Oh shoot... You are right. Should be mm/min. for the speed. I have pinned this comment. With videos this long I usually miss something obviously like that.
I hope you guys enjoy my CNC plasma cutter build. It’s a long video, but there is a montage of some of my completed projects at the very end that you won’t want to miss!
Great vid! But i nearly swallowed my tongue when you said they wanted 1000 dollar for a custom tray. I work in a metallshop in Sweden, we do steel, iron, stainless and alu parts in plate, small series or oneoff´s. A pan that size in stainless would cost us at the tops 200 dollar to manufacture. Add for profit and such, but at the end it would not be more than 300-350 USD for the customer. Thank you for a fun video, and enjoy your new toy!
Klas, Yah I live in a bit city that is rapidly expanding. Half of the fabrication places I contacted wouldn't even consider doing a one-off part... I wish I lived near your company
@@DrDFlo Even here there are not many oneoff workshops left, but we have a steady customerflow of regulars and newbies wanting everything from decorative metallcoverings to staircases and machineparts... Hope you find a metallshop in your area that likes to do a bit of trade with others than big firms, and doing it for a reasonable amount of money and not that 1000$ bullshit! That´s just greed!
while school systems are different, consider speaking with your local tech school welding instructor. that is an easy welding project, and if you supply the materials, most schools will do the work at a very reasonable rate. (last time I had some welding done (sealing the leak in the fuel tank of my motorcycle) about 6 years back, it cost something like $10 per semester to have "access" to the programs, and $5 per project to get the welding done. (I am planning to put them to work again in the not too distant future building several parts for my truck such as a custom cargo rack (mostly because I do not have a tubing bender, or a place to set up a welder)) the students generally enjoy getting to work on real world projects instead of just welding scrap metal. Keep in mind some projects might involve more than one "work order" meaning they can tack on additional fees for services... this can add up fast when you need something like the machine shop making 5-10 parts for your single project. --------------------------------------------------------------------------------------------------------------------------------------------------- it is true that most fabrication shops do not want to mess with one off parts...it is a lower profit margin. after all... that might take someone on a metal brake all of 10 minutes to make the 4 bends in a cut piece of material. someone is going to spend maybe 5 minutes per corner setting up and cutting the corner notches, then someone is going to spend another 5-10 minutes with a tig welder sealing up the corners. on the other hand if they are doing batch work, they might be producing 10 parts an hour for someone ----------------------------------------------------------------------------------------------------------------- on the other hand, the you's and me's, we would just cut 4 strips and weld them to the base, maybe even weld in a train bung if we make the pan out of mild steel, we maybe have $80 US invested in the materials, and we spend an hour or two laying out the pieces cutting them, welding them, testing for leaks, fixing the pinholes we missed :)
I had the same problem with the noise resetting my controller. I tried everything you mentioned, at the end i made a box for all the electronics separate from the machine on the left side, and the plasma cutter on the right side. No such problems since. I use a cheap plasma cutter. I Hope this helps someone to not go through a month of troubleshooting and fried electronics :)
Great video. We built a 4'x9' CNC Plasma cutter about 4 years ago. We thought about building our own controller or going with one of the cheap ones. We went with a commercial quality controller and did not have any problems with EMI. The torch cable is in the cable tray/chain. We milled the torch holder out of aluminum and use a couple of magnets to hold the torch holder to the z-axis. The HyperTherm 45 will cut up to 5/8" decent but when we cut thicker stuff, we can mount an oxy-acetylene torch on it. The HyperTherm has voltage sensing so the torch height is adjusted automatically while cutting after the initial pierce. After the initial build, we made the gantry higher and mounted a chuck on the table so we can cut up to 6" diameter pipe. I guess we are going to build a CNC router now.
you could have a ground loop issue. Make sure each piece only has a single path to ground. Unplug all the power cords and make sure there is no continuity between grounds on different plugs. Otherwise, current can run through one cord, through the wall, and back up another cord. An often forgotten path is through the USB cable and through the computer, so buy a USB isolator. Another possible unwanted ground path is on that torch trigger, so check that circuit. Elsewhere, use plastic or ceramic mounting hardware, or relays and optocouplers
You should still construct a water tub as a spark arrester below the bed. It should be slightly slanted so that the metal residue collects in a corner and can be easily cleaned.
Put your ground rod outside wherever possible & run the lead to the machine. You are the first person to discuss the need for a large air compressor & the duty cycle of the plasma cutter, WELL DONE!
Hi David, I just wanted to say thank you for sharing this video. I appreciate your honesty in sharing your mistakes and subsequent learnings. You have saved me from making many of the same mistakes in my own build. Well done and keep them coming!
Okay so I just bought my first plasma cutter. Nice unit but it was broken when I bought it. The gentleman I bought it from could not figure it out. I took my time downloaded all the manuals and tracked all the systems. It only took me two weeks to get this one working. Plasma cutters are usually pretty simple if you happen to buy a used one that's not working. Make sure it at least turned on and air flows before you buy it though. If it does not turn on most likely the main board is bad and it's not worth the used price. I picked up mine for $200 and it's a PCM 750i so this is the one I will be incorporating into my table! I'm glad to see a lot of people building these things. I have access to one all times it's such a massive machine though that I'm not cutting any steel that size I've seen him cut upto 4 inch thick Steel whereas my machine will only do about an inch-and-a-half maxed out! Another word of advice is run about 40% of the duty cycle to keep the longevity of your unit
HI all, Really nice build. I had problems with the noise also. When I went and used a laptop instead of a desktop, no noise problems and I have been up and running for about 6 months now. Hope this helps anyone that is stuck. Stay safe out there.
I think this may be related to the fact that the desktop is sharing the same electric network as the plasma cutter and the noise is coming from there, maybe adding a UPS to the desktop might help filter the noise
I had the exact same issue with a Mach3 home made cnc and a cheap ebay 50A plasma cutter, I had no previous experience with plasma cutting and spent about 6 months trying to isolate the circuit from the EMP noise, never could, ended up selling the CNC machine with a spindle cutter, now thinking on getting back on the plasma cutter thing with a more robust and expesive plasma cutter, Thank you!
Wow you should be an instructor. Clear, concise, thorough with the right amount of hands on and theory. It would be exactly what I need if it had a rotary axis for notching pipes. As it is I have been looking at JD's Garage DIY build. Your build is much nicer but lacks this 1 critical feature that is a must have me and lots of other people.
This is the absolute gold standard on how tutorials should be made. Informative, thorough, well researched, clear, easy to follow. It shows what went wrong and how it was fixed. It shows different alternatives and the factors to consider when evaluating alternatives. I don't think I have ever seen a better designed and presented tutorial. Well done.
Regarding a 3D printed torch holder: Consider printing a mold for the part, or casting a urethane mold from the original PLA part (well, an unmelted version of the part). Cast a new part in hard silicon. The silicon part (being a thermal set rather than thermal plastic) will be much more robust mechanically and thermally.
Wow, I don't usually comment on videos but this build was amazingly informative and made me want to hop over to my CNC and start tinkering. Keep up the great work, you've got me sold on modding my workbee!
Awesome video. You might want to try placing the controller in a metal box that's grounded. It depends on weather the noise is radiated or conducted. I would try that first and move on from there. Keep up the great work!
Noise mitigation Easy - have a separate cord that just connect outlet ground to the water box. Have your power supply plugged into a filtering power string and the box connected directly to the ground of your outlet (possibly a plug where only the ground is connected). Also separate power supply from frame with an insulator (e.g. mount to wood mounted to frame) Medium - some power supplies have dc- connected to AC ground. This can be tested with a multimeter with everything unplugged measuring resistance between negative terminal and ground terminal of power supply. If there is a connection get a better power supply (I think meanwell have isolation, based on online diagram). Hard - what you really need to do is isolate the ground of the processor from the ground of everything else, which is something that would have to be in your controller. A 5v dc-dc isolator (b0505S-1wr20) is a couple bucks used in CAN bus electronics. But you probably can't do that reasonably.
im sitting 2 hours watching your videos and its 5 AM now, i only can say wow, you are a smart guy and i woul love to try some of the project you do anytime soon! thanks for the projects :)
Hi, I don't know if you still want to try it because it has been some time you posted this video already and you bought another machine but... I am an electronics engineer and have worked with EMI for some time. What I would try first to solve the HF interference issue is: -Route the plasma cutter negative cable as close as possible to the torch cable. If possible moving the ground connection with the cutting head to minimize the loop. -Adding a choke around the output cables instead of everywhere else can be helpful as well. -Try to shield the spark gap in the HV generator inside the machine. If you or anyone else reading this comment try any of these, let me know if it helped. Nice video !!
There are plenty of easy to print heat resistant filaments out there. Annealed PLA is pretty awesome but there are polycarbonate and nylon blends that have desirable properties that can be printed on a relatively ordinary printer. Overture easy Nylon and Polylite PC mainly come to mind but there are others. ABS is a very good option for building mounts, brackets or other structural parts, it's fairly heat resistant, it doesn't deform as bad or has a tendency to crack over time like PLA or PETG does. Then there is ASA which is very similar to ABS but is more UV resistant, and assuming it doesn't go above 100C would probably be perfect for a mount.
hang the cables from the ceiling and use swivel support arms and weights to maintain tension... using weights allows for 'floating' movement instead of like fighting a spring when the changes in head position occur.
Hey, it would be really great to get a follow up video of the cnc mill, maybe show us some cutting footage and talk about how it’s working out? Really nice work on the plasma table too!
Stay tuned. Definitely more CNC mill content coming soon (not the next video but the video after). A couple of quick updates, I’m using the ER tormach tool holding system for the R8 spindle, which has been great. I also switched to the Hallmark ITTP probe which has been way more reliable than my previous probe. I have mostly been buying tools and sharpening my skills over the past 6 months. Let me tell you buying and converting the mill is not the most expensive part... the amount of money that I have invested in tool holders and end mills is way higher than I thought. Definitely an expensive hobby. Still very happy with my purchase. It’s crazy the different parts I can churn out of my garage between the plasma cutter and the mill.
When you start the torch, it will splash water up and conduct thereby bypassing your insulation. Try putting a sheet of acrylic under the pan and then you can connect the ground clamp to the pan and this isolates the pan, material rest, and the material from the frame. Also the problem of air delivery can be fixed with an auxiliary air tank and a regulator on the output. More quantity can compensate for consistent delivery.
You should make a cage anywhere possible which should aid somewhat at keeping RF frequencies at bay however your main issue you are having here is that when a spark is made you get a pulse of EM radiating outward from that pulse. When the pulse stops the EM field collapses which results in it inducing currents in anything conductive in its field. Do this super rapidly and what you get is electromagnetic frequencies in the RF range pulsing through conductive materials in the field around the arc. Normally one would shield using a faraday cage but in this case it is not that simple really. It is likely the Arc the cutter is using is a purposefully created RF frequency making your problem a bit worse honestly. I recommend grounding everything with chokes everywhere, more importantly of a variety of different sizes. If you can find out a rough range for the frequency that your cutters circuit is using it would help you fine tune what chokes you need but it is best if you have chokes that do not resonate with the circuit so that as a magnetic field is pulsed into the choke its collapsing field counters that of the circuit. The Chokes will help anything that is forced to be in the circuit with the cutter or at least close but without a doubt you should place metal mesh around your electronic so that EM fields can not get into them. Perhaps a full on metal box around your micro controller and camera you are filming with to block as much RF as humanly possible.
Absolutely love it! To help you lead screw last longer you can add a dry lubricant as it does not bind with dust and other particles in the warehouse environment and lubes. I have seen these screw painted with a dry lubricant as well, however, I do not know what the process is. Nevertheless fricken awesome dude, you damn well read my mind. Just got to figure out where I can put it lol.
Hi Sir, I am fully new for CNC plasma , | have set a CNC plasma after start to work then CNC X ,Y Z are vibrating ,when Plasma switch off then vibrating is off. For vibrating what to do solved this issue , Please suggest me .
I would suggest using 0.1 microfarad disc ceramic capacitors across any leads going into the microcontroller. If you have a pair of wires that go out to a switch or something just put the capacitor across the input to the microcontroller otherwise you can run the capacitor leg with one to ground IE frame and then the other one to your input. The correct term for this is bypassing and it will shunt any high frequency AC electrical noise to the common ground for the machine. You can also get a bigger ferrite donut like a ft-240-43 toroid ferrite core and wrap the leads of concern several times maybe even up to 20 times if you have the wire length through that core and it will impart enough inductive reactants to that system to effectively choke off any high frequency currents going into the microcontroller. You might also be able to do the same thing with your work clamp lead and if you have a long enough plasma torch and a big enough ferrite core you could even wrap those conductors several times through a ferrite core in order to disrupt that circuit from a high frequency noise perspective. For more information look at just about any ham radio handbook or a specifically the ARRL amateur radio handbook for more rfi/emi solutions. Your other concern about properly grounding the machine might be able to be resolved by buying some wide copper strip or someone in copper braid from Amazon run this wide conductor from the frame of the machine tied into the frame of the plasma cutter through the shortest and widest lead possible, and then run another heavy strap or 1 in copper wire braid from the frame of the plasma cutter directly to the breaker panel behind your machine. Ideally you would tie it into the ground bus bar however you could also clean the paint off of the frame of the cabinet and secure it with a nut a bolt and a set screw type terminal. Wide and flat conductors are required due to a phenomenon called The skin effect whereby the higher the frequency of an ac voltage or current it uses a smaller and smaller cross-section of any given conductor sometimes as low as only a few thousands of an inch. Do this phenomenon a 1 inch copper or aluminum pipe for a grinding bus bar is more effective than a wire with the same cross section. Of course you could also use a strip that is 3.14 in wide to get the same result. I wish you the best of luck if I can help anybody out Google my business name and give me a call I'll be happy to help you out with issues like this.
I don't know if this is elsewhere, re Compressed air, you do not want "contamination" as in oil or particles coming out of the plasma nozzle, it must be clean air. I use a "oil less" compressor. I added a "air filter", "dryer", and "de oiler" at the output of the compressor. I also use a "as short as possible" & "dedicated" air hose, meaning it is only used for the plasma cutter. And if you can use a 3/8 air line from the compressor to the plasma is better (more volume!) Additionally, I added two "portable" 10 gal. air tanks to the line. This addition is way cheaper (about $100) than it would be buying a new and very "large" compressor. It adds volume which increases the duration of "air time" to my unit, it also helps even out the "pulse" when the compressor kicks in and out, making for a more "even flow" of air. I'm using a Lincoln electric tomahawk 625. The air compressor is a Porter Cable 3 HP, 4 Gallon. This setup works easily for a 4 -6 foot cut in 1/4", and up to 8 foot in thinner material, which covers just about everything I do. If required, you pause in a cut and allow time for the compressor to recharge for longer cuts.
Excellent video, thanks a lot sharing it with all of us ! I am thinking building a Plasma CNC cutter since one year and watched tens of youtube videos on that topic. I discover openbuilds with your video and actually, it makes me changing all my initial plans... You brings each key principle of such a project with a very nice structure and simple words. I will watch it one, two or maybe more times ;-)
With regard to the noise issue, there are two ways to induce noise into a cable: Electrostatically i.e. capacitively and magnetically. The insulating of the table from the work piece reduces the electrostatic portion, but the magnetic part is made up of a loop which is bounded by your cutting current cable and your grounding current cable. The greater the area enclosed by the separation of these two cables the greater the inductance and thereby magnetic coupling to other devices cables. It also makes the electrostatic worse because the ground cable becomes more inductive, allowing the workpiece to radiate electrostatically as well. To reduce the inductance run the ground and the cutter cable as close to one another as possible for as long as possible(their magnetic fields will tend to cancel one another). Only let the ground diverge from the cutter current cable for as short a length as possible to connect to the workpiece. It will massively reduce the inductance and that means that the ground cable and workpiece will act more like a ground. Hope this helps.
1) Use the earth from your garage mains electrical system to earth your table, use some nice thick cable. 2) Use twisted pair for your signal wires. 3) Use a braided shield over those signal cables (the sort of braid you get in coaxial cable, in fact you can strip some coaxial if you have any to pull the braid out and use it), grounded to your 24v power supply gorund, at ONE END OF THE BRAID ONLY, insulate the other end from anything else. 4) Move your blackbox away from the cutting area if you can. 5) Make your motor cable runs messy, straight neat cables pick up crosstalk very nicely, so within the cable tracks, make sure things are all a bit higgldy piggldy and mixed up. I know it'll mess with you, and feel wrong, but believe me one of the best ways to avoid cross talk and interferance on cables is to have them in one huge knotted mess, it stops big EM fields from forming properly. 6) Look at your configuration and think "earth loops" - multiple paths from a given point to the same earth are bad, remove as many as you can. Insulate your motors from the frame if you can. 7) Insulate the water table from your frame, not your work piece from the table. 8) Dont use T taps, there are way better options out there. Now sit in the corner and think on your sins. :-) *Your sins are absolved for sharing your hard work with us. Fingers crossed those steps will make your table useable.
You are using a printed clip to hold your slats into your water tank. Cut the slats narrower and shorter with a little bit more material to space them further from the tank sides, then use a non-conductive spacer to isolate them from the tank bottom. This way you can place your parts directly to your slate and do away with the wood isolators. Also isolate the tank from the frame.
you can also twist your dc lines from your e-box to create a small capacitance. Also, when using your chokes on your lines, make a loop in your lines then clamp your chokes on the double part of the line. On top of all that, you can used braided shielding on all your power and ground lines, and take that to chasis ground. Good golly, if all that doesn't do it, you'll have to do something with the source/head. I work with RF and dc in close proximity all the time. Those are the basics without using active circuitry to cancel signals. Do you have a scope to see the noise?
F****** awesome tutorial, very informative and well structured....Prob one of the best tut i have watched on youtube. Wel done, always excited when i get to watch your stuff. Keep it up
The thermal dynamics is a good machine, however if wanting to stay with a cheaper cutter there are many examples of machines similar to your first , that have pilot arc. I use a ltp5000 which is an inexpensive pilot arc plasma. I run the torch lead in the same cable chain as the stepper cables and it works with no noise issues on my diy cnc plasma. Most of these use a SG-55 or AG-60 torch. There are cheap straight versions of these torches.
Recommendation, insulate the water table from the machine fram, print PLA or PC spacers that separate it from the frame itself and you should be more capable of using AC start on that plasma torch, that way you can also just ground the water table to the Plasma generator rather than reply on a clamp to the work piece. I recommend also adding a drain port to the water table
Just go with a hypertherm power max……no electrical noise to shut down computers The biggest thing is grounds everything needs to be grounded ….everything you need a grounded rod in the ground close to the machine outside Large copper ground coming into the building where the plasma table is located Hypertherm sell a machine torch When I built my open build ox table years ago I tried a Chinese plasma table a simadra 50 amp I’m still using it for hand cutting That little $300 plasma is bad ass That being said nothing compares to hypertherm Nice video thanks for sharing
I was curious about the long cuts you make making the fire-pit. Was the dross/cut quality as good as you were hoping for and did the air supply/duty cycle hold up so you didn’t need to take rest breaks to either let the plasma cutter cool or the compressor catch up by filling the tank. Really good video. Not just an excellent tutorial on building the CNC but also on plasma-cutter selection. Looking forward to your future endeavors. Cheers!
I suspect the noise issue may be caused by your DC common being connected to chassis/Earth ground. This is an absolute no-no in electro-mechanical design. Isolating DC common may be difficult as some of the attached components may have the DC common connected internally to its chassis, which when mounted to the metal assembly connects it to Earth ground. If you can't isolate all the components from Earth ground, the next best solution would be to isolate the work surface as you had done. Unfortunately, in this scenario the assembly itself becomes a large antenna, so it may not improve the noise situation sufficiently to avoid problems. Hope this helps!
You could run a ground wire similarly to the way you would for a portable generator. Connected to the outside (usually on other side of breaker box wall) grounding bar that has multiple connection leads or to existing ground wire.
Re: the compressor air flow....could you simply add a second compressor (identical to your existing unit) and 'T' the output lines into the cutter? I have a California Air Tools compressor and I love its performance and quietness, so I understand not wanting to change. Good luck.
FANTASTIC! Thx for putting in all the obvious effort/hard work (on machine, and vid). Have been wanting to get into plasma cutting for a while. This vid is packed w/ valuable practical, and theoretical info. Great stuff. Thx, again. And, keep up the great work! P.S. I'm a regular F360 user (even teach it), so I really appreciate the F360 section at the end of the vid.
Giving your slats a slight bend makes them less likely to wobble with gantry movements. Also makes it less likely to line up a slat with a straight cut causing cut issues.
this is absolutely fantastic build ....some plasma cutters put some kind of leather or a brush or something around the cutting head to reduce splash !!! that thing splashed water all over the place and that cant be good for the bearings and bolts and stuff you can see it in the plasma cutter of colin furze ......also i think you should add more water coz there is so much fumes coming from underneath the work piece and the thermal mass of the water you have will heat up fast and evaporate ...more thermal mass will surely help im sure you will bow our minds off in the next video ...keep up the good work
So I'd honestly try gluing a rubber mat to the outside of your water pan. On the bottom and up the sides. I'd also probably bend and weld my own pan. A DIY press brake to bend thinner sheet metal is an easy thing to make.
Can you use a Y adapter to have 2 different air compressors work together? Each compressor supplying ~4cfm at the same pressure into the adapter, giving you ~8cfm out the adapter. Or is this not a thing?
Yes. Depending on the type of pump (s) you might need check valves so as to not have the teed pressure overload one pump or the other during startup, but most respectable compressors will have a head relief valve that blows off that head pressure so startups are softer...
float all the grounds on your electronics, and at the source ground, use filter capacitors big enough to filter out what HF noise you get (one on each neg and pos line should do it). These caps have to be on DC lines into your electronics, presumably after your rectifier which is after your step down transformer in your e-box. Don't use electrolytic caps as they are polarized.
Looks like it would be easy to run some tape (maybe metal tape used on insulated foam boards?) over the channel the lead screws are in to avoid any further chance of debris getting on the lead screws? Thoughts?
I'm only about half way through the vid but before I forget. If your entire frame was the noise conductor? Why would you not just mount the blackbox with an non conductive buffer material like a 3d printed block and spacer blocks for your wiring and shielding?
Depends on your table size and weight of the gantry. For really long and heavy tables, the belts will stretch slightly, which will affect accuracy. If are making a small table then it should be fine.
I’m no expert in high frequency but what I would do is use wood to separate the power supply from the frame. Make a faraday cage around the black box . Also I think you need to put a loops in your wire with the faraday shunt. Use a rubber mat between the water pan and the frame. Keep the plasma cutter wire separated from the frame. Maybe some kind of shield on the plasma cutter wire. Do not wire tie to frame.
A simple solution to the problem of the positive connection to the ground terminal remove and insulate the controls and power supply and use a better insulator for the wires such as the plastic tubing that they use for AC units but you can use the flexible insulation for areas that have to move a lot this would keep the computer from becoming shorted by the startup and will make it more or less likely for lower cost plasma cutters to be used. And just based off of what I seen the the power supply was connected to the frame which made it have a short which caused it the trip by separating it keeps the short from ever happening unless it shorts through a motor that is my functioning
David, I have a question for you. Since the water pan, slats and job material sit directly on the frame extrusions and the ground clamp is connected to the work material, do you think that insulating the water pan from the frame would help with EMI? I recently purchased Bestarc 50 plasma cutter, it has a pilot arc start and the euro style torch connections but is not LF. I can only afford one plasma cutter and decided to go the cnc route after I purchased the Bestarc. Thanks so much for your content..
I had a similar problem as you have with the machine shutting down when the plasma starts. The solution for me was to forget about the homing cycle and disconnect the limit switches and wiring, BUT also turn off homing in GRBL as well because if not turned off there is still the chance of the same problem happening. In my case it made no difference until I turned off homing and limits in GRBL, after that I had no problems whatever. I am using GRBL 1.1h
Hey Reg! Thanks for the advice. I unplugged all these wires AND turned off homing in GRBL (and hard limit switch). Was still having tons of problems. The problem with EMI is that it effects everyone different...
Instead of insulating the work from the water table, why don't you insulate the water table from the frame? I don't think that will fix your EMI issue though. Perhaps one way to do that would be to electrically insulate the control module from the frame, and put shielding around every single wire in the system. Make sure the shielding is electrically insulated from the frame as well. Plug the plasma cutter and the control module into different outlets, or preferably different circuits completely. That way the only way the EMI can jump from the frame to the ground of your control module (apart from capacitative coupling) would be to travel up a power cord, through the wiring in your home, and back down another power cord. It wouldn't hurt to put big, beefy ferrite beads on both power cords, either.
What about shielding your pan from the frame with plastic spacers and shielding your sacrificial slats from your pan with the same idea. That way your circuit close is only on the material and slats?