Servo torque to weight ratios

0.16 Hitec HS-805BB 24.5kg-cm / 152g
0.16 HS-645MG 9.6kg-cm / 60g
0.18 HXT900 1.6kg-cm / 9g
0.19 Hitec HS-805BB @7.2V maybe 29kg-cm / 152g
0.30 Robotis AX-12+ 16.5kg-cm / 55g
0.37 Hitec HS-7955TG 24kg-cm / 65g
0.42 Futaba BLS152 31kg-cm / 74g
0.44 Hitec HS-5990TG 30kg-cm / 68g
0.46 JRPROPO DS6311HV 36.5oz-in / 80g
0.52 Robotis RX-28 37.7kg-cm / 72g
0.55 Robotis RX-64 64kg-cm / 116g
0.68 Robotis EX-106 106kg-cm / 155g
0.66 VSTONE V-SERVO VS-SV410 41kg-cm / 62g
0.68 VSTONE V-SERVO VS-SV1150 115kg-cm / 170g
0.85 VSTONE V-SERVO VS-SV3310 327kg-cm / 387g

The nice trend is that as you go up in torque your torque to weight ratio also goes up. So bigger robots should actually be easier to make than smaller robots. However what I also should really factor in is how much weight is required to power the servo. You could run a entire HXT900 biped on 2 A123 cells, but you might need on average a cell for each VS-SV3310 ( just a guess. )

Another thing going for the bigger bots is that the computer and censor bits take up a much smaller percentage of the over all mass.


Now for laughs.. http://www.z425.com/robot/diy-servo-project/
1.63 BILLY CRAZY DIY SERVO 594kg-cm / 365g ( no electronics )
I came up with those numbers years ago. Soon I'll finally get around to building one and see what it actually ends up at. I think I want to swap out the motor with a smaller one for physical size and amp reasons. The ratio should still end up above 1.0 with electronics. Which would be awesome.

Computer naming conventions

My computers are currently named after GIJoe characters. This works pretty well because there are a lot of them and they are usually interesting names. I just work my way down the list.

I'm thinking that it might make sense to have a convention for drives as well with a sub-convention for partitions. Having all of these gives an absolute path to the existence of a file. While usually not important I'm running across the issue when syncing stuff across the cloud.

The question is what convention. The computer names are already long enough. Adding in another long drive name and another long partition name is a bad idea. One thing that also is useful is to include the name of the OS for the drive. That fails if you have more than one OS per partition, but that's really not all that big of a deal to limit things to one OS per partition. Probably a good idea anyway.

For partitions I could name them letters. This maps well for Windows, but does not make as much sense for other OS's. Though it does fit the convention of being nice and short. If I do go with drive letters I should start them down the alphabet enough so I can map them myself and not have to worry about windows reordering them.

So what happens when a drive ends up living in another computer? Does this live on via the naming convention? Maybe the real absolute path is the name of the drive and whatever computer it happens to be parked in at any give time really does not matter.

Yet another plan

In February Janel's income goes up and it'll be right on the border of enough to support the family. We have money in the bank, but it'll be burned by then ( we've been burning it for awhile. ) So we're finally taking some of the "free" school money. We'll be proud Americans. With Janel covering the financials I will no longer need to both take care of the kids and pull $20K a year out my ass.

Which is good. Since I got back from Australia all the jobs I have taken were just to have an income. Shit. Most of my projects in that time have been directed towards being a way of having an income so that I could work on real projects. I'd get a job for a year or two and then spend the next year failing at making money and then have to go back to a job again. Taking a job always just ended up being a set back. The Ageia gig was particularly a bad plan. At the time I had the fastest and most robust physics tech around. Not that I would of built a company around it because that takes money ( though I should of), taking that Ageia job killed any of its potential. Of course when Havok was sold for $110M to Intel and Ageia for $35M to nVidia that was particularly "funny."

Anyway.

So the new plan.

1. Ship ShapeManiac for the iphone.
2. Give up.

I am no longer going to pretend that I have time enough to get anywhere on anything. It's just the same damn cycle over and over again. It just pisses me off.

I'm not really sure what I'll do with all those superfluous hours. I can always do nothing and go to bed at 9pm. Maybe I could take up the Wisconsin lifestyle and drink a 6 pack after "work." Maybe I could be ambitious and do both. I like to be ambitious.

Two weeks off

Janel finished up 3rd year and had two weeks off. That gave me some time to work.

I spent the first half of the first week on random crap I needed to get done. Stuff like finally getting our data backed up properly.

The second half was very productive and I got the iphone version of ShapeManiac to a playable state.

The second week didn't go so well. I had more random crap and over-sleeping to take up my time. Janel also needed to start taking blocks during the day. About Wednesday I started feeling that my time was nearly up and that was completely demotivating.

So the end of the 2 weeks all I accomplished was getting iphone ShapeManiac mostly done. I thought I'd be able to kick out 2 apps, but I was not even able to kick out 1. FAIL.