Single Axis Zero (x Axis)

[Realstevesmith2]

Is it possible to zero JUST the X axis between cuts? Preferably automatically but if we had to do it on the program with hot keys on the keyboard or a button on the screen that would be fine.

Installed program options are...
Advanced Machine Control
Basic height control
Customizable size

We have a plasmacam go torch Z-2 that has been expanded to 24" wide X (standard) x 289" Y. we built a railing system to support the side rails and lined up all of the gear tracks. we are about to expand the Y another 48". we have a conveyor running the length of the Y and use this setup to cut holes in Ibeam.

This set up works fairly well but some of the beams are bowed causing the holes to be off center. We are having to measure centerline on each set of holes. Clamping the beams is not a viable option. What I want to do is add a laser eye or IR sensor in line with the torch but on the back side of the gantry to avoid slag. I would wire it into the limit switch for the x. I would add another limit switch and a relay close to the home position of the Y so that the machine would zero off the original limit switch when close to home but would zero off the falling edge of the beam when down the length of the beam a ways.

Any help with my original question above would be appreciated!

[weldguy]

Sounds like a tall order on a Plasmacam but I have little to no experience on a Pcam. I am confident one of the Pcam gurus here will chime in shortly

[Joe Jones]

Holy cow! A 24"x337" machine?!? I'd LOVE to come to see that one in person!

You have the GoTorch mounted to a conveyor, and it runs the machine along the beam? That is GENIUS LEVEL out-of-the-box thinking!

Hmmm... So ... as a former AWS certified structural steel fabricator (with Los Angeles Earthquake Standards endorsement, no less!) ... I fully understand your problem. A long steel beam, or extrusion of any type is never "arrow straight."

(Hint:) Have you ever watched a KEY CUTTING MACHINE?

So here is my Right-Brain, out-of-the-box, super duper cool suggestion!

Rather than have the GoTorch attached firmly to the conveyor belt and UNABLE to shift its X position on that belt as the curved beam snakes beneath the 0,0 position of the machine, create a sliding platform on the X axis and mount that platform to the belt. The GoTorch will be mounted to this platform, and the conveyor will move the platform along the beam as normal in the Y direction.

Now all you need is some sort of wheel that runs along the top or bottom flange of the I-beam. The sliding platform would be either weighted or spring loaded, to pull the wheel against the steel extrusion regardless of WHERE the X contact point of the extrusion falls over the conveyor at any given point. As the torch is moved along the beam, the wheel will ride against the steel, which will establish the new 0,0 at the new X (thus the new X/Y HOME position ) at any given point. The carriage will always return to the same position on the machine, but the 0,0 home POSITION of the machine will change in relation to a stationary X/Y point on the floor with the curve of the steel beam.

In this way, the beam could curve as much as a few inches, and reverse its direction several times along the length of the beam. It will not matter, because the machine's side surface tracking wheel and the sliding platform will ALWAYS position the 0,0 home coordinates in the same place, in relation to THAT POINT along the beam's edge.

You would pull the GoTorch to extreme X, drop the beam onto the table or sawhorses or whatever .. and let the springs pull the GoTorch wheel against one side of the beam. From there, any Y movement of the torch will automatically adjust for any X position of the side of the extrusion.

I hope this helps.

Joe



.

[Realstevesmith2]

I've attached an isometric view of the machine to help visualize what we have. I haven't taken any photos or video but we have been running it for a couple months. The beams are quite heavy, approx 400 lbs. . the plasma rails are attached to a Braced U shaped leg system that is bolted to the floor. The conveyor is a roller conveyor that is bolted to the base plates. the conveyor is somewhat independent from the rail supports this way so moving the beams on the conveyor does not move the plasma rails. the conveyor wiggles a bit during transport.

I think I understand what you are suggesting. I think in order to do what you are suggesting, i would have to have a floating x on the gantry that allowed for a few inches each way. I could use cam followers to follow the edge of the beam. I would loose maybe 4 inches each side on my 24" X but the beams I cut are 6 to 8" wide so that would work. I have to be careful how much force i put this all through because the stepper count will likely suffer. I previously thought of a similar thing to what you are suggesting except i was going to move the ibeam I'm cutting from side to side but I thought it would be too heavy to move without causing other issues.

Your suggestion could work if there is not a more simple programming/sensor solution.

p.s.
- I built this all with scrap/left over metal so "No cost" there.
- The big gap between legs you see in the picture is to push a C frame 4'x8' plasma table in so I can still cut 2'x8' on a flat plate then index the table over 2 more feet. I don't cut plate though so didn't make the table...yet.

[Joe Jones]

Wait a minute! Did you buy TWELVE 48" rails and gear racks from PlasmaCam to make this thing? Now you are going to extend those rails another 48 inches to 337 inches? Frankly, I am SURPRISED they sold them to you!

I thought the whole GoTorch machine as factory assembled ran along a conveyor, but it looks like the GANTRY rolls the 289 inches along insanely long rails. Am I right?

Okay ... What about this? Securely mount another gantry tube (B) on glide rails on top of the stationary gantry tube (A).

Gantry (A) moves rigidly along the rail system you have built and does NOT move +/- X. You do NOT use the gear rack on (A). Gantry (B) slides +.- X on glide rails on top of (A), and shifts left or right on the X axis based on a spring loaded wheel that follows the profile of the extrusion. The carriage moves along the X of (B), and the X limit switch on (B) remains in the correct position for stepper count, despite the entire (B) assembly riding atop (A) and shifting on X by following the curve of the extrusion.

You would need to lower the machine torch by the height of (A) to return it to the proper cut height. Would that work?

Joe


.

[Joe Jones]

You could create a rolling HOME position brace that would follow behind the gantry tube. It would slide easily in +Y, but it would LOCK in -Y. This would allow you to "drag along" the HOME position +Y at every point along the beam. It would work like the thumb trigger on a caulking gun plunger, or a sliding C-clamp. A solenoid could be activated to release the thumb lock to move the gantry (A+B) back to the REAL HOME position at the end of the rails for the next beam via ALT-V or by typing in coordinates.

By doing this, you could quickly initialize the machine at any Y position without needing to roll the gantry 289" or less back to the end of the rails. Once you cut a set of holes, the NEW Y home position is right there, AS If you were cutting the first holes at the end of the beam.

Joe


.

[Joe Jones]

More on this ...

[Joe Jones]

It occurs to me that in order to run a torch along a beam 337" long, the plasma machine must also travel along with the gantry tube. The PlasmaCam controller box would also have to move with the gantry, OR you have figured out a way to extend the main wiring harness to work over very long distances.

I would really like to see some photos or video of this thing you have created! I am curious about exactly how you managed the logistics of a GoTorch gantry running along 337" of rails!

Joe


.

[adbuch]

Is it possible to zero JUST the X axis between cuts? Preferably automatically but if we had to do it on the program with hot keys on the keyboard or a button on the screen that would be fine.

Home Carriage.jpg

[adbuch]

If you have both the Advanced Design and Advanced Machine Control upgrades, then you could use the Customize feature to create a keyboard shortcut to Home the carriage via the jog button.
David

Home Carriage 3.jpg

Home Carriage 2.jpg

[Joe Jones]

You have a long extrusion on a conveyor belt. I will assume it is an I-beam. you need to cut holes along the flanges from one end to the other.

How many holes do you cut along one beam? I ask, because it will make a big difference in the efficiency of the process. If you are cutting six or eight holes, then a manual process is workable, although slow, and it requires someone to "babysit" the machine. However, if you are cutting 50 holes (key slots?) along each side of the flange, then you want something AUTOMATIC.

David, he knows about the controller button. The problem he is having is that the beams are curved. Which means when the carriage returns to the home position at the limit switch, he cannot send it four feet further down the beam to cut holes spaced (n) inches from the edge, because the edge of the material has moved in relation to the rails, and the center line of the gantry tube.

He is looking for a way to arrive at any point along the beam, and the either mechanically or optically "find" the edge of the beam, so the software can then plot the holes at the proper offset from the edges AT THAT POINT. So simply sending the carriage to the home position on the gantry does not solve that problem.

This is why I suggested he fabricate a second floating gantry on top of the fixed gantry which is spring loaded and rides on glide rails, with a cam follower that rolls along the physical edge of the beam to shift the gantry +/- X.

Assuming the holes in the array on the computer are in a straight line, but the beam is curved. Lets say he places the beam on the conveyor so that the end of the beam is dead center beneath the center point of the gantry. When those holes are cut, they are in the correct places. However, several feet along the beam, the gantry and the home position of the carriage have NOT moved, even though the center of the beam is now two or three inches to one side of the center of the gantry tube. This is causing the holes to be cut at the wrong distance from the edges of the beam.

I would love to ride out there and help him fabricate a floating gantry with a cam follower to track the beam's edge. I think it would solve the problem.

Joe



.

[adbuch]

Is it possible to zero JUST the X axis between cuts? Preferably automatically but if we had to do it on the program with hot keys on the keyboard or a button on the screen that would be fine.

You will need to be more specific for me to understand what you are asking. To me, "zero the x axis" means to send the carriage to its home position relative to plasma table. If you are asking about a method to find the center of your curved beam and then zero the table on that, then that is another question entirely.
David

PS - also MSEE

Business Card.jpg

[Joe Jones]

This is the issue he wants to resolve.

He calls the gantry X and the rails Y.

In order to place the holes at the correct distance from the edge along the curved beam, a +/- X spring loaded sliding gantry tube atop the stationary tube will be required, so the centerline of the gantry will ALWAYS be over the center of the beam, controlled by a cam follower that pushes against one side or BOTH sides of the metal flange as the gantry moves in the Y direction along the beam.

The beam is straightened in the construction phase, but he cannot straighten the beam on the conveyor belt, and EVERY BEAM IS DIFFERENT.

Joe

CURVED BEAM ISSUE JPG.jpg

[Joe Jones]

You can't move just the carriage, because the encoder on the carriage motor will lose its place. you cannot slide the gantry gear rack along WITH the carriage, as that would be monstrously complicated. You cannot plot the holes along a "fit to path" because every beam is different. Some of them are C-shaped. Some are S-shaped. Some are straight at one end, and curved at the other.

The easiest solution is to place another gantry on top of the stationary gantry, and let it slide +/- X via a spring loaded pull against a large cam follower that tracks the beam's X variances. This will always keep the centerline of the gantry over the centerline of the beam, regardless of which X direction that centerline moves. I see no other solution.

Joe

[Joe Jones]

Actually, there IS one other solution.

Mount an alternate Carriage limit switch to the carriage itself, and extend a feeler wire down. As the carriage moves to home position, the stiff wire hanging down will hit the far side of the beam and trigger the carriage to believe it has hit the home position. So the holes would be plotted AS IF the edge of the beam was at X=0

That would work too, I believe.

Joe



.

[adbuch]

zero x question.jpg

[Joe Jones]

I have already told him how to resolve the problem.

Joe


.

[adbuch]

I have already told him how to resolve the problem.

Joe


.

He just posted this question today at the Owner's Community Forum, so perhaps he didn't understand your methods or is looking for a different answer.
David

[Realstevesmith2]

Hello Guys. THANK YOU FOR YOUR SUGGESTIONS. I had posted my question in both forums the same day because I wasn't sure what response I would get and if it would be the same pool of people. I'll update the other forum as well.

Unfortunately we had a death in the family so I haven't been able to respond sooner.

Joe, you have put a ton of effort into helping me. Thank you so much! I like your idea about the secondary gantry and agree it would likely work well. I wonder if it would make my holes more egg shaped but maybe not since they are fairly small ay 3/4". Each beam has 30 to 40 holes. half or more on the top side then we rotate the beam 180 degrees and cut the rest on the bottom. I attached the controller to the gantry with a bracket I made and I'm using a cable carrier for cable management. When the beams are straightish, it works well. IF I was able to hit a key stroke to do a machine zero (Not program zero) I would install a IR or laser sensor in line with the torch but on the back side to protect from slag. I would use an additional limit switch and relay to only utilize the stock limit switch when I'm within 2 or 3" off the true zero. Otherwise the laser wired normally closed would be open when the torch is over the beam and would close when it fell off the side of the beam. The sides of my beam are on the program zero. I can adjust the offset for zero in the program if I want to account for bow in both directions. I'm exploring this electronic approach first since it would be faster and cheaper but I am on the same page with you on the double decker gantry. Thanks for your video and illustrations. the feeler gauge idea is similar to my laser idea but i want to just tell the X to zero. when I press go home or initialize it zeros everything. I will try to get some video and pictures tomorrow.

David, thank you for your comments and to clarify. I want to send just the X to machine zero. I will be in front of the machine mid day tomorrow and will see if I have the customize option and see if I can send it to X zero there. I see where you highlight to jog the X in the negative direction but getting it to zero when it sees the sensor close would be the trick. if there was another hot key to set machine X to zero I would install a light on the sensor and do it manually if I had to and that would be quicker than measuring. See above on my thought on how to use a sensor to catch the edge of the beam. I do not have advanced drawing. I use a separate CAD package. If I had a sensor set up and a hot key to zero just the machine x I would do the following...
- Initialize at the start of the beam.
- cut the first set of holes (they are in groups of 4)
- use the move command to move the machine to the center of the next set of holes
- hit the hot key to zero just the machine X
- cut that set of holes
- repeat for each set of holes.

This is still simi automated because I have to control each set of holes but it is way better than what I am doing now and I don't see any way to program a move to the next set of holes and auto zero the machine X.

Thanks again everyone. so sorry for the delay

[adbuch]

- Initialize at the start of the beam.
- cut the first set of holes (they are in groups of 4)
-

use the move command to move the machine to the center of the next set of holes
- hit the hot key to zero just the machine X

- cut that set of holes
- repeat for each set of holes.

You are confusing machine zero with program zero. Sending the machine to it's X-zero position (home position) is not the same thing as re-zeroing on the actual part (in this case a group of holes). What you need to do is to use the "torch snap" function to move the cut paths for the group of holes (as shown on the screen) to the new physical location of the torch.

So your process would be

- use the jog function to move the machine torch to the center of the next set of holes
- use the move command with torch snap to move the set of holes as shown on the screen to the new location (Program zero)
- cut that set of holes
- repeat for each set of holes


David

[Joe Jones]

If I came out there and spent a day or two with you, we'd have that think signing checks!

Think about this... If you create the floating gantry tube, which would keep the carriage (X) in sync with the (x) movement of the gantry, you could create a simple leading or trailing set of wheels that follow along the edges of the beam as it moves (+/- Y) The four holes would be cut, the stationary gantry in the rails would move along the beam to the next four-hole position but the floating gantry would shift (+/- X) with the bow of the beam. The gantry would retain its centered (X) position over the beam as before EVEN IF (Y) changed, and the four holes would be cut again.

I would set it up so you could lay the beam onto the conveyor, and set the gantry at the start of the beam. The entire series of holes would be cut in the correct places all along the beam from one end to the other, without human intervention. I will ASSUME that the bow in the beam is not SO severe, that the four holes would need to be rotated +/- (n) degrees to retain their parallel relationship with the beam's edges.

So you attached the controller to the gantry! Smart, but now I must ask, does the PC also move with the gantry, or have you magically worked out the communication issues with extremely long parallel cables? It would seem that the PC and the plasma torch would BOTH need to make the journey from one end of the beam to the other.

Tell me ... do you have ANOTHER PlasmaCam table there, or some way to design and fabricate brackets on the fly? I would want to come there, and create some alternative components on the spot, to test them right there.

I absolutely MUST come out to see this machine in action!

Joe



.


.

[Joe Jones]

I have thought of another way to deal with the issue of BENT BEAMS. I will draw something up tonight.

Joe



.

[adbuch]

Is it possible to zero JUST the X axis between cuts? Preferably automatically but if we had to do it on the program with hot keys on the keyboard or a button on the screen that would be fine.

Installed program options are...
Advanced Machine Control
Basic height control
Customizable size

We have a plasmacam go torch Z-2 that has been expanded to 24" wide X (standard) x 289" Y. we built a railing system to support the side rails and lined up all of the gear tracks. we are about to expand the Y another 48". we have a conveyor running the length of the Y and use this setup to cut holes in Ibeam.

This set up works fairly well but some of the beams are bowed causing the holes to be off center. We are having to measure centerline on each set of holes. Clamping the beams is not a viable option. What I want to do is add a laser eye or IR sensor in line with the torch but on the back side of the gantry to avoid slag. I would wire it into the limit switch for the x. I would add another limit switch and a relay close to the home position of the Y so that the machine would zero off the original limit switch when close to home but would zero off the falling edge of the beam when down the length of the beam a ways.

Any help with my original question above would be appreciated!

By the way, all 4 of my other cnc tables do have the ability to re-zero on the part. Mach3, Centroid CNC12, and Candle are the control software and each one of these makes it easy to set the program zero anywhere on the table with the simple push of a button. So I can move my spindle/torch to the location I choose, and then simply re-zero either x or y (or both). Plasmacam does not offer this option. Instead, the cut paths must be moved in the x/y coordinate system of the table to align with the desired program zero location.

In addition, all of these other control software programs offer the ability to control separate I/O lines for monitoring external inputs or controlling external devices. Plasmacam does not offer this capability. So to do what you are asking, a separate computer/controller would be required for running an additional gantry and monitoring any laser activated inputs.

Another option would be to ditch the Plasmacam controller/software and substitute something like the Centroid system with the existing servos or perhaps Mach3 with stepper motors.

Or perhaps do as has been suggested and add an additional "gantry" independently controlled by a separate dedicated system (Mach3, Centroid, etc.) or perhaps even an Arduino/GShield/GRBL/Candle combination.

David

[Joe Jones]

Waaaayyy too complicated, and overkill. All he needs to do, is to find the edge of the beam, to place four holes through the flange at any given Y. The gantry does not move in the X direction. The carriage home limit switch position does not move in the X direction.

Normally, from the home position, you set the PlasmaCam to cut, and it moves to the first cut path, presumable the top left hole. It cuts that hole, then proceeds to cut the other three. After that, it has the option to return to HOME, or to move (n") Y and cut four more holes. However, this will not help him, because he needs to move (n") farther along Y to cut the next four holes, and those four holes are at a different X.

He cannot lay out the cut paths as an array in DesignEdge, because each beam is curved differently. So the center X of the four holes MOVES +/- X as the beam curves.

I would want the entire thing to work automatically. However, I believe he is MANUALLY moving the gantry to the next Y position, or he is using the keyboard to {Alt-V] to a new Y coordinate. Then he still has to deal with the four holes not being where the last four holes were, because the beam either curves toward, or away from the carriage limit switch as the gantry moves along Y.

He wants to move the gantry to the Y of a four-hole set, [Alt-C] and cut four holes in the right place, but he needs the machine to determine where the edge of the beam is at X, so the holes will fall correctly at that point. This is why I suggested a floating gantry tube. Once the stationary gantry is at the correct Y, and the floating carriage centerline is aligned with the beam centerline, it will cut the four holes as normal. Two large cam followers rolling along the outer edges of the beam will guide the floating gantry to the correct X with the carriage then in position to cut the four holes correctly. The floating gantry would not move, as it is being held at X by the cam followers against the edge of the beam.

I am really thinking of heading to his place, to see this things in action, and to resolve the curved beam issue. I would not be happy unless ALL of the holes along the beam were cut automatically, with no human intervention.

Joe


.

[djreiswig]

I think it could be done fairly easily in Mach3 with a couple of macros, but I don't think PlasmaCam allows this sort of customization. Guess that's the tradeoff of an all-in-one program.