Time off can really get in the way of fun!

This weekend was supposed to be quiet and reflective, just the time to use my new soldering station and build a test power supply. WRONG! This weekend was the busiest for months and I only had about ten minutes to myself apart from chatting to Russ. Even now, I still have some major work to do in the house that will keep me away from the soldering iron. sigh.

All the parts have arrived and I'm ready to go................................but I can't yet.

Silence is golden.

It certainly is in this case. Sitting and pondering, just cogitating and thinking, hell, just doing nothing for a few days can bring rewards. I have now got my hands on a better power supply circuit! The older one works well but the small and low cost transformer was just too powerful for the basic circuit and I had to bodge it a bit. The new circuit, has slightly more expensive components and takes up more room but is almost a perfect fit for the transformer and can handle over three amps. I have most of the parts already and now await the larger strip board I think I need to make the first one. I am also waiting for my new soldering station which I have bought in readiness for my retirement.

Another breakthrough!

I have added two new links to the 'Suppliers' list on the left. These are Minibloq, a method of programming Arduino and a few other boards in the same way that 'Scratch' works and a source for mechanical robotic component parts. Both of these are open source projects and both look very useful. The hardware even includes a custom Arduino clone especially designed for robotic needs and the software has a very clever method of interfacing directly with it.

Or to put it another way,

no I won't. The whole thing has suddenly got much easier, and I will explain why. This is a simple, cheap and very clever board made by Evil Mad Science. Add just five components and 5 volts and you have a working Arduino clone board. All you need are two capacitors, a crystal, IC socket and ready programmed IC. Build your interface on the areas at either end and away you go. Simples.
However, I have had to put everything on hold whilst I work in the bedroom. Hey - Ho.

gottit!

Well, I managed to sort out all the problems I was having with the control board. A wild combination of the wrong components and stupidity. Indeed, some of the wrong components were down to my stupidity so I doubled up there. The switches had seemed just what I wanted when I bought them but turned out to be not quite as robust or well designed as I had assumed. I have had to make several changes to my board in order to make them work.
So
As this is a diary of work done, this is what went wrong.
My first circuit was built on a breadboard and was, basically, only a single channel version to test the combination of hardware and software. It worked well so I moved on to the next stage.
Which was a hard wired version built on a small PCB version of the breadboard. This was for two reasons, firstly I didn't have any standard  Vero board at that precise moment and secondly, I was keen to try out the other type of board. Well, "More haste, less speed" as I made a simple mistake with the resistors and, as they were very close together, it was difficult to spot.
The wrong component problem was that the switches were ideal on paper but not so great in practise. They have a built in LED that has separate leads for the LED and switch connections and that was what was needed for my circuit. They were also cheaper than the only other version I had seen which is sold by Maplin.
So I bought five on the basis that, as I needed four, I would have a spare to cover me damaging one or, hopefully, one to play with after finishing the project. Well, they all arrived damaged due the seller not protecting them for the post. I managed to straighten out all the leads but one of the LEDs was not working so that left me with just the four working items.
My final problem was more to do with inexperience as I hadn't used modern PCBs before and still feel happiest with standard sized Vero board.
I assembled the PCB for the switches and began to check it out, nothing worked so I had to remove some wiring and start again. The copper simply peeled off with the solder so I had to bodge the connections when I reconnected correctly. The wires on the LEDs were too short to reach the other side of the board so I had to wire through the board and then solder at both ends to get a connection. At this point, I was beginning to get somewhat perplexed as, my hastily built PCB test board was working correctly but the carefully wired and set up switch board wasn't and I didn't know why. This wasn't rocket science either, just some very basic connections that should just work. I eventually came to the conclusion that the PCB was compromised and no longer to be trusted. This was a shame, but I still had my original test built on the other board so I wired that in. Success, all the switches worked and two LEDs did. But not the other two. So I began yet another round of checking voltages. I soon sussed out that the two failing LEDs were simply not getting enough power to light them so I traced back to the board and then discovered that my eyesight had let me down. The resisters on the test board were wrong. Only two of course but enough to stop their associated LEDs from lighting. They had been tested on very small low needs LEDs and now I was asking them to power larger and higher needs LEDs, they couldn't cope. So, out with the soldering iron again, swap them for the correct value and, finally, all is working. Now I have to shorten all the wires between the two boards and fit the whole thing into my bench.