Head moving but Torch not engaging after first cut

[Gavopian]

Hello everyone!

Sorry that this my first post but I am a beginner, quite lost and over my head and am looking for help if people are willing.
I got an internship at an electrical engineering company local to me. They originally hired me because they have a Premier Pro CNC plasma cutter that they had no idea how to operate (they didn't know what a DXF file was or how to operate software). I ended up just looking through the manuals and figuring out how to draw in CAD and upload to Sheetcam and export as G-code into Mach3. And to cut rectangles, the table works incredible! However I was asked to cut a stencil to spray paint logos on larger equipment and I have had a myriad of problems.
I start, I will list my specs
Premier Pro Plasma CNC 5'x10' table
I use [inkscape for text] > ViaCAD > SheetCAM > Mach3
Hypertherm 65 Sync with 30-45amp finecut nozzle

Settings (as of mild steel 45 amp fine cut cartridge LOW SPEED)
16GA, 40Amp, cut height: 0.14, pierce height 0.14, pierce delay set to 1000ms, cut speed 120in/min, 94 arc volt (set on THC auto sensor device on control box), 0.038 kurf width.
Additional Info:
This table was bought and assembled 8 months before I arrived and sat unused the entire time. I have no idea how anything is assemble in the internals and wont pretend like I have the knowledge to understand it. Although I would like to, it scares me at $18/hr to disassemble a multi thousand dollar machine of which I don't know a thing about.

To describe my issue in as much detail as I can provide:
I am trying to cut text, using a stencil font I have procured from DaFonts (I checked multiple to ensure they would work with plasma cutting)
when cutting out this text the head will start moving before the torch engages.
Originally I had a test piece I would use (6x6 square with lines spaced 1 in apart) to test settings. I ended up using the cut chart settings for slow cut, with fine detail tip for hypertherm 65 sync, and adjusting the THC voltage to 75(what cut chart said) instead of 200 (likely was set by some bozo mesing with the machine who had no idea what he was doing). This fixed the test piece but made the problem worse for the stencil cut. The first pierce will always be perfect, then it will get worse and worse as the piece progresses. and yes I saw on about a million forms to increase pierce delay, I had it set all the way to 4000ms (just because I was frustrated) and this still did not fix the issue.
when my THC sensor is set to 75 volts it improves the problem slightly but not enough for me to believe this is the issue.

I do not believe this error is cause by sheet cam as I have uploaded test pieces consisting of squares, circles, and lines which cut perfect as well.

I am just lost onto what can be causing this issue and how to fix it if anyone has any suggestions as to what to look for. Not sure if it THC sensor, internals, or something I cant even think of

Thank you anyone who read this!
let me know any other info I can provide

[weldguy]

When you say things get worse and worse as the piece progress? What gets worse?

[acourtjester]

Ok lets get you cutting and producing good products. First off you can use any Art of CAD programs the product a DXF or SVG file to import into SheetCam. When in SheetCam you need to use the Post Processor the is created for your table. If using a THC there are Mach 3 post processors in SheetCam with that lable. You also need to have the tool in SheetCam with the cutting parameters for the metal you are cutting. These are in the Hypertherm manual, if you don't have one download one from the web. In your above statement you show cut and pierce height at 0.140" these are never the same. Most likely the pierce height is 0.150" and cut 0.060" for 16 Ga. (if you have a sync type cartage that may be slightly different). Other parameters are also listed in the manual like cut voltage and pierce delays times. So you understand after the cut starts the after the pierce delay and the torch starts moving the THC will control the torch height via the voltage feed back to the controller. I Mach 3 you need to have the THC box green on the main page, just click on the THC box to make the box green.
Just keep coming back to here when you have question member will help.

THC button.JPG

[Gavopian]

When you say things get worse and worse as the piece progress? What gets worse?

Typically the movement between the head and the torch on will get worse. It always cuts the first shape perfectly, and then starting at the second cut will miss the first bit of a letter, progressively it will start missing more and more parts of letters. It is not consistent enough that I have found errors in the G-code, but by the end of the cut there are typically less the torch is cutting out of letters.

Sorry if the explanation is confusing

[Gavopian]

Thank you for the help!
Currently I do have a post processor but I see no options in SheetCAM for THC settings, I also do not see any option for THC labeled in the Mach3 controller. I will do some investigation as to what post-processor SheetCAM is using. For how I adjust voltage, there is an external switch that adjusts the voltage setting for the machine, I assume it is THC controls as it is labeled "external THC".
Additionally, I will fix the height settings and I will get back to you on how the tests go.

[weldguy]

Ok so when you set the pierce delay to 400ms (4 seconds), the torch just sat there still for 4 seconds, then began to move, and then fired an arc while it was moving, missing a portion of the cut?

[Gavopian]

Yes that is correct, the torch would sit still and then move, missing a portion of the cut. for subsequent parts (as it was a collection of letters) it would cut out some letters perfectly and others would miss the beginning of them.
I tested with multiple other pierce delay times and still the same results.
What would likely be perfect, as I saw in other discussions, is an "Arc Okay" setting being enabled so the machine does not move without the arc being sufficient. However, after much digging, I do not believe I have this setting.

[acourtjester]

As Weldguy points out 4 second pierce delay is too long, my book says 0.4 seconds for 16 Ga. You will end up with a big hole when the torch does not move fast enough. Now a few things also, attached are 2 screen shots of SheetCam's post processor list for Mach 3 the first one is PP I have modified for different reasons. I am showing it as any added or modified PP will show up at the top of the list for selection. The others are what SheetCam shows as normal for Mach 3 down the list. It has been a long time since I used Mach 3 so these other thing may not help. In the configuration page there are 2 things (arrows pointing them out) one is the look ahead. This is for how far Mach is looking in the G-code for how it should prepare for the operations. and the other is if it should used ms in the time for the G04 delay. The last is for the THC to see Arc voltage signal enable which triggers the delay, not sure if that is for pierce delay or arc voltage stabilization delay. You said the table sat for 8 months did anybody monkey with the software before you started with the table??? The THC you are using is it a Proma, if so which on a SD or 150 model????

mach post 1.jpg

mach post 2.jpg

mach post 3.jpg

mach post 4.JPG

[weldguy]

Yes that is correct, the torch would sit still and then move, missing a portion of the cut. for subsequent parts (as it was a collection of letters) it would cut out some letters perfectly and others would miss the beginning of them.
I tested with multiple other pierce delay times and still the same results.
What would likely be perfect, as I saw in other discussions, is an "Arc Okay" setting being enabled so the machine does not move without the arc being sufficient. However, after much digging, I do not believe I have this setting.

OK, I find this very odd. I would keep an eye on the Hypertherm LCD screen when this happens and look for an error code. I am beginning to think it may be an air pressure issue which might sound strange but a simple check of the LCD screen when this problem happens would let us know.