CNC Plasma Table Build

Dude, huge difference hell yeah!

 

2"x2" square with two 1/2" radius corners an a 3/4" circle in the center.

16g mild steel
Pierce height 0.15
Pierce delay 0.5s
Cut Height 0.06
Feed Rate 110ipm

one side is really close..




the other side is off, by more



the circle is in sad shape looks more like a lemon. Jim Colt of hypertherm has a post over on plasmaspider talking about cutting good circles, so following his tips I will, go through and check an make sure everything is square, and make sure the motor couplers are tight. I suspect I will find issues with both.

I will also try to slow down the circle cut to 66ipm (60% of cut speed) an add an overburn of .150 to .200

 

The oblong shape is definitely telling you something is out of square.

As for cut speed, I don't suspect you'll need to adjust it at all until you get to below .4" diameter.

Think I posted this in the other thread but the outside cut, and holes larger than .4", utilize a cut speed of 100ipm.

Below, are the different speeds used to practice a .25" diameter.



60% does seem to follow close to the rule you mentioned, but tuning it up just a tad faster can ensure you get the tiniest bit of dross possible.



Looking forward to your results!

 

Also, this might be an indicator as to what is either loose, or binding up even?

Let's assume the bottom part in the picture you attached is your lead-in and lead-out which means lateral movement.

The same lateral movement that occurs while cutting the circle seems to be at play here.

 

Pierce height seems high to me.

I haven't played with my table for some time, so I'm just going by memory. I'm more than likely wrong.

 

.15 pierce is default, and what I use for everything.



Although I think there's a discrepancy with one of his axis movements.

Here's the test piece with the 4 holes.





Ignore the two hundredths difference from the 2.5" drawn, I have it set that way on purpose for my dino kits I cut and sell.

 

I checked the set screws on the motor couplers, they did move a bit. so they had come loose on me.

when I get time I will verify the motor tuning on the X axis. If I recall I just used the same data as the Y axis

I am starting to wonder if the steps per count is off for one of the axis..

 

That was going to be my next question.

You know way more than me about this DIY CNC stuff, but I've watched other videos of people building their own tables and a few have had to go back and forth ensuring that motors were all moving in the exact same increment value IE having two X axis motors and one is slightly faster than the other.

 

If 2 motors are used for 1 axis, 1 needs to be slaved to the 1st motor. No adjusting the second motor is needed.

 

Was just an example!

I can't recall the wiring setup and all used in pyro's table, you guys are the DIY masters here!

 

looks like the X axis motor tuning was off. measured the actual line drawn short by 1/16. the Y axis steps per count still appear to match what I originally figured them to be.

I ran the steps per unit wizard inside Mach4 and retested with the new motor tuning value, and the line came out nearly exact, least to the limits of measuring a sharpie line with a steel rule and my old eyeballs. I can always set up a dial indicator and do several tests to verify.

@Kwikvette Do you think this small amount of loss in the X axis would lead to wonky cuts you saw?

 

That's EXACTLY what I assumed.

You have specific sides of any geometry coming in perfect, and as soon as lateral movement comes into play, that's where your discrepancy shows IE the red arrows I pointed out in your test cut, and the spots in your bottle opener.

Glad you found the cause.

 

thanks..

I been out in the heat all day, and I decided to stop at this point, instead of rerunning the test square, which was off because of the X steps. but, There is one more thing I need to add after the gantry plates, an that is a spacer plate so the home switches will detect the new gantry location. this will allow me to square the gantry in relation to the Y axis. I am hoping between the issues I found, it solves the cutting problem

 

I think, I have the problem nearly solved.

I used sheetcam to generate the g-code, and right away there was a slight improvement, Sheetcam is my controllers preferred CAM package as they have several different post processors for it. but it was still off.

I put a dial indicator and check to verify the steps where right. they where nearly dead on in both in X an Y Axis, with a bit of backlash that will need to be adjusted out.









During this time I Double check the setscrew on the motor couplers, and while doing this I notice a Squeak in the -X direction. So I cleaned everything, and lube the ball screw again. Still a slight squeak, So I cracked the bolts loose holding the ball screw bearing block in place, and ran the machine all the way down to the end, and retighten, and repeated with the motor end. Soon as the motor end was loosen, the squeak was gone, and you could feel it running smoother.

After running the X axis back an forth a couple times, I cut the test piece again.

 

best cut so far on the table.



I wanted to see how bad the pierces look after cleaning up the part.



Room for improvement, but much better than previous attempts. Overall, I am pleased with the improvements.

This was a previous Fusion360 generate G-code. Now I know that my workflow was not to far off. With a bit of tweaking I should be getting some really nice cuts.

 

Is that 16 gauge, 11 gauge, or 12 gauge?

Regardless, looks way better.

 

16g

 

Makes sense, yeah absolutely looks good!

Was gonna say if 11 or 12 gauge you should have to do no clean up at all

I'd test it since it's running much better

 

It's a pretty amazing feeling when it starts doing what you want, isn't it?

I respect those that bought turn key packages, but me, with my limited computer background, I needed to build it from the ground up so I know what makes it tick.

 

very much so...

I have always been the type who will build it over buying it. It has been a challenge in many ways, would it be cheaper to buy? probably... depends on how you value time, over knowledge gained. but a machine like mine same size and style of construction runs about $10k.

For me the building of the machine is sometimes more fun then the finished product. If my mistakes and lessons learned help the next person. then it was worth all the frustration, and hours spent trying to figure out why something wasn't working the way I expected not to mention the expense of having to do something twice.

At this stage it really is time to stop putting off building the water table, and slat tray.