Fox you're a great inspiration, and role model for precision. I'm looking forwards to your next anodizing video. I'm implementing what you are showing in my shop.
I think if you are worried about the run-out of the ball screw assembly affecting the plane stability of the plate, you can try turning the rigid link between the ball screw nut and the plate to a flexure joint that can take up all sideways and up-down wobble, but can still transfer the longitudinal motion to the limiting accuracy of the ball screw grind.
This is precisely what i was worried about. Flexture sounds like a great idea, i will keep it in mind if i face similar issue. For this z axis there is not much space left to put a flexture. But who knows. I will give it more thoughts and for sure i will test it once is installed in the cnc, before fixing anything ;)
Those plates look absolutely stunning with the matte black on them. You attention to detail both from a technical and artistic standpoint gets me exited every time.
hey! thanks :) i wish i had more time to burn because the concept was better but i had to focus on functionality. maybe i will add some laser engraving to it if i find time.
Pretty good job Fox! Thanks for sharing. 2 remarks from my side: -The precision of the alignment of the rails and the ballscrew do not correspond with the quality of the shaft coupler. If we are talking about microns of precision, precise shaft coupler should be used instead of the plastic inserted one. - The positions of the rails and the ballscrew's suppor bearing should be fixed during the lifetime of the machine. In the industry it's solved with "constaining the ground surfaces of the rails and the bearings to a pre-machined, defined edges". But for us, hobby guys there are special materials exicting, which were developed for this kind of job. My purpose is for example the Diamant DWH. Maybe it cost a bit more in New Zeeland, but it worth it. You just to have to dispense it to the side of the rails and bearings, and when it bonds, you'll get a proper, solid joint, which fixes the rails.
Oh boy, we got a serious and amazing bit of machine built today! And so soon after the last video! What a treat! Thank you for sharing your designs in such beautiful form too. When you started lapping you reached a level of perfection where I won't be imitating you, but the ideas, the methods and the results more than anything, are a joy to watch! Can't wait for the explosive Finale when I think your cnc will split atoms (in my mind) 🤯
@@PiotrFoxWysocki haha, nice! But don't get carried away with the theme and make us hold our breath til we go blue in the face please. Pretty sure 2020 put us pretty close to midnight, it's gotta be 2021! (Hope im not getting my geek lore wrong)
Doh. I got curious and asked Google... and you were obviously talking about the actual Manhattan project, not the cartoon education my mind is filled with😅
@@PiotrFoxWysocki ah nice, that makes me happy! its cool talking to you; you're a bit of a superhero (not only to me but in the cnc facebook groups too!). i knew the manhattan project was a thing but you know...these days are more netflix than reality. i was thinking of mr manhattan from the watchmen tv series in case you're looking for something entertaining while i'm away...as its 1:30am here and time for me to keep dreaming/have nightmares about my own cnc build :-D talk soon buddy! keep up the good work!
Nice Work! I can feel your pain when trying to get everything aligned ultra precise. I'm currently building a CNC machine which should be as precise as possible. Sometimes it's just frustrating to see the indicator needle move "a lot" 😅
Greetings from Melbourne across the pond. Poitr You are up there with the great You Tube engineers, your attention to detail and precision is outstanding. I look forward to your next video. Regards John
Another great video i very much enjoy watching your progress. i will give you a tip on finding the high spots buy some canode blue spotting ink. put a thin layer on the plate then rub the part and it will highlight the high spots with blue ink, when you are finished it just wipes off your granite plate with some glass cleaner. this is what we do for hand scraping to get the whole surface under 1-2 um accuracy. for the thickness i think you also need to look at the parallelism between the rail mounting surface and the screw mounting surface as you know its important, the screw needs to be aligned perfectly in left to right and top to bottom (vertically & horizontally) to the rails. Also don't use these couplings with the rubber they are not accurate enough due to the flex under load, instead use Huco Oldham Couplings they are significantly better for holding tolerance on a direct drive and will protect you if the machine has a crash because the black plastic slip joint in the middle will break before the screw/bearings, they also remove the issue with having the perfect motor alignment. you don't normally see the accuracy improvement just measuring without cutting but you will see it on the finished part tolerance from my experience. have a look at their website. www.huco.com
Awesome video series. Addictive. It is so weird to see the super precise lapped plates, to be then cut with all the pockets and 1/3 of the lapped surface removed. Hehe. But of course, the other surfaces will maintain quite a bit of flattness. Pozdrawiam.
This build is an inspiration. My build is VERY far from this, but if I have gone this far, is because of you Piotr! Just wanted you to know that... Thank you so much!
Thank you :) I'm happy my build has become an inspiration, but with this, I feel bad because I have encouraged people to spend lots of money and time on something that might not work as they expected. So lots of pressure here :D On the other hand, I'm glad I have inspired someone else to go and do some cool stuff :)
Good job. Don't worry for the 20 um of screw wobble: it is not a problem coz the screw will be guided by the nut after assembling. You don't need a "better one".
thanks for cheering me up :) 20um leaves a bad taste after all the work i have done. the wobble is particularly visible in Z axis when you start 3d milling. when the screw turns fast its will shake the whole z axis. i guess 20um on a screw won't do much, considering it is only 10um from the center of the rotation once screw is bolted. ok now i feel much better :D thanks :D
Excellent work, as always! It is really nice that you guide us through the process. If I can recommend, you should put some affiliate links for the parts on your website. So, people can also support you when they buy the parts using your links!
Great suggestion! most of the parts i got from ebay. most of places don't have affiliate program. but i will pay more attention to in next time. thanks again :)
I have made a video where i have showed my setup and a bit of the process: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-bayPkLSqmrY.html check it out :)
Super robota, miałem dzisiaj pytać kiedy następny film a tu jest :-) Szkoda tylko że zamieniłeś miejscami wózki z prowadnicami... Widziałem pod spodem w komentarzu czemu tak zrobiłeś.. szkoda sztywności. Ja kiedy tylko mogę to podkładam coś pod detal żeby mieć zetkę jak najwyżej. Co do przeciwwagi to bez problemu załatwiłbyś to małym siłowniczkiem z buforem powietrza. Nie mogę się doczekać kolejnych filmów, aż sam się wziąłem po roku za modernizację swojego złomka :) też przeróbka osi Z + dojdzie stół z rowkami teowymi. Działaj tak dalej 👍
@@PiotrFoxWysocki Niemożliwe żebyś nie wiedział :) to taki stół jak w fabrycznych frezarkach. Stół teowy, nakrętka teowa - hasła do wrzucenia w google. Dużo wygodniej montuje się detale do stołu. Mam teraz tak jak Ty czyli gwintowane otwory w stole i zawsze tak wypada że brakuje otworu żeby coś nietypowego zamontować na maszynie..
So there was mean thinking my Z axis was done :) I will be making a few design changes tomorrow to allow me to adjust everthing as it should be and I never thought about the wat the ball screw could push the plate in and out like that, guess ill be polishing the plates tomorrow
Hi, nice work, I see you mount the sliding block to close in order to gain travel on Z axis, and I guess you gonna mount the spindle offset to Z - , I think this will amplify the litlle play in sliding blocks. Currently, I am building a CNC and I allowed 200 mm between the sliding block with 350 mm sliding rails so will be a 150 mm travel on Z axis and the spindle clamp will be 50 mm lower than the block if you look from the side. Did you consider to reduce the distance between the blocs and spindle clump?. THANKS
i use roller guides so there is not much play to start with. yes i wanted to have more travel looks like its 190mm. spindle clamp is mounted 50mm below last carriage. so just like in yours :) im considering installing smaller clamp higher on the spindle. its not to support but to remove few microns from the deflection. currently it looks like z plate has a deflection of 10 microns for every 17 kg, from the carriages to the spindle mounting point.
@@PiotrFoxWysocki Thanks for the info. I see the good guides will do the job. Can I have the link from where you purchase them? Just now I downloaded fusion 360 file and the design looks ok just the Y-axis the motion from servo to actual axe look a bit too long but I can be wrong. 10 microns you tested on the machine or separated from the machine?
Yes y asis look long because this was the lenght i could get. Beside i didnt wanted travel to be from end to end. Beecause in case of any miss aligment it wont bend permanently or the increase in load will be minimal. All the sources are in video descriptions of epic upgrade 1 and 2
Hi Fox, have you considered including a counter weight to compensate for the weight of the spindle? The ATC spindle is quite a heavy beast, so I think that a counter weight could be beneficial. What are your thoughts on this? Thanks, Eric
i thought about it. but i thought to add a gas strut, counter weight would add 15kg, and inertia, strut weight almost nothing. but currently I just use lock pin. after use while machine is on I just home z up, and put the pin in. so Z doesn't drop when you power it off. when its on, there is no problem for servo to keep it up. and because im lazy and it works like this, I can not force my self to change it :)
@@PiotrFoxWysocki Your honesty always gives me a big laugh :) But we as all know you are not lazy, instead you are selective and pragmatic about how and where to spend your time. That is why you are so effective and get so much done :) To get back to the topic, yes I was also thinking about adding a gas strut instead of a real counterweight. My previous machine ( a Gerber Sabre 408 with a heavy 5.5kW spindle that I had converted to LinuxCNC) had springs installed, which also proved to be quite effective, but I think a gas strut is a more elegant solution. I think I am also lazy on my end and putting a pin in to prevent the Z from dropping is something that would keep annoying me and I would probably forget to do it anyway and destroy a lot of router bits because of the spindle plunging down on the table or work piece. An even more simple solution to prevent the Z from dropping would be to add a break to the Z-axis servo motor, but that would still not eliminate any of the other disadvantages of a 15+ kg weight pulling on the Z-axis constantly. I think that a gas strut together with a brake on the Z-axis servo would be the best solution. Any additional thoughts on your end?
@@EricBow I'm glad i can bring happiness to your life :) I looked up the prices quite quick. gass strut 20$ servo brake 150-200$. in my opinion its no brainer :) gas strut, 2 holes and you are set. servo brake, quite possibly you will have to buy a new servo with the brake, and what is worse you have to refresh your memory how to wire this bloody thing. and since you want to use strut, there is no point for the brake. if your z weight is 15 kg, and your strut is matched for 15 kg. then there is no way for the z axis to move. cause it will weight nothing. even if you get a strut 10kg or 20kg it wont move because you have to break holding tourqe of the ballscrew with servo. even if its turned off. use of the brake is only justified when you use it, for example, in a actuator to lift something heavy and keep it lifted after you turn off the actuator. its more to hold unknow amount of load, once you lift engine and the other time a car. then strut is useless , but brake is perfect. but when you have the same load all the time strut is the winner. makes sense? right ? :D stay lazy :D
@@PiotrFoxWysocki Yes, absolutely makes sense. I was thinking along the same line. Since I am building a new machine from scratch, I'd like to verify my own thoughts against experienced people like you, so your feedback is very much valued and appreciated. The next challenge for me will be to get all the parts I need, specially the ball screws and linear guides. I recently moved from Europe to the Philippines and Holy Christ on a pony..... there is nothing decent available here. I guess I will have to import stuffs myself. I used Star Rexroth and THK in the past, so will see if they can ship from Singapore or Japan or so. Since I have made machines before from epoxy granite, I was tempted to do that again, but given the attractive prices of Grade 00 granite surface plates and straight edges from China, even made to customer specifications, I think I will stick to that for now. I expect it will save me a lot of time achieving high accuracy, allowing me to stay more lazy and enjoy my time in more fun ways :D
Nice black alu. How thick is the whole Z plate? The middle was 15 mm. Maybe total is 25 mm? How would one think when choosing where to put the fixed and floating block on the Z axis, fixed at the top or at the bottom?
@@PiotrFoxWysocki The lead screw seems to have bearings in both ends. Usually people go with the BK/BF type, fixed in one end (angular bearing) and floating (deep grove bearing) in other end. Usually I see that the angular bearings (the fixed side) is mounted at the top with the motor. Here it looks like you have the fixed one at the bottom. But that is just a guess since I have no idea.
You are right. By mistake o have ordered the wrong type ;) but it all worked out as i was needing more spacers. It does exactly the same thing eclxcept this one is more expensive ;)
You showed the linear roller curving in a previous video. did mounting the linear rail change the flatness? if so by how much? Great video series btw it is really entertaining and fascinating!
Linear roller had very small bow. Around 10 microns. Once its bolted to the plate it will straighten up. It will also bend the plate but not noticeably. If the bow would be around 40 microns i would consider straightening it up first. You know. Everything with in a reason ;)
I might be mistaken and may have missed that part but pretty sure this dude is on the up and up with standard metrology practices and standards. Anybody that can get submicron precision on an entire machine in their basement from parts ordered on the internet.... probably knows what they're doing. Like I said I haven't watched the whole thing through but as a professional matrologist myself I can't imagine this dude doing anything crazy or wrong. So much of the surface plate he has used and sculpted has to be conditioned sometimes and he always brings it back down to weigh weigh weigh lower than any standard or tolerance needed
Also, he's got everything dialed and so precise that the warmth from his hands touching the linear rails will do them and move them more than any other tolerance on his machine. I never seen someone make a DIY machine that is basically machining a diffraction grating type finish on aluminum parts and even stainless
In my opinion surface plate is a tool, not a trophy behind the glass. Tools are meant to be used 😉 I have couple of surface plates, some are good and some are for stuff like that. You have to break few eggs to make a omelet 😉 Btw surface plate on this video cost about 160usd so in my opinion it's a consumable. Like a sandpaper. Or endmill.
Nice work like always, just don't like the design where spindle plate overhangs. The way to avoid that is to mount Z rails to spindle plate. At home there is no overhang at all and progressively increased as spindle goes down toward table. Can you send the link for ballscrew and LM guides set please?
i mentioned before, my rails are bolted from the top, not the bottom, assembly in this case would be very tricky. and from my personal preferences i don't like the idea of variable rigidity according to the depth. so i don't have different finished at the to and the bottom. sorry but i don't have the links for the ball screws and rails as i bought them on ebay.
@@PiotrFoxWysocki I've never mill any 90deg wall higher than longest tool length - 75mm. I did not noticed any changes on surface finish on that wall towards bottom where rigidity would be lower. Fair point, my 1st cnc had this idea and somebody mention to me therefore since I do opposite. I spent fortune on precision tools after watching your stuff:)
No you made me feel bad :) its the last thing on my mind to encourage people to go and spend their life savings ;) In my case i want to mill plates mounted vertically in the vises. As far I remember new Z Axis has 185mm travel.
@@PiotrFoxWysocki : You shouldn't feel bad... if you encouraged someone to spend money on useful things, you may have stopped them from spending it on things they drink, snort, or inject! Keep up this wonderful social service, you may save the planet! :o)
white lithium grease. found it being recommended by some manufacturers, not sure if its the best grease. If someone would like to comment on the topic i would appreciate it greatly :)
Why didnt you put ballscrew and linear rails on the z plate? That way you could more seperate linear rail carages and have linear rails on the bottom of z plate which increases rigidity. You have to mount spindle much lower than those rail blocks are so why not encrease rigidity with those rails? Only con would be that z axis motor has to be mounted on the z plate then and it would be a little more tricky to assemble it, but it can be done.
i guess its my personal preference. but i did think about putting rails and blocks flipped. 1. my rails are fixed from the top not from the bottom. kinda bummer when it comes to assembly and adjusting. 2. I'm not the fan of variable rigidity, when you mill at the top the z axis is the stiffest and as you go lower it changes. so i prefer to have the same finish all around. 3. z axis has 2 plates, front and back. they both bend. so loading one makes other more flexible. and than the weight, with the servo and 20kg hanging i'm a bit concerned of servo holding it up when its turned off. i don't want for the Z to drop as soon you turn off the machine. Im trying to avoid building a counter weight at the back of the gantry.
Have you considered adjustable gas filled rods as an alternative to a counterweight? That's what I'm planning to use for mine. Looks good in my CAD model but yet to try it. Amazing videos by the way 🙂
@@PiotrFoxWysocki fair enough. My thoughts are that it would also take unnecessary load off the ball screw. But that's maybe not important with one the size you're using. It's quite large!
@@PiotrFoxWysocki A pneumatic cylinder with a large feed line to a large pressure tank with a regulator will give you an adjustable - but more importantly almost constant pressure -> constant force.
Hi to everyone , I'm kinda a newby when it comes to mechanics, does someone knows the name of the stone he used to take out the high spots ? And what about the plane in which he was rubbing the guides ? It's a reference plate ? Or it's something else. Thanks in advance if someone is going to reply
Stone is called a toolroom stone. But knife sharpening fine stone will do. Not ideal but in case if you dont have enything else. Plane is a granite surface plate. I have 2 of those. One for measurments and the other one for those kind of activities;)
@@PiotrFoxWysocki ohh ok that makes sense ...I didn't understand why you were scratching a reference plate ...now it makes sense, thanks for the reply and love your work
A bellows style coupler would help you a lot between your motor and ballscrew. The mechanical hysteresis of a lovejoy style coupler you're using now will show itself in 3d surfacing style toolpaths and ball-bar results, something I learned from an old fadal engineer. I'm assuming you had to do more lapping on the plates after you did all the pocketing right? or did you get all the stress out of it from just the facing?
Coupler works fine. Not complaining on finish. Plates have open up but when I bolted it together they cancel each other out. It wouldn't, I would have to redo them. But everything turned out just fine 🙂
Great question. I havent done it yet because its not attached to the gantry and it doesnt have spindle clamp. Those things will change the amount of flex it will have. I will keep it in mind when i bolt it in. Thanks for the suggestion
This is my hypothesis. Steel is much more dense than aluminium. And the tone you get by striking the same size aluminium and steel. Steel has much higher pitch and resonates longer and louder. In addition with high speed spindle steel would resonate through the roof. Where aluminium would dampen the vibration much better and quicker. Besides its easier to machine and lighter. Forces operating on cutting can not be bigger to bend an endmill. And i did some tests on endmills and 20kg of pressure would break 4mm endmil and distort any other endmill beyond reason. So perhaps on a roughing, forces are much higher but on a finishing passes its barely anything if you want to have your parts in spec. Hope my explanation makes sense ;)
G'day Fox. Like your work, admire your build, look to do similar... But you dooo frustrate me at times. Talk in microns but don't have a set of mikes, use a didgital aproxometer for critical dimensions! Get with it dude. But I do really like your design concept and that flood coolant set up is the most effective I've ever seen. Keep up the good work!
haha :D thanks for the honest feedback :) really appreciate it. i will get more mikes and indicators. i got brand new 1 um indicator but the spring is so stiff that it was marking the surface i was measuring, so 10um indicator was better, because more sensitive. but i will look in to this one and see if i could do something with it. or i just get new one. i used calipers just to see roughly what it is, and for the video to show the number. later on i switched to mike and got proper reading. not sure why you are angry at this. this 15mm dimension is not critical at all, it could be 20mm as well. critical was the consistency in the thickness, which turned out to be spot on.
Fox, not angry, frustrated. I well understand the cost of tooling. Pharq! I have spent a small fortune on tooling!! And continue to do so. The 0-25mm Mike is the cheapest of the lot, just sayn'.. Keep up the good work though.
@@PiotrFoxWysocki If you want to check the motion of the final system for straightness and accuracy, you'll need a laser interferometer. Like this: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-WtZfg7OGXG8.html They are cheap and easy to get now. I can help if you need some pointers.
Only precise piece of the screw are the ground groves. Its really hard to indicate the center of the groove. Nut on the other hand is attached to the screw with balls which are in the ground groves. The nut will be attached to the plate. So this is the part you should measure. Btw im not rotating the nut, just holding it in the same orientation and measuring from the same point. Could you please explain why i shouldn't relay on the ball screw ?
@@PiotrFoxWysocki Put the indicator tip on the ballscrew groove and turn the leadscrew forwards and back. This way you will be able to check all direction AND backlash!
@@PiotrFoxWysocki The thoroughness and care put into this machine has made it the most impressive homebuilt cnc machine I have ever seen, certainly precision wise. I'd imagine a machine like that would be a watchmakers dream! Take care and keep building - Dean.