is being written on January 18th, 2011)
January 4, 2011 § Leave a comment
The algorithm improved. Now it has a button that increments the number shown if pressed; this is useful to set up the clock on time. This sounds simple, but there are no interruptions in the Basic Stamp, so the cycle had to be reduced to 250 ms frequency checking if the button is pressed or not and a whole recalculation needed to be done to know when a full second or minute occurred so a change of state is required.
I am planning to contact Hack a Day to see if they are interested in publishing it.
I got samples of flip-disks from one of the companies (finally!!!) It seems like these are not the types that keep the state when electricity is not been provided, but I will have to test them. Nothing from the other company yet.
In the video you can see how the clock behaves normally first and then (from 0:12) goes faster while I press the button in the center of the test board with a screen driver. The video can be found here in Youtube.
December 30, 2010 § Leave a comment
I had to modify the Excel file to display up to the digit 7 since digit 8 and 9 use to extra dots per face in the domino tab, and eliminating those two dots per face makes possible to sue cheaper microcontrollers.
So at the end of the day, the algorithm is almost finished, the tab of what it would be the minutes changes accordingly. The logic to minimize the number of state changes when switching digits in one of the tabs works beautifully. You can easily notice that some changes are not optimal, but that is what I wanted from the beginning, or else you would be looking at the same tabs again and again. The only thing that is missing is the logic for the “set button”, that if pressed, the number will advance one digit, so you can put your clock on time.
In the following video, I set the clock to seconds and I let it ran for two rounds, 13 seconds each (from 0 to 12). In the first round the states takes typically 1 or 3 changes, with the exception of the change from 9 to 10 (at 0:07) which moves from 4|5 to 3|7, taking 5 changes in this case. In the second round, the exception is when changing from 8 to 9, taking the clock 7 moves to switch from 1|7 to 5|4 (at 0:19).
Both rounds of state changes, take a total of 26 moves to represent a full 13 seconds (purely coincidental!). The video can be seen here in Youtube.
December 29, 2010 § Leave a comment
I had to go to Akihabara yesterday to buy everything required to start making the prototype: resistors, LEDs, jumping cables, bread boards … etc.
Today I managed to connect in one board what it would be one tab: 14 LEDs with their respective resistors, all connected to the ports of the microcontroller module. Now I can seriously start testing the program that will do the change of state for each of the tabs, using the Excel file with the probabilistic algorithm as a reference.