Monday, April 13, 2009

Finished the relay board PCB design

I've been spending much of the last two weeks designing a PCB to hold the relays and associated drivers for the 24VDC power signals to the engine valves and the 220Vrms power to the igniters. (This is the first time I have designed a complex PCB. The reason for the new board is that the current board doesn't have enough 24V channels even with additional relays glued onto the top of the board.) I finally sent off the file to Sunstone Circuits on Saturday.

Some time ago I read an entry in Wayne Hales' blog about the origin of "burning rocks" issues and how they still give the Space Shuttle Flight Director sleepless nights.

Well, I can tell you that "burning rocks" anxiety is a real factor, even with such a trivial matter as sending a PCB file off for manufacture. In my case, if I have made a mistake, in the worst case it would cost around $300 to get the boards remade. More likely I would have to do messy board mods. (possible because the boards are only 2 layer) In the event that the board could not be modded the lost time waiting for the faulty board, finding the problem and fixing it, then possibly getting the board remade would be a significant issue. It could conceivably cost up to 3 weeks, so I was understandably cautious.

Based on previous experience, not to mention having burning rocks on my mind, I was not completely unsurprised to get a sudden anxiety kick at the point when I was prompted to enter my credit card details. Of course I closed the window straight away and put the anxiety to good use, trying to think of things I might have forgotten. (such as component clearance off the edge of the board) I don't think it's an exaggeration to say that having got so close to committing myself, I could feel my mind straining to think of something I had neglected to consider seriously.

As it stands I don't recall making any changes to the PCB between the first and final submission attempt, but when I finally did pay up and submit the PCB design I had very little anxiety about my design, and was more concerned about using the 10% discount coupon.

Burning rocks issues are real. I think, if it hasn't already happened, there needs to be some serious studies into the psychological mechanisms behind it and whatever is learned needs to be integrated into engineering courses.

In any case, the next test will be to find out if their absence implies a clean design, as I will find out when I get the PCBs in a week or so! In the mean time, to avoid burning rocks issues I recommend some imaginary brinkmanship. Pretend that the design you are submitting three months before the final article is launched is actually going to be used in action in two weeks, or one week - or tomorrow. Hopefully those little issues your mind blanks out so easily will be forced out into your consciousness :-)