Kit versus full DIY with expert help

LowRider CNC Advice
1

I am a software engineer considering my first subtractive CNC. I’ve operated a Prusa 3D printer for a while and am very comfortable with that process.

But I don’t know about general CNC.

I’ve been reading various resources and saw a few kits that seemed interesting at a variety of price points, scaling from the LowRider up to the AltMill. I’ve been experimenting with a shaper origin, but I’ve been having some issues getting it to be as accurate as I would like.

I shared this with a friend who is very experienced in manufacturing, not as a CNC operator, but as an engineer; he has a PhD in mechanical engineering and knows a lot more than me about CNC; at work he has access to crazy CNC machines, and he has a few CNCs at home. I know one of his is a Tormach.

The largest bed size he has at home is 30 inches, and he’s told me for a while that he’s thought about building a larger one, a larger CNC capable of 48x48” and a fairly long Z of about 8".

He has offered to help me build a CNC he would design and spec parts for, saying it would take him about 30 minutes. He spoke of how CNC is simple kinematically and getting good accuracy is mostly going to be a matter of going slow enough for the material.

I showed him a few designs I had seen (LowRider and LongMill). He was not enthusiastic about EMT or extruded aluminum and instead started thinking immediately of parallel linear rails (Vevor for ~$100), motors and controllers, he said it would probably cost about seven or eight hundred dollars to build it that way and have a much more accurate and capable system.

…

On one hand, I am intrigued. Having expert help from a friend plus ending up with a better system for the money sounds great.

On the other hand, I feel like there must be things that I/we are missing. Features that are being overlooked. Maybe safety features or maintenance or maybe missing features like z touch, homing, axis limits, better handling of backlash…

If the difference in performance were as simple as specing a set of $100 linear rails instead of using $20 of EMT, I imagine that would be a common upgrade for the LowRider.

What am I missing? What features of a CNC machine are more difficult than expected to design?

1 Like
2

If your mate suggested those, you should not take anything he suggests for granted. Those Vevor rails are bottom shelf quality (as well as the spindles).

Also, the linear rails do not replace the EMT or extrusions… You still need something to fix them to.

He also does not seem to consider stiffness/twist on X with a big machine. I built the PrintNC and its beam is a 100x80x4mm stainless steel beam.

Designing a CNC in 30 minutes is possible, but it’s going to be bad. I’ve seen a lot of custom designs that in the end get scrapped because the creator actually didn’t know what was important.

2 Likes
3

I have a PhD too and I built a lowrider. A community that can both rally behind you and push you along as you learn the art of cnc as well as your machine nuances is often overlooked or undervalued. I find that community here. Having a completely bespoke system that you designed and are proud of can be nice (yay diy), but at some point the cnc product is the goal not just building a machine, so a kit or directions will get you there much faster.

The lowrider isnt going to cut steel easily, though someone did it. If you need more accuracy than a lowrider, then a 30 min diy system probably isn’t the answer either. You will probably need to spend a lot more to get a couple more zeroes in your tolerance and as you mentioned, good tolerances are about mastering the process for the material, so really more money buys more speed or just shinier parts. I wont begrudge you for taking the diy route. However there is a well established path here to build and use a cnc and some decent human beings that hang here that could help you out with your build or cheer you on with your adventure.

You have a knowledgeable friend though. I didnt know anyone with a cnc when i got my first kit, so you have an advantage many of us didnt, yet we got them working and have made some cool stuff. Keep at it.

2 Likes
4

What are you planning to produce with your CNC? If you are wanting to use it for wood, a 48” X 48” LR4 is very capable. If you are going to be working with 8” material you will want to use some form of a drop table to lower your stock to work on the surface of the material. I don’t know how you would be able to make an 8” cut if that was your intent. I have not seen any bits that will work with the router or spindles (65mm) the LR4 is designed to use.

A 48” X 48” LR4 would definitely not be a good choice if you are planning to machine metal though.

2 Likes
5

So my take on it is pretty simple. You can’t beat the performance per dollar of a LR4 for a larger format machine in wood and plastics. Not even close.

With that said it depends on what you are planing on making, how often, and what materials you will be using, and how much room you have (the LR4 comes of the table easily) on what machine you should invest in.

4 Likes
6

Another pro engineer here, I don’t have a PhD but as the only guy in an office of 7 people without one, I know just how little that means in the real world! Unless his PhD was directly applicable to CNC/motion control platform type stuff, I wouldn’t necessarily read too much into this.

That sounds like some epic hubris… There are SO many designs for CNC routers out there and a lot of them are woefully compromised or overbuilt in specific areas while lacking in others. Maybe he is some kind of mechanical engineering savant or has some specific amazing take on how to make one of these machines, but I’d suspect not.

I’d actually be pretty interested to see where he gets to in 30 minutes, my guess is rough sketch with some basic parts specced and it’ll be some typical fixed gantry variant using linear slides and ball screws which will end up having to use bottom-tier shitty parts to be remotely affordable, as well as huge motors to handle swinging around all the mass needed to make something like that rigid enough to cut well.

I don’t mean to sound so jaded but I think if there were some magical recipe out there to make it that easy, one of the many hundreds of CNC kickstarters would have stumbled across it by now.

To echo the rest of everyone else’s sentiments: Start with what you want the machine for and work backwards from there.

5 Likes
7

Thank you all for your responses.

I know my friend has built motion systems as part of his work/research, but I believe those have generally been to sample data in one to two dimensions. There is a significant difference between that and specializing in motion control.

His take is that aligning the linear rails accomplishes this. It’s how he dialed in his Tormach, by adding shims until the rails ran smoothly. And while stiffness can be an issue, he suggested that running slower will reduce deflection to match. Speed is nice, but as a hobbyist I’m not as constrained by speed, more by material costs.

And thanks to everyone who re-iterated in a more direct way what I have read but didn’t truly apply to me - first specify your goals.

My friend suggested what he would want in a machine, which I took as ideal specs - and maybe they would ultimately be that for me too, but I don’t really know at this stage.

This a chicken and egg problem.

Prior to purchasing a 3d printer (and other expensive tools), I have tried to come up with a list of at least 5 projects that tool would significantly enhance or make possible, as well as longer list of minor projects that would be fun to try/use to learn. Once I have the tool I try to use that tool ~daily for about a month.

Historically I have worked mostly with wood (sheet goods, 2x4/6 etc and some S4S lumber), a little with foam and clear polycarbonate, and done little with metal, in part due to it being harder to accurately process. If I had the ability to easily machine metal, would that change? Yes. But is it worth the cost? Perhaps not.

I believe the LR4 could handle most of what is on my current CNC list, so long as I build a table that can work on the ends.

There is still so much I don’t about CNC operation.

I think this is a great summary. Ryan is obviously a clear and concise communicator, otherwise this project wouldn’t exist.

1 Like
8

And that’s perfectly fine! The key thing is to be clear with your reasoning, not that you need to have a specific A or B reason. In your case, maybe the reason is because you want something to play with and learn about CNC, see what it can do for you etc.

That’s a great perspective to start from. I would say it can be a little prone to scope creep, but it answers some clear questions.

One thing I really like about the LR4 approach is that the main components don’t change that much as you change size. You need longer rails, wires, longer belts and to cut new strut plates but other than that the rest stays the same. If it were me I’d make a small one and use that for a while to see how I get on before committing to the space for a bigger one etc. If you find later that you need more size or have a specific project, re-configuring the machine to a larger format is quite straight forward and could even be done on a project-by-project basis with a bit of care. A smaller machine will always outperform a larger one in pretty much every way.

My eventual plan is to make one for my ~800x600 workbench but with belts/wires for a full 2450 x 1250 machine and then make another spare gantry in case I ever need to use that size. I don’t have space to have that set up long term but I frequently move the car out and cut down full size sheets just on a sheet of polystyrene on the floor etc.

I think the key here is that there are lots of options and lots of ways you can go.

2 Likes
9

As a retired toolmaker with 50 years total experience and 40 years of machining with cnc. You have to decide what kind of tolerance and speed your looking for. Next comes your pocketbook how much you want to spend. There different levels people are going to settle for. What kind of control FluidNC or a siemens cnc controller sinumerik, $100 or $5,000. I believe most of us here like to tinker / modify things not deep pockets. As for the Lowrider4 Ryan has quite a design I hate to think the amount of hours he has in it, all. The angles and tapers on parts to get them to 3d print a lot of work and a lot of failures along the way. I LOVE what Ryan has done. So if you want to design your own machine go for it and remember to share – rest of us might want to build one to.

2 Likes
10

Well You liked my sales pitch, so let me know give my opinion.

I like your friend, he sounds fun, I have certainly said I can design a machine in my sleep. If you want to hang out with a buddy and have a great time. I would absolutely design a machine together. It would be super fun, and it will probably work really good (big secret 3 axis CNC’s are easy to design). If you enjoy that time spent as much as the machine…pick a budget and build a machine together, and please share your journey here. There are a few fully custom machine in the forums. Two ways to build a good machine, a well-designed solution, or throw money at it.

Selfishly I would like to see you guys build a LR to have both your minds looking at our machine in action to build out or collective hive mind, :grin:.

I do think you are actually being sneaky and doing what I tell people to do at the trade shows we go to, “ask questions to the machine communities you like and see who you want to have in your corner when you hit the inevitable hole in your knowledge and need some help”. Hope you like us, relaxed, sometimes a little direct, but regular people using machines, not just a company designed to sell stuff, I much prefer making things to having to deal with the business side, but the business side lets me spend most of my time making things.

1 Like
11

We have cookies and barbecue! And goats!:goat:

1 Like