October 1, 2010
4-26-12: As with many of my projects this one got stalled out. Not only was I busy but the bot was being difficult so I put it in solitary. That'll teach it to be so sensitive.
The main problem is that when the solar panel heats up in the sun it eventually stops making enough current. I'm pretty sure if I remove the diodes in the SEs that it will be ok. Problem is the diodes were never meant to be serviced and are thus buried in the freeformed jungle. Some disassembly will have to take place.
The second problem is the feelers. It turns out I didn't leave enough room for their placement. Pair that with the fact I put them on my Hexaspider and we're stuck again. I'm going to see if I can add some accomodations for them when I tear down the bot to remove the diodes. This is very low on my priorities however.
Every now and again I pull it out and give it some time in the sun. I smile and then I frown. Gotta get rid of the frown portion of the experience.
11-25-10: Got the feelers and photodiodes in the mail today to give me some new stuff to mess with over the holidays. I already attached the PD's and they do help to some degree. It seems the bot is a touchy little bastard to tune.
I'll work on the feelers over the next couple days. Without those I'll never check the "Complete" box. I hope to have this sucker done by the end of the weekend, dammit.
11-5-10: The bot is nearing completion. It's working ok but I want it to work better. I also don't have the damn tactile feelers on yet which is a must. Crumbot blog covers the cluster-F about the feeler situation. I also found that when outside with the photodiode eyes I use I think they get overloaded and the bot won't work correctly. I'm planning on getting the IR photodiodes from Solarbotics that they use in their Photopopper kit. I think that will eliminate that glitch.
The only niggling thing about the overall design is the cell is a lacking in voltage a little bit. When it gets hot in the sun it can drop under the trigger voltage of the 1381s. The change in photodiodes might help some however.
Close to checking the "complete" box!
10-6-10: Made my first mold out of cardstock and poured my first silicone casting last night. I didn't think it though very well so it's still curing this morning. I took some pics that I'll post on the main bot page. I can't wait to pull it out and see how it fits.
10-2-10: I realized last night I have the perfect silcone mold making material at my disposal: Parafilm. Hard to describe if you've never seen it. Imagine what feels like a thin sheet of wax that can stretch to membrane-thin thickness while having all the physical properties of parafin wax (water and air tight, sticky, pliable, etc). First I'll make a wax prototype of the aka to get the strange shape I want. Then I'll make a negative mold around the wax, probably made of some sort of air curing clay. Then melt the wax out when it's cured, line the mold with Parafilm, and fill with sticky goo. The Parafilm should make for easy separation of the mold and piece and the Parafilm itself should just peel away. It may not make for the cleanest looking products but for a rough prototype piece it should suffice.
Here's to my work providing me with a great deal of my robotics supplies! :clinky:
10-1-10: I had this guy on the back burner but I thought I'd throw some of my hard earned US dollars at it while in a component frenzy for my other projects. I tested out a 9V 80mA solar panel on it and the setup seems to like 40ish mA as far as firing frequency. I bought a 5V @ 60mA panel and a 3300uF @ 6.3V cap with plans to use higher rated 1381s.
Also considering designs and materials to use for the aka to join the two. I was going to just make a simple mold and use some silicone as the material. I have the real flimsy clear stuff right now but the piece will probably support the weight of the cap and possibly the solar cell. I was thinking of using the black automotive RTV as it's significantly stiffer. There is also gray RTV but it verges on flexible plastic. All suggestions are welcome.