[epichutch]

Hi all,

i have recently purchased a 4x4 plasma table and it has been nothing but a head ache from day 1!

i purchased the machine used for $6000. seemed to be in decent shape and came with a good plasma cutter. i have made a few small upgrades to it to make sure it was fine tune. (torch mount, a few small mechanical components)

the machine is a 4x4 percision cutting systems table (which happens to be a local company but there is no life on the other end of the phone line). it runs a hypertherm 1250. the computer came with some old software - viaCAD 2d , lazycam , and mach3. all older versions.

the issues at first were some serious bevel to the cut, even on a 1/8 material. and the accuracy of the cut. a few mechanical corrections were made and also i started downloading the newest version of the mach3, and combined it with sheetcam. after some major complications were sorted out, i was able to get the machine working again and cutting better. although there are still a few issues with the display screen on mach and i think it is related to the back up .xml files. since the company is uncomplient i am not able to get the peramiters to calibrate the machine, so i am stuck with the back ups until i have to learn how to reverse engineer the calibrations.

an issue that persist is the smoothness of the line (choppy, but not cut quality) and the ability to cut a small circle accurately. i see pictures and videos of guys cutting half inch holes exceptionaly well. mine is just unacceptable.

i am getting to my wicks end! so i am turning to you guys for a little insight.. a few thoughts i had and a few things i heard - i wonder if a lot of it is related to the CAD software? i also heard that it is not a good idea to run the CAD software on the same computer as the cutting software?? (currently a pentium4 with 4gigs) i am confident the stepper motors are in good working condition. if i need a new CAD program, any reconmedations? (i am not the most computer savy but i can figure stuff out) trying to figure out draftsight right now.

the purpose of the machine is to build accurate components, if i have to dress all the pieces and drill the holes by hand i might as well go back to my angle grinder.

again just looking for some insight before i turn to the tank of gas and a match.

thanks, Sam.

[alan_3301]

with the stepper motors powered up, grab the torch holder and try to move it with your hand, in all directions.
This will show you how much backlash and flex you have.

Now disconnect the drive mechanism from each axis, and move it through the range of motion by hand, feeling for roughness or binding.
If it is excessive, that will have to be fixed.

Also post up some pictures so we can see what you are working with.

Mach3 will accept gcodes in the MDI option. with this you can input gcodes directly to control the machine, this bypasses cad and cam software to eliminate that as a cause.
If you are not famaliar with gcode, look it up and become familiar with it, it will help when you have problems.

[epichutch]

i did those checks and everthing seems to be good.

here are a few pics to give a better idea of the issues. all cuts are made with a 3/8 arc lead in. none of these pics are of the starting points

Attachment 253724the bottom two are the same file, this is why I'm questioning the CAD
this is a half inch hole,
Attachment 253722this may be picky but this one is 1.5" I would like to think it could be almost perfect at that size

[jimcolt]

Smooth edges and round holes are a function of all of the components of a cnc plasma system. poor speed control, rough motion, inaccurate height control and and older technology....or improperly set-up plasma cutter will produce cuts just as your pictures show.

We would be happy to help.... but need information. What plasma cutter make and model? Does you machine have height control? What make and model? Pictures of the drive mechanisms on your x and y axis...more detail about the mechanical condition of the table.

Jim Colt Hypertherm

These are holes from my cnc plasma. Plasmacam 4 x 4 table with Hypertherm Powermax85 plasma cutter

Attachment 253758

[epichutch]

the general condition of the table seems pretty good. the motors seem to be tight, and there is no play in the main arm or the guides.

as for the height control, it has the the components but like someone diconnected the wires some side cutters, also it is missing the micro switch.

the plasma cutter is a Hypertherm Powermax 1250, G3 series.
the peramiters for the cuts shown above were:
- 3/16 material,
- 75 psi,
- 60 amp consumable,
- 50 imp.
- stand off height at 0.1875"
- Kerf - 0.07
Attachment 253764
Attachment 253766

[jimcolt]

You need to open up the operators manual for the Powermax1250. There are multiple pages of cut charts. I recommend using the 40 amp shielded process....follow the cut specs exactly from the cut charts. You are using too much power (60 amps), your cut speed is too slow, and your standoff height is far too high. These alone will create terrible cut quality! If you do not have the operators manual...you can download it at Choose your region and language | Hypertherm , click on downloads library, choose product (Powermax1250), then choose manuals, then operators manual.

Jim Colt

[LeeWay]

I thought it was pretty high on amps and cut height. I wasn't familiar with that cutter though. When all else fails, consult the manual. You can put three exclamation points behind that statement when it involves a plasma cutter. better yet, ask Jim.

[jimcolt]

Here is the cut chart along with the correct consumable part numbers to use....look at the 3/16" material thickness. You will see the cut speed, the cut height, amperage, etc. Cut height is 1/16". If you get your torch height control operating then your cut quality and consumable life will improve even more.

Jim Colt Hypertherm

[epichutch]

the parameters i used were from messing around a bit to try to optimize the cut.

i switched to the 40 amp consumable, and changed to suggested settings as per the manual.
3/16
40 amp
63 imp (optimum travel speed) (max travel 97)
1/16 torch-to-work

these are the results. the cut barely cut al the way through, the bevel was worse. and the accuracy of the hole is no different

Attachment 253804 1/2"
Attachment 253806 3/4"
Attachment 253808 1 1/2"

[LeeWay]

I can't run a PM45 with only 75 PSI. I need at least 100 PSI and a little more is good. I don't know the exact PSI inside the machine. I feed about 120 PSI to the unit.

[jimcolt]

Next step.....can you mount a pen in place of the torch and do a tracing on paper, lets see if the machine motion is very bad...

Jim

I can't run a PM45 with only 75 PSI. I need at least 100 PSI and a little more is good. I don't know the exact PSI inside the machine. I feed about 120 PSI to the unit.

[epichutch]

okay, here are the results. (the marker was mounted pretty rigid, no flex in the pen)

without the torch runnning i can watch it much closer. it appears that there is a stumble at the top of each circle, this would lead to a inaccurate file? leaning towards the CAD. also a thought - i was once told it is not recommended to run the CAD program on the same computer as the cutter ?...

what do you think of leeWay's comment about the air pressure?

[LeeWay]

My manual mentions 80 to 100 PSI required and a max of 135 PSI. 110 to 120 is good and will hold the machines settings through cycle times of the compressor.
Yours apparently does not require as much. 70 to 75 PSI.
Here is a link to your manual.
https://www.hypertherm.com/Xnet/libr...464&format=pdf

[jimcolt]

Unlikely that it is the CAD....more likely it is backlash and or drive tuning.

-Could be the drives are set too hot...and are overshooting
-Could be that the drives are too small for the mass of the moving parts

You can try another CAD program....but I'm willing to bet that is not the issue. Your machine should be tracking within .003" to .006"....if it doesn't you will have ugly cuts. Further.....if the speed is not being maintained (motors out of tune...or undersized) expect a very wide plasma kerf and dross when going too slow....such as on holes and cornering.

How about drawing one circle with a lead in and lead out......and one rectangle with the same.....trace those. That will help to show if we are seeing the lead ins on these multiple cuts....or if it is actually overshoot.

Jim Colt

okay, here are the results. (the marker was mounted pretty rigid, no flex in the pen)

without the torch runnning i can watch it much closer. it appears that there is a stumble at the top of each circle, this would lead to a inaccurate file? leaning towards the CAD. also a thought - i was once told it is not recommended to run the CAD program on the same computer as the cutter ?...

what do you think of leeWay's comment about the air pressure?

[epichutch]

first of all i just want to say thanks for all your help! my frustration is high, but a little more manageable with a little help in my back corner.

here are a few more pics. the small circle is 0.5", larger is 1.5" and the rectangle is about 3" x 1.25"
i do think the kerf if pretty big... a friends machine cuts 5/8 plate with half the cut size...

Attachment 253834
Attachment 253836

[scorpion]

first of all i just want to say thanks for all your help! my frustration is high, but a little more manageable with a little help in my back corner.

here are a few more pics. the small circle is 0.5", larger is 1.5" and the rectangle is about 3" x 1.25"
i do think the kerf if pretty big... a friends machine cuts 5/8 plate with half the cut size...

Attachment 253834
Attachment 253836

I've had a couple of different kinds of plasma tables over the years and I've noticed that many times a circle won't be round because of the pinion gear in the rack and it's mostly due to wear or foreign material being in between. Some software let you account for slop by setting backlash if that's the problem. In some cases you need to tighten the pinion to engage deeper into the rack.

What I used to do was pay very close attention to how the pinion meshes with the rack as I'm cutting a circle. Almost always an imperfection like that will be extremely obvious during observation. In the case of the pictures you shown the circles look to be consistently deformed in the same place.

So, what I would do next is program a bunch of circles and observe the pinion on the long and short drive gears. Also inspect the Pinion and rack to ensure they're not excessively worn and you have proper engagement.

[epichutch]

given the advise, i will rule out the CAD software. i just thought it was odd that the stumble is consistent no matter how big the hole or where on the table it was working. (my 2 cents)

i ran a test of about 40 circles at different sizes and locations on the table, i closely watched the pinion and gear on the x and y axis. not sure what you mean by long and short gear?? my table runs one gear for each axis... anyway, the x - axis is tight with about 0.003" play,
the y - axis seems to be in the 0.025" range, the play is not related to the mesh of the gears but seems to be loose in the stepper motor. suggest replacement?

at this time i will disassemble all the drive mechanisms and clean thoroughly. i am considering putting some graphite on as a lubricant during reassmbly, thoughts on this??

[Mike 1948]

As for accuracy some of this looks like Mach settings. All the holes look small. Maybe the number of steps per inch is FUBAR. Tell the machine to move 1 inch and measure with a calipers if possible. If it only moves .9 then you know you have to change the number of steps by approx. 11%. Check both the X and Y axis. The steps per inch are found in the motor tuning areas in mach.
Also in the settings area make sure your are using " constant velocity" not 'exact stop". Exact stop cause mach3 to pause ever so slightly at each node in the line.
There is also a setting for "line look ahead" set this to about 200. This helps mach3 to kind of anticipate its movements.
Can't help any with the torch but Jim can.

Mike

[epichutch]

the pics with the holes and the green tracer, i used a slighty bigger reference hole to better display the issue. i was cutting 2x2 squares earlier and was able to get it within a few thou, mostly by adjusteing the kerf.

as for the settings in Mach, i found that it was already set to constant velocity, and the "look ahead" was set to 20, which i changed to the suggested 200. this did not improve the accuracy.

[epichutch]

thought i would shed a little light on to where i am headed now,

while observing many many traces. and working through check. while tracing the circles i found that the stumble appears to be when the y - axis changes direction. leading me do believe there is excessive backlash in the motor. i am currently hunting for a local supplier of a replacement motor..

since i'm buying parts i will be getting the THC working as well.