Thanks for sharing the tip on marking the coupler. I have had to tighten the screws on the red brackets holding the servo and lead screw on the left side a few times as they started to work themselves loose. Now I check them every week. Didn't think to check the couplers for any slippage. Thankfully, I don't often need to worry about precision holes but this was very informative in comparing the different variables and using the overlap feature. Your efforts are very helpful and greatly appreciated. I often go back through your videos and catch little details that I missed before or maybe didn't understand when I first started working with my table.
Very informative! I usually cut small holes slightly larger than needed to compensate for the ball screw backlash and easy assembly. I will try adjusting some of these settings on my next project. Thanks for making this video! 😊
Thx for the kind words glad the video was useful. Since that video I have purchase a hydraulic punch and now just do a pierce at the hole location which marks where the punch is made. Cheating but solved my round hole problem 😜
@@timblack33for small holes I bought a cheapy off Amazon. Name on it is “Amzcnc” bought an inline foot switch so I don’t have to pull trigger. Not sure how to add pic on these replies but I’ll try. For larger holes I have a Vevor CH-70 punch tool and Vevor model GYB-700A hydraulic pump to power it. No problem punching up to 25mm dia holes in 3/8” mild steel.
Thank you for visiting this. A rinse and repeat on some .250 or .375 plate would be nice, or post your Gcoode. A few obserbvations... Coupler slippage - you need to mark the Stepper motor shaft also. There are two points to watch - though if you are slipping, it is obvious on all cuts curved or otherwise. Your Y would be interesting to verify if it is stepper motor loss, which I doubt, or estep error which would be repeatable, which yours is, and see what the total is over 18" of travel. If it is backlash, since there are two Y motors, I would check left right and center...maybe 3 x .250" holes each for sample. Might be a clue to point to something. Your entry into the circle at the start with a sharp angle (30* or 90*), change your lead in radius to about .03 or so with the default lead in angle. This way it is not trying to make a hard turn and that little dwell is not there.
Good point about marking the stepper shaft as well as the lead screw as either end of the coupler could be loose. Based on Langmuir’s response that the circle “roundness” is worse for thicker material (due to bevel) I didn’t think repeating the test with thicker material would reveal anything additional. Another viewer requested the Gcode so it is been posted on my Google drive: drive.google.com/file/d/1ECI2y88yZjm6EFf2t7sIfFWKFoJ3aEhr/view?usp=drivesdk drive.google.com/file/d/1lrXKx3sLUxwVSX-3sR95IrZMvz7Jgvat/view?usp=drivesdk drive.google.com/file/d/1bPsZr39EZe9KCijZ4gzKT5zEa17DzzJd/view?usp=drivesdk drive.google.com/file/d/1jKZ2EDQbMWfU0QUr3REyLuFUSmxHlVu7/view?usp=drivesdk Another viewer had suggested the error might be due to stepper motor losing steps. After considering this my reply was “As for the cause being due to missed steps I’m not convinced that is the main cause. Many of the items I cut as wall art include hundreds of pierces and complex curves per part and I will cut multiples of that part out of a sheet in one program. If even one out of 10 of those cuts lost a step then by the end of the hundreds of cuts the locations of the cuts would be off by a large amount. The control program doesn’t know the step was lost so the true location is not the programmed location. Assume you lost a step each .25” circle and you had to cut a pattern of 300 circles. If 100 of those circle lost a step the actual torch location compared to the program location would differ by a significant amount. I’ve cut over 100 small complex shaped items out of a 4’x4’ sheet and they overlay each other with no noticeable differences in edges or shape locations. If I was losing steps consistently I just don’t think that would be the case.” I could be missing something in that assessment so am open to rebuttal! Might look into the lead-in radius…thx
@@CustomCreations3 I agree that it's highly unlikely that you're losing steps. Now if the e-steps for the motor are off, then at your home position would be 0,0 but the farther you got away from zero on the y-axis, which seems to be the one in question, would gradually increase so possibly at the end of a full 33-in travel you might be off almost a few hundreths, then back to nothing at the home position...though .003 every .250 is a lot. But worth considering.
I did a cursory search on Langmuir and the first machine I found showed it has step motors and I am afraid you are finding out the big disadvantage of step motors compared to servomotors, missed steps. There can be many causes the most common is that there is just too much friction in the axis. The lead screw in that picture showed acme screws instead of a ball screw, easier and cheaper to make but they have more friction and arent quite as accurate as an actual ground ballscrew. Both of those comprimises are cost saving decisions in the design, otherwise the machine would have cost 10x or so what you likely paid. No worries because the mechanicals of that motion system should be far beyond the reasonably expected variation of a plasma cut anyway. There is nothing you can do about the design of the machine, it is what it is. What you can do to reduce friction is to keep the rails and lead screws cleaned and well oiled. I would suggest throwing a tarp over the machine when not in use to keep those moving parts clean but that is probably not a good idea with a tank of water sitting underneath you could get condensation and rust. So just to keep the dust in your shop under control and clean the machine before and after using it. I would make sure there is not any water on your lead screws when your are done with the machine too. So if you dont have grit or dings on your lead screw or rails but you are still getting missed steps mmm now you are getting into a more complex problem. Check that your axes are still running square to each other, that everything is tight and solid including the levelling feet. Step motors are bad about having certain resonant frequencies, certain feed rates where they could jump back out of a commanded step position. At what speed you encounter those resonances is a product of the motor itself, the friction of the axis, the inertia or weight of the axis and the power amp feeding your axis drive, and even the moving structures. Complex stuff. The higher end stepper drivers will have a tuning routine that goes through a sequence of moves and it watches the current draw for hints of likely risk of missed steps and maps out some offsets at different speeds to avoid missed steps so that the drive can vary current at different speeds based on that offset map. Unfortunately when you get in to a sophisticated stepper driver like that it costs as much or more than a servo and you are way better off just going to a servo system at that point, and eliminate the whole missed steps problem altogether. Bottom line, keep it clean and oiled and if you continue to have missed steps you are just going to have to figure out under what circumstances, most likely certain speeds, you get missed steps, and dont do that. Listen closely and you might be able to hear sort of clunking noise when it does it. Oh and next time, get the machine that has servos. Stepper motors are a false economy for anything that needs to be reliably accurate because too often you get in to a situation where they are finnicky and it can be frustrating to figure out where the problem really is and solve it. If anything changes slightly you could later develop the problem again. The working height from the nozzle to the workpeice really needs to stay as consistent as possible, the whole point of having THC, but the flatter you can keep the workpeice, the better, and if you could ensure the workpeice is always a perfect plane you wouldnt need THC. I really dont like that the THC is ineffective in under a certain cut length, I would dig in to that to see if there is a way around that, maybe a precut move to establish active THC before the pierce? I'm coming at this from a laser background, same general principles though, and have no idea what custom control software you have there. I notice in your cut routine your head comes down and presses on the material prior to each peirce, mmm I would rather the head never touch the material and I can see how that could push your axis enough to move out of step. But that does not explain missed steps DURING a cut.
Thanks for such a detail discussion. You are correct in the conclusion that the designers choose economy over accuracy and thus achieved an excellent table that includes torch height sensing and control plus the benefit of a water table delivered to your door for under $3k. As for the cause being due to missed steps I’m not convinced that is the main cause. Many of the items I cut as wall art include hundreds of pierces and complex curves per part and I will cut multiples of that part out of a sheet in one program. If even one out of 10 of those cuts lost a step then by the end of the hundreds of cuts the locations of the cuts would be off by a large amount. The control program doesn’t know the step was lost so the true location is not the programmed location. Assume you lost a step each .25” circle and you had to cut a pattern of 300 circles. If 100 of those circle lost a step the actual torch location compared to the program location would differ by a significant amount. I’ve cut over 100 small complex shaped items out of a 4’x4’ sheet and they overlay each other with no noticeable differences in edges or shape locations. If I was losing steps consistently I just don’t think that would be the case.
Hello Dan, I have a original CrossFire table with the XL kit and Z axis up grade. I too have been wondering about how to get round holes in my plasma cuts. One thing you have not looked at is the variation in drag for the X and Y tubes and the amount of dirt buildup under the bearings on both axis over time. I suspect there is a greater build up on the X axis tube that increases the drag on the movement in that direction. I just disassembled my table and cleaned the surfaces of both the X & Y tubes. After 3 years of use in my hobby shop there was a lot of build up in the tracks the bearings roll in. I use 20 mule team borax laundry soda to keep the water table fluid at a ph of 8 - 9 preventing rusting of the slates, but it does add to the buildup on any and all surfaces that water splashes on. I to had some problems with the drive couplers loosening and you need to mark the lead screw as well as the motor shaft to be sure there is no slippage. I'm going to try the longer overlap on my machine to see how that works for me. Thanks for your work on this!
That a great point and anything that restricts free motion in x or y could exacerbate the roundness issue. I probably didn’t think of this since I’m pretty anal about keeping the rail reasonably clean. I use a water soluble cutting fluid mix in my water table that does a really good job a rust prevention but the splashing of the table fluid does leave a film that collects vaporized metal particles and I have to clean it off regularly. Fortunately it cleans easy with just water. So because the roundness issue seems to be the same when the rails and bearing blocks are clean or dirty I didn’t think it was a major contributor. Still I should have mentioned it as I’m sure other users may not be as anal about cleaning the rails and I can see how a significant buildup could definitely increase drag on the xy motion and make the hole roundness worse. Thanks for your response and excellent insight. Hopefully others will read this thread and consider it for their application.
I clean them after every use and often use oil and flat stone the gantry aluminum tubing. The roller couplers often drag and create grooves because the way they are designed metal bearings on metal and you obviously want them tight. A linear rail probably would have a been better design by langmuir. The couplers they provide are completely garbage and prone to slipping out of the gate the set screws are way too small and the relief tangs aren't enough. I ordered and installed some better ones. I can get down to about a .200 hole accurately anything smaller I have to fiddle with the lead ins and outs. I have a separate tool I create for holes that usually runs at about 50% less feed.
I like the idea of using oil and flat stone on the gantry. I agree a linear rail would be much better…not sure how the water and cutting fluid mix might affect a linear bearing but would be far more accurate. I have to keep reminding myself that Langmuir consciously chose to go with a low cost design to appeal to the masses as opposed to a more accurate (and thus costly) otherwise I get frustrated with some nagging issues.
@@CustomCreations3 Yeah I've had to do quite a bit of mods on my machine although I figure with any kind of machine kit today that's going to be the case. Essentially it has to be looked at as a project in itself and for the price it's an amazing base machine. I've done a few improvements built a nice shelf underneath isolated the controller a bit better so my laptop doesn't freak out from emi (tried the ferrite choke) did some cable management inside the frame added a splash guard aluminum slats ect. I was looking to replace the lead screw nuts with more high quality to help with accuracy (because there's a cheap plastic guide inside) but the sizing is odd so means I'd have to replace the entire lead screw assembly. Another good thing for people to know is to shut off the controller when you're not cutting as I don't think the controller is designed to cut power completely to the steppers. Stepper motors become worn out when they idle (basically they want to move) and a signal keeps them from moving all the time. Idling constantly can reduce the accuracy over time as they become worn out. Same principle for 3d printers and such.
For circles and other “inside” cuts the controller defaults to left compensation meaning it will move the jet position 1/2 of the kerf diameter to the left I.e. inside the cut path. So for a kerf of .050” the jet will be position .025” to the left of the path. Reducing the kerf setting will increase the circle diameter. A kerf setting of zero is the same as choosing “no compensation” which moves the jet directly centered on the path geometry. Hope that helps. Thx for watching
I was looking into this problem as to why my holes were not round in my langmuir cnc plasma table and I stumble across the z axis where the torch is mounted , it had alot of slack sideway and that affected the cut quality by alot, try making sure the Z axis has NO play so you get accurate holes.
Scott, I've uploaded the files to google drive and given anyone with the link access. There are 4 g-code files that should run on your Langmuir. The hole test sheet, a 2.5" hole with 1/4 overlap, 1/2 overlap and full overlap drive.google.com/file/d/1lrXKx3sLUxwVSX-3sR95IrZMvz7Jgvat/view?usp=sharing drive.google.com/file/d/1bPsZr39EZe9KCijZ4gzKT5zEa17DzzJd/view?usp=sharing drive.google.com/file/d/1jKZ2EDQbMWfU0QUr3REyLuFUSmxHlVu7/view?usp=sharing drive.google.com/file/d/1ECI2y88yZjm6EFf2t7sIfFWKFoJ3aEhr/view?usp=sharing let me know if you have any questions thx Dan
Using these cheap little lasers (note: you might need to mess with the lens on the front to get it focused but once it is done haven’t had any problems. www.amazon.com/gp/product/B072XF49LK/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1 Holder is: www.amazon.com/gp/product/B07F1T97V4/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1 The nozzle “shield” is a collapsible silicone funnel that the center is cut out to slip over the torch nozzle. www.amazon.com/gp/product/B089LR8VD2/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1
Sarah, are you a member of the Langmuir forum? If so there is a signup sheet for the die set. Here is the forum link: forum.langmuirsystems.com/t/pan-drain-dimple-die-mod-2/13000/120 Alternatively, the drawings are available on the forum that allows you to either turn a set for yourself or 3D print a set. I understand from the forum discussions that a 3D printed set using backup washers are good for a couple of dimples!
I don't see a finishing overlap in fusion 360 for plasma cutting. I did see it for milling........ Can you tell me where to find overlap for cutting? Thanks!
I have it in the passes tab labeled “finishing overlap”. I can’t figure out how to send a pic but you can email me at ccllc@drbmetal.com and I’ll send it to you
