may i… All black!
Maybe you should run a poll? 
Might as well keep with most of my others and go all black.
Thanks
Kurt
Hey, you know me…
All my bots are black, I’m in for black again! 
ps. This sure goes fast! I should free up my agenda for the next weeks! 
Happy to work as a team on this guys!
Xan
Xan, would it be advised to use the same code math you used for the hybrid (a code im still dying to look at) or would you recommend a rewrite. i remember we had a few hills to climb regarding the Math. have you made any developments to that?
I should be building a hex again soon, so it will be great to test TA on that.
Kurt, did you start a thread on how to assemble the code that can be used with different DOF leg designs?
The hex i have planned is for a 4dof on the back and middle with 3DOF for the front. 8)
The idea is to give the hex the ability to walk with 6 legs and also with 4, then using the two front legs as “arms”. 8) 8)
Im also thinking of adding a pair of small grippers to the “tars”. 8) 8) 8)
Nope - I think Zenta started the thread
But I did put some code up there that may work for multiple leg configurations… I think it is on this page: viewtopic.php?t=5670&p=68067#p68067
Kurt
Aha there it is. thanks.
Hi Jonny,
The hybrid code is (almost) as universal as the phoenix code. So porting it to the T-Hex will be no problem. The hybrid has some limitations since the tibia is very long referring to the femur. The legs of the T-Hex are more balanced so this will solve that problem. The idea is that I take the hybrid code and develop it further on the T-Hex.
Kurt is willing to join me on this quest. I think things will move pretty fast, having someone with his programming skills working on the same project.
The code will be shared once it is more stable. Currently I’ve got some problems with the TA part combined with the gait engine.
To be continued…
Xan
Ok thanks for the update xan. Thats good news. Things are looking up. Im sure jims new design will make this easier. Will look forward to further updates on this. Thanks again.
Update… Well I got sick and have missed a week of work… I never miss work! James was able to get all the legs done, and I was able to get one fully assembled bot before everything went all wonky. Give us a little more time to finish these up.
Nice to let the pro’s work, i dont even understand the basics…
I understand that somehow i’ll use WKP-interrupts using either ASM or BASIC
(viewtopic.php?p=68122#p68122)
But if I telll the leg to go down using a HSERVO, and my footsensor suddenly strikes the floor raising a interupt , how do i stop it ?
With Hservo, once you get the interrupt, you can do an HservoPos, which will return to you the current position (not the end position). You can use this to try to stop quickly or what I was planning to do was to know where I was going from and were I was going and how much time that was supposed to take, then take when the interrupt happened plus some fudge factor for response delay, guesstimate where I think it happened and then issue a new HSERVO command to tell those servos, to go there…
Hope that makes some sense…
Kurt
ok, i get it (i think):
You do a new HSERVO to the same position you allready are in … that should work, dont it ?
Yes, thats what he is saying. Do an hservo [p0\hservopos(p0)] to stop the servo at the current position of the servo.
Understood 8)
I have read some examples of TA, and my idea is to simplyfy it a bit by only using TA when the fot is in air moving towards ground, and when the footsensor indicates “this is ground” the movement is stoped, it fixates the Z position and disable the sensor. This way only one sensors are active at the same time and the risk for “shaky legs” is much less than when the sensors are active all the time (because i only activate TA in a “carefull gait” only moving one leg at the time)
If the foot dont reach any ground the hexapod goes into “pit mode”, the sensor is disabled and the leg moves up again to its top position so it wont me an obstical for the hexapods movement forward. If this happens a second time (when the hexapod allready is in “pit mode”) it stop forward movement and indicates by sound “I cant go forward”
The next time the rised leg in question moves, it moves whith the same Y, and the “pit mode” isnt reseted until the leg is supposeed to be lifted again the next round
Everytime a full “good” round is over whithout the hexapod has gone into “pit mode”, the Z possitions for all legs is recalcylated threedimentional around a “Ideal Z” so there is no risk of the legs runing out of Z-freedom, for example if the hexapos suddenly goes uphill or even sidehill
Of course this simplyficaton have some problems, for instance if someone liftes the hexapod, or if the ground moves away from an allready fixed Z-position, but i think this may be indicated by a accelerometer or something …
Is this a good idea (is it understanable inspite of my bad english), and what have i forgotten? 
Some TA-ideas
Hope you are feeling better now!
Have you had any time to try out the new bot?
Kurt
Yes we did have quite a little distraction there. 
Both bots are mechanically assembled. One is getting the electronics installed and wired now. The other is soon to follow. We are making progress again. Hang in there…
Hi guys,
Sounds great! I’m exited to try the T(A)-Hex!
@Kurt, I’m documenting my TA findings to give you a quick start. I hope to post the information this weekend or the next. I think that working on this project with a couple of people will give it a quick boost! Looking forward to that!
Xan
Update and an idea… James has the bots up and walking. We are planning to ship them Tuesday or Wednesday of next week. If you guys think it’s do-able I thought it would be a good idea to test the foot switches thoroughly before they leave here. Would it be terribly difficult for one of you guys to write a simple program to test the switches? With the bot standing lift one leg, enable the speaker, lower leg, short beep indicating switch closure, then repeat on the other 5 legs. This way we will know for sure both bots are identical in every way and that the switches are working properly. If not enough time we can test them manually. Thanks! 8)
This would also be a good opportunity to get a standard for the switch pins set up. Right now, I have them set to pins 3-8. This is to help preserve as many of the analog pins as possible for other use. Let me know if we should change this or not.
Right front: P03
Right middle: P04
Right rear: P05
Left front: P06
Left middle: P07
Left rear: P08