Help - Get interference from Plasma on my PC

Hi, I'm new to CNC Plasma cutting and need some help.:confused:

I've build a 3-axis CNC machine to run a router on it and it worked very well.
A while ago I decided to put a plasma cutter on it. I'm using a Thermadyne Inverter type with high-frequency arc start.

The problem is, as soon as I "trigger" the plasma arc my computer pickup interference and either hangup or start with alarm beep and lose control of most functions.

I've tried to isolate the cutting table from the gantry but it doesn't help. Even with the plasma cutter removed from the cnc machine and positioned 10' from the computer the problem remain the same.

I assume the problem is the high-frequency arc, if so, how can I get arround this problem?

I've had a little experience with a plasma causing one of my PC's to reboot. I'd say that the motherboard had a design weakness that made it susceptible to the RF noise.

Best to try a different PC. Before you actually go to all the work of changing over the PC in your control, you can just set up the prospective PC near the same location, and fire up the plasma and see how the PC reacts.

Run the PC through a battery backup type UPS (surge protector). You should already be doing this.

Grounding for noise and grounding for AC safety are two different topics. The HF noise from older plasma cutters is a real problem. If it reboots you computer it most likely will cause problems in the motion control electronics.

Make sure your computer is not grounded locally to the table or the plasma side.

It's ground needs to be from another source. Depending on the motor drives, power supply and control electronics you may not have the option to run a two ground system in which case you may have to put your PC inside a metal enclosure not connected to the table. We have several customers running MACH on a PC with our Digital THC with old HF start plasmas but they had to go to extreme grounding techniques to make it work.

TOm Caudle
www.CandCNC.com

After some testing, it seems that "extreme grounding" should solve my problems.

Thanks for advice and info.

Quote:

---

Originally Posted by nicjac

After some testing, it seems that "extreme grounding" should solve my problems.

Thanks for advice and info.

---

In the wonderful world of plasma, there is no such thing as "extreme grounding"

Very common problem with HF plasma systems. The first thing you should do is go to a plasma manufacturers web site, such as Hypertherm or Thermadyne and research the proper grounding techniques for providing a proper ground for plasma. One thing to remember is that just because a ground rod is driven into the ground, does not mean that it is 'connected' to earth ground. This is usualy the most common fault of plasma interfernce. People think that their ground rod is making earth ground just because it is in the ground.
Once a good earth ground is established, follow the plasma manufacturers recommebdation for grounding the plasma and the table.

Quote:

---

Originally Posted by nicjac

After some testing, it seems that "extreme grounding" should solve my problems.

Thanks for advice and info.

---

NicJac, If I may ask, what did you end up doing to solve your HF interference problem? I have a problem with an ESAB PCM 1125 hf start, it's interfering with the MP1000thc from CandCNC.
--Mike

First thing was to isolate the cutting bed from the rest of the machine, either on rubber or other isolating material or by having the cutting bed as a loose table not connected to the machine.

Equip the cutting bed or loose table with an earth cable connected to it's own earth-spike.

The last thing I did was to use USB mouse & keyboard, the PS2 mouse and keyboard seems to be more acceptable to interference than their USB cousins.

I still get a 1-step "jump" each time the plasma arc start, but through the speed reducer 1 step on my stepper motor is only a fraction of a millimeter and does not bother me.

I hope this will help you.

Most RF (high frequency) noise issues can be solved with by making the following changes:

1. The electrical noise is generated by the high frequency circuit (as much as 15,000 volts at 2 mhz) in the plasma.....and the high frequency travels through the torch leads (20' or more) to the torch. Consider this the antenna that is transmitting the RF noise.

On industrial systems the torch leads have a tinned copper braid that covers the leads...and is connected to the plasma system chassis on one end....and not connected on the other end. This shieded torch leads packeage is then isolated from all other wires and cables on the machine....either by running it through a plastic powertrack by itself....or by hanging it from an overhead boom. If there is excess length....do not coil it...rather, keep it layed out straight.

You may be able to simply isolate the torch leads without shielding them....try rerouting them temporarily by hanging them from the ceiling...and keeping them away from other cables as much as possible. Try cutting with the machine....if this solves the problem than you now know where the noise is coming from...and you can work to isolate it.

2. Grounding. A copper coated steel, 8' length ground rod should be driven as close as possible to the cutting machine. Short lengths of heavy gauge stranded cable (flat braided wire is actually best as high frequency travels on the outside of a conductor)...and this has the most surface area. The use of #12 single strand building wire is practically useless...use heavy gauge multistrand wire for grounding). Connect the cutting table to the ground rod...short wire. Connect the chassis of the plasma directly to the ground rod...sort wire. Connect the gantry (moving parts, both x and y axis) to the ground rod....short as possible wire. Do not "daisy chain the grounding...every compponent needs to be connected to a "star" ground directly at the ground rod. This will probably help.

Jim Colt Hypertherm

Thanks nicjac, I cannot isolate the cutting bed on my machine because it's all an integral table, but I'm working on it today.

Jim, What about the plasma chassis ground that connects to bldg. ground? The ESAB unit has the power cord ground attached to the unit chassis inside, so if I run another ground to the star point on the table, (or directly to the 8' rod then it will be grounded both through bldg ground and my plasma table earth ground?

Thanks for the input guys!

The chassis ground that goes to electrical ground can create a "ground loop"....which can be problematic for noise......but safety concerns must overide that. I would suggest leaving the electrical ground as it is....but still ground the chassis to the ground rod. Ideally the plasma power supply should be very close to the ground rod.

The goal is to get stray RF to follow a path of least resistance (large surface area conductors) directly to the ground rod. If there is any daisy chaining....it won't be direct!

I have seen many sites over the years with noise issues (on big industrial machines) that had an overly complicated grounding system.....by simplifying it we were often able to eliminate the problem. Isolating the torch leads is the #1 best cure!

We run high frequency plasma cnc machines at every trade show....with absolutely nothing but the electrical ground....with no problems whatsoever. We do pay special attention to routing of all leads and cables on these machines. Motor and encoder cables are low voltage, and the RF from the high frequency loves to inductively couple to these cable....wreaking havoc with the computer.

Many of the newest industrial machines today use sercos drives all linked with fiber optics. noise problems eliminated!

Jim Colt

One other comment.....I would not attempt to isolate the table from the rest of the machine.....could be dangerous. All metal surfaces need to be at the same potential as the ground rod. Plasma systems can have open circuit voltages of 400 vdc.....if the table was isolated...and some sort of electrical failure ocurred within the plasma system...you could have some lethal voltages present.

jim colt

I have a 6awg stranded ground that runs from the Z carriage to the gantry, then one from the gantry to the table . . . maybe that daisy chaining is part of the issue. I'll get those running right down to the 2/0 ground cable connection point on the table. That 2/0 cable is 5' long going directly from the table leg to the 8' ground rod.

Thanks Jim for the info. I'm still working on it.

I would run the each directly to the ground rod. If you are picking up noise on the carriage...a very likely place because the torch is mounted there...then you are now carrying this RF to the gantry by daisy chaining these connections.

I would clean up the grounding a bit...then isolate the torch leads by pulling them completely off the machine....you could even have someone hold them up away from everything (I always bring a helper!) while you fire the system.

Another fix that occasionally works.....find the spark gap assembly in the high frequency circuit of your plasma system. The gaps are probably set at .015" to .020" ....remove the tungsten rods, clean them up and file the ends (only with a diamond file....a steel fill will contaminate the tungsten) perfectly square and sharp. Re-install the tungsten rods and try closing the gap up to less than the manuafcturers specs....this will diminish tha starting voltage. If you close the gap too much the torch will misfire...so you want to balance this carefully. If the tungsten gaps are worn and rounded....your starting voltage will be very high (because RF travels on the outside of conductors)...and may be causing your problems.

Good luck, Jim

Hello,

We have recently retrofitted a 6m by 3m S.A.F. Plasma cutting table.
The machine works great however, the arc transfer creates noise which is interfering with our control and P.C.

I will list components used and current grounding techniques below;

Plasma Source : MaxPro 200
Control : Usb Version of Flashcut CNC Controller
Motors/ Drives: Clearpath Stepper Servos (SDSK) type.
Pc: Standard consumer grade Lenovo.

The cnc controller and PC is located in the control cabinet which is located on the gantry.
The Plasma unit is located beside the table aproxomatily 1m.
The plasma hose runs in a cable chain distance of 150mm from other signaling cables.
In this cable chain a usb and hdmi cable are also run to a standalone control console.
At the end of this cable chain (gantry side), the plasma hose is physically 800mm from all other cables on it's route to the torch holder.
I currently have the plasma hose shield grounded on both the Z axis frame and on the plasma unit side to copper grounding rods. The machine is grounded to hypertherms reccomendations. I had this grounding checked be by a third-party company locally. Incase I was missing some (simple) grounding.

Currently the Flashcut Control is losing it's USB connection to the PC as soon as the torch fires.

Any help would be greatly appreciated.

My next steps are
To physically distance the plasma hose completely by hanging it down from above.
but beyond this Im not sure what else I can do.

Is Flashcut the issue as its USB connection?
Would an ethernet controller be a better solution?

Thanks

Andy leahy - Ireland.

Not sure if you noticed but this thread is from 2009. You probably would be best to start a new thread. The USB cable you are describing, what does it connect too? the signal generator? If so this is your problem. I believe the max200 is a high frequency start plasma unit, and causes a lot of interference. You might try removing the grounding from the computer that runs the cnc software to see if that changes anything.

I maintained a couple of the commercial tables that used a PC based system with a Galil motion controller card, the company installation included a ground rod at the table, the service and enclosure ground was also bonded to this stake.
Never experienced any spurious re-booting etc.
Al.

Quote:

---

Originally Posted by andyleahy

Is Flashcut the issue as its USB connection?
Would an ethernet controller be a better solution?

---

Flashcut shouldn't be your problem.... I have used Flashcuts USB version on more than one Plasma Table without issue, and always with Hypertherm units. My current machine has the PC outside of the Cabinet with power supply and drivers, fed with a 10-15 foot quality USB cable. I'd expect problems with that length, but I have none currently. I always ground the plasma cutter right to the plate I'm cutting, and the drives, limits and motor wiring all use common star grounding, with all cable work ONLY grounded in/on the cabinet. Cabinet is bolted right to the machine frame.

First, I'd try a quality USB cable. If that does not help, Flashcut is easy enough to install on another PC or laptop. Install it, copy the Setup file to the Laptop, open Flashcut and load the setup file.

Try running the machine on a longer cable away from the machine. Certainly worth a try as Flashcut is not an invasive program and any laptop will do for a test. If you get better results, You could first swap in a newer or better PC power supply in that Lenovo, reseat all the PC's boards and memory cards, and even use a DIFFERENT 110v power source for it.

One more thing... make sure you have a really good 9v supply for the SIgnal Generator. I've found that it needs a good solid 1 amp or higher wallwart. I've had isues when the PS was not up to snuff.

Lastly, you could give Flashcut a call.... they have put pretty much all their energy in the last few years directly towards Plasma cutting specifically and there is no way they would be successful if it did not work.

Good Luck !