Due to other commitments, and the fact that I spent most of Sunday up to my elbows in ferric chloride sludge, progress on my two current projects, PIC programmer and new clock has been rather slow. However, I’ve managed to etch one of a set of four double-sided PCB’s that I’ll need for the PIC programmer. The board has now been drilled and populated.

Over this week, I’ll hopefully get some time to hook it up and prove that my idea for this new programmer will actually work – what could possibly go wrong :(

For those that are interested, I'll also post details on how to make double-sided PCB's in the home workshop. It's been a bit of a learning curve but it's not rocket science if your careful.

I have a Mega PCB developer, wash and bubble etch tank, and the bubbler bar in the etch tank has given me nothing but problems over the last year.

The idea is that air bubbles are forced through the ferric chloride etchant and float up over the board being etched. The bubbles agitate the solution and give a faster and more even etch result, and when the darn thing works, it works well.

Around a year ago, I noticed that the bubbler wasn’t working as well as it used too and that bubbles were only appearing at one end of the bar and even these were very tiny and intermittent and since I’d not emptied and cleaned the tank for a couple of years, I suspected that it was just in need of a good service. The old ferric chloride was removed, neutralised and disposed off and the tank cleaned with caustic soda solution and rinsed several times. I filled the tank with cold fresh water and switched on the air pump… nothing except for a tiny little stream of bubbles at one end again.

On dismantling the tank I found the pump to be in good order so decided it must be the bubbler bar holes that were blocked.

Removing the bar is a simple though messy job and once out, I attempted to unblock the 12 or so holes using a pin. After fiddling around for half the morning, I reassembled the unit and hey presto it worked. I think the problem was caused during a long period of inactivity. The sludge from etching process seems to have settled in to the bar and blocked the holes.

Fast forward a couple of months and this morning, the bubbler bar is blocked again.

I of course don’t have a replacement bar and whilst I’ve been putting up with the slowly worsening bubbler performance, today I’d had enough; time to fix this for good.

The tank was drained, washed out and the bubbler bar assembly removed; it was indeed blocked with sludge again, but instead of attempting to poke a small pin through the holes and unblock them, I re-drilled out the holes using a 0.8mm bit making them slightly bigger than they were and on forcing water through the bubbler bar, all sorts of rubbish and debris started to spew from the holes.

The unit was re-assembled and again works perfectly even with the slightly larger holes. The hope is that as the holes are a little larger, they won’t block so quickly in the future.

So whilst that’s £8 saved on a replacement bubbler bar, heaven knows how much I've spent in latex gloves, kitchen role and time fixing the ruddy thing.

I’ve been having intermittent problems with my PIC programmer recently. It was quite literally thrown together in an afternoon several years ago and whilst it’s served me well, it’s time a new one was constructed. Now, you would be right in thinking what’s the point in building a PIC programmer when they are available for peanuts and well supported, and you would be right. However, nothing in my world is ever simple and I have a couple of additional requirements that a standard off the shelf programmer doesn’t support.

My existing programmer can, at the turn of a switch, be connected to one of three different PICs and this makes developing multi-PIC projects much easier; I don’t have to spend time unplugging and re-plugging cables all the time. My first requirement is to upgrade from a maximum or three, to four PICs.

My existing PIC programmer also allows for serial data to be routed to and from the select PIC to aid debugging via a dumb terminal. Second upgrade is that I want to be able to support two serial ports and select which PIC uses which serial port. This will allow me to, if required, connect the programmer element to one PIC, and the two serial ports to two other PICs.

My final requirement is to support a much wider range of PICs than my current programmer BUT, I don’t want the hassle of having to write and maintain the software so the actual programmer element will be based on two Microchip PIC programmers; a PICKIT2 and a PICKIT3. The reason for using two different programmers is because whilst I prefer the PICKIT2 as it’s MUCH faster than the PICKIT3, there are several PICs that I use that aren’t supported by the PICKIT2. One useful feature of the PICKITs is that you can connect several of them to a single PC.

So, my current clock project got moved to the back-burner and this afternoon I started work on my new programmer.

Once all the buttons are labelled up it will be much clearer, but there are four push buttons per PIC channel. One selects which programmer is connected (PICKIT2 or PICKIT3), and the lower two push buttons select which COM port the serial data is routed to. On the right of the panel are three additional switches. Two of these allow for serial data to be read/written directly to the serial ports via the 4mm sockets, and the third push button is the master reset. Those square buttons have in-built LEDs so it’s possible to see which options are selected. The LEDs on the right are power monitors, and status indicators from the PIC programmers.

Hopefully, I should be able to get this completed during the evenings next week.