Thank You for making these videos and sharing test results! This is the first time (here on RU-vid) i've seen someone take buckling into consideration.
I work for a big company that today are switching from Mazak Smooth's HCN 6800, to the newest Mori Seiki NHX 8000, but i tell you that i prefer my G550 T GROB machine center with 5 axis. One day my friend i will be able to setup my own cnc too, btw you X axis is THICC. IF anyone is courious the only machine that i was able to do 8mm depth pass for pre finish milling with a 250 mm diameter face mill on cast IRON was the MAZAK Nexus center 6800, all the mori seiki give me axle overload alarm and spidle overload, even heller and some other machines didn't handle it, the surface finish wasnt the best but you get the point, 8mm on iron is insane dude, with aceptable pre surface finish and precision, i think we was running seiko Inserts on some national toolin with dual contact zone on cat 50 spindle taper. if anyone complains about precision we run about 10 3d centers inside the company that measures every product that leaves the machining area.( only one is mitutoyo, all the others are ZEISS), now imagine for someone having 10 3d centers on the plant, how many cnc do they have on. fuckin hell
Looks good and definitely going to avoid damaging components. I would think the force settings could be higher in tandem with encoder and current monitoring or things like that so the machine can know its crashing and vary servo force output. Assuming that's what high power mills do for safety.
forces could be different on the cut and as long servo is rotating counter is counting. when the torque goes up to the set threshold it stops. i have plan at some stage to connect warning to mach3 to turn on the E stop. so all the axis are stopped at once.
Hey Piotr! I started a machine built exactly like yours! Already have the base and both y rails and the modular y axis’s complete. For the gantry, I poured epoxy with sand around an aluminum skeleton with threaded inserts attached for rails and ball screw. Can send you a video once complete!
I would love to see it. I'm curious if people would be interested if I would make a something like a monthly or weekly videos showing and talking about viewers cnc machines builds designs and what not. Just hope people are brave enough to submit their projects 😉
@@PiotrFoxWysocki Can it be seen any other way ? Micron precision, granite gantry/base, beefy and sturdy rails. Your machine and way of thinking about whole process is top quality.
I guess. But this was the most visual way. I know that presentation was scary to watch ;) but i knew it will handle it just fine. And as a side effect you can get the idea about the strenght of the granite. It wont break that easly ;)
Most of the cutting force should come from the spindle anyways, no? Maybe you could even save your tools this way? I mean if the force get's too high sideways your probably need to increase rpms, no? It would be probably worth a test to limit the servo force to an amount that reduces the cutting speed and see what happens. I imagine the finish could suffer, and depending on the servo settings you could get some ugly vibrations, but that should be possible to reduce by some dampening. Thanks for that inspiring video! 🙏❤️
i don't think that interactive reducing of the servos power during the cut is something worth doing. fusion 360 does it for you and tool paths are optimized for the chip load and speed, etc. when there will be a another variable which will be changing those parameters than you will end up with blunt tools and lots of wasted time because your settings would be too sensitive or something. beside it would be really complicated to make and maintain. but this is my point of view after a quick thought. i just can't find any benefit of doing it.
@@PiotrFoxWysocki I agree, I didn't expect it to be practical used this way, but still would be interesting to see what happens. Of course I understand if you don't want to risk getting into unstable states. I still think it could be a pretty valuable if you are able to for example use the expected chipload calculation and set the limit dynamically with a safety margin, that maybe can save the tool from breaking. But that of course would need deeper software integration and that's probably not that easily available. OTOH I am surely not the first one to think of this 🙃 Thanks for your reply 🙏
@@PiotrFoxWysocki I was checking the news and it appears a popup that your had a new video so, well like I like your videos and work, I jumped on the occasion ! I finished to watch it and I have a question about your setup, nice setup i should say. In mach3 ( i think you use it) what acceleration do you use ? I am finishing a big machine, big for my standards and curious what you use. ( above 4000 in accel my machine to move across the room :D)
sorry for late reply but i had to check. i set it to 500 which is 0.5g. my table under the cnc is a bit flimsy so i don't want to push my luck too far :D with the gantry weight of 130kg. beside high accelerations will ruin your ballscrews. 500 works well for me
@@PiotrFoxWysocki you use metric units mm/sec2 ? don´t worry for the ballscrews, they are 32mm and really I don´t use 4000, I use 450 almost like you, but it doesn´t give me 0.45G but 0.045G
The travel speed should not be affected much since the servos do not loose their torque with increased speed, but I am curious what happens to the acceleration.
sorry i don't get the question :( i would recommend to look at any servo graphs in google. for me as long it works, Personally I don't really care how does it work as long it works :D If it wouldn't than i might start looking around :) otherwise its one less problem to be worry about :D
@@PiotrFoxWysocki You will need the most torque when you change the direction of an axis. Imagine your z axis coming down at full speed and then, without stopping, going back up. If the torque limit kicks in while you change direction, you might end up cutting inaccurate. There should be a note and a diagram somewhere in the limiting torque section of your handbook. If you use autotuning, you should do it again after you set all your limits.
@@Christian-ti7wh well it wont cut inaccurately because it has encoder. and when it will hit the torque limit it will just stop. so no problems so far. and if you are going through the cut inertia works a bit different because you are getting resistance from the cut. so to change the direction doesn't require as much power and changing direction without the load on the spindle. i hope my attempt of explaining makes sense :D
@@ramous5182 well it wont cut inaccurately because it has encoder. and when it will hit the torque limit it will just stop. so no problems so far. and if you are going through the cut inertia works a bit different because you are getting resistance from the cut. so to change the direction doesn't require as much power and changing direction without the load on the spindle. i hope my attempt of explaining makes sense :D
Very interesting, as always. As you are limiting the power, Do you think you can build (amost) the same CNC spending much less if you buy smaller, less powerfull, servos, like 100-150W ? Or that cheaper servos would be lacking some needed features, like weight, speed or precission ? I can see in your links "motioncontrolproducts" only have 400W and 750W kits. Thanks for sharing !
Not sure how much less it would be for 150w servo. I guess for my current machine i would feel more comfortable with 200w. So i have some spare power left just in case. Speed wise they are all 3000rpm minimum or more. Precision should be very similar as well for cnc. Servos have different resolution encoders, but most of them should outperform whatever resolution you will set in your cnc software. I can't talk about all other servos out there because i did the test on 400w servos in this video. In my machine so you can draw quite accurate conclusion what those setvos are capable of. Im guessing that other servos should be similar in performance to be safe we could assume 20% +-. So if 400watt gives 500kg than any other 200w servo should do minimum 200kg. Obviously there are calculators for it. But ball park you can use this video to draw general conclusions.
@@PiotrFoxWysocki saw you joined the cnc group on Facebook watching your videos made me want to build my own machine so thanks for the inspiration maybe I can show you my progress
Just out of curiosity, you never mentioned in your video when you tested the stepper motors what size motors were on your Y axis. I am looking at using 425oz-in Nema 23 for my Y axis. I am hoping that will be enough to move it at a reasonable speed.
i mentioned in video description "X - Dual Shaft Nema 23 Bipolar 1.9Nm (269oz.in) 2.8A 3.2V 57x57x76mm 4 Wires" and this one was going over 3000mm/min but i limited it there to be there so it won't stall at higher speeds.
@@PiotrFoxWysocki ah I somehow missed that. Thank you for the quick reply. One last question, how heavy was the gantry? I'm curious as to what weight limit I should keep in mind when building mine.
One thing I'm always interested in with these home built machines is, given the current state of the machine, and the investment, would you have bought one for the same price you invested or would you still have built it yourself.
well.. at the current state i would have to spend 100k+ on a commercial machine to have similar surface finish quality. even than its not guaranteed it will be as good. So i prefer to make it my self.
Maybe I missed it, but what controll software/hardware are you using? Mach3? Do you have a control board you use? Are you using fusion 360 for the CAM? Thanks! Love you video's!
Heya Fox, still in the research/prep/courage building phase on making my own, smaller version of your router. It will have less beefy rails and screws but once I get back to China, I aim to start sourcing secondhand Japanese hardware. But judging from your numbers and given that my carriage will weigh, perhaps, half of yours, wouldn't you say I could likely get away with "just" 200w servos?
most definitely! if 400w servo is capable of 500+kg on 5mm pitch ball screw, than 200w should do 250-280kg. only one way to stall it would be to have lets say a gantry which weight 100kg and you will accelerate it with 2.5g than it might stall because it will require this amount of force to change the direction. and im sure you wont be running it at those accelerations because your cnc would leave the building :D
@@PiotrFoxWysocki Awesome:). Though I am locked out of China for now, one huge benefit of living there is how easy it is to get the materials to sustain this growing hobby of mine. Imagine needing some 25mm rod of Gr5 titanium - just order it in lengths of 10cm and it will be at your door the next day for next to nothing. Same for tooling inserts. Huge market (as you know) for the Japanese secondhand machine parts and obviously, all the Chinese cloning brands, too - often at 15-30% under AliExpress prices. I think I may even be able to find a surface plate manufacturer to cut my pieces to custom sizes and possibly even drill holes affordably. Now, I wouldn't mind living in NZ, either, but that's another talk:)
Fox, you know I am building a P&P machine, as it will be "a bit fast" and heavy, I could not decide yesterday in the shopping cart to go with 660W or 400W, I can sleep now... :D
hey, im sure you exactly know which one :) weight of your gantry * g force = servo power :) so if your gantry weight is 45kg for 400watt servos you could run it with 10g acceleration on 5mm pitch ball screw and 5g acceleration on 10mm pitch ballscrew. very scary when you think about it :D
just in case bolt it to the floor :D when i was setting mine servos i put some stuppid number in to acceleration and run autotune.. holly hesus... seriously i didn't know where to run. house was moving! and it was only my x axis :D
I think you should go alot higher. For example my 5.5kW lathe has a 400kg longitudinal trip feed and 200kg in the crossfeed. Also never put your hands in squeeze zones even if you have a scale in your hand.
Im really happy with 85kg as i said, i can always dial it up. It takes only a minute. Thank you for your safety concern ;) Maybe it doesnt show in the videos but i'm really capable of thinking for my self, where to put my own hands and where not to. And i really, really like to have autonomy in that matter. I have quite few friends who where saying that i shouldn't do this or that. But somehow they are the ones with missing fingers.. To clarify they where saying it, before they put the finger in the wrong place. Almost like in this famous sentence "hold my beer" 😉
@@PiotrFoxWysocki lol... I'm just saying things don't grow back (had a spine surgery a year ago and maybe headed for a second). Don't do what your friends did!
@@PiotrFoxWysocki Some lessons are written in blood. "I have quite few friends who where saying that i shouldn't do this or that. But somehow they are the ones with missing fingers.. " I wonder why!
Check ODrive servo driver ...it is around 120$ 2 motor driver , 50$ for encoder and around 100$ for 80100 or 6080 bldc motor up to 4kw of power yup 4kw each motor with all settings u need for it....super powerfull and cheap servo motor setup :)
Funnily enough, i considered it. You need hell of a dc supply, or some way to supplement the battery. Was thinking of driving the spindle with it though
They get really realy hot 🔥i used one 2kw and you would get burn if tou touch it. I run it with 52v 30A power supply. Problem would be the cooling. But there are also some with water cooling. Max rpm 5k
yes , promising but yet time will tell . people already tested on 3D printers few hundred hours in 1 go without issues DIY CNC the same but crashes etc still exist. For DIY CNC good idea if u go "on budget" and want super speed
may i ask Which controller do you use for this cnc？and if i want get a controller with 5 aixs rtcp and atc function，what is your proposal？thanks a lot！！
Oh man,i would never purposely test equipment like that,,especially special equipment like that. Alot of force on the ball nut! Least now you might not kick a vice in a crash tho?!
haha :) i build it so i guess i have a bit more faith in it as well :D In addition i have been using it for years. i guess its better to do this in controlled manner than discover it after some disastrous crash. now i'm not really scared of crashing it anymore. Even if, for some reason i will lose one of the y axis, it wont have enough of power to snap my gantry in half :) or move vises, or destroying the spindle
I would be very interested in the stiffness of your upgraded machine. Have you ever measured it? Like measuring spindle deflection while applying a force with the scale to the tool.
I am curious, what's the max rapid movement that your machine achieve with that servos? I'm planning to build a similar machine machine and I'm extremely interested to buy that kit.
Its very hard question to answer. It depends on your controller speed. Resolution you want and travel speed. Also how you run it. Analog or step and direction. I set mine to 2800 because i have ot send double amout of steps to 10mm pitch X axis. To have the same resolution as my other 5mm pitch ballscrews.
@@PiotrFoxWysocki I am using MASSO controller, it can generate 100k pulses per sec. Is a nice stand alone unit and doesnt need a pc to run, just mouse, keyboard and monitor, i have made a crude cnc out of wood and one of my axis can go up to 4500mm/s in rapid with nema 23 steppers. It can even process aluminium, although for now I'm still trying to develop a design to fit in my tight space and have the maximum possible work volume. I will build the "big boy" with steel, alluminium extrusions and epoxy granite.
@@PiotrFoxWysocki i have configued it in a strange way. The stepper drivers are set to 1600 divisions, the cnc controller is set to 800 steps and counts half of turn as a full turn and i have 5mm pitch screws. So tecnically is 1600 steps per rev.
Wise to limit the servos. Have you tried doing some very fast rapids to see how they behave under hard acceleration and deceleration with the new settings? Usually the torq isn't required in the cut it's more critical for fast controlled rapids.
Looking at your hardware I think when your comfortable with more speed you would be able to easily manage 10m/s and keep precision. I think maybe even more could be possible, my old torqcut 22 with the old dc servos 32mm rails and screws that is much heavier than your machine was easily doing 10m/s. thetorq cut models with Siemens ac servos were even faster.
Well my gantry is 130kg it would be terrifying to run it ar 10m/min 😱☠ wit this speed i would kill my spindle in a month time. Im not running massive runs so every second counts. I can wait few minutes longer. And be sure nothing is goung to explode. Any mistake at those speeds is catastrophic.
@@PiotrFoxWysocki 130kg isnt heavy ;-) my spindle motor and transmission weighs more than that without the casting and its only on a 32mm 5mm pitch screw and 32mm rails, no counterbalance 😂 but I agree it is scary at first. Spindle should be fine as long as you don't rapid too fast at high rpm with tools heavier than the rating. If your only doing small runs maybe a tool changer next is your biggest efficiency something like you see on cnc routers would probably work quite nice on a gantry style machine like yours.
I have around 40 tool holders. So there is no way i can fit them on the table. And changing them by hand is much quicker than table tool changer. And much much simpler.
2.5KW ATC spindle link here: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-h9q6j7POgOI.html i thought i will have to get something bigger looking at the size of those servos and the new size of my machine :D
@@PiotrFoxWysocki that spindle is a is a beast and I suppose one has to stop at some point :) this is why I now have parts for two machines lol.. cheanged my mind a few times on the build
I have put all the info and prices in video descriptions of epic cnc upgrade series. Part 1 and 2 I think. Accuracy is more than nessesary and surface finish is perfect. Check my instagram for some results.
After watching this vid the 4th times. I wonder if servo is this strong, do we even need counter balance using pneumatic piston or spring? like some commercial machine does (like Shopsabre)? Thank you and have a nice day!
Servos are strong. And you won't need piston to keep it up once is powered. Cut as soon you turn of power it will drop the z axis. Stepper will hold much more when unpowered but won't be able to lift as much and won't be able to operate at full speed.
@@PiotrFoxWysocki Thank you for the great info! Appreciate it as usual! well, given that might need to install piston just for the sake of holding z up i guess
