Yeah same......so sad to you use rollers instead of proper rails......kinda aims besides the target if you spend all that time and money on a CNC why not spend 300$ more and get something decent? Also the 300 bucks will be spent on broken endmills real fast
@@Basement_CNC He described the reasoning behind his choice in the video. If you think it's such a bad choice, go make your own. I've got a Shapeoko with rollers and it works great
I probably would have done those holes by just marking them with a spotting drill, take it to the drill press, and then cordless drill to tap. Easier to CAM and much less stressful than watching the machine do it. These days I do that for any flat piece I can fit on the drill press and which doesn't require precise depth holes.
my old cnc router had vbearings, and they were just ok. but that was also 10 years ago when linear rails were $$, last year i upgraded it to linear rails and wow, for 100 bucks or so, its 1000x better. since you are doing this from scratch, did you price out the differences?
I did not because my goal here was mainly to explore improvements to an existing architecture with a simple change. At work I can play with a 4x4 linear rail + ballscrew Shapeoko 5 so I don't really crave the need for outright power or performance. Oh, and I also have the Neo to play with... And as kind of a thought exercise, I wanted to see what an affordable Shapeoko could be. If you assume the linear rails for X and Y could be had for $150, plus labor and machining time, by the time you add retail markup you're going to have a Shapeoko that's ~$500 more expensive. That's basically the price delta you see between Shapeoko 4 and Pro. To me it would be less fun to try and build something that's essentially already been explored.
@@WinstonMakes for a production model sure, but this is more of a one off thing. dont get me wrong, your work is awesome and have watched every video lol. just figured if building something from scratch why not.
Can you explain how the middle wheels are going to take any load? Maybe I just don't understand it. Is it just that as the carriage gets loaded and the outer wheels take some load, the plate flexes a bit and then gets slightly stiffer when the middle wheels start seeing load? I think It will get less than 2x as stiff when the middle wheels engage because they have less mechanical advantage. If I do understand it correctly, then I agree with the person who said they think that moving the wheels outwards will do more good than adding the middle wheels.
That's what I was thinking too. If the wheels are soft enough then maybe they will all engage, but otherwise two will just be going along for the ride.
Well, i`ve builded few CNC routers with linear shafts/rods (SBR16). These shafts are very cheap in comparison with Hiwin type profile rails and they are much better than wheels. I can agree with you- it`s possible to replace pricey ball screws with teeth belt. GT2 10mm wide tooth belt can do the job good, even during aluminum milling.
Hey Winston, what resources do you recommend to learn more about speeds and feeds? I’m getting back into CNC and that’s been one of the most challenging parts trying to cut metal.. Loved the video, I especially like how you leave in your mistakes. It’s much more human and also gives valuable insight to help others not make similar mistakes, thank you!
So speeds and feeds is a potentially deep rabbit hole. I think it's always a good idea to have a good high-level understand of the theory, and what the terms actually mean. There's some good videos about it (NYC CNC and the like), and lots of websites do a decent job explaining it. It really depends on how you learn best (visual, written, etc). From that baseline knowledge, I think you need to make a judgement call. You can totally nerd out with programs like gWizard, or spreadsheets, and deep-diving machinist forums to try and find the "optimal" feeds and speeds. But unless you're using an industrial machine, the theory can fall apart pretty quick. There's always exceptions and ultimately you need empirical testing to figure out what works best for your setup. I personally don't like going overboard, nerding out on these things. What I do recommend is to use Speeds and Feeds knowledge as a guideline. Ex. If you're cutting an unfamiliar material, but you know there's a suggested range of surface speeds, you can work backwards to find a ballpark RPM to try first. You don't necessarily need to get every last cutting parameter perfect though. And remember, these are guidelines, not immutable laws of machining. Keep a couple cheap endmills on hand, and don't let the fear of getting something wrong paralyze you.
It's a fun build. Who cares if it's not very stiff, it'll be light and that's what matters right? And maintenance is fun and it'll be nice for content.
One mistake that I made when building my CNC router was not making it easier to square up the X -Y axes. If you want .001" -.005" accuracy in your parts, you are unlikely to get that without making your X axis slightly adjustable.
Unfortunately on this architecture there's not a good way. I ended up using shims between the gantry extrusion and its endplates to square it up. There's also provisions for dual Y-axis homing for auto-squaring if I end up using a Shapeoko 5 controller to power this thing.
Please lose the V wheels. I built the original Shapeoko Kit around 10 years ago. It was so poor I built my own keeping the controller, and V wheels as part of the new design. My only regret was keeping the V wheels. They might be great for some applications like lasers but not where you're creating dust which can build up on the wheels and rails. At best they can cause erratic behavior at worst stall the machine. At this point linear rails are cheap enough and better in all possible ways but cost. IMHO the should be considered the only reasonable choice. A little background. I'm an electro-mechanical engineer with many years of experience and in an earlier life was a tool and die maker. I loved my home built CNC router but hated the problems I had with the V wheels. I tried all kinds of fixes to no avail. Keep up the great work. Small CNC machines have changed the world for makers.
Aligning 4 eccentric nuts a single rail seems like a pain. Would it be too complicated to have the wheels connected in pairs, able to pivot and auto adjust with a single eccentric nut per pair?
It's definitely possible if you throw two wheels on something like a rocker arm. The problem is that every linkage you add is a potential point of deflection. My feeling was that the simplicity of an eccentric nut + V-wheel on a solid plate would be the safest course of action.
In general, weight is good. But the rigidity of a machine needs to be balanced with the sturdiness of the powertrain. V-wheels and belts are at best, "medium duty". Making the frame out of aluminum, or steel, or cast iron will make very little difference in the grand scheme of things here.
Doesn't look ridiculous to me, perhaps a touch over polished for what it is, but no more than that. Don't imagine you or anybody else could make money today selling a tool that is simultaneously a bit too pricey and premium compared to the budget tool and way less performant than the more professional grade machines (unless you already have Apple's brand recognition and marketing department anyway).
All these cutouts at the profiles weaken the structure! This is not a 3d printer, so its not nessesary nor useful, to lower the weight at any part! Have you ever seen commercial routers and mills? All this machines with the footprint of your machine have no such cutouts. Because its a waste of time, material and effort! Sorry, but i have to say when a construction is not good!
I think you missed one of my previous videos. I know I technically weakened the rail, but it's fun and looks cool. Plus, the material removal from the gantry is negligible in the grand scheme of things. The belts and wheels are ultimately the weakest point. It's like saying a pencil should be made out of titanium for precise writing, when ultimately as long as the body is stronger than graphite it will work.
Who says it is not also a 3d printer?!?! Plus with the limitations that will be inherent to the sort of spindle that thing can mount and the wheels it runs on I doubt its had much if any impact on performance while reducing the mass that the machine has to move around a tiny bit, so saving a bit of energy if the machine was going to be used often would make up for the time invested. (Not 100% it won't matter, but it looks like a pretty safe bet it won't have made a meaningful difference with all the other factors that look much more limiting. Also even if it has he gains a great excuse to reinforce it in a lightweight and stylish way) Also plenty of commercial machines do have lightening features in places - not every machine of every scale will, and the method used to have both sufficient rigidity and lightness might not look the same as here - for instance T, C and I beam profiles are a stiffness in multiple directions to weight ratio saving measure, and they get everywhere! Ultimately no machine wants to pack excessive weight where it isn't actually required, that just makes it more expensive to run and likely limits the speed of its operation. So like all engineering its a compromise between cost to build, cost to run, and final performance.
@@foldionepapyrus3441 okay, its like you buy a VW Beetle and you say this my solution for F1 racing, family camping, and your distance for your dayjob. This car fits only for one of this 3 cases, .....the last, and if you ever had done so, you never want this solution back for only one day. Its a compromise, and for shure a very bad! If i have the choice, i always look for the best possible way. You can do 3 d printing with a much more massive machine, ......but slower. You can mill with a machine out of rubber nothing, the flexible parts cause the damage of the endmills. Carbide mills are made for stiff machines and can not handle the flexing, cutting edges break ....or the complete mill. Okay you can instead produce dust with smalest possible chips, but this is not milling! This stupid cutouts are not useful, an you can discuss as long as you want, at the end of the day its the truth.🤷♂️ 40 Years of machining .....👋
@@schwellenzaehler4564 In this case it was always a VW, and thus only suited for a slower pace. Now its just wearing a body kit so it can pretend to be a more impressive car while making no meaningful if any change at all to its performance. Seriously its lots of effort to make it look pretty, might help the machine have faster rapids, but its not going to make a huge difference if any difference at all to how that machine can perform - its only going to be able to mount a baby spindle that can't handle cutting loads significant enough anyway, the wheels themselves are on tiny little axles etc - there are so many factors that limit the performance of this machine to well below what you'd expect that beam to be able to handle!
No worries I got there. I am sitting at 2 CNC mills, one being a Millright MegaV and the other a Langmuir MR1. MegaV showed me that things like a rack/pinion or belt setup are great for fast travel speeds but the router spindles allong with tesselation can be a bit of an issue. Ball screws and linear rails along with a legitimate spindle really helps with being able to remove material.
Excited to see the rest of the build! Nice shirt too :)
Yeah same......so sad to you use rollers instead of proper rails......kinda aims besides the target if you spend all that time and money on a CNC why not spend 300$ more and get something decent? Also the 300 bucks will be spent on broken endmills real fast
@@Basement_CNC He described the reasoning behind his choice in the video. If you think it's such a bad choice, go make your own. I've got a Shapeoko with rollers and it works great
@@BPSspace Excellent product placement!
I for one believe the LGM-30 is a perfectly reasonable upgrade to your car.
Ah, someone who appreciates fast and effective transportation. 👋🏼
I love the POV shot at 6:40 that made me know how aluminium plates feel and what they see before being machined haha
Love the video Winston! I love watching you build and upgrade machines!
Awesome video. Can't wait to see the finished machine up and running.
Nice to see a new video, and digging the tee shirt.
Its great to see creators supporting each* other! 👍
Love the look of the machine, can't wait to see it in action!
Love the process.
I probably would have done those holes by just marking them with a spotting drill, take it to the drill press, and then cordless drill to tap. Easier to CAM and much less stressful than watching the machine do it. These days I do that for any flat piece I can fit on the drill press and which doesn't require precise depth holes.
So great to see you!
The wider stance of the top wheels are going to do WAY more for ridgity than extra wheels within the existing plate. 👍
hit the nail on the head , extra wheels waste of time
Looking awesome!
my old cnc router had vbearings, and they were just ok. but that was also 10 years ago when linear rails were $$, last year i upgraded it to linear rails and wow, for 100 bucks or so, its 1000x better. since you are doing this from scratch, did you price out the differences?
I did not because my goal here was mainly to explore improvements to an existing architecture with a simple change. At work I can play with a 4x4 linear rail + ballscrew Shapeoko 5 so I don't really crave the need for outright power or performance. Oh, and I also have the Neo to play with...
And as kind of a thought exercise, I wanted to see what an affordable Shapeoko could be. If you assume the linear rails for X and Y could be had for $150, plus labor and machining time, by the time you add retail markup you're going to have a Shapeoko that's ~$500 more expensive. That's basically the price delta you see between Shapeoko 4 and Pro. To me it would be less fun to try and build something that's essentially already been explored.
@@WinstonMakes for a production model sure, but this is more of a one off thing. dont get me wrong, your work is awesome and have watched every video lol. just figured if building something from scratch why not.
Oooh, now I'd love to do that to my Shapeoko 3....
Can you explain how the middle wheels are going to take any load? Maybe I just don't understand it. Is it just that as the carriage gets loaded and the outer wheels take some load, the plate flexes a bit and then gets slightly stiffer when the middle wheels start seeing load? I think It will get less than 2x as stiff when the middle wheels engage because they have less mechanical advantage. If I do understand it correctly, then I agree with the person who said they think that moving the wheels outwards will do more good than adding the middle wheels.
That's what I was thinking too. If the wheels are soft enough then maybe they will all engage, but otherwise two will just be going along for the ride.
Well, i`ve builded few CNC routers with linear shafts/rods (SBR16). These shafts are very cheap in comparison with Hiwin type profile rails and they are much better than wheels. I can agree with you- it`s possible to replace pricey ball screws with teeth belt. GT2 10mm wide tooth belt can do the job good, even during aluminum milling.
An early shapeoko 2 mod was bolting the 2 rails together to greatly increase rigidity.
Hey Winston, what resources do you recommend to learn more about speeds and feeds? I’m getting back into CNC and that’s been one of the most challenging parts trying to cut metal..
Loved the video, I especially like how you leave in your mistakes. It’s much more human and also gives valuable insight to help others not make similar mistakes, thank you!
So speeds and feeds is a potentially deep rabbit hole. I think it's always a good idea to have a good high-level understand of the theory, and what the terms actually mean. There's some good videos about it (NYC CNC and the like), and lots of websites do a decent job explaining it. It really depends on how you learn best (visual, written, etc).
From that baseline knowledge, I think you need to make a judgement call. You can totally nerd out with programs like gWizard, or spreadsheets, and deep-diving machinist forums to try and find the "optimal" feeds and speeds. But unless you're using an industrial machine, the theory can fall apart pretty quick. There's always exceptions and ultimately you need empirical testing to figure out what works best for your setup. I personally don't like going overboard, nerding out on these things.
What I do recommend is to use Speeds and Feeds knowledge as a guideline. Ex. If you're cutting an unfamiliar material, but you know there's a suggested range of surface speeds, you can work backwards to find a ballpark RPM to try first. You don't necessarily need to get every last cutting parameter perfect though. And remember, these are guidelines, not immutable laws of machining. Keep a couple cheap endmills on hand, and don't let the fear of getting something wrong paralyze you.
Great video!
Would you be able to share some info (or vid) on cleaning that bit with the sodium hydroxide?
I would go as wide on the positioning of the V-wheels as you can... but it is nice to have a few mm spare for changes etc.
It's a fun build. Who cares if it's not very stiff, it'll be light and that's what matters right? And maintenance is fun and it'll be nice for content.
Which software is used to render those models? Looks awesome!
One mistake that I made when building my CNC router was not making it easier to square up the X -Y axes. If you want .001" -.005" accuracy in your parts, you are unlikely to get that without making your X axis slightly adjustable.
Unfortunately on this architecture there's not a good way. I ended up using shims between the gantry extrusion and its endplates to square it up. There's also provisions for dual Y-axis homing for auto-squaring if I end up using a Shapeoko 5 controller to power this thing.
Please lose the V wheels. I built the original Shapeoko Kit around 10 years ago. It was so poor I built my own keeping the controller, and V wheels as part of the new design. My only regret was keeping the V wheels. They might be great for some applications like lasers but not where you're creating dust which can build up on the wheels and rails. At best they can cause erratic behavior at worst stall the machine. At this point linear rails are cheap enough and better in all possible ways but cost. IMHO the should be considered the only reasonable choice. A little background. I'm an electro-mechanical engineer with many years of experience and in an earlier life was a tool and die maker. I loved my home built CNC router but hated the problems I had with the V wheels. I tried all kinds of fixes to no avail. Keep up the great work. Small CNC machines have changed the world for makers.
Aligning 4 eccentric nuts a single rail seems like a pain. Would it be too complicated to have the wheels connected in pairs, able to pivot and auto adjust with a single eccentric nut per pair?
It's definitely possible if you throw two wheels on something like a rocker arm. The problem is that every linkage you add is a potential point of deflection. My feeling was that the simplicity of an eccentric nut + V-wheel on a solid plate would be the safest course of action.
may i ask, why make everything lightweight? isnt the point of a cnc to be super big and hefty and heavy and solid and ridgid and sturdy?
In general, weight is good. But the rigidity of a machine needs to be balanced with the sturdiness of the powertrain. V-wheels and belts are at best, "medium duty". Making the frame out of aluminum, or steel, or cast iron will make very little difference in the grand scheme of things here.
⭐🙂👍
Doesn't look ridiculous to me, perhaps a touch over polished for what it is, but no more than that. Don't imagine you or anybody else could make money today selling a tool that is simultaneously a bit too pricey and premium compared to the budget tool and way less performant than the more professional grade machines (unless you already have Apple's brand recognition and marketing department anyway).
1st viewer 🎉❤
2nd viewer, and a bunch of critisism....
All these cutouts at the profiles weaken the structure!
This is not a 3d printer, so its not nessesary nor useful, to lower the weight at any part!
Have you ever seen commercial routers and mills?
All this machines with the footprint of your machine have no such cutouts.
Because its a waste of time, material and effort!
Sorry, but i have to say when a construction is not good!
I think you missed one of my previous videos. I know I technically weakened the rail, but it's fun and looks cool. Plus, the material removal from the gantry is negligible in the grand scheme of things. The belts and wheels are ultimately the weakest point.
It's like saying a pencil should be made out of titanium for precise writing, when ultimately as long as the body is stronger than graphite it will work.
@WinstonMakes ignoring the science is not wise.
But if your goal is the cool look, that is no point.
All the best👋
Who says it is not also a 3d printer?!?! Plus with the limitations that will be inherent to the sort of spindle that thing can mount and the wheels it runs on I doubt its had much if any impact on performance while reducing the mass that the machine has to move around a tiny bit, so saving a bit of energy if the machine was going to be used often would make up for the time invested. (Not 100% it won't matter, but it looks like a pretty safe bet it won't have made a meaningful difference with all the other factors that look much more limiting. Also even if it has he gains a great excuse to reinforce it in a lightweight and stylish way)
Also plenty of commercial machines do have lightening features in places - not every machine of every scale will, and the method used to have both sufficient rigidity and lightness might not look the same as here - for instance T, C and I beam profiles are a stiffness in multiple directions to weight ratio saving measure, and they get everywhere! Ultimately no machine wants to pack excessive weight where it isn't actually required, that just makes it more expensive to run and likely limits the speed of its operation. So like all engineering its a compromise between cost to build, cost to run, and final performance.
@@foldionepapyrus3441 okay, its like you buy a VW Beetle and you say this my solution for F1 racing, family camping, and your distance for your dayjob.
This car fits only for one of this 3 cases, .....the last, and if you ever had done so, you never want this solution back for only one day.
Its a compromise, and for shure a very bad!
If i have the choice, i always look for the best possible way. You can do 3 d printing with a much more massive machine, ......but slower.
You can mill with a machine out of rubber nothing, the flexible parts cause the damage of the endmills.
Carbide mills are made for stiff machines and can not handle the flexing, cutting edges break ....or the complete mill.
Okay you can instead produce dust with smalest possible chips, but this is not milling!
This stupid cutouts are not useful, an you can discuss as long as you want, at the end of the day its the truth.🤷♂️
40 Years of machining .....👋
@@schwellenzaehler4564 In this case it was always a VW, and thus only suited for a slower pace. Now its just wearing a body kit so it can pretend to be a more impressive car while making no meaningful if any change at all to its performance.
Seriously its lots of effort to make it look pretty, might help the machine have faster rapids, but its not going to make a huge difference if any difference at all to how that machine can perform - its only going to be able to mount a baby spindle that can't handle cutting loads significant enough anyway, the wheels themselves are on tiny little axles etc - there are so many factors that limit the performance of this machine to well below what you'd expect that beam to be able to handle!
The downside to these mills is the lack of a true spindle. Using routers is an ever loving pain and they develop runout as the hour count goes up.
May I suggest reviewing the last 10 seconds of the video...
No worries I got there. I am sitting at 2 CNC mills, one being a Millright MegaV and the other a Langmuir MR1. MegaV showed me that things like a rack/pinion or belt setup are great for fast travel speeds but the router spindles allong with tesselation can be a bit of an issue. Ball screws and linear rails along with a legitimate spindle really helps with being able to remove material.