Here is another possible “mandible” from some other supplier. 
I think it only uses one servo.
http://www.budgetrobotics.com/shop/images/279.gif
Looking at your existing mandible design, with the fixed distance between pivot points, it seems that it would be a fairly straightforward matter to remove one of the servos, and add a geared arrangement to the pivot points of the chompers, similar to that used by the Lynx-A] gripper , to run both jaws from a single servo. Though I haven’t taken a look at them since Revell acquired and repackaged the line, the older green VEX gear sets were great to work with, and the 24-pitch teeth are rugged and fairly tolerant of slop in their mesh. Depending on the distance between your pivot points, you may or may not need to move the mounting block farther out from the chassis, in order to clear the arc for the gears.
Great work, Zenta! I found this post last night and didn’t stop reading until I reached the end. Your routines are so smooth. I can’t wait until you are satisfied enough with your excel file that you release it for to try.
Keep up the good work!
Hi! Thanks!
The PEP (Phoenix Excel Program) will hopefully be available on Lynxmotion soon. I still have some work to do and making a “how to” manual. But as usual I have been busy with my job and family. But there is some progress… 
Meanwhile, here comes a little Phoenix teaser… Just some new steps and moves.
youtube.com/watch?v=IqWKGh4XLbk
I think PEP + GP sequencer would make a great team!
WOW! The manueverability of this robot is beyond any of the other Hexapods IMO. That is super cool movement! 8)
Thanks!
Yes, I like flexibility… That’s the main reason for why Phoenix does not have any fancy spikes or flames on her legs… 
They would easily inhibit her manueverability. 
… Flames crashing into her body etc…
Wow that video is nice! You are doing some very impressive moves with your Phoenix! Nice work! 
Nice moves! Very fluid. Can’t wait to see the finished product!!
Correct me if I am wrong, but it appears that the gain in range of motion is a result of both design and programing, yes?
The Phoenix has a “long” slender femur whereas the typical hexapod on this site has a more stubby, wide femur. I’m not sure if the spacing between the “hip” joints lends a flexibility advantage or whether the tibia design does either, but it does look like the femur design adds to the flexibility.
Second, I wonder whether Zenta’s programming approach using Excel provides an easier way to design fluid motions.
Thoughts?
Yes that’s true. The Phoenix design allows the knee to be positioned directly above the vertical hip servo and still have the leg positioned downward. The SES leg design does not allow this. In addition the PowerPod created hexapod control program has several built in limits to the range of certain movements due to the complex IK math involved in keeping track of the walking gait.
The smaller body is probably more aesthetic than anything, but the long femur is the real reason for the added flexibility. The short femur on the SES leg design provides a stronger leg. That’s why the Phoenix robot can’t carry the payload the AH3-R can carry.
I wouldn’t say it was easier. (edit) I mean to create the spreadsheet program that is. The resulting program definitely does make creating the sequences easier.
The two methods of controlling a hexapod robot couldn’t be more different. CH3-R is controlled from a microcontroller / servo controller combination where the microcontroller calculates the leg positions in real time as the robot is walking. The properties of the walking are gathered also in real time from several optional sources such as PS2 controller, serial commands, or direct from the microcontrollers program using sensors. Zentas approach is to use a spreadsheet to calculate the leg positions, then send them to the servo controller to create motion. This is fine for a presentation, but the robot is acting as a puppet, and is not able to change it’s behavior on the fly. The SSC-32 does have the ability to store these canned sequences in EEPROM, and if created (set up) properly you can select which sequence to run, as well as the speed and direction. However the real power is when the two are combined, evolution will make this happen in time. If you offload the special case sequences, step onto box, push box away, any posing or posturing sequences into EEPROM and fine tune the PowerPod walking routines, you end up with the best of both worlds. This is an exciting time to be building hexapods. 8)
Hi Fish,
Just hardware, I think. A sequence designed in SEQ can move any and all servos through their entire range. Zenta’s exquisite sequences are a wonderful example of the moves the hardware can make. I’ll bet he could make a CH3-R get up and dance as well!
IF, on the other hand, you are comparing Zenta’s sequences to the code generated by PowerPod, Then you are comparing apples and oranges! When a PowerPod generated 'Bot program runs, all the movements (other then “flying” and “Attack” which are “canned”) are generated by code that has a few limited states with which to move the legs and create motion, a “gait engine”, if you will. Along with the heading and range inputs from the joystick, it uses “ride height”, pitch and roll parameters (also from joystick inputs) to alter its ambulation.
With the addition of the EEPROM, and some code changes in the SSC-32 (as I believe are planned), the program running on the Atom 'Bot Board should be able to “call up” a “canned” sequence from the EEPROM, and thus give a presentation of one or more of Zenta’s fine sequences.
What will be interesting to see is how Zenta’s elaborate productions (variations in gaits) can be incorporated into a gait engine!
The delay routine(s) in Atom BASIC currently appear to be limiting the speed of a servo move to a rather limited number of choices. This seems to be determining how various gait speeds are implemented. Maybe this will change. Zenta’s gaits may be difficult to work into a Atom BASIC program for this reason.
Simply adapting the code for Zenta’s arrangement of legs, on the other hand, would not be difficult. But you’d have LM moves, not Zenta’s.
It all boils down to range of joint motion, and lengths of the limb components, as you’ve observed. When the tables of joint angle limits and limb component lengths are developed, we should be able to directly compare them with, say, a CH3-R.
It would appear, what with the smaller size and non-symmetrical arrangement of the legs on Zenta’a 'bot, that the Phoenix would be more limited! (my concerns) However Zenta appears to be doing quite well with the constraints! Perhaps the coxa (HipH) motion is not all that limiting. Welcome news!
Hard to say, and I admit I’m not at all versed in Excel (much to my loss), or I could examine the algorithms in his spread sheet(s), and get some clues as to how his sequences are generated! I for one could certainly use that information! A good reason to learn more about Excel.
EASIER, Apparently so! Certainly would beat the painful process of determining and setting the position of each joint manually! I hope to get a hang of Zenta’s process (and spread sheets)!
Perhaps Zenta knows a DANCE instructor! That’s my problem, Engineers can’t dance!
Alan KM6VV
Hi!
Interesting subject.
I think both Jim and Alan has pointed out the essence here.
When it comes to flexibility it think the ideal leg/body construction is a design that alows the servos to move in their full range (+/- 90 deg or a total of 180 deg). A small body (short distance between each coxa) gives you a very acrobatic pod, just take a look at our fantastic nature! (insects and spiders). But a small body gives our servobased legs some limits and obstructions. Take a look at my setup sheet in PEP:
http://i214.photobucket.com/albums/cc6/ZentaOlbaid/Setup.jpg
As you can see the tibia has 167 deg of freedom, femur 180 deg, rear and front coxa (HipH) has 180 deg and finaly the middle coxa has 137 deg. But since Phoenix’s body is rather small you have to be aware of the risk of collisions between the legs. (oh… pardon my bad english… 
)
When it comes to fluid motions I think the answer is rather easy: Invers Kinematics (IK). Making fluid motions with Forward Kinematics (FK) is very hard. Creating walking gait sequences with Visual Sequencer is FK. Another major key to fluid motions are short IK steps, for every step in a sequence the tars (foot, end of tibia) doesn’t move more than 0,5 - max 2 cm. 8)
Wow! A great set of responses! Thanks to each of you for taking the time to explain your views - I’ve learned quite a bit. Perhaps “easier” was not the best word to use!
I still like both bots and understand that one is not necessarily “better” than the other. I’ll be keeping my CH3 (and my polypod) and look forward to obtaining a Phoenix as well!
The other thing I would point out about the “more natural” design - anatomical designs in animals are usually rounded and rarely include perfectly straight lines. Even the more square-shaped bones (such as the cuboid) are rounded. Perhaps the curvy nature of the Phoenix body parts and femur remind us of this.
Hi,
I have been busy lately working with my new project Oxyopus, a octapod robot. I’m just finished milling coxa and femur parts, but there is alot of work left. I will post more info and pics about this project in a new thread in the projects forum, or should I use the multi leg forum?
Anyway, I’m also working with some documentation for the PEP program, the progress is a bit slow but there IS some progress.
You can post here or in multileg which ever you prefer. Looking forward to seeing the doc’s for the spreadsheet.
Hi Kåre!
Another 'bot already? You are fast! Maybe it’s the winter weather? ;>)
I’ve kicked around some ideas for a tetrapod (quad), and also an octapod of my own (need servos for additional legs); I’ve even designed a new body for it already, but I’ve got a lot to learn about my present hexapod (Shelob). … Although that name is more appropriate for an octapod!
You’re milling coxa and femur parts similar to the Phoenix? So far, I’ve only milled tibia parts. I should investigate milling some coxa and femur parts as well! What do you mill them on/with?
Looking forward to seeing your new pixs!
Alan KM6VV
P.S. Saw you on the “list”, welcome!
Since I don’t have any fancy CNC milling machine I’m using a XYZ drilling machine we have at work. Here is a picture of the machine while I’m making the femur parts. One (almost finished) femur part lies on the vise:
http://i214.photobucket.com/albums/cc6/ZentaOlbaid/XYZdrillingmachine.jpg
But I think I’m going to make the body and tibia parts of the same plastic material I used on Phoenix.
I’ve never seen one of those before. At my highschool machine shop we have a CAM that is guided by computer software in which you design your cut-out on. Is this a manual way of doing that (as in without the computer aided software)?
Hi Kåre,
OK, I milled the body plates too. That proved to be a chore, as I had to re-setup to do the 2nd half of the body (I can only handle 4" x 8" at a setup). Next body for an octapod will require 4 EXACTING setups! No time now. I’m still cutting Polycarbonate for the body; I haven’t graduated to aluminum (cheaper until I get it right)!
I don’t have a good design for the Femur and coxa parts other then some early (bulky) sheet metal designs I did; I’ll have to give yours another look!
Whew! That must make it hard. Did I see an arc? I hope you have a rotary table!
Alan KM6VV
Yes, you do see a arc…
It’s quite easy. The vise has a rotary socket. Take a closer look at the picture