Single Axis Zero (x Axis)

[Joe Jones]

Okay ... this will work also. Rather than a gliding gantry tube, set up the torch on a X glide rail, with an edge finder and a solenoid lock to prevent movement of torch in relation to carriage. Once the carriage moves to extreme X, the torch will be over the top left hole. The torch glide car is locked with a solenoid. The edge finder is retracted via a solenoid. Four holes are cut in the correct position. Move to new Y position. Rinse and repeat.

Joe

GLIDE RAIL TORCH.pcm

.

[Joe Jones]

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.

Nah! You just have to think outside of the box!

Joe

GLIDE RAIL TORCH JPG.jpg

[adbuch]

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.

I agree that this would be easy to implement if his table was running with Mach3, Centroid, etc. This is not necessarily a trade-off of an all-in-one program, but is due to the lack of customization that Plasmacam implemented with their hardware/software design. There are many features which Plasmacam could add, but I think they have no financial motivation to do so.

For example, it has been mentioned on more than one occasion that it would be nice to have the ability to add an auto tracing attachment to the Plasmacam table - something like the one that ArcLight Dynamics offers as an optional attachment for their cnc plasma tables. This uses the Scanything software developed by Les Newell which uses an inexpensive digital camera/light source combined with his software to automatically trace a pattern, drawing, template, etc. and output as a dxf file. I use this setup on one of my cnc router tables running Mach3.

But short of hacking the Plasmacam hardware/software to make modifications, this is not a possibility for the Plasmacam table owner. The op talks about adding additional limit switches and/or laser to edge find, but just not doable with the Plasmacam table - as I expect he will find out as he investigates further.

David

[Joe Jones]

I take it all back! A floating gantry or a floating torch on a glide rail WILL NOT give you the results you expect. The holes would not be perpendicular to the edge of the beam!

The only accurate way to cut these holes along a beam of unknown and variable curvature is to create a crawler that will attach the GoTorch gantry DIRECTLY to the beam, and follow the beam's contour, cutting holes as it reaches each Y point.

I have made a video to explain this. I will attach it here when it uploads to You Tube.

Joe



.

[adbuch]

I would just attach the entire go-torch directly to a "carriage" (trolly) that rolls along the length of the beam. Then just roll it to the "Y" position along the beam, lock in place, and press "Start" - either directly on the keypad or using a custom keyboard shortcut from the pc keyboard (wireless or corded). So the beam is stationary and the go-torch moves along the beam, as opposed to the OP's method of keeping the go-torch stationary and sliding the beam thru the go-torch. If budget permits, I would also consider shrinking down the size of the go-torch to something just large enough to do this job. So instead of 24" x 24", maybe 12" x 12" or some other size just large enough to accommodate the job at hand. Spare parts (gantries, rails, etc. are pretty cheap) so just customize it for each job.
David

[Joe Jones]

This is the ideal solution. It keeps all of the hole sets parallel to the side of the extrusion, regardless of which way the snake turns.

Joe

[adbuch]

trolly.jpg

[Realstevesmith2]

UPDATE with Pictures of the machine. I didn't have time to take a video of it running. I've only been in the shop for a few minutes a day.

Big thanks to everyone helping on this especially you David and Joe. Joe you have put a ton of effort in this and I hope I can return the favor some day.

Accuracy. The accuracy doesn't have to be super high for these beams so the minor rotation due to the bow in the beams is not a problem. It bothers me but not enough to make a rotating feature on the gantry unless it made sense with a re design. accuracy is 1/8" or so on these.

I don't have comm issues with the longer parallel cable. Approx 25'. I don't get noise errors.

FYI, i make hundreds of these beams at a time.

We are currently moving the holes on the screen to match the bow but this is dangerous from a quality standpoint.

This is the only table i have and it has morphed over time. I did design into this whole assembly a c frame table that can push into the large gap in the legs on the 1 end and give me a flat table to put plate on but we have not built the table yet.

I am going to do a cost study on converting to Mach 3 vs the double decker gantry. weight is a concern on the gantry but I may can cut down the weight with aluminum hollow quad rails from mcmaster carr and the floating slide can be aluminum too. I do various other parts like rectangle tube so i would have to build a lot of adjustability into it. I am familiar with Mach 3 and have used it a lot in the past on a Mill and on a cnc plasma. Ultimately that may be the best option for expandability in the future. Time, cost and future needs will play into it but I'm leaning towards mach 3 so I can escape the bonds and limitations of plasmacam.

My software options do not have the option to "Customize" so I can not use hot keys. would cost $1000 to upgrade to the drawing package. Not interested in doing that.

I have an immediate plan to make my guy's life a little easier for the rest of these beams due to a discovery I made today while looking for the customize option. There is a Diagnose screen that allows you to press a button and zero each encoder separately. I can aim the torch on the corner of the beam and do what I plan but I will mount a cheap laser pointer in line with the torch to make it quicker. The steps will be....
- Move the torch with the move function in the program to the center of the 4 holes.
- On the pendant, use the dial to move the X axis until the laser point is on the edge of the beam.
- in the program screen open "Diagnostics"
- press the "zero" button for the X encoder.
- Highlight the holes and cut them.
- Rinse and repeat
- we already reinitialize the machine at the start of each beam otherwise it seams to act weird. I don't have a detailed description on the weirdness.

I need to cut 200 more beams before I make any major changes.

[Joe Jones]

If the rotation of the holes is not critical, then look at this file. Rather than a sliding gantry, why not just give the TORCH the ability to slide (+/-) X?

The carriage would always move to the same position, to cut holes in a straight array, but the torch would be held back by the edge of the beam, and then locked via a solenoid. The "beam finder" would retract, and the four holes would be cut as normal, even though the torch is now in a different X position in relation to the carriage assembly.

Joe

GLIDE RAIL TORCH.pcm

GLIDE RAIL TORCH JPG.jpg

[Joe Jones]

So, if I come to visit and to see this magnificent machine, I will bring a 4x4 plasma table and a computer with me, so we can experiment a bit with making different brackets, mounts, etc. I can pull the controller box off of my 510 easy enough, and mount it to the 4x4, and bring along the shop PC tower to run it. No one runs my 510 when I am away from my shop anyway. Then we can brainstorm a solution to FULLY AUTOMATE this "Holes Along a Curved Beam" process

I am excited to see that you can actually put two or three beams on the rollers, side-by-side. this opens up a whole new process, cutting three beams at once, and centering on EACH beam as required. It may be faster to just do one beam at a time, and give the GoTorch the ability to auto-find the beam's centerline every time.

So when is a good time to head out?

Joe



.

[Joe Jones]

The beams in your photo are SHORT. What is the longest single beam you need to cut holes through, end-to-end?

Joe

[Joe Jones]

If the controller moves WITH the gantry, then it seems the only things inside of the cable track are

1. Controller AC POWER Cord
2. Parallel cable
3. Torch cable

What else?

Joe



.

[Joe Jones]

RESPECT!

I struggle to mate up TWO gear racks on a 510 table so the transition is smooth. You have connected ... FIVE SETS of 4x4 rails, and aligned them all? AND the gear racks are shorter than the rails, so you probably needed six, or even seven gear racks per side. Or did PlasmaCam bless you with 510 gear racks for the extra length?

I really like the idea of the controller moving with the gantry. It solves the extended main wiring harness issue! Then you only need a good parallel cable that can traverse the table from one end to the other, with the PC in the middle. I suppose one could argue that the controller could "run away" as a rapid travel dart to one end of the table, so you would need to keep the speeds under control.

I am still trying to figure out how to program my game controller that I purchased for the purpose of controlling the table while AWAY from the PC or the controller box. I haven't worked out those kinks YET.

Now you have answered my questions about expanding my 510 to a 5x20 table. I guess I now have ANOTHER project on the books.


Joe



.

[adbuch]

- Move the torch with the move function in the program to the center of the 4 holes.
- On the pendant, use the dial to move the X axis until the laser point is on the edge of the beam.
-

in the program screen open "Diagnostics"
- press the "zero" button for the X encoder.

- Highlight the holes and cut them.
- Rinse and repeat
- we already reinitialize the machine at the start of each beam otherwise it seams to act weird. I don't have a detailed description on the weirdness.

I need to cut 200 more beams before I make any major changes.

I believe that zeroing the encoder will set the current position to Machine Zero, and not Program Zero. I would recommend testing this to make sure so you don't crash your machine. I would recommend adding a short "fence" with rollers (maybe 3 or 4 feet long) for the beams to align with and some rollers with springs on the opposite side of the fence to push the beam up against the fence. You might also consider using ball rollers instead of directional rollers to facilitate movement of the beam towards the fence.

As long as the edge of the beam is always in the same location relative to Program Zero for your hole pattern, then there would be no need to re-zero anything. Just slide the beam to the next cut location and press the "Start" button on the Plasmacam controller panel.

Short of doing that, then your best method would be to do as you have been - use the Move command to relocate your hole pattern to align with the adjusted torch position using the Torch Snap.
David

[adbuch]

We are currently moving the holes on the screen to match the bow but this is dangerous from a quality standpoint.

Why is this dangerous? I assume you mean possible operator error where the operator fails to move the hole pattern correctly?
I assume all of the beams are of the same approximate width. If so, then again I would recommend using short fence adjacent to the cutting location with a means to force the beam over against the fence. I expect so sort of hydraulic or electro-mechanical actuators would be the most efficient method, preferably controlled via foot pedal or hand push button by the operator.

If you can add these several features to your existing table, then I see no need to switch to Mach3, etc. I would just try to keep it as simple as possible.

David

[adbuch]

I just now realized that you are apparently moving your gantry the entire length of the beam. I wasn't paying close enough attention, and somehow thought you were sliding your beams under the table at a more or less fixed location. So forget what I said about the fence and actuators, as implementing these features would be much more difficult with your current configuration. So in this case, I think it is either have the operator reposition the cut pattern on the screen before every cut, or switch to Mach3 or equivalent where you can conveniently reset your Program zero before every cut.
David

[Realstevesmith2]

Good Morning,

Joe,

I like your beam finder idea. This would be a way to speed up the current process quickly and cheaply but it would not allow for the full automation I desire of pushing a button and not touching it until the beam is cut since the operator has to flip a switch on and off for each beam movement and cut.

We don’t cut more than one beam at a time. There are two on the conveyor because they roll one in, cut the top set of holes, roll it out and roll the second beam in. while they are cutting the top set of holes on beam 2 they deburr the top set of holes on beam 1 and flip it over 180 degrees to cut the bottom. They do the same thing for the bottom.

I’m not sure when would be good for a visit. I will be out of town the majority of the middle/end of the month and I’m out on location every day right now finishing up a project (I have multiple jobs/businesses). My main job is Designing and building manufacturing lines from concept to completion to mass produce metal railroad products. The possibilities are endless when using other people’s money lol. I should be done by the beginning of September. I work on improvements at my metal fabrication business at night and on the weekends for now.

The beams vary by the project. On this project, the shortest beams were 15’6”. The longest are 20’6”. The longer ones have the issue typically. The next Job has 27’ beams.

Your assessment of what is in the cable track is correct. There is enough room for small air line.

Since the gear track is shorter than the rails, I had to order an extra gear track for each side. They were 4 or 5 feet. What I did to line them up was, I cut a 1” section of the gear track off one end. I tacked it to a thin piece of metal. We butted up the gear tracks, sat the 1” piece on top (all inside the slot using our fingers to hold things…it sucked) and the 1” piece spaces the teeth out across the gap. We then tightened the screws all the way down taking car to push the track up along the way. There are no bumps. The legs had oversized holes and we shimmed each leg set and laser leveled the rails. They are perfect… I installed stops on each side so the gantry can’t come off without permission lol.

It would be awesome to have a remote pendant/game controller to use. Great idea!

David,

My machine zero and my part zero are one in the same. Since the holes are 1.5” away from zero I don’t get any issues where I’m touching the home switches. I have 2 stops that we clamp the beams up against. The clamp is light pressure.

Yes the beam is stationary and the gantry moves above it.

I thought about a clamping apparatus bolted independently to the floor so it doesn’t mess with the rail alignment but haven’t progressed any further. I think it would still allow for some inconsistency and the beams are W6x20 to they are pretty stout. The beam size could change with the next order. I also cut 3/8” wall 6”x8” rectangle tube. Hydraulics would be ideal and I have a couple extra power packs.

When I say dangerous, I mean the guys could move the holes to the wrong location or accidently forget to undo a previous move for a fresh bar. It takes a lot of concentration. My desire is to further automate to reduce the opportunity for error or inconsistency.

I’m strongly leaning towards a mach 3 conversion for the future guys. I will test the rezero encoder idea when I go to the shop mid day.

Thanks for the help. I will post an update on the encoder zero idea this afternoon.

[Realstevesmith2]

wasn't able to make it to my shop today but will try again tomorrow.

[djreiswig]

You might want to look at Linux. It's the new, open source program a lot of people are going to. I think it's probably as flexible as Mach, or probably more flexible since it's open source. There are a lot of guys here that have built tables using it. If you are switching your hardware, you might as well check it out.

[bubbagsj]

It would be awesome to have a remote pendant/game controller to use. Great idea!

If you're on plasmacam's owners site. There are a couple of threads about using game controllers on these machines. Just do a search for "game controller"

I have this hooked to mine(ZX-2):

https://www.amazon.com/gp/aw/d/B003VAHY ... asin_title

[adbuch]

It would be awesome to have a remote pendant/game controller to use. Great idea!

If you're on plasmacam's owners site. There are a couple of threads about using game controllers on these machines. Just do a search for "game controller"

I have this hooked to mine(ZX-2):

https://www.amazon.com/gp/aw/d/B003VAHY ... asin_title

That's a great idea to help possibly simplify the operation of this Go-Torch setup. Thanks for sharing that.
David

logitech gamepad controller for Plasmacam.jpg

[adbuch]

Here a few excerpts from the Plasmacam Owner's Community Forum game controller search for those without access to that forum.
David

wireless game controller.jpg

wireless game controller 1.jpg

[cutnweld]

I think too that Linux CNC could quite easily be configured to do a X probe, zero reset after every X move. Someone like Rodw would know.

[adbuch]

wasn't able to make it to my shop today but will try again tomorrow.

Steve - did Joe ever make it to your shop, and if so was he able to help you? If not, did you reach another resolution to solve your problem?
David

[Joe Jones]

Good Morning,

Joe,

I like your beam finder idea. This would be a way to speed up the current process quickly and cheaply but it would not allow for the full automation I desire of pushing a button and not touching it until the beam is cut since the operator has to flip a switch on and off for each beam movement and cut.

??? Not as I drew it. You let the carriage move along the rails in a straight line, and the edge finder (a large cam follower?) would stop the X movement of the TORCH at the edge of the beam as it floats on a horizontal glide rail and the edge finder is spring loaded against the beam's edge, even if the X movement of the CARRIAGE continued for a bit. Once the carriage is over a four-hole cut pattern, both the gantry and carriage STOP moving. The Z movement of the backing plate could be used to trigger TWO microswitches, mounted a few millimeters below Z MAX UP. One switch would automatically locks the torch 's floating X movement with a solenoid, and the second switch would retract the edge finder. Now with the torch down above the beam at cut height, it will cut the four holes in the correct location as normal, even if the beam's edge is slightly +/- X from the last four holes.

When the fourth hole is cut, the the backing plate will begin to rise to rapid travel height. The two microswitches DROP the edge finder and unlock the torch floating X glide rail car. The edge finder will snap back to the edge of the beam and gantry will move farther along the beam to the next set of four holes, as the torch floats X to remain in alignment with the beam's edge, and the process would repeat.

1. Lock torch X movement
2.Raise Edge finder.
3. Cut four holes.
4. Raise backing plate which lowers edge finder and unlocks torch X glider to allow edge finder to adjust torch X to (n) distance from edge as carriage moves along beam.
5. Repeat the process, automatically.

This WOULD work, and the operator would NOT need to "babysit" the beam AT ALL.

We don’t cut more than one beam at a time. There are two on the conveyor because they roll one in, cut the top set of holes, roll it out and roll the second beam in. while they are cutting the top set of holes on beam 2 they deburr the top set of holes on beam 1 and flip it over 180 degrees to cut the bottom. They do the same thing for the bottom.

Perfect! So done correctly, this should be fairly simple to set up.

Joe