Well I've been in Texas for a few days and its been pretty nice. I've had time to lay the electronics out and work on things as well as build a decent prototype/test stand. The new rod I got in Santa Cruz feels totally beefy and there is hardly any flex. The stand did feel a little vulnerable though because the friction of the bicycle seat post clamp isn't really up to the wind torque that can be exerted from the forward position of the mirror assembly. I've got some heavy boards making a 6'x8' cross and the center of gravity is pretty close to the middle, while the mount is to the rear. Since it's been really windy here, I used some rope to tie the front down so it won't move. I'm going to leave it outside for a while just to observe how it handles the winds. I've got to say, I like the way it looks. It's probably just a parent's bias though. Someone else came up, looked at it, and said "I give you an A for engineering, and a C for asthetics." Ouch.
One thought about the stand is that a forward mount with some sort of provision for yaw adjustment (i.e. north/south alignment) would help solve much of the stress issues. Parking in a horizontal position also greatly reduces surface area for wind resistance and reduces loads. The final model needs to go into Park mode under high winds, probably by sensing excessive vibration rather that wind directly.I spent all day yesterday working on the i2c bus and ds1307 clock module. I had missed a solder point on the chip socket leading to about an hour of fiddling with the software before finding the missed pin while double checking the circuit.
A significant portion of the previous day was spent building headers for the i2c cable (cat 5), and in the process of hooking everything up to get both boards and the clock all talking over the i2c bus, I realized that I had to add a reset line for the mirror/motor control module so it would properly sync up after the ui/command module booted.
The Arduino IDE has been a headache lately, I learned the hard way that multi-line parameter lists for function definitions completely throw the IDE line count off and compile errors become very hard to line up with the causative line! I'll try the new rev and see what the guys on the forum have to say about it.
My FSM for the command module needs a lot of work yet, not to mention the rest of the code issues. Slowly though it is coming together. Today I finished a routine for setting the date/time for the clock which gets triggered on boot up if the clock has lost power too, which mine does since I lost the darn battery!
Have I mentioned that this stuff takes a lot of time? No, really -- lots of time, everyday. Four months of work now. So much for that original two month schedule! Still I feel like my basic assumptions about the design are all sound, and production issues should be minimal. What is becoming more and more clear is how hard it is going to be to make people understand the significant differences between heliostat arrays and parabolic type reflectors, and the advantage of heliostats. The heliostat is just an enabling device for the target device, and as of yet I have no official target device built, so it just looks to some like I'm shooting into space. Multiple targets have to be built to demonstrate the full concept.