You probably already know this, but just throwing it out there as general knowledge. Only ground one side of the cat-6 cables. You’ll get ugly ground loops otherwise.
Well shoot I didn’t know that!
I really hope it works… I have been wanted to try a plasma cutter too. (actually have an old one that’s just sitting in a corner).
Ferrite cores on the cables perhaps ? (I have no idea if that would help in this case 
)
There are a handful of videos of some really long plasma cuts with an MPCNC. I never looked to far into it but It must work. There was some talk of a tiny grounding wire from the nozzle to the work before to help with the HF start but I don’t think they needed it.
I would say If you want to get it going see what they did. I have a few linked in the gallery.
Tried again today, still no success… That’s really annoying, I wonder why I’m the only one having that much trouble…
I shielded everything I could, motor wires, motors, screen wires, etc.
I removed all the unnecessary wires, so the only ones I was left with were the motor wires, screen and PSU.
Same results, it works for a while, then as soon as it starts a new path, poof, the Arduino either freezes, slows down dramatically or restarts.
I’ve seen a method supposedly effective for getting rid of the HF start, by running a wire from the tip of the torch to the ground clamp. I tried that, didn’t seem to change anything.
I tried to manually fire the torch after removing all the wires: seems to work 80% of the time, crashes 20% of the time.
As soon as I plug any of the motor wires, it crashes about 80% of the time I fire the torch.
Only the firing phase seems to pose problems, if I never release the switch between different paths, then it continues to cut perfectly fine. I just need to get rid of this stupid torch firing up HF interference issue to make it work.
I really don’t want to give up on this, but I’m not confident at all that it will work one day, I think I’ve done a lot of things to shield and protect everything for almost no discernible result, so I wonder if spending more time and resources on this will actually make any difference. It really sucks, I’m super disappointed. I wasn’t expecting it to work perfectly, but I was expecting it to be much better than yesterday…
But yeah, for the shielding I know that only one end of it must be linked to ground (but thanks for the reminder), and I watched a lot of videos, forums and website to try to understand what I’m doing wrong, but so far I found nothing really helpful.
For instance, on the few videos I’ve seen about plasma cutters MPCNC, people didn’t seem to have taken any particular precaution to shield anything (the Arduino is just left outside in the open air, motor cables are connected with dominos without shielding or protection… My machine is way, way more shielded than that!), but for some reason, their machines work and mine doesn’t.
This one seems to be shielded pretty good: http://forum.longevity-inc.com/showthread.php?1971-High-Frequency-shielding-for-CNC-Plasma
Have you confirmed 100% that it’s radio interference and not some power/ground issue? Can you completely electrically isolate the Arduino/RAMPS and confirm that it still happens when it fires? If you’re running off of a battery, and you don’t share a ground, etc. A large spike in the ground would look like a drop in voltage to the Arduino, which would cause a reset.
Can you test another plasma cutter? Maybe there’s a shop that would have one (like a car repair shop) that would let you bring your Arduino near and fire it a few times to confirm that it’s not something wrong with your cutter?
W.r.t. the foil tape, is the adhesive conductive? If it’s an insulator, then you’re not getting a great faraday cage.
Are there any wires that you can twist? Twisted wires have a lot less RF area, because each loop cancels out the next since it goes through the loop the wrong way. This only works for differential things, like power, but it could help.
Don’t give up. You’re about to crack it. You might just need a calming bike ride to stew on it, but you’ll get it.
1 Like
I tried with the Arduino directly hooked up to a 12V battery, but it did the same thing, so I’m pretty sure this is some radio inducted noise at this point. The Arduino crashed while being connected to the battery, with no cable connected except for the screen and the two battery wires. So I’m pretty sure this is some really nasty radio noise here.
I have no idea where I could find an other plasma cutter, I’ll see if any shops have that around, but I have little hope.
For now, I’ve decided to rebuild all my wiring. I will get rid of all the ethernet wires and go for shielded 4 wires of higer gauge. Hopefully that will help. I will also use some copper breaded wiring to put everything to ground, instead of my skinny wires and aluminum tape.
I will also put the Arduino in a proper metal enclosure.
Any other suggestions are welcome, especially regarding possible modifications to the plasma cutter itself, since it sounds better to me to tackle the EMI issue from the source.
The tape adhesive is indeed not very conductive, but that was the best stuff I had on hand… I’ve used a ton of it, for nothing.
I have little hope that this will work anyway, the fact that the EMI is so intense that it can shut down the arduino without anything being connected shows that the level of protection I’ll need will be quite important.
When you did the battery test, you didn’t have the grounds connected in any way? Something like grounding the case to the Arduino and the earth ground? Or grounding both sides of some shield?
I don’t know if the motors wires are the most susceptible to noise. I would look hardest at the Arduino itself and the Arduino power. What would cause it to reset? Either the power is dipping, the voltage difference is dipping, or there is noise on one of the important signals leading to, or inside the Arduino itself. Would current flow on the motor coils cause it to reset? I would definitely try twisting the power cables. Is the power supply shielded?
You could also try without the LCD connected, although how would you know if it reset…
On a program I worked on, we had AC inverters in these vehicles we were using all the time, and we had radio communications (just normal, consumer band radios). The AC inverters had GFCI plugs built in. Turns out, when the radios turned on, there was a large EMF burst, and the GFCI would trip. Most of the users didn’t know it was a GFCI, so they would call in the engineers to fix the AC inverter, and by that time, the correlation to the radio was not noticed. It was months before someone figured out the correlation. The solution? Replace the GFCI with normal, unsafe outlets.
1 Like
For the metal case, Can you just throw it in a metal toolbox?
No, the battery was connected straight to the ramps terminals. The Arduino wasn’t connected to its faraday box, but the Faraday box was connected to earth ground
I had the issue even without any wire connected to the Arduino, except for power and screen. The screen wires were shielded directly to the faraday box using metalic tape.
From my tests, I would say that they are. I had far less issues with the motor wires disconnected. I tried to disconnect all the wires then connect one after another to see if it was coming from one particular motor, but they all crapped out after 2 or 3 attempts of firing the torch, no matter which motor. This occured far less with all motors disconnected, but still occured.
In addition, the fact that the extruder driver died yesterday seems to indicate me that this may have to do with the motor wires. I have one more test I would like to try tomorrow in order to make absolutely sure.
I tried that. But I haven’t shielded it, so maybe I’ll try that to.
Yes
Yeah, that’s the problem. I could hook up my computer but this is too much of a risk right now.
Well, I’m in China, unsafe outlets are everywhere XD. I’m pretty sure the pugs are standard, el cheapo stuff.
Well, I could, but I don’t have one. I’ll purchase a metal enclosure, it shouldn’t be very expensive.
Twisting the motor wires would be good too. Making sure each motor coil is on it’s own twisted pair. Maybe you’ve already done that, IDK. You have to be getting a s***load of current to be resetting the arduino from the motors. I suppose the current could be going back into the 12V supply, which could reset it. Man, that’s a lot of EMF though…
1 Like
If your plasma cutter is controlled by a relay, the relay coil could be feeding back into the ramps via the control or power lines. You can try putting a metal oxide varistor (MOV) or transient voltage suppressor (TVS) across the coil (or relay power inputs if there’s a separate power and control inputs).
1 Like
Motor wires are already twisted, I paid attention to use the correct wire pairs in the Ethernet cables.
I tried with and without the relay, made no difference, so I guess the relay is not the main issue.
Anyway, I’ve bought the following:
-Shielded motor wires: https://item.taobao.com/item.htm?spm=a1z0d.6639537.1997196601.659.91b1f81SKCjbu&id=528099322225
-Better connectors: https://item.taobao.com/item.htm?spm=a1z0d.6639537.1997196601.222.91b1f81SKCjbu&id=529658070871
-A small enclosure, I will also put the screen inside just to make sure: https://item.taobao.com/item.htm?spm=a1z0d.6639537.1997196601.4.91b1f81SKCjbu&id=13056873632
-Braided copper wire, to make a reliable connexion to the earth ground: https://item.taobao.com/item.htm?spm=a1z0d.6639537.1997196601.539.91b1f81SKCjbu&id=529307086803
I will make the motor wires much longer, this way I can put the Arduino enclosure at least 4-5 meter away from the table to, hopefully, limit the EMI sensitivity. If this works, I think I will use two Arduino: one fully enclosed for plasma cutting and another one for 3D printing or CNC milling. I will just have to disconnect the box and plug the other one, it should take less than a minute and make things way easier. Kind of the same principle I use for switching Z axis. This way I can also have different firmware settings, since plasma cutting seems to necessitate different jerk and acceleration settings from what I’ve seem so far during my brief tests.
I just hope I’m not wasting my time doing this, that would really be a bummer if all those efforts were for nothing.
Have you checked to make sure your outlets are actually grounded. If you are in a garage type space and they added any plugs not everyone thinks to ground them, I have seen this sooo many times with the plastic boxes.
Just make sure you only ground one end… 
Not really, I’m not sure how to do that… I’ll check on the Internuts to find out, good suggestion!
Hey guys, by the way, I created a small instructable about how to build a water cooled 3D printing nozzle for your machines, the same way I did it for mine. I’ve been using it for a while now and I’m now sure it is reliable enough, so you can have a look here and build your own if you want, it’s super easy:
Hope it helps!
4 Likes