I came up with a ripple gait, and wanted some feedback on it. One thing I’m not sure about is if I’m going forward fast enough. It just doesn’t seem as elegant as other ripple gaits (like the ones on the main Lynxmotion site, or Zenta’s Hexapod, Pheonix.
I’ve uploaded a Youtube video as well as the sequence file from the SEQ program. If anyone tries it, the servo wiring is the same as the official assembly guide. romshark.com/misc/hex%20-%20 … orward.csv
(Honestly, almost all the SEQ files available for download are for the BRAT. I like the BRAT, and own one, but we need SEQ files for the other robots also.)
Here’s the video. I know it goes off at an angle, but this is the first test. I’m ordering a new (longer) body shell anyway. The green wires on the legs are for lights that I haven’t hooked up yet. youtube.com/watch?v=cnI_PZru424
A section in the forums for people to share SEQ files would be nice, but Beth will not be able to even consider this at this point due to other responsibilities. But it would be nice if people could attach SEQ files to share with others without having to go to another link.
Take a closer look to the Visual sequencer manual on page 10 :
“When you export a project it creates a ‘shp’ file too (with the same filename), it contains the robot shape and params from the sequencer module.
You’ll have to provide both csv and shp files in order to share a project!”
We need a copy of your .shp file too.
Also pay attention to this warning:
“Importing your home-made csv is a risk to crash the internal database!
To prevent this, close the software, go to installation directory and backup the two files ‘MovesSeq.dbf’ and ‘MovesSeq.mdx’ in a secure place. If you crash the database, you’ll just have to close the software and restore these two files.”
…
I guess I need to read the manual. Thanks for pointing out about backing up the databases.
I’ll upload the other file later. I’m not at my computer with the files on it.
I tried your sequences on my Phoenix…
Your gait is almost a ripple, it seems like you got the sequence correct. But one fault is that only one leg are raised or lowered per step. In the ripple gait there are always one leg raising when the previous leg is lowering.
See this link: oricomtech.com/projects/leg-time.htm
Another thing is that the legs are not moving in a strait forward direction, meaning that the tars (foot/end of tibia) have to move in a parallell direction relative to the body. When you alter the coxa angle you also have to alter the femur and tibia angle. One tip is to place a ruler lengthwise to the body and then make small correction to the femur and tibia angle when you alter the coxa angle. You can call this manual invers kinematic. hehe
I’d like to study the Atom BASIC (?) code that generated the “Ripple Forward” gait illustrated below. I haven’t come across a reference to it in any of the tutorials or build guides, has anyone seen it?
Laurent will be working on updating PowerPod for the Atom Pro processor after he is finished with RIOS. This update will include a multitude of walking gaits. I know it’s taking him longer than expected with RIOS though.
I think I (and others) would be more interested in code for the new Atom Pro processor!
I did just get some insight to quadraped gaits, I posted them in reply to Zack’s thread on quads.
I did notice that quad legs are vertical and inline, rather then cantilever. However that makes for a tall 'Bot. Kinda like putting 4 biped legs UNDER a platform. Might be interesting.
We have been in business as Lynxmotion, Inc. for 12 years! We operated as Technical Services and Solutions for a couple years before that. We are celebrating our 12 year anniversary this Halloween. Make sure to visit the website the week of, as we may have some specials, nudge nudge, wink wink…
