Pull out the multimeter and start checking The Fun with Plasmas is to avoid shorts, ground loops, EMI etc after all those cheapies put of 3-5kv of dirty HF start prearcs. Those wreck havoc on usb signals Throw Ferrite cores on every wire Space inverter far from controller Make sure you only hook up the relay (its isolated) Get some mains filters before the PC and BlackBox PSUs.
I am having the same problem with my system. It is DIY nano based controller using GRBL 1.1h. I have EMI filters on the power in for the UPS and ferrite cores on most wires and cables. All cables to limit switches and motors are shielded and only earthed at one end. The controller is built inside a metal case that is also earthed. All inputs are opto isolated. The earthing has been done similar to the image shown in this post The plasma machine is a Chinese CUT50P that has supposedly CNC compatibility. Pilot arc start. As soon as the plasma fires, the controller freezes. I was thinking about getting a blackbox to try and solve this problem, but it appears to have the same problems I have. I have purchased another more expensive machine that has low frequency pilot arc start. I am waiting for it to arrive later this week and will send the results back to this forum if successful as it may help others that are having similar problems that I feel are based on the dirty EMI that radiate from many of these Chinese machines.
Pilot Arcs make a lot of EMI. So with any controller you have to spend the effort to mitigate that if you buy a Pilot Arc (hf or lf both). What you really want is a Blowback Start like the ones from Hypertherm That said, its not more than a days work, and the cheap inverters make it a no brainer that its worth the time. Couple ferrite cores, good grounding, and clever thinking to avoid running high power cables near any low voltage stuff, its not impossible, in fact, its quite easy, just takes a little elbow grease
I have fitted more than a dozen ferrite cores on all the cables that I can attach to, including the USB cable from the computer (it has 4 ferrite cores on that cable). The mains power supply to the UPS has an EMI filter and also ferrite cores fitted. The electronics are enclosed in a steel cabinet, all stepper motor cables are shielded and the shield is connected at the case end only. The case it earthed to a common point where all other earth connections are made and that is a bus that is fitted to the table frame. The plasma cutter is distanced as far as I can get it from the electronics. All inputs are made through opto isolators and the outputs are through relay modules that are also opto isolated. I can't think of anything else I can do to fix this problem, except changing the plasma cutter for a different model which I have ordered and am waiting on coming.
When you say the controller freezes, what actually happens? - Grbl goes into Alarm state because of a false limit trigger? - USB disconnects so the gcode cannot stream to the controller? it helps to diagnose what is being affected by the EMI Just to check on behaviour as well, as you air-starting or drag-starting (placing dragtip on the material when starting (typical with floating head) cuts the pre-arc distance a lot)
I tried using the CNC for the first time without the plasma connected at all. I just wanted to check the movements and also check the settings were correct for size, etc. All was perfect, directions good, sizes good, limits all working, homing in correct direction. I thought now I can test the plasma. I homed the machine, set the nozzle at the work surface and zeroed out the machine and then lifted the nozzle up away from the work piece. I had made a simple testing program in sheetcam to do just a cut out of a 50x50mm square and the settings were for a 3mm high start and a 1.5mm cut height. I tested first without the plasma turned on and all went well, no touching of the nozzle on the surface and all the movements correct. I tried again with the plasma turned on and it went though the first motions as expected, but when the pilot arc started there was just a quick flash and then everything stopped. The display showed that an alarm was triggered. No indication what the alarm was for. I had to do a soft reset and send a $X to clear the alarm state. I tried again and the same thing happened. I thought maybe the limits were being triggered, so I disconnected the limits completely so they were no longer a part of the system, I also disconnected the probe , even though I was not using it yet. Made another test and the same thing happened, just froze as the plasma fired. and went into alarm state. I disconnected the plasma from the CNC machine completely , but left it powered up. I ran the program once again with the nozzle just moving in the air well above the surface of the work piece. While the machine was going though its movements and would normally be cutting, I pressed the trigger on the plasma cutter by hand while the plasma cutter was completely away for the CNC machine and the same thing happened again. This time, the plasma arc would continue while I pressed the trigger, but the alarm was set off on the GRBL control, just like it had done before in the other tests. This proved to me that the plasma was radiating a huge amount of EMI to be able to have done that as it was not actually a part of the system any more, and could only cause the CNC to freeze up due to electo magnetic interference that had to travel across the room. The only common connection would be the earth from the power points as both the CNC and the plasma were using the same source of power from the mains. .
I disabled both homing and the hard limits and tested again. It is cutting OK with that setting, no alarm and cut OK. Need to adjust the speed a bit, but a good first cut, just a bit of slag. So what do you think is the problem with the limits? Maybe I have to re-wire them in some way?
Good test! Now we know what fails. Which limits are you using, we designed Xtension Limits with built in EMI RC filters: Xtension Limit Switch Kit An unplugged port is just as bad (the pin sitting there unconnected still acts as an antenna) so lets wire up some Xtension Limits, with a ferrite on the plug near the board. Make sure to wire gnd, signal and power for the filter to work. Or if you want to run without limits, add pullup resistors to keep the open pins in a known state
I am using just micro switches for the limits, but I do have several proximity sensors that I could try and use. I also have some micro switches that look similar to the Xtension limits switches that you have shown. Now that I know where the problem is, I can try several things to overcome the problem, finding the actual problem is the hardest and you have done that for me, thanks. I only have a few days to work on this project as I am going overseas for about a month, so I must try and sort it all out before I leave or I will not be able to keep my mind off it ha..ha.. Maybe I will have to do the same with the probe switch?
That is good to know, it makes things much easier. When I come back from overseas I will re-wire the limits. I need to get some more shielded cable and being in a remote country area it means travelling to the big smoke to get anything. I will purchase what I need on my return while I am in the city.
Peter, Im having the same issue he was having in this thread. How do the hard limits affect this? I have not tested turning off the hard limits yet but I will tomorrow. The plasma will trigger, then the BB will disconnect. Ive attached the Log. Im using a Hyperthem 1000 plasma. Everything is running smoothly otherwise. Any thoughts?
Sounds like EMI. See docs:blackbox-x32:faq-emi [OpenBuilds Documentation] Are you sure your plasma is CNC compatible (Blowback/LF)? In particular make sure your USB is protected well (short as possible USB cable, computer and controller as far away from plasma/plasmacabling as possible, desktop computer typically better than a laptop, make sure to use back USB ports directly on motherboard, not the front panel USBs, use high quality shielded USB cable, No USB hubs, extensions - counts for all USB devices, mouse/keyboard too - etc)
Yes sir. It’s an older machine but it is a Hypertherm and came with the controller port. Since the blackbox only triggers with relay1, and after research, I had to use the backend method to trigger the plasma. I did take precautions with using filters on the cables. I will try the limits and let you know.
See Post #9 in BlackBox X32 V1.2 Relays ? Please elaborate, lets see if that could be part of the issue Yes, but what is the torch pilot topology: HF, Prearc, LF, Blowback? See video below from around 2min30s onward
I read up on the V1.2, I will have to dig deeper into this to see if this would be a benefit. See attached, I spliced onto the trigger wires of the handheld torch. The trigger is normally open and doing this, the BB is able to close the circuit with the relay and trigger the torch. I know this isn’t ideal but it definitely works. Any thoughts? After really getting a better understanding of my machine it is indeed a blowback start. After working on it this afternoon, the BB has complete control over the torch. I definitely did notice EMI when the torch fired, it caused my monitor to get waves during burning. I’m installing ferrites on everything tomorrow and maybe that works. I will update you on the status after that.
Good then your relays are fine, no need to dig deeper into that That will be the main difficulty, squashing the EMI... Lots of trial and error ahead, but yet that will be cause disconnects etc
I spaced the electronics apart from one another. See photo. Far enough apart you think? It’s better but I’m still waiting on my ferrites to come in to install these.
