CNC Plasma Pierces and then shuts off

[mikel92]

I am currently working on setting up a Plasma Table at our shop, and I cannot get the table to execute a full cut. Currently, it will pierce the work piece, and the torch will then shut off (with the gas still running). The problem seems to be in the G-code because I can cycle start, and it will trace out the cut path at a safe Z height. I have notice that the "ArcOK" signal does not activate. I am somewhat of a novice when it comes to Plasma tables so I feel like maybe I am missing something simple. Has anyone else had this problem?

Setup:
- Northern Plasma NV-1 Table
- Hypertherm Powermax105
- SDS Quickcam Pro
- Mach 3

[rodw]

Can you give a bit more info about your setup. Eg software, plasma machine, software and wiring.

You could have a wiring fault that means you are not getting an ArcOK signal
Your sensing circuit may have a minimum current draw that is less than the mimimum spec for the relay in the machine
or it could be that your machine is not establishing a valid cut and shutting down.

You have not given enough info for us to help.

[mikel92]

I listed my setup at the bottom of the post, but here it is again for convenience sake:

Table - Northern Plasma NV-1
Processing Software - ShopData Systems Quickcam Pro
Control Software - Mach3
Plasma Cutter - Hypertherm Powermax105

As for the ports/pins what would you need to know exactly?

As I said previously I have noticed I have not gotten a ArcOk signal during the cut, is this required even if the THC is deactivated? I have tried cutting with the THC on and off, and the results are the same.

[rodw]

Sorry, I don't know anything about your hardware but I think this is expected behavior. Normally, if the ArcOK is not received within a specific time period after commencement of the pierce, the control software will stop the job. So you need to troubleshoot the error. Its frustrating I know but you need to trace your wiring on the Arc OK pins on the Hypertherm manual. There are dry contacts (eg a relay) in the plasma cutter that closes when a valid arc is established. There will be a circuit in your control box that sends current down one wire and back to an input pin when the dry contacts close. Depending on your hardware, this circuit will be either on the V+ supply rail or the -V rail.

I did list some issues I've experienced with ArcOK.
I did find on thicker material, extending the time the machine waited for the ArcOK signal solved the problem as the crappy Chinese machine was very slow establishing an arc..

Just a thought. Does your plasma have an ArcOK indicator on the front panel?

[mikel92]

Thanks for the advice, the table was shipped, so something could've come lose during that. The ArcOK indicator is only displayed on the Mach3 interface.

[Beefy]

The problem is not the Arc OK signal.

Mach3 only requires an Arc OK signal to allow motion, when THC is enabled.

Seeing as you said you have the same problem with and without THC enabled, then that eliminates Arc OK as the cause.

The first thing you need to do is split the system into 2 parts and see which side has the problem. So you have the control side which tells the plasma cutter to turn on the torch, then you have the plasma cutter itself.
Put a normally open push button switch in parallel with the relay output contacts that tell your plasma cutter to turn on the torch. Disable THC in Mach3 for testing. When the torch fires, immediately press and hold the button. You are now manually keeping the Torch On signal active to the plasma cutter. Does the torch still go off, if so the problem is in the plasma cutter. If not the problem is coming from the control side.
Do not use very thin sheet for this test. Use something a little thicker that requires maybe a second to pierce. This is so the torch arc has some metal to burn, and also gives you time to press the pushbutton. Thin metal is pierced very quickly and if the torch doesn't move immediately the arc could have nothing to electrically connect to, and the plasma cutter could shut off. So you want to eliminate that possibility by using thicker metal and the smallest nozzle that can pierce that thickness metal. Of course your cam software needs to put a Pierce Delay value in your gcode so the torch is stationary while piercing.

You said earlier the problem "seems" to be in the gcode. I can deduce from that you don't know how to read basic gcode. In Mach3 normally M3 turns the torch on and M5 turns the torch off. Maybe post your gcode here so others can see, and can confirm if the gcode has anything to do with this issue.

[rodw]

Beefy, I know you know far more about Mach3 than I do but that does not sound right to me. This video of cutting with THC turned off clearly shows the ArcOK signal coming on in Mach3. Watch the onscreen LED.


But I agree, it would be good if the OP could share his GCode and any subroutines called by that code.

[Beefy]

Hi Rod,

the Arc OK LED is only an indicator to inform the operator of the presence of an Arc OK signal. The THC being enabled or not will not change that.

However, the reaction to the Arc OK signal is different depending on whether THC is enabled or not. Example when I wanted to do a dry run of my gcode, I'd turn off THC (and the plasma cutter) and run the file. If I attempted a dry run with turned THC on the code would get as far as M3 but after that it would halt, i.e. it was waiting for an Arc OK signal before motion could commence..

By the way I responded to your Sheetcam post earlier. Hopefully that's the answer you are looking for.

[rodw]

Hi Rod,
the Arc OK LED is only an indicator to inform the operator of the presence of an Arc OK signal. The THC being enabled or not will not change that.
.

But the OP said the THC signal never shows in Mach so it sounds like its more likely to be in the physical connection. But in reality, if this is a new table, he should seek guidance from the manufacturer. So many ways to skin the THC cat as you well know.

Thanks, I'll check out the Sheetcam forum as I did not seem to get subscribed.

PS. Writing this on my LinuxCNC box in front of my plasma table... Upgraded my Everlast to a 120 amp machine and got it running pretty sweet.

[Beefy]

Hi Rod,

you know the bit where I told the op to use some thicker plate. I'm thinking a potential cause of this issue could be if the op is using thin plate and there's a small pierce delay in the gcode. Maybe (just maybe because I'm clutching at straws here) the torch fires, and immediately makes a big hole, then a flameout occurs, thereby not giving a chance for Arc OK to appear.

Mind you I'm at a complete loss what is meant when the op says, "I can cycle start, and it will trace out the cut path at a safe Z height". Definitely need his gcode file to try and make sense of that one. I agree, it might be time for him to chat with the manufacturer.

[rodw]

Keith, that could make sense. If thats the case, he could wind down the power. My new neighbour at my workshop who is very experienced with plasma said there is a rule of thumb he goes by that the cutting amps should be about 10x the plate thickness so 2mm plate should be able to be cut at 20 amps. So the full 105 amps might be overkill!

PS, I did cut some 2mm at 80 amps which I knew was too hot. I was supplied 20 and 30 amp consumables with the machine but I have not tried cutting that low yet.

[mikel92]

I have contacted the manufacturer, and they are not much help without costing a substantial amount of money to help fix/setup. I understand the basics of gcode, but I have only dealt with a CNC router before this, so it's all kind of new to me.

I am using 3/16" aluminium plate for my test cuts - the pierce hole is roughly 1/8". Based on your posts I tried reducing the amperage to see if that would help. The manufacturer of the torch recommends using it at full bore (105 amps), and the software I am using allows 40, 60, 80, 100, and 120. I ran it at 80 amps, with the same results.

Here is the gcode for the part I am trying to cut:

(Plate test)
(WIDTH: 0)
(LENGTH: 0)
( 80.000 AMPS AIR_AIR )
F180
N2 G70
N4 G90
N6 G54
N8 G00 X12.075 Y5.842
N10 F180
N12 M04
N14 G01 X12.078 Y5.812F90
N16 G03 X12.138 Y5.818 I0.030 J0.003
N18 G03 X11.919 Y6.016 I-0.219 J-0.022
N20 G01 X10.919 Y6.016
N22 G03 X10.699 Y5.796 I0 J-0.220
N24 G03 X10.919 Y5.576 I0.220 J0
N26 G01 X11.919 Y5.576
N28 G03 X12.138 Y5.774 I0 J0.220
N30 G03 X12.138 Y5.787 I-0.050 J0.005
N32 G01 X12.132 Y5.824
N34 M05
N36 G00 X13.200 Y6.592
N38 F180
N40 M04
N42 G01 X13.203 Y6.562F72
N44 G03 X13.263 Y6.568 I0.030 J0.003
N46 G03 X13.044 Y6.766 I-0.219 J-0.022
N48 G03 X12.824 Y6.546 I0 J-0.220
N50 G03 X13.044 Y6.326 I0.220 J0
N52 G03 X13.263 Y6.524 I0 J0.220
N54 G03 X13.263 Y6.537 I-0.050 J0.005
N56 G01 X13.257 Y6.574
N58 M05
N60 G00 X14.374 Y5.296
N62 F180
N64 M04
N66 G01 X14.074 Y5.296
N68 G02 X13.794 Y5.016 I-0.280 J0
N70 G01 X8.972 Y5.016
N72 G01 X12.032 Y8.076
N74 G01 X13.919 Y8.076
N76 G02 X14.074 Y7.921 I0 J-0.155
N78 G01 X14.074 Y5.296
N80 G01 X14.216 Y5.155
N82 M05
N84 G90
N86 G00X0.0Y0.0
N88 M30

[rodw]

What happens if you add an M3 in your code at the start of cut? I would have thought it was required.
What gcode turns on your torch?

I am still leaning towards there being a fault on the ArcOK wiring. so Beefy's advice might be a way forward.