New and having trouble

[catlingm]

I recently finished up building a 4x4 CNC plasma table. I think it's a pretty generic build, I'm running Mach3, MiniTHC, and a Powermax 65 cutter. I have decided to use Fusion 360 to make my drawings/models to take advantage of the CAD/CAM feature as well as future use on a Bridgeport Boss 5 Series II CNC milling machine that I hope to start retrofitting electronics on in the next few weeks.

As the title states, I'm new to CNC work and I'm having trouble. Searches to resolve the problems do not yield any really useful information. It may have a lot to do with the words that I'm using to search for answers, but I'm not sure of any other words to use to properly describe the issues.

I have created a basic shape that I would like to cut out, but the THC keeps driving the torch into the material. Which is really annoying because it throws things out of whack and I have to fix things before I can try again... Which is followed by the same problem...

I have the THC set to the voltage that is in the Powermax 65's manual for the material and thickness that I am trying to cut. So, not sure if I need to adjust that.

The delay in the g-code was initially set to 0.2, I then changed that to 0.7 thinking that it was seconds... I saw that other places have it set at 800, so I'm now not sure if the G4 P is in seconds or milliseconds. I suppose that the THC might not behave properly if the material is not pierced, so this could be contributing to the problem.

Also, I have Mach3 set up for inches and I see in the post processor code, that it's set for millimeters. Every Mach3 post processor that I have found for Fusion 360 is in millimeters, will this cause a problem when I'm trying to use g-code generated by a post processor in millimeters, cutting on a table that is setup in inches?

I figure that, once things are running well and I have a better understanding of what is going on, this should be pretty fun and very helpful in my hobby world of doing dumb stuff to cars and trucks, but right now it's a pretty frustrating process...

I do appreciate any help that can be offered.

Thanks,
Greg

Here's a pic of my table...

20180429_145209_small.jpg

[weldguy]

Be sure you can cut the part properly with THC off first. This will confirm your pierce delay and speed is working well before introducing THC into the mix.

Experiment with your THC doing simple straight line cuts across the width of your table to minimize waste until you get it working properly using the speed and pierce delay you already determined is working for the material your cutting.

Set your THC voltage much higher than the manual calls for if your diving into the plate. Better to be too high than diving in and crashing things. Slowly reduce your THC set point voltage until you find the voltage that provides the proper cut height. Typically cut height is .060" off the plate.

Once you determine the proper voltage document all your settings for future use on this material type and thickness. Hope this helps.

[beefy]

Have you got the THC DELAY enabled.

See my post in this thread to understand the THC DELAY:
http://www.plasmaspider.com/viewtopic.php?t=21222

[catlingm]

Ok, so it appears that it was the THC delay. Denis, from MiniTHC, got me set up really quickly and now I have made a number of test cuts successfully... Well, to some extent. I made a bunch of circles that cut pretty nice. Then I tried a number of cuts that I'm hoping to get running better to use as a key chain. The profile that I hope to use as a key chain cuts very rough. I suspect that it's the way the gcode is written, but I'm not totally sure. I'm running Fusion 360 to draw and post process and Mach3 to cut. Here are a couple pictures to show what I'm talking about.

20180512_145129_small.jpg

20180512_145156_small.jpg

Circle gcode:
N10 G70
N15 G90

(2D Profile1)
N20 M6 T1
N25 G54
N30 M9
N40 G0 X-0.019 Y1.043
N50 G31 Z-100
N55 G92 Z -0.2245
N60 G0 Z 0.15
N65 M3
N70 G4 P 500
N75 G0 Z 0.06
N80 G1 X0.151 Y1.142 F175.
N85 G3 X0.214 Y1.25 I-0.063 J0.108
N90 G2 X2.286 I1.036 J0.
N95 X0.214 I-1.036 J0.
N100 G3 X0.151 Y1.358 I-0.125 J0.
N105 G1 X-0.019 Y1.457
N110 M5
N115 G0 Z 0.5

N125 M30


Part of the key chain gcode:
N10 G70
N15 G90

(2D Profile1)
N20 M6 T1
N25 G54
N30 M9
N40 G0 X1.7371 Y0.9954
N50 G31 Z-100
N55 G92 Z -0.2255
N60 G0 Z 0.15
N65 M3
N70 G4 P 1000
N75 G0 Z 0.06
N80 G1 X1.6683 Y1.0228 F80.
N85 X1.6601 Y1.0163
N90 X1.6517 Y1.0093
N95 X1.6429 Y1.0017
N100 X1.6336 Y0.9933
N105 X1.6237 Y0.9841
N110 X1.6133 Y0.974
N115 X1.6021 Y0.963
N120 X1.5902 Y0.951
N125 X1.5775 Y0.9379
N130 X1.5638 Y0.9236
N135 X1.5492 Y0.9082
N140 X1.5335 Y0.8915
N145 X1.5167 Y0.8734
N150 X1.4987 Y0.854
N155 X1.4794 Y0.8331
N160 X1.4586 Y0.8104
N165 X1.4384 Y0.7886
N170 X1.419 Y0.7675
N175 X1.4003 Y0.7473
N180 X1.3823 Y0.7279
N185 Y0.7278
N190 X1.3651 Y0.7092
N195 X1.3486 Y0.6915
N200 Y0.6914
N205 X1.3328 Y0.6745
N210 X1.3178 Y0.6584
N215 Y0.6583
N220 X1.3035 Y0.6431
N225 Y0.643
N230 X1.29 Y0.6286
N235 X1.2772 Y0.615
N240 Y0.6149
N245 X1.2652 Y0.6022
N250 X1.254 Y0.5903
N255 X1.2539
N260 X1.2435 Y0.5793
N265 X1.2434 Y0.5792
N270 X1.2338 Y0.5691
N275 X1.2337 Y0.569
N280 X1.2249 Y0.5597
N285 X1.2248 Y0.5596
N290 X1.2167 Y0.5513
N295 X1.2166 Y0.5512
N300 X1.2093 Y0.5437
N305 X1.2092 Y0.5435
N310 X1.2027 Y0.5369
N315 X1.2025 Y0.5367
N320 X1.1968 Y0.531
N325 X1.1966 Y0.5308
N330 X1.1917 Y0.526
N335 G2 X1.1914 Y0.5257 I-0.026 J0.0263
N340 G1 X1.1873 Y0.5217
N345 G2 X1.1868 Y0.5212 I-0.0257 J0.0266
N350 G1 X1.1835 Y0.5182
N355 G2 X1.1823 Y0.5171 I-0.0251 J0.0272
N360 G1 X1.1675 Y0.5046
N365 G2 X1.1658 Y0.5032 I-0.0239 J0.0283
N370 G1 X1.1501 Y0.4916
N375 G2 X1.1484 Y0.4903 I-0.0221 J0.0296
N380 G1 X1.132 Y0.4795
N385 G2 X1.1303 Y0.4784 I-0.0204 J0.0309
N390 G1 X1.1132 Y0.4685
N395 G2 X1.1114 Y0.4675 I-0.0186 J0.032
N400 G1 X1.0937 Y0.4584
N405 G2 X1.092 Y0.4576 I-0.0169 J0.0329
N410 G1 X1.0737 Y0.4494
N415 G2 X1.072 Y0.4486 I-0.0152 J0.0337
N420 G1 X1.0531 Y0.4413
N425 G2 X1.0514 Y0.4407 I-0.0135 J0.0345
N430 G1 X1.0321 Y0.4342
N435 G2 X1.0304 Y0.4337 I-0.0118 J0.0351
N440 G1 X1.0107 Y0.4281
N445 G2 X1.009 Y0.4276 I-0.0101 J0.0356
N450 G1 X0.9889 Y0.4229
N455 G2 X0.9873 Y0.4226 I-0.0084 J0.036
N460 G1 X0.9669 Y0.4188
N465 G2 X0.9652 Y0.4185 I-0.0068 J0.0364
N470 G1 X0.9446 Y0.4156
N475 G2 X0.943 Y0.4154 I-0.0051 J0.0366
N480 G1 X0.9222 Y0.4134
N485 G2 X0.9205 Y0.4133 I-0.0035 J0.0368
N490 G1 X0.8996 Y0.4122
N495 G2 X0.8979 I-0.0019 J0.037
N500 G1 X0.8769 Y0.4121
N505 G2 X0.8753 I-0.0002 J0.037
N510 G1 X0.8543 Y0.4129
N515 G2 X0.8526 Y0.413 I0.0014 J0.037

Is there anything in the gcode to point to for the rough cutting of the key chain?

Thank,
Greg

[beefy]

I know what that is, and Dennis will too. Contact him again and he should be able to explain how to adjust the parameters to get rid of it.

The file runs fine in my system.

[catlingm]

I'll shoot him a message. Thanks.

[plasmartist]

Not sure if others mentioned this already, but just because you have 65 amps to work with doesnt mean the 65a nozzle is the right way o go every time. That was my first mistake when i got into cnc with a pmx85. I now find myself using 45a for a ton of stuff. Also Ill share a little secret for you voltage tests that I havent seen anyone else do yet. I draw a tight scroll that takes up about 3 square inches but has a total cut distance of 36 inches. The scroll is a vector so it's a continuous smooth line. Since it is so small, deviations in material height dues to dross are minimal and thus the arc voltage is much easier to get an accurate reading. Furthermore it gives you the ability to see how the cut is performing in all directions, which can tell you if your nozzle is going bad, if you torch angle is off etc. This trick has saved me thousands in material and headache and you're the first I am sharing with.

[beefy]

Also Ill share a little secret for you voltage tests that I havent seen anyone else do yet. I draw a tight scroll that takes up about 3 square inches but has a total cut distance of 36 inches. The scroll is a vector so it's a continuous smooth line. Since it is so small, deviations in material height dues to dross are minimal and thus the arc voltage is much easier to get an accurate reading. Furthermore it gives you the ability to see how the cut is performing in all directions, which can tell you if your nozzle is going bad, if you torch angle is off etc. This trick has saved me thousands in material and headache and you're the first I am sharing with.

Great tip, thanks very much.

I'm actually banging my head off a wall after reading that. Seems such an obvious thing to do once someone spells it out. I've always done straight line test cuts, taking the time to ensure the plate is flat, so that the readings are accurate. Your scroll idea is just great. And many guys have experienced the lean in a cut in only a specific direction.

I would hazard a guess that it's best to start from the inside out, so that the outer bulk of the plate does not allow any expansion movement of the scroll ??????

[plasmartist]

spiral.PNG

Sheetcam always starts mine in the center. You can draw the scroll easily in inkscape, then define the number of turns in it. Then convert it from a spiral to a path ctrl+shift+C, then at the outside node I draw a line to connect into itself so you can easily lift the cut out and analyze it.

[beefy]

Sheetcam always starts mine in the center. You can draw the scroll easily in inkscape, then define the number of turns in it. Then convert it from a spiral to a path ctrl+shift+C, then at the outside node I draw a line to connect into itself so you can easily lift the cut out and analyze it.

I'm a Coreldraw user. Idiots like me pay for things

Sheetcam is not actually starting your cut in the centre. Well, to explain better, you can control which end Sheetcam starts at by 2 methods (that I know of). From what I've observed, Sheetcam starts the cut in the same direction as the curve was in the cad program.

1). Change the direction of the curve in Inkscape then export. When you view a curve you should see a starting and ending arrow. Those arrows swap ends when you reverse the curve direction. Then when you export, and subsequently import into Sheetcam, the cut direction should also have changed.

2). Just use the "Edit Start Points" feature in Sheetcam to make it start at the other end.

I'm still friggin kicking myself for not seeing your trick ages ago. Just played with it in Coreldraw. I make the outline the same width as the plasma kerf width so I get a "real" view of what the cut will be like. I did a 9 turn scroll with a width/height of 92mm and the Sheetcam job report said the cut length was a whopping 1.4 meters long. And based on the kerf width I could have closed the gap between the spirals.

Thanks again for that tip.

[plasmartist]

Thats interesting to know about sheetcam. I guess I was just lucky that its defaults worked out in a way that was beneficial. The biggest money saver for me is that I have learned from my test cuts when my nozzle is close to failure since there will be spots in the cut that dont cut as well. ie the nozzle is out of round/worn (not much of a problem at 45 amps, but on 85 amps my machine eats nozzles for breakfast)

[jimcolt]

When you get the height issues dialed in that allow for exactly correct pierce height, pierce delay and physical cut height (controlled by arc voltage) then you will get excellent consumable life. Incorrect piercing with an 85 amp nozzle will cause shorter life than with a 45 amp nozzle .....more power to melt itself! Height control is the most critical component when it comes to cut quality and consumable life. Jim Colt Hypertherm

Thats interesting to know about sheetcam. I guess I was just lucky that its defaults worked out in a way that was beneficial. The biggest money saver for me is that I have learned from my test cuts when my nozzle is close to failure since there will be spots in the cut that dont cut as well. ie the nozzle is out of round/worn (not much of a problem at 45 amps, but on 85 amps my machine eats nozzles for breakfast)

[plasmartist]

Jim I have mine dialed in with a feeler gauge at .15”. If that’s not the correct pierce height should I go up or down?

[jimcolt]

Pierce height is listed in the Hypertherm cut charts, it varies with material type, thickness and amperage level. Jim Colt Hypertherm

Jim I have mine dialed in with a feeler gauge at .15”. If that’s not the correct pierce height should I go up or down?