AddenDUMB: I Broke My CNC Plasma Table! | Learn from my mistake | Langmuir Crossfire THC
HTML-код
- Опубликовано: 2 окт 2024
- When I put the monitor mount on my CNC plasma table, I totally broke the torch height control, without realizing it. Today, we're going to figure out why and fix it.
Tools used in this video:
*This site contains affiliate links for which I may be compensated
EEVBLOG Brymen BM235 Multimeter (Amazon*): amzn.to/2YKFSEk
Milwaukee Brushless 3/8" Impact (Amazon*): amzn.to/3uxwHJ4
Cable Raceway (Finger Duct) (Amazon*): amzn.to/3CFUskr
Metal Power Strip 12' Cable (Amazon*): amzn.to/3kZNktm
CarbonX Carbon Fiber PETG (Amazon*): amzn.to/3BEsoh3
CarbonX Carbon Fiber PETG (MatterHackers*): bit.ly/3l2grMx
ASUS 21.5" Touch Monitor (Amazon*): amzn.to/3B7MbW4
Beelink T34 M Mini PC (Amazon*): amzn.to/3nwjy1m
Hypertherm Powermax 30 XP Plasma Cutter (eBay*): ebay.to/2JgTrK6
Rhino Cart Welding Fixture Table (Amazon*): amzn.to/3za8aez
Raw Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 3.0 License
creativecommons...
SAFETY NOTE: I do not recommend disconnecting any safety grounds. The connections that I isolated were connections between the plasma table frame and the earth ground that I inadvertently added to the system with my modifications. The plasma cutter, CNC electronics, computer, monitor, and plug strip are all still grounded to the safety (protective earth) ground through their power cords, as they came from the factory. I am aware that there are various documents recommending that plasma tables should be separately grounded. The Crossfire plasma table does not come from the factory with provisions or instructions to earth the table frame, and in fact doing so prevents the torch height control (THC) electronics from reading the arc voltage. If you are concerned about this, I'd love to hear about it in the comments, but I can't do anything about that design, as it is inherent to the Crossfire THC.
I guess a GFI (RCD) would take care of most of those concerns.
The solution (potentially) would be to opto isolate the signal feed/s from the table. But without seeing a schematic of the CNC height control it's hard to be specific.
@@boblewis5558 that is a very interesting idea. I don't see any reason why that couldn't be done, though it would require very careful matching of an optocoupler to stay in the linear range.
@@spehropefhany I bet a GFCI in this situation would be tripping out constantly from stray currents induced by the plasma cutter, especially now that you have two ground returns from the plasma cutter and the controller!
Telling the internet you did something wrong. Oh i'm putting on the popcorn and coming back to the comment section! Thanks for sharing and educating those open for that sort of thing..
As soon as you mentioned you went back to the carbon-reinforced materials, I had the concern that carbon fiber is a quite decent conductor. Evidently, the level of integration in your feed material, it's density, and discontinuous orientation of the cut strand is isolated enough that you didn't form a circuit with your part. Again the carbon version, like the metal is NOT necessarily - "more better". Running a composites shop, we are extraordinarily careful to keep fiber strands and dust out of the air as they eventually and inevitably find the magnetic fields of circuits and create shorts. Plain plastic is definitely more predictable isolation. ceramic, phenolic or even end-grain oak can provide modulus if you need them:)
Metal is more better has been my musical philosophy for 30 years 🤘🤘🤘. Great video as usual James
Love all the haters out there. I made a silly mistake with my table this summer.... I set an exhaust manifold on it to weld a broken bolt. This resulted in damaging the transistor on the motion control board for torch fire control. Didn't know I did until months later making Christmas presents. Guys at Langmuir helped diagnose and really took care of me. I love my little crossfire. Only wish I bought the pro for size!
Thanks for sharing the mistakes, I think I learn more by seeing them and it also helps reinforce the fact that people in videos are actually humans that make mistakes just like the rest of us 😁
Well you ironed that out really quickly. And I thought I was the only person who could improve things so much that they stopped working.
Fix it 'till it's broke.
Hopefully it didn't fry anything...
Just in case something (some sheet or else you are cutting) goes shorting the table to the post (which isn't far), I would insulate the monitor from the post (opening the holes in the mount and adding some plastic spacers should do the trick)
Don't bother making mistakes! Sure you learn things but we do too :) even more when you make the mistakes before we have the chance ;)
This is not a bad idea.
Well... I am glad I watched this .. lol I was wondering why I need plastic washers for .
"because Metal is more better" 6:34 yep, generally true... electrical isolation can be a pain though.
Carbon fiber is a great conductor!
Merry Christmas, James & family! Nicely done diagnosis, too.
I doubt it would ever cause problems after your fix, but you might want to insulate that mount at the monitor. The water in the tank probably gets conductive between cleanings, and that monitor stalk looks pretty close to ground. 73 de W4YBO
Splashing enough water to conduct is possible. Worth considering.
One thought: the printed sleeves; do they contain considreable amount of carbon fiber ? Carbon fiber is conductive in tself, so you might need to keep an eye how this evolves. Personally I would have used PE for the sleeves as enything containing carbon fiber can easily be conductive.
I was thinking exactly the same.
Same here, I've been toying with the idea of 3D printing some cover plates for a saildrive that I'm modifying to be electric. Would be cool to do them in CF-PETG, which I have an un-opened spool of on hand.
But, they would be bolted in to the saildrive housing against bare aluminum. It's on the inside of the boat, so not going to see much water, but definitely some moisture and salt.
The plates will cover holes from old hardware, so will need to seal against the housing with o-rings.
I'm worried that I'll get a corrosion problem between the CF and the aluminium and the stainless bolts..
I have made mistakes on projects in the past, but I am a narrow-minded ass with a big ego, so I am not sharing. Happy holidays!
Most problems are caused by solutions to other problems.
How did all of us genius-tell you what you’re doing wrong, know it all’s, not see this coming?!?
When I was a noob at electronic, I was so confused seeing the starter of my car having a single wire. I was like how the starter is working with a single wire. I didn't know the hole car body was the ground. :)
Bad chassis grounds are the bane of very many trailer lighting setups, where it's substantially harder to get a nice solid conductive path in say, a folding trailer....
I wish to you and your family Merry Christmas.
Finding those kind off short circuits in elevators always make you rethink your capabilities.
It seems at least one other person has spotted the same issue I did. My youngest brother is the senior electrical engineer at a top F1 team and been there for 27 years! The problems he has experienced and had to resolve due to carbon fibre composite conductivity are many and varied. Without going into great detail I should advise anyone thinking of using ANY form of carbon composite material to consider fully and test experimentally any possible issues that may occur. An example of one "trivial" decision such as to cut two separate holes to pass high current cables through a carbon fibre sheet rather one much larger hole is the potential for LARGE eddy currents WITHIN the carbon fibre to occur (just as in a steel sheet) & generate a LOT of unintended & unexpected heat and even fire.
James, you may be lucky with the carbon filament you have used but it MAY just be that there is NOT full isolation but sufficiently high resistivity to "isolate" the problem. A proper high voltage isolation resistance test from a tame sparky might be wise as a precaution.
There isn't enough carbon in this filament, and it isn't continuous. The good news is if it eventually breaks down and conducts, the torch will stop firing.
It is interesting since most pro level tables ground the enclosure to everything through star grounding. Hypertherm has a 12 page manual in proper “star” grounding of everything so does Lincoln. All main plasma systems call for a single busbar to each individual component including Cabinet, gantry and the Z to an earth ground/ Rod! It’s common to ground all these points then run it to a wall ground.
Langmuir machines are a horrible quality product.
They don’t use ANY (absolutely no) double shielded cable or drains and in fact.. they use standard printer cable with no db9 pas through EMI filters. The noise issue is exactly why they don’t implement safety products like limit/ home switches or e stops.
They don’t use any ferrites
They split wire all power lines, coms and grounds! They use no busbars or distribution blocks!
They use all (too long) 22 gauge non silicon wire with agin no shielding or ferrites between the controller and the cheep $10 Chinese drivers.
The don’t use proper terminals or ferrules, they just tin the wire! (Even in the PS!
They use cheep, unbranded and “Noisy” power supplies!
They use usb power to power the 5v and 24v lines of their older Chinese boards....pre Arduino uno board.
They shipped units with improperly rated torch relays for months! The original relay was also inadequate!
They don’t coat the control box in EMI/EMF protective coatings.
They inadequate ground the entire machine.
They use extremely cheep stepper motors.
They use undersized ball screws that whip like it’s a hurricane and many times arrive bent. They use no preload bearings/nut on the ends. They use cheep backlash nuts.
They use non EMI filtering AC panels mounts.
They use extremely cheep panel connectors know to arc, rot, crack, and melt with no ability to drain EMI!
They have customers use T-splice connectors on their plasma cutters.
Their marketing is the only thing they do well...like no other cnc plasma company that came before. They have a business model that sells inadequate products then UP sells continuous UPGRADES to its consumer base. Their forums are filled with idiots who’ve completely rebuilt their machine many times over but are still loyal to the brand like a cult. They used to falsely claim that their linear bearing system was proprietary but wasn’t.
Their water pans / water table can’t even fully drain, yet they know this, still produce it the same and suggest that each and every customer modify their own pan.
They white labeled “fire control” and utilized open sourced projects for their THC and main board firmware.
I was extremely disappointed to when you started pushing their machine. You know quality components and engineering! How many sub par machines do you think your channel sold? You fully comprehend EMI protection. You know the value of these sub par components don’t justify the MSRP if a 2x2 2 axis machine. Your peddling junk.
There are serious quality control issues with this brand! n in arguable fact.
Ok lets say you are right about the quality. Yet they do seem to work for most people. And we can afford to buy it. You forgot to mention what CNC machine we should buy. All your complaining and you left out the main point.
Question: Who sells a metal power strip that is not double insulated? Mind blown
Everybody. If the case is metal it gets earthed as its safety. If it's plastic it can be double-insulated and not earthed - since it can't be. The fact the case is conductive means it MUST be earthed since if a live wire came loose inside the unit and touched the case, it would make the case live and then the user would complete the circuit to ground next time he/she plugged something in. With the case earthed, the earth connection should be a much easier circuit to ground and hence take all the voltage. If things like an RCD are fitted, it will also trip as soon as the current flows to earth.
@@Cenedd Yeah. all well and good. until you get fried. Metal extension leads just seem like stupidest idea ever.
@@Cenedd Possibly "WATER-COOLED" would be even stupider... doubtless that is an optional extra
As others also stated… watch out for the conductivity of CF-reinforced filament. Depending on print settings it can become conductive. If i remember our tests correctly, the conductivity of the 3dxtech CF-Petg increases with print temperature. Stratasys CF-reinforced filament prints ARE conductive to a higher degree… i would replace these CF parts for pure plastic, to avoid indeterministic future issues.
( I would not trust the parts to keep their high impedance when exposing them to transient voltages, as the plastic so easily can alter its behaviour due to print settings.)
That said… i am a heavy user of 3dxtech esd-petg for electronic projects… as normal petg generates quite big levels of electro static voltages when breaking them free from the metallic build plate of the Prusas we use.
It's always interesting to see what sort of signals people put through the chassis. I've been rebuilding a cheap 3D printer that suspiciously tied the logic board to the frame, but fortunately it appears that was just a "safety" ground. (Note that the printer takes 19VDC in, and no grounding, so it's not earth ground.) The nozzle, thermistor, heater cartridge, and bed heater never connect to the chassis ground, so nothing ends up being mysteriously polarized, fortunately. (You have to check, because if one end of the heater ends up connected to the case and that end happened to be wired to the control board's positive terminal, then you blow a fuse, which is annoying. But, luckily they isolated everything, so it wasn't a problem.)
For welding and plasma cutting, often the table is used to pass the arc current. That's all that's happening here. Though it typically isn't a huge safety issue, in this case it's a problem of offsetting the small divided voltage being fed into a microcontroller analog input pin.
Something to keep in mind is that the heaters are generally negative switched, because N-channel MOSFETS have a low RDSon and are cheap, and it's also cheap to not use any kind of special gate drive magic circuit; so when the board connects GND to the frame, you're one fault away from the heater switching being bypassed and the heaters being stuck on uncontrollably. Mind you since there is nominal heater resistance present in the path, fuses, even if present and well-chosen, won't fire either. It's a sneaky fire hazard. The only way around this is to have PS_ON control, so an extra relay that the board can switch that would unpower the whole machine, and ideally test that it also works with watchdog timer.
Well with a Langmuir it’s a ton! They are riddled with EMI/EMF noise as they use no methods of insulation or isolation. If there’s a manual on how not to build or wire a cnc Langmuir wrote it.
James obviously "solved" his problem, but this video provokes more questions than it answers! Part of the confusion appears to be between electronics signal ground versus chassis safety ground and the implications of connecting the chassis ground of the control box to the chassis ground of the table. In principle, no arrangement of interconnected grounds should cause the plasma voltage detect signal to get shorted and read zero... unless.... OK, so the plasma table is basically a positive-ground system -- that is the table carries the positive side of the plasma power, and the torch voltage (to be measured) is negative relative to that. And yet Langmuir's detection interface "Voltage Input Module" (VIM) appears to consist of a simple voltage divider, possibly with some filtering. If it has no active component, then it will produce a negative output voltage (relative to the table). But that would require the Torch Height Controller board to accept and digitize a negative input voltage, which most analog-to-digital converters will not do.
That could be easily solved with an inverting op amp circuit on that control board -- but is that what Langmuir implemented? The alternative would be to float the electronics and attach the electronics signal ground to the negative output wire from the VIM, and feed the VIM positive output (ie: possibly the table frame) to the ADC positive input. If _that_ is how it's implemented, then of course an unanticipated connection of the electronics ground to the plasma table will short the VIM output signal, and make the TH control board see zero for the plasma voltage. That would be a very poor design, so I don't want to jump to conclusions, even though it explains the symptoms.
It would be super interesting if James would measure the signals from the VIM to the TH controller, and determine what their voltages relative to the table, and relative to electronics ground!
Meanwhile, judging by forum posts, some owners of this gear have other problems relating to noise propagation of one sort or another disrupting the controller MCU(s?) and/or comms with the PC. These may entail grounding or shielding issues, but are not the same as the controller running fine but seeing a zero plasma voltage. Further, connection of industrial gear to a PC over USB is a known issue in general, and greatly eased by providing an isolated USB port, at the cost of a ~$1 part, (though supply of those has dried up during the chip shortage).
FYI, you can measure negative voltage with any ADC using a resistor divider. You only need a positive bias.
@@MaX271 You indeed propose an alternate way to avoid having the ADC see a negative voltage. And that method is going to require a positive voltage reference and also depends on knowing pretty exactly the source impedance (in the separate VIM box) in order to combine it with one or more additional resistors to get a known divider ratio. Unless you add a local buffer, in which case you might as well make it inverting. My point was that Langmuir would have to implement _some_ strategy, however trivial, to deal with this. One of those strategies (electronics ground connects to the VIM negative) leads to the symptoms James saw, but again I'm reluctant to think they did that without further evidence
Whilst this is working the use of carbon fibre reinforced plastic raises a little concern for me. Makin the sleeves with a conductive element seems like a risk for some future hard to debug issue. I've not done any work with that material yet in 3d printing, mostly know it in terms of electrolysis risk in composite form where the structure is quite different.
No harm…no foul…good discussion
Merry Christmas, glad you solved your problem, no harm, no foul. A while ago, I fried my multimeter while diagnosing a plasma cutter; I mistakenly connected my volt meter to the high voltage output.⚡
"What things have you screwed up trying to be clever"... Think about most of life so far... Shuffles one foot, nervously....
Shit. A ground loop the long way....
During university courses in electronics and measurements we learned about ground/earth free measurement situations, and the cautions. .
Why a screen, computer would be have chassis to ground/earth? Poor design I would say.
Pooh, nothing got damaged. Very nice to hear. You were lucky having equipment that just shut down.
It's nice to deal with electrical stuff being "fully isolated" meaning having to protective ground/earth...
To Your final words....
In the end of the 1970:ies we built hobby computers, Z80, Texas 9900 etc. One hard thing was to get a monitor. Some guys used televisions and they had the chassis connected to one of the two mains pins..... That called for turning the mains plug to the right position to get the chassis connected to the neutral......
As soon as you started describing the problem my brain went, the vesa mount screws! You just solved the problem higher in the chain. Very nice, thanks for sharing, and like you I am impressed nothing broke. Given how I haven't seen any signs that say "do not ground" on the metal frame, I'm a little surprised the box doesn't do a similar continuity test on startup. (I'd be curious how good the isolation is that you have now: what at least the British call a "Megger" - for measuring mega-ohms resistance vs thousands of volts, which is part of at least their workplace electrical safety routine testing - would be an interesting test, since the voltage used for continuity checks is pretty low.)
The voltage used for cutting is also pretty low (under 100v) so it's probably not an issue. They give a cutoff in the docs of something less than 100 ohms. Don't remember the exact number.
I have a similar problem, maybe you can help. When I check continuity from the USB to machine frame It show continuity, only when the Plasma cutter is plugged into the DIV input. Do you know what would cause this. any help would be appreciated.
Great find and troubleshooting. Hopefully no damage will pop-up later.
Thanks for sharing the "Lessons Learned", hopefully it will help the rest of us not over look these small details.
My current project had a similar issue with a USB connector and a touch screen. The entire aluminum frame and enclosure is connected to mains ground. Turns out the OEM touch screen mounting pads were connected to the electronics ground which in turn meant the electronics ground was connected to mains ground. Had to use plastic mounting screws and washers to isolate it. Next was to make a custom USB enclosure surround to keep the USB ground from touching the case. No damage from being grounded but testing with the meter was kinda baffling at the start.
Whenever USB is present, you can assume that it will be earthed sooner or later, for example via an ATX PC.
Deductive sleuthing at its best !!! And how do I fix it with innovative reasoning. Good Job......
Any issues with the carbon fiber making the sleeves somewhat conductive?
I noticed you are using a 30xp and the thc upgrade. I am using the same table and cutter. I purchased the thc upgrade but I am not sure how to hook it up inside the 30xp any tips? Specifically what color wires to hook to.
do you think the manufacturer could feed the height voltage through an isolating adc/comparator ? instead of simply feeding it to the control box, which is the root of the problem ... afaik
That was amature level dumb, you're gonna have to dumb it down a notch or two, to make professional level stupid mistakes like I do....
Great Info. Thanks very much. Did you really say that metal is "more better"? LOL Just had to give you a bad time.
2- things i need to know from you.. your powermax 30xp, have you cut aluminum with it
and what is the max thick you have cut with it (any material) ..?
I skipped checking all corners on a cut and over an empty space, the THC started diving. Coulda been bad but the empty space wasn't that big.
Merry Christmas to you and family James. A year ago I built my own 3-axis Mach3 control cabinet and also built my own Z-axis for implementing THC on the Crossfire. In doing so, it encountered the same grounding issue, but in the form of crippling EMI at the controller. Once the box was isolated, problem solved.
thank you so much really appreciate the video having the same issue, hopefully it's the same fix
A useful gotcha to be aware of. Thanks for sharing
As a retired healthcare professional, kudos for your excellent application of diagnostic process.
Nice Christmas gift of advice from a bit of honest self-criticism. Thanks!
Hi James there is not enough lines on you tube to list all the screw up's I have made in my life time but that one is a very easy one to do and yes I have done it not on a plasma table as I don't have one, but it can happen on many different machines that have a computer attached. Often it is allot harder to track down probably why I don't have any hair left.
Not a big one but today I was finishing mounting a DRO to my drill press (long project) and after several hours work I turned on the display, moved the axis, but the reading didn’t change. I was worried I had damaged something during the installation. Eventually I realised I had not plugged it into the display! All working fine now.
Nice. Any chance on sharing the G-Code for the spacer?
Well, I'm just glad nothing expensive got baked on you in the process. I think we all have some ideas that turned out not so great in the long run, lol, we're all human.
dude , you're a fabricational rock star ...
Solid find and fix James! Thanks for sharing.
I can see how this is easily overlooked, I guarantee you’re not alone on this one James!
Yeah, for sure. There are a few forum threads about the same issue.
A video out of a couple of plastic washers? 8) still you are good at it and happy 😁
To answer your last question, Life. 😊
It's a matter of remembering where your wires are - and the fact the table is one of them !
Anything in a PC that measures anything is NOT supposed to be referenced to the ground in any way. Differential probe, galvanic isolation, take your pick - anything EXCEPT "well ground is also Protective Earth, is that going to be a problem...?"
You're not wrong.
This is a bit confusing. If what you say is true this means in operation anything connected to the return lead of the torch is not at ground potential. This sounds potentially dangerous. I think the problem really is related to the way the torch height voltage is measured. Got to look at how my torch is wired tomorrow. Planning on mounting my plasma torch to my CNC router table.
On this table, the cutting circuit is floating, and not ground referenced. In practice, the table probably ends up floating around 5V above the earth ground, due to the way the torch height control divider circuit works. As others have pointed out, this wont be a safety problem unless something else goes wrong. But it could become dangerous if the torch lead were shorted to ground, for example. I would not have felt comfortable designing it this way and selling it.
I spotted a multimeter that I also own
Suggestions for your plasma table from experience with mine. Put it in a separate room and/or build a downdraft vacuum system to evacuate the particulate while cutting. Even with the downdraft you WILL get top spray that will accumulate in your shop unless you set a vacuum nozzle at the cutting tip. For best results put it in a separate room and build a downdraft for it. Just saying. Enjoy your videos
Yeah, if I had more than a bay and a half of a residential garage, I would totally do this. I would also have a separate room for grinding. As it is, the table sees occasional use, wheeled out of the corner to cut something, drained and wheeled away to store.
@@Clough42 I was just going to edit what I wrote earlier. I had a 2100sft shop and noticed even with the downdraft it was getting over spray in the surrounding area. I realize you will do occasional cutting so please disregard my previous post. On a side note, I agree with your mistakes video. I should have a PHD by now from the mistakes I've made and learned a lot. I just resubscribed to enjoy your content. Have fun.
ah, yes, the goods old ground through a data cable. Learned that one the hard way on my first project at my first job. We made a data capture box for Buses (that they could plug in to the bus PLC, capture acceleration/deceleration/speed to analyse driver performance), spent 4 months on it, worked flawlessly. we deliver it to the customer, they come back saying it didn't work. We try it, it works, ok, check all the cables, everything is fine. I go, with my boss to their office, we plug it in, it doesn't work. I get my laptop out, it works. raaaah. turns out, their laptop power bricks passed ground over to the positive pin to the computer. When plugging in the serial port, while the capture box was connected, the ground from the db9 between computer and plc would create a ground loop and crash the electronics.... that was a fun one.
Wow. That's pretty bad. I guess as long as everything you connected was floating, you'd never know the whole laptop was below ground potential DC.
Just found out I have a McMaster-Carr in Atlanta and its only a 30 min drive for pick up. Thanks for introducing them to me. They are awesome!
Indeed. Their shipping is pretty fast and reliable, too. Not cheap, but when you need something, it's good quality stuff.
I'm really curious now if that did in fact fix the problem because our plasma table at work (which has THC too) has the controller box grounded though the table and through the ground pole of the power plug.
One thing to note though that might be a factor here is that per the manufactures instructions, our table is directly grounded to earth via 10ft of ground rod driven into the ground next to one of the legs of the table and its connected to the ground connection point on the table where the plasma work clamp ground connects. I know that there was a note in the install instructions that said that without that true earth ground, there will problems with erratic THC control behavior.
The THC design is probably fully isolated on that unit.
Thanks for this video !!!
thank you for this video
Isn’t carbon fiber typically conductive?
The fibers are, but the composite of chopped fiber in PETG isn't.
Excellent tip. I should be getting the THC upgrade to my older CrossFire in a few days and I'll try not to repeat your mistake. Merry Christmas. Ciao, Marco.
A nice, simple fix to close out 2021 😊. Happy New Year, James! 🥳
Cheers!
Any reason you went with the Avid CNC router instead of the Langmuir mr-1 ?
The MR-1 didn't exist a year ago when I placed my order.
It's true, metal is more better.
So not only electric chairs, but also electric tables. Interesing. PLEASE measure voltage between metal and ground while you are cutting.
Are you asking about safety? This concern would apply to any plasma cutter, and the plasma cutting process in general, independent of the CNC controller. The work clamp will always be in electrical contact with any welding or plasma cutting table, car body, etc.
I'm about to add THC to my crossfire. Thanks for the heads-up. I'd much rather avoid a problem than spend a day diagnosing it as I pull out my hair.
Curious what the resistance is of the Carbon printed parts.
My thoughts exactly! The carbon fibers are definitely conductive, whether they're close enough together in the PETG to cause a low enough impedance connection is the question he answered in this video😋
It reads open circuit. The fibers are not continuous.
I thought the same thing, while simultaneously thinking carbon fiber embedded in plastic likely not a problem. The conductivity of the carbon fiber is probably measured in nanosiemens and is irrelevant. Though I'd keep it in the back of my mind if things get wonky in the future.
So proud of your honesty and willingness to share ... I am living proof that dumb mistakes only get worse as your age increases ... sorry ... But maybe you will break the trend... Thanks for Sharing ... Stay Safe and Merry Christmas ...
"What kind of things have you screwed up by trying to be clever"... might be the best question I've heard all year😆. Great vid as usual James!
Oho so sad 😔
..to answer the last question, what kind of things I've screwed up trying to be clever .. I'd quote my good friend Thanos; "Everything"...
And I thought you were going to show us a Fusion add-on for Pareto Analysis… maybe next time?
Two givens for my projects> 1. Its always Something… 2. And usually not exotic.
When I was very very young my parents had some decorative candles for the Christmas tree. And, in my mind, candles should burn, no? Otherwise they wouldn't be candles. So I "fixed" my parents' mistake... And then I couldn't sit right for a week...
No way would i mount a power strip right under the edge of that table. overspray, condensate etc will drip right onto it.
The ground gremlin strikes again - thank goodness it didn’t damage anything. At first glance, I thought you had epoxied your hand to your forehead - hate it when that happens 😀. Really like the subjects you cover.
Um, as an electrician, I sincerely hope you’re misinterpreting Langmuir’s instructions. Earth bonding (grounding) is a safety practice, and a large body of metal like the plasma frame being unbonded can lead to electrical shock in case of ground fault. In your shoes, I’d see if you can properly bond the the frame to ground, and then perhaps isolate the computer ground from the controller with a USB isolator.
Admittedly I don’t know the code in the US, but Langmuir’s instructions on how to float a laptop on the arc voltage sounds like something that’d not pass code in any residential household, according to what code I’m familiar with. I’ve never done industrial work, where this MIGHT be allowed, but then again this is a residential garage.
No. I'm not misinterpreting. The computer and metal electronics enclosure are grounded. The table is part of the cutting arc circuit, so is floating. Here's the documentation explaining how to make sure the table is floating: www.langmuirsystems.com/thc/guide#grounding-section
@@Clough42 Is the controller earth-bonded? They want you to check continuity from the USB shell to the gantry, but there's no word on whether the gantry is earth bonded. It'd make me REALLY uncomfortable in a residential setting to have that whole thing unbonded to ground, and since you CLEARLY have ground creeping in someplace, you're now floating your entire machine frame on the arch voltage during a cut. I dare say this wouldn't pass muster on any residential code I'm aware of.
Note that arc welders and - I assume - plasma cutters are galvanically isolated from mains, so you can have your machine frame bonded to ground while running the plasma current through the frame back to the plasma ground connection. Laptops are generally also powered with class-2 power supplies, so they're galvanically isolated from the mains (and ground) as well. The safe way to do this would be to bond the machine frame to ground and measure the arc voltage from the tip of the plasma cutter. That way there's no shock hazard anywhere...
@clough42: You definitely did nothing wrong assuming the computer/monitor chassis can be bonded to the metal frame of the PAC. In fact - generally they should be explicitly tied together (your “local” code may vary - I’m not that much on par with UL). This is either an anomaly or a shortcut on the voltage divider unit designer side - which in this application should incorporate a differential measuring circuit. Grounding is indeed one of the more headache-inducing fields in EE, but assuming potential separation between two metal boxes in close vicinity in your design is not only near-sighted, but wrong and potentially(ha!) dangerous. “Over here” all accessible conductive surfaces must be equipotentially bonded (and bonded to PE) and I assume the regulations are more-or-less the same in the US.
Merry Christmas
I dont want to be negative, but i think this is the wrong solution to the problem. The THC electronics should be produced in a way where this wouldnt be a problem. Now the machine frame is (theoretically) not grounded, and is a potential hazard if mains power shorts to it somehow. I dont know how regulations in the states apply but in my country there are actually specific rules for machine construction that say the machine frame is to be grounded to the same ground that supplies the machine. Not saying youre in the wrong here, just that this might be a potential hazard and maybe the design flaw lies with Langmuir? Again maybe things are done differently in the states.. Otherwise great video!
Yeah, this appears to be the way they designed it. I see your concerns--especially in the case of an equipment failure of some kind, but there's not much I can do about it if I want the THC. I want the THC.
The THC is a must of course! After all the odds of something going wrong is minimal, i would probably done the same if it was a private/hobby machine, but at work we follow the rules. In this case i doubt there will be a big problem. Happy new years!
High frequency of plasma unit can be pretty bad on electronics. Good thing you didn't fried anything because sometimes it does that to boards....
I fried the PCM on one of my customers Jeep once. I was using a Power Probe Hook to test a crank sensor and needed both hands to place the leads. I had wedged the probe between the heater hoses so I could see the screen. I had a coworker start the engine. As engine started, it shook the heater hoses and depressed the voltage button. It sent 12V to the PCM via the signal wire. $800 later, I learned to never use a Power Probe to check for sensor signal.
This looks like a nightmare to find the short, but fortunately you didn't break the controller with this high voltage. I once connected 20V to USB data through a faulty stepper driver and it broke a few USBs on the motherboard and the computer has some trouble booting now
How long did it take to chase that down? Myself, I'd have pulled most of my hair out before finding that issue.
James Marry Xmas to and the Family.
On my plasma system I put the work clamp to the sheet that I am cutting I did the table thing and get ark spikes that would shut the computer off it was isolated system I seen my friend system in his shop and ask him about how he was connected the work clamp to the steel directly and then he said the spikes would reset or shut down his computer. The light bulb 💡 on yaaaaa since then my table been running flawless for 10 years.
Great info
Great troubleshooting James - thanks so much for showing us the problem and solution! Merry Christmas
finally got lined out on my plasma cutter i have been 3 weeks, burnt up one vim box and your videos saved me. i was ready to give up and sell my machine.
At first I thought you were going to say that the grounding issue fried the computer. That would have surprised me somewhat since industrial computers are more resilient than the consumer PC, but there is always that one instance where something goes south and the magic smoke leaves never to return again.
I am sure that this video will help people well into the future.
repeating Mathieu B, I would use something other than the carbon fiber petg, it is not conducting now, but I think you maybe lucky, and wear and tear may eventually make electrical contact.
There isn't enough continuous carbon to conduct well enough to be an issue. It would need to be under 10K ohms to be an issue.
Happy christmas, and i din't know that earth thing thanks for the warning🎄🎇✨👍😎
I would have thought the plasma cutter voltage would have had a full isolation measurement.
I think this is a case of it being a simple hobby-grade controller.
I think that would be ideal. Even a non-isolated differential probe would probably solve it, also.
Not that I’ll ever own one of these, I’ve been a very long time retired however, a video well done sir, and thank you, I enjoyed watching it and learn something
Mistakes the world’s greatest teacher . Thank you for teaching us through your mistake.
To me this "fault" is more to do with table design. The manufacturers have no control over the numerous environments likely to be encountered up and down the Country where inadvertent grounding conditions might be encountered, as you found. Thankfully with no lasting consequences. Electronically, no circuits should ever be engineered where conflicting grounding conditions may be encountered; the potential (pun intended) for accidental circuits is enormous.
You're not wrong. :)
Very good point, but I'm not sure what the solution would be. As far as I understand it, Plasma cutters have to use positive "ground".
But, only the slats that the material is placed on really needs to be connected to the plasma "work ground". If it could be isolated from the rest of the frame, the frame could be grounded. But, that would of not be feasible since you would then have a positive from the plasma and actual ground easily accessible on metal parts next to each other.
So when the plasma is running, you could easily get shocked by high voltage, high current, would easily kill you. My hand held plasma runs something like 250V and up to 50A if I recall correctly, at least its a lot higher voltage than say a TIG welder.
So I'd imagine that it is a fairly hard problem to solve..
@@Robinlarsson83 You could touch the ground and 250v of the plasma without problems. the plasma cutter is isolated so the 250v are floating voltage. if xou touch it together with ground the 250v are now 0v and on the other side -250V. Best regards