It’s been a busy few months in my life, but I’m pleased to announce that the project is still alive and well.

FIrst off, I left my cozy job to join a friend’s new startup. We’ll be doing really cool things & I’m excited about it. I’ve been cleaning up the loose ends at the old place and coming up to speed at the new one, but eventually I should get more time to work on the watch.

Besides the many friends at my old job (who I’m keeping in touch with), one of the perks were the tools available to me. Specifically, the equipment needed for fine-pitch soldering came in handy. My new company deals primarily with software, so I’m buying new tools as I can find them on ebay. I recently got a stereo microscope ($120, list price $1500) and am awaiting a nice Metcal soldering iron.

Second, I’m in the process of changing the microprocessor. TI has some nice processors, but the ones that come in small packages have only limited memory. I filled up all 2K in the MSP430F2013 and got just the minimum functionality I wanted.  I’m looking at the C8051f411 processor (full data sheet) from SiLabs that fit 32kB in just 27mm^2 -- that’s 16 times the memory, in a package smaller than the one I’m currently using! The smallest TI part with the same memory required 45mm^2 of board space. The fact that the part also includes a built-in voltage regulator helps seal the deal -- I haven’t tried the development tools yet, but there is a low-cost ($25) programmer and a free compiler available.

It’s a little painful changing the processor twice, but I think it’ll be worth it.

Third, I’ve been working on building up a bill of materials.  I can’t say why ;-)  All I can say is that a custom-made case produced by a real machine shop is spendy.

Lastly, my prototype broke. I’m bummed. One moment it was working fine, the next it was dead.  I was wearing a wool sweater in the dry colorado air (a recipe for disaster!), and I must have shocked it. When I got it back to the lab, I found out that the processor was working fine, but it thought that the batteries were low and turned off the display. The low-battery signal had been zapped to a permanently-low condition; I’ll replace the processor as soon as I get the new soldering iron.

There were two design changes I can make that will make the watch more robust. First is the plastic case: unlike a metal case, it does not provide the protection of a Faraday cage.

Second is the physical location of the wire that got zapped (see below), and it takes some explaining. The ground wire goes to a terminal on the outside of the plastic case, and I suspect it received the static shock from my sweater. The circuit is designed to tolerate static on the ground wire, so the majority of the watch survived.  But, when two wires run in parallel, the electric noise in one (the ground) can couple on to the other wire (the affected low-battery wire) and cause damage. I suspect that’s what happened. The solution is to rearrange the circuit board so that the power and ground wires do not run in parallel with susceptible signals.

Oh well, the new processor will require a new circuit board, so I’ll incorporate the changes then.

Next up (besides the new job and getting new tools), is to try out the new processor development tools.  My friends and I have got some things other than Ping Pong that we want to run on the watch. And, I’ve got to meet with the machinist to see if I can make the case cheaper.