Which CNC Router?

[steve355]

Morning

Reading some of the previous posts, it sounds as though there are some people on here who do know about CNC, so I may as well ask! I am looking to buy a small, budget, CNC router for engraving and milling brass. Basically I want to engrave some small brass plates and cut them out of a (say) 2mm sheet. I’m not interested in CNC milling wood.

I have a moderate amount of hobby machining experience, owning a lathe, a couple of manual mills and surface grinder. I have been through the learning curve of trying to get small underpowered machines to cut metals and get a decent surface finish, so I am not completely naive.

My budget is - budget. Let’s say 2k absolute max all in.

It seems that, in the last couple of years, manufacturers have made quite an effort to install ball screws, linear rails, VFD, spindles, etc on budget machines. This is welcome. However, the rigidity of the budget machines looks extremely flimsy at best.

I’ve watched James Dean Designs videos, which gives an idea of what’s on the market, however, the bloke is more of a sales agent for the companies who make the machines, and doesn’t really have a Scooby Doo about machining. However, he does show some budget machines doing some metal cutting, getting fairly questionable results, but I’m sure with some tweaking I could do a lot better. Which makes me think that for my purposes, one of those machines might not be a bad idea.

A few, I’ve been looking at are:
The new Genmitsu 3030 prover max (moving bed, gantry looks extremely flimsy with 4 screws holding it on)
Foxalien Vasto - flimsy bed, moving gantry.
Then the JFT 3040 on Amazon which looks more sturdy.
Anolex.

Otherwise, it’s an area I’m completely completely new to and I don’t really know what I’m looking for. I’m fully aware that these machines are basically toys.

Any thoughts anybody has would be very welcome.

Cheers
Steve

 

[Ollie78]

For 2k you should be able to find something that will do the job.
The same principles apply if its for cutting wood or metal.
The frame must be strong and stiff in all directions.
You want profiled linear rails (or round ones at a push) not wheels running on an extrusion.
Ballscrews are the way to go, rack and pinion is OK as well. Not belts.

The machines that you mention are not that great if you look at how they are made.
I am not against these machines as such and people can do great work on them but I also know a lot of times they will soon find the limitations and upgrade.
Having a machine that requires constant fiddling can easily put you off.

A way to look at how good a machine is in the spec sheet is the maximum cut speed and rapid speed along with the repeatable accuracy figures.
If the manufacturer is happy to let the machine go fast then it is probably built well enough to cope. You may not need it's max speed but it is a way to compare machines performance.

A fixed gantry design can be stiffer but they are generally smaller as they become impractical in bigger sizes, so if you only need small area this might be a way to go.

You can find second hand machines with pretty good spec, the issue can sometimes be the older control software. It is possible to upgrade this to a modern motion control system with UCCNC and an ethernet controller. Depends on how much messing about you are willing to do to save money.

Ollie

 

[steve355]

For 2k you should be able to find something that will do the job.
The same principles apply if its for cutting wood or metal.
The frame must be strong and stiff in all directions.
You want profiled linear rails (or round ones at a push) not wheels running on an extrusion.
Ballscrews are the way to go, rack and pinion is OK as well. Not belts.

The machines that you mention are not that great if you look at how they are made.
I am not against these machines as such and people can do great work on them but I also know a lot of times they will soon find the limitations and upgrade.
Having a machine that requires constant fiddling can easily put you off.

A way to look at how good a machine is in the spec sheet is the maximum cut speed and rapid speed along with the repeatable accuracy figures.
If the manufacturer is happy to let the machine go fast then it is probably built well enough to cope. You may not need it's max speed but it is a way to compare machines performance.

A fixed gantry design can be stiffer but they are generally smaller as they become impractical in bigger sizes, so if you only need small area this might be a way to go.

You can find second hand machines with pretty good spec, the issue can sometimes be the older control software. It is possible to upgrade this to a modern motion control system with UCCNC and an ethernet controller. Depends on how much messing about you are willing to do to save money.

Ollie

Thanks Ollie

The reason I was looking at the Foxalien Vasto is that it has the ball screws/ linear rails etc and this seems like a good start. It seems like the manufacturers are listening and there’s a lot of competition that seems to be driving improvements.

However the rigidity is likely to be questionable - but it may be easier to strengthen it up if necessary than try to make an older machine compatible with modern standards.

The thing I can’t get my head around it the crazy RPMs. I spent ages trying to get my horizontal mill to turn at 80rpm, so it didn’t make burnt chips when cutting tool steel. Then these CNC machines run at 10,000 RPM, with the kids getting excited about how fast they can make them go. I suppose I use a 4” diameter slitting saw whereas the CNC uses a 1/8” end mill. I’d better get the speeds and feeds calculator out.

We’ll see. I am quite keen on opening a box and having a machine to work out how to use rather than another refurb project….

Steve

 

[Steve Mehrlich]

great advice from ollie and to be fair the fox alien vasto looks to a be a fairly decent option, linear rails, ball screws etc.

i bought myself a queen bee pro 3 years ago which is based on c beam ali extrusion with lead screws and a 1.5kw spindle which im now in the process of upgrading to ball screws and swapping out the x gantry. these desktop type cnc's suffer from torsional twist at the spindle when making deep / agressive cuts.
the gantry set up on that fox alien looks to be a decent design. the vertical gap between the top and bottom rail should reduce that torsional twist. the only thing id question on the fox alien is the controller and spindle. i spent years faffing with a little 3018 pro and again added linear rails and beefed up the gantry plates and x axis but always had issue with the spindle and its optimum speed depending on cut depth, feed rate and material, the twisting of a potentionometer to get the speed and subsequent noise of a cut just right gets boring, that combined with brushes going made it hard work.
the openbuilds website is worth a look at as is ratrig but 2k may get swallowed up pretty quick.

definitley steer clear of v wheels, lead screws are okay but require a little maintenance of the delrin nut blocks. definitely linear rails and steer clear of those moved by belts. outside of that download the hobby version of fusion 360 and have a play.

the problem for me is what i thought i would be machining and at what size that would be when i bought the machine are very different now and the more i do it the more i love it, sad but true. if there were a platform where i could volunteer/preach the joys of cnc id be there in a flash so if you need any more advice let me know.

cheers

 

[steve355]

great advice from ollie and to be fair the fox alien vasto looks to a be a fairly decent option, linear rails, ball screws etc.

i bought myself a queen bee pro 3 years ago which is based on c beam ali extrusion with lead screws and a 1.5kw spindle which im now in the process of upgrading to ball screws and swapping out the x gantry. these desktop type cnc's suffer from torsional twist at the spindle when making deep / agressive cuts.
the gantry set up on that fox alien looks to be a decent design. the vertical gap between the top and bottom rail should reduce that torsional twist. the only thing id question on the fox alien is the controller and spindle. i spent years faffing with a little 3018 pro and again added linear rails and beefed up the gantry plates and x axis but always had issue with the spindle and its optimum speed depending on cut depth, feed rate and material, the twisting of a potentionometer to get the speed and subsequent noise of a cut just right gets boring, that combined with brushes going made it hard work.
the openbuilds website is worth a look at as is ratrig but 2k may get swallowed up pretty quick.

definitley steer clear of v wheels, lead screws are okay but require a little maintenance of the delrin nut blocks. definitely linear rails and steer clear of those moved by belts. outside of that download the hobby version of fusion 360 and have a play.

the problem for me is what i thought i would be machining and at what size that would be when i bought the machine are very different now and the more i do it the more i love it, sad but true. if there were a platform where i could volunteer/preach the joys of cnc id be there in a flash so if you need any more advice let me know.

cheers

Hi Steve

The spindle will be the first upgrade. For the most part I’ll be cutting shapes out of and engraving brass and ideally O1 if I can do it.

I used the free version of Onshape for a while, but I’ll be needing CAM so I recently switched to FreeCAD. The learning curve is very, very steep but I’m getting there.

What are you using yours for?

Steve

 

[Steve Mehrlich]

steve

i can highly recommend fusion 360 which covers both the design and gcode for your toolpaths. i started on vetric but even that seems archaic when you get the hang of fusion.

i tend to design in fusion and create my toolpaths and then export the g code to open builds cam which then sends it to openbuilds blackbox which controls the stepper motors and spindle.
if you can find yourself a rigid second hand machine i cant recommend enough the combination of fusion 360 an openbuilds black box controller and then openbuilds cam. both programs are free, the blackbox will set you back £170 but its pretty much plug and play and ive never suffered any emi with limit switches etc.

work wise anything from chopping boards, relief models, making new plates for the new gantry in phenolic to half tone wall art, egg cups, salt sellers anythign i can avoid buying. ive made a couple of pairs of speakers recently .

 

[steve355]

Very nice Steve.

The thing I really, ideally want to be able to do is to cut plane blades out of O1 tool steel. Damn things take me ages to make by manual milling machine. Annealed O1 is only 30% harder than brass, as long as it doesn’t work harden!

What do you think of this one? Too expensive for me, but…. You’ve got to haggle.

EBay machine

 

[Steve Mehrlich]

scousers love a haggle. mechanicaly that looks mustard and he looks to be getting good cuts with the what look like clamps he's making. anyone with a dial indicator probably knows what he's doing!! and its always nice to see a bag full of bits, just like my garage.

id question what kind of controller he's using currently, size of stepper motors etc. steppers look like nema 23's. the power they use will make a difference to the type of controller you use but if its in use now then it looks pretty much ready to go, may not need to change anything apart from the space in your brain that allows you to learn a new skill combining 3d modelling and feeds and speeds

 

[steve355]

scousers love a haggle. mechanicaly that looks mustard and he looks to be getting good cuts with the what look like clamps he's making. anyone with a dial indicator probably knows what he's doing!! and its always nice to see a bag full of bits, just like my garage.

id question what kind of controller he's using currently, size of stepper motors etc. steppers look like nema 23's. the power they use will make a difference to the type of controller you use but if its in use now then it looks pretty much ready to go, may not need to change anything apart from the space in your brain that allows you to learn a new skill combining 3d modelling and feeds and speeds

Interesting, isn’t it. It looks to me like he might have tried upgrading it, put a big meaty bed on it and a plate on the back to strengthen up the gantry. It seems to have one access with an upgraded ball screw, but the other two axis are not. For some reason he stopped upgrading it.

Here’s another one…. EBay machine 2

 

[Steve Mehrlich]

i think you can just make out ball screws on all axis on the first one but not sure about linear rails, yes on the x but not sure on the z or y.
second machine looks to have linear guides, round bar, rather than rails. im open to correction but id think linear rails would be a better option than the guides.

2nd machine z axis looks a little flimsy. the weight of a spindle combined with whanging a bit though steel is going to give you some deflection , chatter headaches in my opinion.

 

[steve355]

i think you can just make out ball screws on all axis on the first one but not sure about linear rails, yes on the x but not sure on the z or y.
second machine looks to have linear guides, round bar, rather than rails. im open to correction but id think linear rails would be a better option than the guides.

2nd machine z axis looks a little flimsy. the weight of a spindle combined with whanging a bit though steel is going to give you some deflection , chatter headaches in my opinion.

2nd machine , Z axis does look flimsy, but the machine is a third of the price of machine one! All for me to upgrade that isn’t a problem. And it’s in London rather than Liverpool, which is definitely a bonus for me.

All of those 6040 machines that I can see have warnings in their descriptions about software incompatibility, needing 32 bit operating systems et cetera, and old parallel port, drivers, or USB drivers. I imagine that’s why machine two comes with its own computer. Would that openbuilds control box that you mentioned solve that problem?

 

[Ollie78]

Thanks Ollie

The reason I was looking at the Foxalien Vasto is that it has the ball screws/ linear rails etc and this seems like a good start. It seems like the manufacturers are listening and there’s a lot of competition that seems to be driving improvements.

However the rigidity is likely to be questionable - but it may be easier to strengthen it up if necessary than try to make an older machine compatible with modern standards.

The thing I can’t get my head around it the crazy RPMs. I spent ages trying to get my horizontal mill to turn at 80rpm, so it didn’t make burnt chips when cutting tool steel. Then these CNC machines run at 10,000 RPM, with the kids getting excited about how fast they can make them go. I suppose I use a 4” diameter slitting saw whereas the CNC uses a 1/8” end mill. I’d better get the speeds and feeds calculator out.

We’ll see. I am quite keen on opening a box and having a machine to work out how to use rather than another refurb project….

Steve

You are correct in the fact that the machines being offered in this market sector now are superior to those from only a couple of years ago.

I think mechanically they are not too bad but it is important to check the likely weak points in the chain from design to machining.
The motion control software/controller does make a difference and many of the cheap Chinese stuff uses stolen mach 3 which is now unsupported or grbl which is not the best.
I upgraded my machine to UCCNC and a UC300 eth which is faultless.

I agree with the someone above who recommended Fusion 360, it is so much easier to use than Freecad and much nicer than Sketchup as well.
I actually use vcarve pro to do the toolpaths but Fusion is the best modeling software I have used. One day I may upgrade to Aspire and try the modelling in that.

You are right about the spindle speeds, this can indeed a problem the spindles designed for running at higher speeds like 24000rpm will lose torque in the lower rev ranges to quite an alarming degree. I was advised not to go below half the max speed so 12000 rpm on my spindle which is fine because I am doing wood and mostly run at about 18000rpm.
I have only recently actually started to pay proper atttention to the recommended chip load of the endmills and as a result have ended up using more single flute cutters and increasing the mm/min in order to actually get to the sweet spot.
Some of the multi flute cutters need incedibly fast linear speeds to get the chip load right.

I don`t know much about metal milling but you may need a spindle with a lower speed higher torque than is commonly supplied.

Ollie

 

[steve355]

You are correct in the fact that the machines being offered in this market sector now are superior to those from only a couple of years ago.

I think mechanically they are not too bad but it is important to check the likely weak points in the chain from design to machining.
The motion control software/controller does make a difference and many of the cheap Chinese stuff uses stolen mach 3 which is now unsupported or grbl which is not the best.
I upgraded my machine to UCCNC and a UC300 eth which is faultless.

I agree with the someone above who recommended Fusion 360, it is so much easier to use than Freecad and much nicer than Sketchup as well.
I actually use vcarve pro to do the toolpaths but Fusion is the best modeling software I have used. One day I may upgrade to Aspire and try the modelling in that.

You are right about the spindle speeds, this can indeed a problem the spindles designed for running at higher speeds like 24000rpm will lose torque in the lower rev ranges to quite an alarming degree. I was advised not to go below half the max speed so 12000 rpm on my spindle which is fine because I am doing wood and mostly run at about 18000rpm.
I have only recently actually started to pay proper atttention to the recommended chip load of the endmills and as a result have ended up using more single flute cutters and increasing the mm/min in order to actually get to the sweet spot.
Some of the multi flute cutters need incedibly fast linear speeds to get the chip load right.

I don`t know much about metal milling but you may need a spindle with a lower speed higher torque than is commonly supplied.

Ollie

>>>I don`t know much about metal milling but you may need a spindle with a lower speed higher torque than is commonly supplied.

That’s an interesting topic. On my manual horizontal mill, I ended up replacing the motor with a 10-pole motor that could deliver low rpm high torque. VFDs can’t do it. Historically, low rpm, high torque power was often delivered using gearing. Unfortunately, my old horizontal mill lost its geared motor years ago.

 

[--Tom--]

For O1 you may well be better with a smaller metal cutting bandsaw for blade profiling that can then be tuned on a linisher.

Or developing CAD and outsourcing to a waterjetter if you’re doing any volume of complex shapes

 

[Jamesc]

You don't mention how big your workpieces are, but I would have thought you would be better off looking at a CNC mill rather than a router.
There a ex school and college Denford and Boxford mills regularly on Ebay. Many are good solid machines I personally like the look of some of the Denford VMC range. Older ones regularly go for around £1000 - 1200. as the electronics are proprietry. There are also a number of companies advertising a conversion service to Mach 3 or LinuxCNC for around £700 so could be a way of getting a well built machine in your budget.

 

[Ollie78]

Out of budget of course but it is pretty cool regardless.
Interestingly max spindle speed of 6400rpm, 0.005 precision. Looks like a really nice little machine.

 

[steve355]

You don't mention how big your workpieces are, but I would have thought you would be better off looking at a CNC mill rather than a router.
There a ex school and college Denford and Boxford mills regularly on Ebay. Many are good solid machines I personally like the look of some of the Denford VMC range. Older ones regularly go for around £1000 - 1200. as the electronics are proprietry. There are also a number of companies advertising a conversion service to Mach 3 or LinuxCNC for around £700 so could be a way of getting a well built machine in your budget.

I would say that mostly I’m looking to process 10” x 18” flat stock or something like that - typically up to 1/8” thick, routing shapes into it. If you imagine a whole bunch of plane blades laid out on a flat piece of metal, cut out with a router leaving dressups. Then I’d grind the taper on them using a different machine. That’s one use case anyway but the others are similar. It’s routing rather than vertical milling. I already have a vertical mill (non-Cnc).

Another interesting one on Facebook marketplace…… Machine 3

Exeter!

 

[Ollie78]

I would say that mostly I’m looking to process 10” x 18” flat stock or something like that - typically up to 1/8” thick, routing shapes into it. If you imagine a whole bunch of plane blades laid out on a flat piece of metal, cut out with a router leaving dressups. Then I’d grind the taper on them using a different machine. That’s one use case anyway but the others are similar. It’s routing rather than vertical milling. I already have a vertical mill (non-Cnc).

Another interesting one on Facebook marketplace…… Machine 3

Exeter!

I wonder if you have considered a water jet cutting machine. Flipping amazing at cutting out shapes from almost any material. I have a vague "might get around to it " idea of making my own one. Looked into it and found some smart people making them with pressure washer components and clever nozzles. The most famous hobby type machine is called Wazer.
Yesterday I watched a video where a couple of guys built their own wire EDM machine with a 3d printer and a bunch of home made circuit boards.

Ollie

 

[steve355]

I wonder if you have considered a water jet cutting machine. Flipping amazing at cutting out shapes from almost any material. I have a vague "might get around to it " idea of making my own one. Looked into it and found some smart people making them with pressure washer components and clever nozzles. The most famous hobby type machine is called Wazer.
Yesterday I watched a video where a couple of guys built their own wire EDM machine with a 3d printer and a bunch of home made circuit boards.

Ollie

I tried looking for a second hand one and I found a Sodastream

 

[steve355]

Well….. after loads of reading and watching YouTube videos, I think I’ve worked out the best way to go if I want to mill metal. It’s to get a Chinese 6040 frame with linear rails and ball screws….

Frame

Then some closed loop motors ……

Motors

A controller ….

Controller

And a spindle ….

Spindle

Then replace the aluminium bed with a lump of plate steel. That seems like the best way for my budget of about 2K to get a CNC machine that will mill soft metals properly, and might stand a chance of doing my tool steel plane blades.

It’s a project! I need to decide whether to take the plunge. The consumer routers, by comparison, are noddy.

Steve