New plasma CNC has interferance issues causing USB to drop out

[johnharvey]

Using a G3 and 4.02 OS.

wired up and fired up a CUT40 HF plasma cutter for the first time. Right now the first issue is that the USB gets knocked out, I loose the keyboard, mouse control (freezes) and the thumb drive also once in a while whenever I turn the torch on. I can e-stop it.

MASSO will run a small test program fine it seems, but the keyboard Icon goes away and mouse control is frozen during the run. I have to unplug and re-plug the mouse and thumb drive to get them back but not the key board?

M3

G1 X15 F50

M5

The torch break and touch off switching is done through relays to isolate any HF in the signal to MASSO and they function OK.

Torch on signal is through a relay per MASSO spec.



The MASSO is in a large Elec. enclosure with all the drives etc. The Plasma is outside with nothing crossing the internal wiring of the machine.

This machine is normally a router so for the plasma box, I tap off the 220V before the main switch of the machine since the VFD is not used at this time.



Any clues as to how to fix the issue????

Cheers,

John

 

[Ross]

Sounds like you have a bad case of electrical noise getting back to the Masso, Your running a cheap plasma like me so don't kill your Masso!

Iv'e done a few videos in how i shielded, grounded and ferrite beaded my machine. You can find on this forum or YouTube.

All the drives and power supplies, i would have at least half a meter from the Masso with a metal shield plate (divider) between the Masso and the drives, power supplies.

Plus you can lookup Hypatherm on shielding.

Ross.

 

[johnharvey]

Thanks, like you wrote, it is definitely HF noise. Going to try some shielded wire to get the torch break and touch off conductors from the outside into the main electrical enclosure then ground the shielding. This is the only way I can see the noise getting anywhere near the MASSO. everything else is outside the BIG thick metal enclosure. Never had issues with drives or PS's. about 2' away, did I say its a BIG enclosure? lol

I'll do some more searching.



john

 

[airnut]

John,

I would also put a diode across the coil on the relays that are controlling anything. On all aircraft I work on we do that on the starter solenoid to take care of the spike when the coil collapses. With out it, it will take out the avionics in some instances. According to your post you are using relays. If the control side does not have a diode shunt it can feed back to the Masso. While I do not have a plasma to deal with yet I have experienced it with my Masso and mill and it did the very same thing. A diode fixed it. it simply gives the circuit a place to get rid of the energy without going back through the Masso. less than a buck.

Steve

 

[johnharvey]

The way I have the relays wired for torch break and touch is this way. I send 24v from the machine supply up to the common on both switches at the torch head in the same conduit for simplicity as the plasma lead and ground for the HF just as they do for the start stop switch with hand held torches.

From there each switch is normally closed, this in turn allows the 24v down to the coils on the relays then to 0v maintaining a NC condition. This relay in turn is HIGH "NC" which maintains continuity with the separate contactors for the MASSO torch and break circuits. When the switches on the torch head OPEN, power is cut to the relay in the enclosure which in turn then breaks the contactor for the MASSO circuit. So I believe the MASSO circuits are isolated from the torch head switch wiring and noise, I think?

What do you guys think?

john

 

[airnut]

John,

Not sure I follow you on this, like I said I have not dealt with the Plasma system as of yet. All I am saying is when there is a relay involved you should use a diode across the coils to eliminate any source of back EMF from the circuit as this will cause a spike on the control side of the line. Here is a simple video I just found that shows this and there are several more to watch on the subject. Payy attention to the polarity as he shows.

I found this to be a problem in my mill prior to my converting it and it caused all sorts of issues like for example Coolant coming on at start up. When a relay activated or opened up that spike will KO a computer or trigger a unwanted result. Its a real cheep place to start, I would consider it mandatory when working with relays and it may fix your issues. The old Fanuc had lots of "shunts" both AC and DC everywhere to prevent the spikes.

Beyond that I have no idea. Steve

 

[fablab]

I had a fair amount of issues with noise when I first built my plasma table, here is a list of things I did, that I believe made the most difference:

• All cables are shielded
  
• All cable shields connect to a star ground (copper bus bar) which connects to its own ground rod
  
• In the electrical cabinet be sure power and data wires cross at 90 deg's if you have to run them in parallel keep that distance as short as possible
  
• The plasma torch cable is in it own drag chain and kept separate from all other cables. I have the torch cable running down the left side of the machine and the stepper & limit switch cables running down the right
  
• The plasma torch cable has tinned copper metal braid sleeving over it
  
• The stepper and limit switch cable bundles also have tinned copper metal braid sleeving over them
  
• The metal braid sleeving is also connected to the star ground
  

I believe the last 2 item made the biggest difference but I did make several changes at the same time I added those so I am not 100% on that.

The tinned copper metal braid sleeving I used can be found here:

Sleeving



Let me know if you have any questions and I hope that helps.

 

[johnharvey]

Thank you guys. the diode will come in handy for my mill also as it burps the coolant pump randomly. The shielding and separation is best practice. I built a 4 x 2 ft combo router and plasma table 5 years ago for a client. I used a Hypertherm XP45. Everything for the gantry ran in the same energy chain and never had issues so that is why I was a little baffled at the noise. but it is a cheap CUT40 we are talking about! Fortunately it is only knocking out the USB communication and not the machine movements at all. And this was while striking an arc without it grounding to make a cut which will shut off the HF I believe.

I'll keep you posted to what solved it once I add the suggestions.

Airnut, the diode will come in handy for my mill also as it burps the coolant pump randomly.

Fablab, what plasma machine where you using?

Pics of plasma machine build a while back, and my single seat aircraft design I did 24 years ago that I'm getting around to offering plans and components for. Hence the new Plasma capability. The router is an old COSMEC 250 retro fitted with NEMA 34's hybrids. 4 x 8' the plasma water table is covered right now between the gantry and the console. There is an arm that reaches back through the gantry that holds the torch head. Have a routing job this morning so had to take it off for the moment. I'll post some pics when the arm and torch are on it. It's awesome to be able to switch between MASSO OS builds with modified machine parameters.

 

[fablab]

Quote from johnharvey on April 4, 2021, 12:50 am

Fablab, what plasma machine where you using?

When I started out it was a cut 40 same as you are, I am now using an Everlast.

I can remember when I was first testing & had all kinds of noise problems I was using my oscilloscope to try and isolate where the noise was coming from.. that was basically a fools errand as just holding the probe in the air would show a ridiculously high voltage when an arc was struck. Back then I was using an Arduino & a CNC Shield, I never did get that to work very well.. Once I switched to Mach3 I had a lot better luck still had some issues but I was able to work around them. Switching to the Masso was a big improvement as well, the machine is more responsive & the THC works considerably better (I use a Price THC)

Using a plasma with a blow back strike (or something similar) is much, much, much better than high freq. that's for sure

You can see pics of my machine here:

https://www.masso.com.au/forums/topic/hello-from-wisconsin/

 

[johnharvey]

Nice setup Fablab, I like the see through enclosure.

Did a bunch of shielding and moving the plasma box a few feet further away from the MASSO, still no joy. So I decided today that the HF problem is a non-starter and that puppy is getting returned. I've read that the Cut50DL and similar "low" freq. plasma's seem to work OK. I'll give one of those a try before dropping almost a grand on a box.



I just posted a bug. Can you confirm what I'm writing is correct? Its not up yet but the parking position does not update between plasma and router, it stays to whatever it is set at, not good for my setup.



john

 

[Ross]

John, I tried to change from plasma to router, but it through a soft limit error.

So i didn't bother to look into it any further. Just not high on my list at the moment!

I think it might be a case of loading (software plus settings and maybe disconnect and reconnect some wires, eg plasma on/off and THC)

Or use other outputs for plasma and router.

Just save your setting before you start changing things.

Just my thoughts Ross.

 

[johnharvey]

Ive been switching between no problem, nothing like you describe. It's just that the parking position is global and not unique to the MASSO OS of plasma or mill or router, lathe etc.

You shouldn't have to switch inputs for limit switches, I didn't. The M3 and M5 that normally start the spindle will only start the torch on the TTL signal output chosen which is nice, no wiring needed to be changed for the spindle, it just lies dormant.

 

[johnharvey]

So HF plasma was returned and I purchased a "low freq" model from a different supplier that is supposed to play well with CNC, well, it did a tiny bit but it is still unusable. USB drops out randomly after using it a few times and then my cycle start button doesn't work. I can ctl+S to run it but the something else will fail. I guess I'm going to have to drop a few grand on a real plasma cutter? Heck, USB is dropping out randomly in router mode also.



john

 

[cncnutz]

Hi John,

After reading about the USB drop outs it seems that since it happens in both Plasma and Router the problem is probably something other than the Plasma though a plasma will accentuate any weakness.
Could you post a copy of your printable settings file here.
You will need to zip it as the forum doesn't accept TXT files.

• Have you tried a different flash drive?
  
• Is the flash drive plugged directly into the MASSO USB port or are you using a extension cable in between?
  
  Cheers Peter
  

 

[johnharvey]

Thanks for checking in, I can get the zip file of settings in a week, vacation time. What is random is that I'll come back to the console after loading material or whatever, and the keyboard icon is gone and the keyboard is off line, or its the mouse, or it is both. Could be fine all day, or not. Have to cycle the usb plug to bring it back, AND when it is really bad, the cycle start button will not start the program. CTRL+S will start it, BUT When that happens I have to power cycle the machine (MASSO) to get the button push back. I put EMI rings on both ends of the USB devices, didn't seem to help? I can't find a repeatable situation that causes it. A bit frustrating.

Cheers,

John

 

[airnut]

John,

Because you are seeing the symptoms on both the plasma and router hook up I would not think the problem is in the plasma radiation. I still think I would install the diodes across all of the relay coils and especially the motor contactors or any relays that are independent. For a dime a piece, A In4007 on each relay or coil should make a noticeable difference if this is the problem and it it doesn't then what are you out?

Regards,

Steve