@ Aram
It is an electrical plasma torch, see here:
http://www.kjellberg.de/pdf_files/prospekt_hf80i_en.pdf
The feedrate for cutting is around 5000 mm/min.

The machine is a self built one, made of cast aluminium profile.
But for these accelerations only alu profile is not stiff
and heavy enough. I will add some heavier steel stringers to
the bed to make it heavier and stiffer.
/@ Aram


Back to my question...
I think I will try to solve my dual axis problem with a new hal
component. It should provide a virtual home switch for
those 2 axes. This virtual home switch output will be high
when the first edge of the 2 real home switches rises and
low when the second edge falls. This way I get the homing
abstracted to the motion controller. I can assume the 2
home positions are close to each other. Further this
component will have inputs for master and slave feedback
and the homed parameter of the motion controller.
The output, besides the virtual home switch, will be
the slave position command.
This way the slave axis is transparent to the motion
controller.

Did I forget anything? I hope not, any suggestions are
welcome.

Thanks

Marc


Am Dienstag, den 23.01.2007, 01:51 -0800 schrieb Kasparov, Aram:
> Hi
> 
> I am also interesting in a body of machine for plasma burner. Do you building 
> from the scratch by buying components or you retrofitting some kind existing 
> unit with EMC2?
> 
> Plasma burner - burning process is very slow process and my question is why 
> do you need so high acceleration such as 300 ms to 30 m/min? 
> 
> Unless you do high speed machining you don't need that high acceleration 
> capability.
> 
> Also plasma burner doesn't need be very accurate. I believe that +/- 0.050 
> inch of repeat ability is plenty for my application. 
> 
>  
> 
> Thank you,
> 
> Aram Kasparov
> 
> 
> ________________________________
> 
> From: [EMAIL PROTECTED] on behalf of Hansjakob Rusterholz
> Sent: Sun 1/21/2007 9:30 PM
> To: Enhanced Machine Controller (EMC)
> Subject: Re: [Emc-users] Gantry kinematics and homing
> 
> 
> 
> Hi
> 
> The system builder also tried the way with an shaft.
> He said you will never get the desired accuracy and dynamics, but for 
> an plasma cutter, this may be not the problem.
> The gantry was about one metric ton and accelerated in 300 ms to 30 m/
> minutes!
> Only the servo amplifiers costed me over 2500 $ each (about 15 Years 
> ago)
> 
> Hansjakob
> 
> Am 22.01.2007 um 05:38 schrieb Kasparov, Aram:
> 
> > Hi
> >
> > I am building similar plasma burning table (10 by 10 feet)  and I 
> > am going to solve problem of second servo mechanically by putting 
> > long 10 feet shaft. That is more reliable and simple way to solve 
> > problem of wide gantry machine. The most important that way will 
> > reduce need in one servomotor and one amplifier and that in my case 
> > equal $950.0 + $550.0
> >
> >
> >
> > Thank you,
> >
> > Aram Kasparov
> > Hi Marc
> >
> > I was working long time on an six axis gantry laser system with about
> > 2.7 to 1.8 meters travelling range.
> > As I understood from the machine builder, they used also two
> > "independent" servos fore both gantry sides, one was the master
> > servo, and the other was tracking to the master. Both axis had linear
> > encoders, home and end switches.
> > For reference, the moved parallel to the home position and the linear
> > encoder was used to adjust the home position final, because the have
> > also some repetitive pattern (Ca all 10 mm) to get home position to
> > about 1/1000 mm precise. If the home position was not found for both
> > axis in some window, an emergency shutdown occurred.
> > If you had bad luck, You had to loosening both gantry sides, and pre-
> > adjust the gantry manually (even the technician moved once the servos
> > controlled in opposite direction to get the machine working again
> > after an bad crash.)
> >
> > I can not tell wich is the best solution for EMC2, but maybe you get
> > some hints from my descriptions.
> >
> > Hansjakob
> >
> > Am 21.01.2007 um 00:08 schrieb Marc Bodmer:
> >
> >> Hello
> >> I am on my way on building a gantry table for a plasma cutter and
> >> setting up EMC. So far i am very impressed with EMC, thanks for
> >> that great work.
> >>
> >> I am thinking about adding a second servo motor to the X Axis so
> >> that it would be driven on both supported sides of the gantry.
> >> However I am not so sure about how I could best do that.
> >>
> >> First I would like to explain my setup so you can imagine the
> >> situation. The servos have planetary gearheads and drive the
> >> gantry supports by timing belts. There are no linear encoders on
> >> this axis, but the servos give resolver feedback to the servodrives
> >> (Baldor Microflex). The drives convert the feedback to encoder
> >> signals wich are fed to a Motenc Lite Card and EMC.
> >>
> >> 1. The first idea is to create a fourth servo PID position loop
> >> for the new X1 axis and connect it to Xpos-cmd together with
> >> axis X0. In the ini file I wouldn't add another axis, but
> >> take the values for X1 and X0 from the X axis parameters.
> >> For the feedback to the motion controller I would only attach
> >> the X0 PID position feedback Xpos-fb. X1 Feedback wouldn't be
> >> connected.
> >> The problem I see in this setup is that first the feedback
> >> of X1 isn't connected and I see no way for homing X1.
> >>
> >> 2. The second way would be to use the gantrykins component.
> >> I had a look at the code, but did not see right away whether
> >> there is a solution for the homing problem. The synchronisation
> >> of these 2 axis however seems to be handled.
> >> Did anybody use this module already? Any experiences?
> >>
> >> 3. Another way I would see is similar to the first way, except
> >> that there wouldn't be a position loop for X1 axis. I have a
> >> master-slave setup in mind where the master axis would be X0
> >> and the slave axis X1. The controller loop would control
> >> position difference between these axis. Similar to the
> >> HAL component "encoder_ratio". However I see no way again
> >> in homing the slave axis.
> >>
> >> The homing problem is the following: Since there is always
> >> a small position error between those 2 axis it would accumulate
> >> up everytime X0 is homed, if X1 is dumbly homed at the same time.
> >> There should be a way in first home X0 (of course both axis have
> >> to move together) then X1 and instead of zeroing X1 it should
> >> correct it's position in respect to X0. This way the error
> >> would be eliminated in every homing cycle and not accumulate
> >> up. I don't know if there are any "professional" ways of
> >> solving this. (Would be interested in it if anybody knows about).
> >>
> >> Did anybody do such a gantry setup with a second servomotor
> >> already? Any experiences? How to best do it?
> >> I am pretty unsure where I should start.
> >>
> >> Thanks for any replies in advance
> >>
> >> regards
> >> Marc
> >>
> >>
> >>
> >>
> >>
> >>
> >> ---------------------------------------------------------------------
> >> -
> >> ---
> >> Take Surveys. Earn Cash. Influence the Future of IT
> >> Join SourceForge.net's Techsay panel and you'll get the chance to
> >> share your
> >> opinions on IT & business topics through brief surveys - and earn 
> >> cash
> >> http://www.techsay.com/default.php?
> >> page=join.php&p=sourceforge&CID=DEVDEV
> >> _______________________________________________
> >> Emc-users mailing list
> >> Emc-users@lists.sourceforge.net
> >> https://lists.sourceforge.net/lists/listinfo/emc-users
> >
> >
> > ----------------------------------------------------------------------
> > ---
> > Take Surveys. Earn Cash. Influence the Future of IT
> > Join SourceForge.net's Techsay panel and you'll get the chance to 
> > share your
> > opinions on IT & business topics through brief surveys - and earn cash
> > http://www.techsay.com/default.php?
> > page=join.php&p=sourceforge&CID=DEVDEV
> > _______________________________________________
> > Emc-users mailing list
> > Emc-users@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/emc-users
> >
> >
> >
> > <winmail.dat>
> > ----------------------------------------------------------------------
> > ---
> > Take Surveys. Earn Cash. Influence the Future of IT
> > Join SourceForge.net's Techsay panel and you'll get the chance to 
> > share your
> > opinions on IT & business topics through brief surveys - and earn cash
> > http://www.techsay.com/default.php?
> > page=join.php&p=sourceforge&CID=DEVDEV________________________________
> > _______________
> > Emc-users mailing list
> > Emc-users@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/emc-users
> 
> 
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share your
> opinions on IT & business topics through brief surveys - and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
> _______________________________________________
> Emc-users mailing list
> Emc-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/emc-users
> 
> 
> 
> -------------------------------------------------------------------------
> Take Surveys. Earn Cash. Influence the Future of IT
> Join SourceForge.net's Techsay panel and you'll get the chance to share your
> opinions on IT & business topics through brief surveys - and earn cash
> http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
> _______________________________________________ Emc-users mailing list 
> Emc-users@lists.sourceforge.net 
> https://lists.sourceforge.net/lists/listinfo/emc-users


-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys - and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to