(NOTE: This page is a mess. I have so many changes I've not added here it's insane. I hope to eventually come back and update this page. Wow has this project evolved.)

This page will list how the ideas have evolved as we go through the process. Ideas folks submit will go here as well. Please feel free to email us with things you'd like to see added here.

• Clock changes:
  • Change from 30deg Nichias to 30degree equivalents from LSDiodes.com. They are slightly dimmer (4800mcd vs. 5000mcd) but will save us massive cash ($0.45/ea instead of $1.50ea!).
  • Swap to a thicker plastic for the digits, moving from 1/8" to 3/16" for better wind tolerance
  • Use snap-in LED holders instead of hot-glue for easier assembly and in-field swaps
  • Change the last digit from tenths-of-a-second to instead show the magnitude of the flare (BIG change!)
  • Make each physical digit mount be for two digits and a colon (88:) that way we can stack three into a sign and simply not turn on the last colon. This makes all boards interchangeable and we can bring a spare in case one takes wind or impact damage.
• Digit board changes:
  • version 8.1: add terminal strips
  • version 9.0: slimmer profile (yield increased from 6 to 9 per batch), better resistor layout
• Distribution Board (A) changes:
  • version 7.3: more even spacing of signal traces
  • version 8.0: moved Digit 4 header to center of right side to give better standoff support
• Distribution Board (B) changes:
  • Version 1.1: Moved GPS to a direct-PCB-mounted serial port. Change laser-board to use Cat5 cable. Add a keyboard that runs via Cat5 cable. Add GPIO pinblocks.
• Laser Board changes:
  • Version 5.0: Change to using Cat5 connection. Added 2-pin SIP connector for laser PWM (couldn't find a headphone jack tough enough). Swap to using 3 lower power 5v regulators instead of one big one and use proper heat sinks.

Ideas:
• Add rotating flasher lights (gumballs) that go off at the 10, 5 and 1 minute mark, for 15 seconds each time.
• Line the outside of the laser table with the spare EL tape we have for better looks
• Track all flares brighter than Mag. 0 (since tests showed that a -1 is still very visible and pretty), notably increasing the number of events shown during the week. Moving to 0 from the previous threshold of -5 increases events from 3-4 a week to over 20 a week. Hey, if we've got it, let's use it!

Update: 2005 (after event)

Wow, did this 'ideas' page ever get ignored. Here's a quick recap with what photos I have of the changes we made before going to the desert:

• We added wind-power via an Air403 windmill on a pole on one side. We also added some solar-powered compact flourescent pathlights by separating their internal panels and putting them on top of the structure, using them to light the info boards on each side of the arch. This also provided a little more pedestrian protection (so folks wouldn't bump into the support tips at night). We also found some 12v spinner beacons that we put on the top and made whirl when a flare was about to arrive.
  
  
• Here's another shot of the pathlights on the sides, with Garth taking a nap under one. They were surprisingly bright and nicely cool-white colored. They ran all night off the charge the little 4" x 4" panels got during the day.
  
  
• We added a serial-programmable marquee sign. The system used it as an auxillary display, scrolling any text we wanted from the gameboy. This way a participant could use the long-cabled keypad (which also had its own solar light) to play with the system and use the text display for feedback on what they were doing (such as scrolling through a list of star names to pick what to point at).
  
  
• Our intrepid testing crew. Eve, Sidd, Amara, Loudog and Learnteaach. They're demonstrating just how Insanely Freaky Bright the green LEDs ended up being when running at 100% duty cycle.
  
  

I don't have any good pictures of the keypad right now but I'll go find one soon. It was a standard 4x4 cross-switch membradne keypad that we rigged into the back of a pathlight to make a little hand-held console. It was cabled to one of the GPI mainboards and scanned by the software in the GBA so you could use it instead of the tiny GBA switches. We put it on a 25' cable and left it out for the public to kick around, locking it down to 'can't mess it up' functionality. During the day we set it upsidedown to charge the light's battery with the solar panel on the back. It worked a lot better than we expected.

Garth also got some video of a flare capture out at the event. When he uploads them I'll put them on the Aftermath page.

_Contacts: Tor Amundson and Garth Minette . Questions/suggestions appreciated! Flames will most likely be snickered at.


This work is licensed under a Creative Commons License.