This video, last time I looked, had only 575 views. It deserves so much more credit than 575 views would indicate. For the amount of effort and engineering that was poured into this project it is criminal that they are not making it today.

Enjoy.

YouTube - G&L Variax Hexapod Mill

Nah, what's more criminal is seeing "Giddings and Screwus" adorning it, since it was actually a Kearney and Trecker patent. Ahhhhhh, another step in the evolution of evil CEO's.

Who is Paul C. Sheldon?

http://www.theroethlegroup.com/pages/PCSvitae.html

does he appear in the K & T "Variax" patent dated October 11th 1994?

http://www.delphion.com/cgi-bin/viewpat.cmd/US05354158__?MODE=fstv&OUT_FORMAT=pdf

does he appear in the G & L "Variax" patent dated February 14th 1995?

http://www.patentstorm.us/patents/5388935.html

and doesn't the 'assignee' in a patent cough up the cash?

I've got a few patents with my name on them as the inventor and I am sure that the assignee that paid the fees would hunt me down if I was to take it to another company. As I understand it, the inventor gets credited for the idea but the assignee owns the patent. Surely this was done above board as an agreement between K & T and G & L?

Thanks for the info SCRAPWOTSCRAP.

Now who wants to build a hobby version? :devious:

Who is Paul C. Sheldon?

http://www.theroethlegroup.com/pages/PCSvitae.html

does he appear in the K & T "Variax" patent dated October 11th 1994?

http://www.delphion.com/cgi-bin/viewpat.cmd/US05354158__?MODE=fstv&OUT_FORMAT=pdf

does he appear in the G & L "Variax" patent dated February 14th 1995?

http://www.patentstorm.us/patents/5388935.html

and doesn't the 'assignee' in a patent cough up the cash?

I've got a few patents with my name on them as the inventor and I am sure that the assignee that paid the fees would hunt me down if I was to take it to another company. As I understand it, the inventor gets credited for the idea but the assignee owns the patent. Surely this was done above board as an agreement between K & T and G & L?

Thanks for the info SCRAPWOTSCRAP.

Now who wants to build a hobby version? :devious:

The issue would be the control (Although wasn't there some talk about EMC being able to do this a few years ago?).

In theory it should be an easy build. Six linear actuators and you're good to go. If I recall correctly, they don't need to be as strong as "normal" machines because the loads are always in tension.

The downside would be the floorspace needed for a given work envelope.

Yes - emc has the kins to do a hexapod.

http://www.electronicsam.com/images/KandT/testing/Screenshot.png

About all I know of is a few univerities that have goofed around with it. I always wanted to make one :)

http://www.imac.unavarra.es/~cnc/

sam

It would be a very cool project.

A black hole for time though....

Hi Samco, Dropout

I have seen the EMC version and had a real quick look but I haven't gone too far into it. The picture above looks like it would be good to move the workpiece/spindle which was placed on top of the moving platform but not in the same way as the Variax. I think the difference is the position of the nodes. The Variax has crossed actuators which allow the moving platform to get much closer to the fixed platform.

Does EMC allow for changing the position of the nodes to suit the Variax configuration?

Sorry Derek, seems I let a little of my bitterness seep into your thread, will elaborate, original patent K&T, Sheldon and his team worked for K&T. Will give you the very abbreviated story. K&T CEO, runs company into the ground (typical), leaves to become CEO of (cash rich) G&L. G&L buys K&T. Now, inquiring minds might wonder if said CEO was on payroll of certain company the whole time. Was an amazing machine. As for the hobby version:

http://www.cnczone.com/forums/video.php?do=viewdetails&videoid=95

Hi SCRAPWOTSCRAP

I checked out the video, cool concept. As for my proposal to build a hobby version, I was thinking more along the lines of a blatent copy of the principle of the Variax. I would love to be involved in a group that built one of these. I have heard rumors that the reason it didn't get off the ground is that they didn't want to promote a machine that effectively made their manufacturing plant of series axis machines redundant.

As for the seeping of bitterness, no problem at all. Vent 'til your heart's content.

For anyone else checking out this thread, get on board, talk PKM, get involved, PKM is the way of the future.

PKM - Parallel Kinematics Machine

Thanks for the video derekj308, I'd only seen pics of this machine a long while ago. I like the Variax machine, good to see PKMs being talked about again too.

I played with EMC very briefly to see if the Hexapod kinematics "worked", and it seems to at least in simulation. It's a very good question whether the apex joints (nodes or whatever they are technically called), can be adjusted in EMC. I fear there are very few EMC people who have actually had much to do with PK models out there now. I'm certainly no programmer and it might take a pretty expert mind to see it in the code. I have to admit the programming side intimidates me a little and I quickly got discouraged. (The 3D modeling has some bugs in there too BTW).
The other thing is I am not sure how to do calibration of the Hexapod kinematics in EMC. What about actuator error mapping, how is that done?
From what I understand on Hexapods, the forward kinematics are difficult and generally require numerical methods to solve. However the inverse kinematics are in closed form solution so there is no problem.
But what about the kinematics of the Variax? It may require access to published papers to get the kinematics sorted out. Does anyone have access to journals?

I put my plans on hold for a Hexapod long ago.:o I still have 6 Haydon actuators, THK rails and a 6-axis Dynomotion KFLOP controller sitting here taking a back seat to my other projects.
I swear I was going to get around to it soon. :rolleyes:

Cheers guys :)

Looking at the kins file

http://git.linuxcnc.org/gitweb?p=emc2.git;a=blob;f=src/emc/kinematics/genhexkins.c;h=1fa4967b6a6efcb9ce9109559efa61548b3c4fe9;hb=b7b3b24502aa7afb95b68a76b66b571274344b33

It 'looks' like it may be possible as the strut end positions are independent of each other.

Yes it does forward and reverse calculations.

sam

Sam,

Thanks for pointing this out!

Yes thanks for that link Sam. I had a quick look over it and I think it is within my grasp to understand it. I am familiar with the maths behind matrices and the documentation terms are also familiar to me. I am hoping that understanding how the EMC2 guys have tackled this will be enable me to apply a similar method to my design.

Cheers