Who's Using or Buying a CNC Router

[jlabrasca]

We are kind of all over the place it in this thread.

Has the OP said what kinds volume/material he expects to run? I don't see it in this thread (but I am preoccupied and scatterbrained).

The OP mentions an upgraded Shark, as featured in the maker videos from this kid



He also mentioned wanting to build his own CNC.

So, lets assume he wants something with the Shark form factor, and wants to build from parts or a kit

As others have mentioned: In this space, at the high end, there is CNC Router Parts. Brian Oltrogge's set up is a pretty good place to start when configuring

https://www.cncrouterparts.com/grunblau-platform-cnc-parts-bundle-p-219.html

Finished and ready to cut, this one will run $4K+

there really isn't a clear bottom to the kit-CNC space. If you shop on Banggood or Aliexpress (or Amazon), you can find very inexpensive CNC routers.

I had a budget, so CNC Router parts and Shopbot were out of reach. I narrowed it down to X-carve or Shapeoko because there are lots of them in the field, the companies appear pretty healthy, and there are active of users's groups to whom to turn when trying to stretch. The folks at Carbide 3D got my money because I wanted a more robust machine than the X-carve.

I find that double-side tape (the crepe-backed stuff) means I don't need to use tabs often.

I don't use D/S tape much anymore. I pencilled it out, and if you buy in bulk/on sale Blue-tape + CA works out to be cheaper than the cheapest D/S tape I could find. I also find it easier to get the work piece flat on the spoil board using the Blue-tape + CA than I ever did using even the thinnest Nitto or Peachtree D/S tape.

I do still tab or onionskin small parts, where the taped surface area is small. I've launched a few postage-stamp sized pieces across the shop when the tape didn't hold.

https://community.carbide3d.com/t/masking-tape-super-glue-to-hold-the-workpiece/1671/4

 

[KC3KNM]

We are kind of all over the place it in this thread.

Has the OP said what kinds volume/material he expects to run? I don't see it in this thread (but I am preoccupied and scatterbrained).

The OP mentions an upgraded Shark, as featured in the maker videos from this kid



He also mentioned wanting to build his own CNC.

So, lets assume he wants something with the Shark form factor, and wants to build from parts or a kit

As others have mentioned: In this space, at the high end, there is CNC Router Parts. Brian Oltrogge's set up is a pretty good place to start when configuring

https://www.cncrouterparts.com/grunblau-platform-cnc-parts-bundle-p-219.html

Finished and ready to cut, this one will run $4K+

there really isn't a clear bottom to the kit-CNC space. If you shop on Banggood or Aliexpress (or Amazon), you can find very inexpensive CNC routers.

I had a budget, so CNC Router parts and Shopbot were out of reach. I narrowed it down to X-carve or Shapeoko because there are lots of them in the field, the companies appear pretty healthy, and there are active of users's groups to whom to turn when trying to stretch. The folks at Carbide 3D got my money because I wanted a more robust machine than the X-carve.



I don't use D/S tape much anymore. I pencilled it out, and if you buy in bulk/on sale Blue-tape + CA works out to be cheaper than the cheapest D/S tape I could find. I also find it easier to get the work piece flat on the spoil board using the Blue-tape + CA than I ever did using even the thinnest Nitto or Peachtree D/S tape.

I do still tab or onionskin small parts, where the taped surface area is small. I've launched a few postage-stamp sized pieces across the shop when the tape didn't hold.

https://community.carbide3d.com/t/masking-tape-super-glue-to-hold-the-workpiece/1671/4

I read the bits about things not being stiff enough and having too much play from the OP. There were some mentions of aluminum, but I might have interpreted it wrong. My bad.

 

[Random Flying Object]

Not to talk you out of building or buying a machine, but you may consider looking for a Maker Space near your location that may already have the tools you need.

 

[JohnCoker]

I also find it easier to get the work piece flat on the spoil board using the Blue-tape + CA than I ever did using even the thinnest Nitto or Peachtree D/S tape.

Neat technique, that's great to know about.

 

[ericm541]

What kind of tolerances are you holding? +/- ten thou isn’t incredibly tight. Is this in aluminum? I figure it’d be a butt pain to cut ply to anything super tight based on the nature of the material.

While routers are designed for side load, they’re not made to handle the loads seen in aluminum (as modest as they are). Not to mention that routers are typically much higher RPM than spindles. Too high of an spindle speed can be as bad or worse than too low. The Dewalt 611, for example, is 16-27k RPM. Going for 280 SFM in 6061 with a 1/8” endmill, you’re shooting for around 8500 RPM. You’ll burn through tooling much quicker than you’ll need to with spindle speeds that high and likely end up with some poor cuts.

It’s not that insane to have a VMC in a garage. When I was in high school I ran parts for a small RC car company based in a guy’s garage. He had a Fadal 914-15 VMC that he bought after fighting with a router for a few years. I’m not saying you need a full machine shop to cut aluminum properly, but a CNC router just doesn’t lend itself to cutting anything more than plastics and wood well. We had an industrial AXYZ router in my high school’s shop class and it still struggled with soft metals despite weighing a few thousand pounds. Having your spindle on a gantry is great for large workspaces and thin materials, but it hurts you when it comes to stiffness.

To reiterate again, I’m not saying routers are bad or worthless. If you’re not cutting anything incredibly hard or aren’t worried about stiffness, they’re fine. The OP seemed to stress the want for stiffness and I’m not sure he’d be satisfied with a router.

The key is your tooling along with feeds and speeds (feed rate and spindle RPMS) the recommended chip load for the material your cutting is a good starting point. You don't need a heavy duty rugged machine for aluminum i've cut aluminum with a hand router single flute bit, i had to use a variable speed plug in and the cut wasn't bad and the bit survived. It was for a piece of 1/8" 6061 aluminum 8" motor retainer plate with 5 holes. Anyway you can get the variable speed plug-in at harbor freight but i would find good router bit.

In my previous post i had couple examples of the aluminum i've cut on my machine, the router bit is single flute and leaves a mirror finish its the same bit i use to cut acrylic. Got it on amazon after checking the manufactures specifications. My cheap tabletop asian CNC cuts aluminum nicely with perfect tolerance.

 

[LOC]

Shopbot PRT Alpha Pro 10x6' table and spindle. The only thing more accurate is the CNC laser... but that can't do what the Bot can do... It is overkill. But, it also has untapped potential I'm hopping to play with. In my industrial experience... you get what you pay for....

 

[ericm541]

Shopbot PRT Alpha Pro 10x6' table and spindle. The only thing more accurate is the CNC laser... but that can't do what the Bot can do... It is overkill. But, it also has untapped potential I'm hopping to play with. In my industrial experience... you get what you pay for....

Get what you pay for totally agree, bought a $500.00 laser engraver, works nice for engraving doesn't cut for sh*t, had to modify it 3 ways too.

 

[KC3KNM]

The key is your tooling along with feeds and speeds (feed rate and spindle RPMS) the recommended chip load for the material your cutting is a good starting point. You don't need a heavy duty rugged machine for aluminum i've cut aluminum with a hand router single flute bit, i had to use a variable speed plug in and the cut wasn't bad and the bit survived. It was for a piece of 1/8" 6061 aluminum 8" motor retainer plate with 5 holes. Anyway you can get the variable speed plug-in at harbor freight but i would find good router bit.

In my previous post i had couple examples of the aluminum i've cut on my machine, the router bit is single flute and leaves a mirror finish its the same bit i use to cut acrylic. Got it on amazon after checking the manufactures specifications. My cheap tabletop asian CNC cuts aluminum nicely with perfect tolerance.

It's not pure RPM that's important, your SFM (surface feet per minute) is a calculation based off of tool diameter and RPM. The 280 SFM I mentioned before is with the assumption that one's using a HSS endmill (in 6061) as they're cheaper and a little more forgiving than carbide. SFM doesn't care how many flutes your endmill has, that comes into play when determining chip load. Too high of an SFM, and you're generating a ton of heat and wearing the endmill to an unnecessary degree. For cutting here and there, you probably don't have to be too close, but if you want your tooling to last it pays off to take this into consideration. It's easy to just look at a tooling manufacture's chart and go from there, but if you have no idea of your spindle speed (which is hard with a router using an external speed control, unless you've got a tach), you're going to struggle.

Here's a better explanation of it all than I can provide.
https://www.destinytool.com/speeds--feeds.html

What do you consider "perfect tolerance"?

 

[Dave A]

What kind of tolerances are you holding? +/- ten thou isn’t incredibly tight. Is this in aluminum? I figure it’d be a butt pain to cut ply to anything super tight based on the nature of the material.

While routers are designed for side load, they’re not made to handle the loads seen in aluminum (as modest as they are). Not to mention that routers are typically much higher RPM than spindles. Too high of an spindle speed can be as bad or worse than too low. The Dewalt 611, for example, is 16-27k RPM. Going for 280 SFM in 6061 with a 1/8” endmill, you’re shooting for around 8500 RPM. You’ll burn through tooling much quicker than you’ll need to with spindle speeds that high and likely end up with some poor cuts.

It’s not that insane to have a VMC in a garage. When I was in high school I ran parts for a small RC car company based in a guy’s garage. He had a Fadal 914-15 VMC that he bought after fighting with a router for a few years. I’m not saying you need a full machine shop to cut aluminum properly, but a CNC router just doesn’t lend itself to cutting anything more than plastics and wood well. We had an industrial AXYZ router in my high school’s shop class and it still struggled with soft metals despite weighing a few thousand pounds. Having your spindle on a gantry is great for large workspaces and thin materials, but it hurts you when it comes to stiffness.

To reiterate again, I’m not saying routers are bad or worthless. If you’re not cutting anything incredibly hard or aren’t worried about stiffness, they’re fine. The OP seemed to stress the want for stiffness and I’m not sure he’d be satisfied with a router.

I'm not looking to cut aluminum that much, maybe thin stock, but mostly plywood. Some of the desktop machines do have more play than others.
I do not care too much if the "thread is all over the place". Just wanted to see what folks are using, how they use it and feedback. You all have given me good responses, thank you.

 

[mkadams001]

We built a CNCRouterParts 4' x 4' pro with spindle. We usually import eps files to VCarve Pro to set tool paths and Mach3 as the controller. I have used Fusion 360 for 3d carving. Anyway, this is great machine we run it nearly everyday for hours at a time. Plus, it was a lot of fun to build.

 

[ericm541]

I don't use SFM calcualtions there are alot of variables, determining a good chip load has been a a great starting point for me from there Inches or millimeters per minute is what is programmed and the machine reads. It's easy to take what the manufactures suggest because good bits aren't cheap you can ruin them quickly from there i do the math to calculate RPM and IPM. When i used hand router with a variable speed plug-in the dial is in percentages, the hand router has one speed , i did the math to get it close enough i didn't struggle. perfect tolerance for me is, i programmed an 6" diameter centering ring i measured it with digital calipers and it came out as 6" oh, the center hole and drill points were perfect aswell.

 

[ericm541]

I'm not looking to cut aluminum that much, maybe thin stock, but mostly plywood. Some of the desktop machines do have more play than others.
I do not care too much if the "thread is all over the place". Just wanted to see what folks are using, how they use it and feedback. You all have given me good responses, thank you.

Yeah i like to cut shiny stuff, should determine how much you willing to spend on a CNC first then go from there. Heres the machine i have, there are issues people have with these machines i got over them and started making my own stuff.

https://www.omiocnc.com/products/x8-2200epl.html

 

[KC3KNM]

I don't use SFM calcualtions there are alot of variables, determining a good chip load has been a a great starting point for me from there Inches or millimeters per minute is what is programmed and the machine reads. It's easy to take what the manufactures suggest because good bits aren't cheap you can ruin them quickly from there i do the math to calculate RPM and IPM. When i used hand router with a variable speed plug-in the dial is in percentages, the hand router has one speed , i did the math to get it close enough i didn't struggle. perfect tolerance for me is, i programmed an 6" diameter centering ring i measured it with digital calipers and it came out as 6" oh, the center hole and drill points were perfect aswell.

You’re saying you make perfect parts? I’m going to leave this one alone. Have a good night.

 

[jlabrasca]

I'm not looking to cut aluminum that much, maybe thin stock, but mostly plywood. Some of the desktop machines do have more play than others.
I do not care too much if the "thread is all over the place". Just wanted to see what folks are using, how they use it and feedback. You all have given me good responses, thank you.

You are welcome?

 

[ericm541]

You’re saying you make perfect parts? I’m going to leave this one alone. Have a good night.

I do make perfect parts, which is the point of owning and operating a CNC machine.By the way the CNC i own is for business, i just use it for personal rocketry stuff

 

[KC3KNM]

I do make perfect parts, which is the point of owning and operating a CNC machine.By the way the CNC i own is for business, i just use it for personal rocketry stuff

The company I work for has a high precision wire EDM that uses wire with a .0008” diameter (https://www.mitsubishi-edm.de/en/mx600.html). It’s used to cut very small parts for watches and electronics. It doesn’t make perfect parts. With such a machine we’re within .00005” or so, and that requires an insane amount of work.

I’m sure there’s something being lost between us here in the language being used, but a desktop machine will not make “perfect” parts.

 

[ttabbal]

I suspect they mean "perfect for what I want them to do". Or what we might call "within tolerance".

There are no perfect parts. Just parts within the required tolerance. For things like centering rings, fins, etc. that's probably something huge like +/-0.050".

 

[KC3KNM]

I suspect they mean "perfect for what I want them to do". Or what we might call "within tolerance".

There are no perfect parts. Just parts within the required tolerance. For things like centering rings, fins, etc. that's probably something huge like +/-0.050".

That’s what I’m assuming, but saying “I make perfect parts” isn’t really a logical thing for a machinist to say.

 

[Dave A]

I originally started the thread for a CNC ROUTER, NOT a CNC machine. I programmed them and used them in college. I've used manual milling machines a fair amount. I fully understand they are the ultimate but I am more interested in cutting fins and rings for rockets.
I am more interested in what CNC Routers you are using. As Nat stated high accuracy not needed because of varying tolerances it tubes and slots but a durable machine that will last longer is a plus.
Just gathering data on which machine to buy.

 

[ttabbal]

For routers working with wood and plastics, you can probably use most anything with good results. Longevity is likely to be better with the heavier, stronger machines. That said, decent maintenance on a small belt driven machine will likely be just fine. I own a Bridgeport mill with a DRO and I'm still buying a CNC router for this sort of thing. When I need more accuracy, or harder material, I'll use the BP. I'll likely experiment with Aluminum on the CNC, but I expect it will mostly see wood and plastic. Maybe some fiberglass once I get dust collection sorted out to my satisfaction.

Part of my reason for the higher end machine is that I know who I am, maintenance might not be 100%. So I want something a little more durable. I am also a bit impatient, so being able to take deep cuts accurately is a bonus. I'll be able to give more info about the Millright machine over the next week or so as it should be delivered today. It's a kit though, so it will take some time to assemble.

The biggest thing with any machine is going to be learning the limits and feed/speed the machine likes to work in, then staying in those limits. It's no different from a "real" CNC machine, or even a 3D printer in that respect. It's a little easier for these sorts of things as most of the time you don't mind if it takes a bit longer as it's automated.

 

[JohnCoker]

There are no perfect parts. Just parts within the required tolerance. For things like centering rings, fins, etc. that's probably something huge like +/-0.050".

Even with plywood, you can definitely notice the fit change with a difference of 0.005". I don't have familiarity with many difference machines, but I believe it's reasonable to expect much tighter tolerances than 0.05" for rocket parts.

Whether you need it or not is perhaps another question, but the satisfaction of having things fit together "perfectly" is a big part of the reason for getting into CNC in the first place.

 

[jlabrasca]

Even with plywood, you can definitely notice the fit change with a difference of 0.005". I don't have familiarity with many difference machines, but I believe it's reasonable to expect much tighter tolerances than 0.05" for rocket parts.

On the CNC forums posts about "tolerances" often go the way posts about adhesives go here. All kinds of "helpful" advice about squaring and tramming, about the best aftermarket collets, the optimal exposed length of different end mills for different surface speeds in different materials...

Probably more than the OP needs to know right now, but I find that cutting multiple copies of a part from a large piece of plywood stock is the biggest challenge for working with my set up. x-y tolerances and repeatability might be ±0.05mm or better, but it gets sloppy in the z.

Small individual parts, cut from brass or aluminum, I can get 3-axis tolerances tighter than I can measure with my cheap indicators and micrometer.

Cutting 8 airfoiled fins spanning the length of a 700 mm piece of 6 mm plywood, I could see differences in the thickness of the cut pieces as they came off the spoil board (I surface the spoil board before I mount the stock, and I check that the top of the stock is parallel to the carriage before the cut). I think the cut plys relax and/or the work holding breaks free in spots, allowing the stock to bow and warp. If I wasn't afraid of the dust, I'd try milling G10 to see if is better for this -- but I am afraid of the dust, so I won't.

 

[ttabbal]

Even with plywood, you can definitely notice the fit change with a difference of 0.005". I don't have familiarity with many difference machines, but I believe it's reasonable to expect much tighter tolerances than 0.05" for rocket parts.

Whether you need it or not is perhaps another question, but the satisfaction of having things fit together "perfectly" is a big part of the reason for getting into CNC in the first place.

I agree. I was being a bit facetious with that tolerance spec. For plywood, I honestly don't see any reason 0.005" or less is unreasonable, even on some of the pretty cheap stuff direct from China. You might have to slow down the feed a bit, but it should be possible.

For someone just getting started, the best bet is probably to be serious about the tolerance and work area you need. Both specs cost money and small changes can make for big price increases. Cut depth and materials matter as well. If you want to do a lot of titanium, you need a much nicer machine than the guy wanting to cut rings and fins out of plywood. For hobby rocketry projects, you're probably looking at mostly plywood, some plastic, maybe a little fiberglass, and aluminum might be a stretch goal. Those are workable with lots of machines. If you drop aluminum, or are willing to go slow and have some looser tolerances, things get a lot easier. I would say most machines you can buy new these days can handle up to fiberglass without much issue. Just be careful with the glass fibers, that stuff is bad news.

 

[vcp]

Ok just to add in one other data point. I have a Joe's 2006 that I haven't set up after my last move (ach, over two years ago now). However, it is proof that you can build an accurate durable machine out of MDF (within limits, of course). It'll do 28"x48" (1/4 sheet). The 'spindle' is just a 1/4 hp router. No linear slides, etc., it's just roller-skate bearings on gas pipe. But the bed is an extremely rigid and flat torsion box. All the MDF is sealed, even the edges and inside the box, and shows no cracking/expansion after almost 10 years and two moves. There is no slop in the axes, and they move butter-smooth. Kit of cut parts was ~$700 IIRC, and I've probably got about $2k into it all together. Mach3 on an old parallel port PC.

Biggest drawback is the lead screw on the long axis - speed is limited by potential whip. Immediate upgrades it could use would be a real spindle - 1500W is big enough for this, and Geko drivers for the Nema23's (right now it's a HobbyCNC board - works but is bare bones).

 

[LOC]

I notice a .001 difference in centering rings. from tight, perfect, and moves pretty free. Simple wearing of the bit can make a difference on our end. I was floored at fisrt as I figured no way would we need a tolerance that tight. But, it's there. And needs to be changed from tube lot to tube lot, climate difference and blah blah.

 

[Dave A]

I notice a .001 difference in centering rings. from tight, perfect, and moves pretty free. Simple wearing of the bit can make a difference on our end. I was floored at fisrt as I figured no way would we need a tolerance that tight. But, it's there. And needs to be changed from tube lot to tube lot, climate difference and blah blah.

When I'm building, if the c-ring or fin slot is snug ok, if a little loose, ok, too. I'm going to epoxy them all sides and with a decent fillet. That way .001-.002" shouldn't matter. As Nat mentioned earlier, size and roundness of tubes may vary anyway.
With all that epoxy, a part is not more likely to "move" just because it's loose. I can see being a supplier you want them to fit "just right".

 

[jlabrasca]

Neat technique, that's great to know about.

Happy to pass it along. There is also fixturing wax.

https://www.machinablewax.com/product.php?product=44

I bought a sampler from Carbide3D. I've only used it once, the blue "strong" formula -- haven't tried the water-soluble -- when neither blue tape nor Nitto D/S tape provided enough hold for machining parts from a scrap of polypropylene cutting board. It worked. It is alleged to be reusable -- although it would have been more trouble trying to separate the swarf from the wax than it was worth to me.

https://shop.carbide3d.com/products/machinable-fixturing-wax?variant=3605767747

The principle advantage of the wax, over the CA + blue tape, is set up time. The CA takes a while to cure, for large pieces of stock I will leave the piece under a pile of bricks for hours to be sure the CA far from the edges has hardened (don't use kicker, the hardened CA makes it impossible to get the piece flat on the spoiler). The wax hardened really quickly. I was ready to cut in minutes.

 

[jlabrasca]

When I'm building, if the c-ring or fin slot is snug ok, if a little loose, ok, too. I'm going to epoxy them all sides and with a decent fillet. That way .001-.002" shouldn't matter. As Nat mentioned earlier, size and roundness of tubes may vary anyway.
With all that epoxy, a part is not more likely to "move" just because it's loose. I can see being a supplier you want them to fit "just right".

This is something you will figure out when you get to work with whatever machine you buy.

Another anecdote: The first time I cut centering rings for a rocket, the first rocket project for my Shapeoko in fact, was kit bashing Estes 2" PSII parts into a WAC Corporal. I used a scrap of 3mm hobby-grade plywood. The rings were perfect on the first try. everything fit together without sanding but fit snugly enough that the assembled motor mount did not fall through airframe or fall apart while dry-fiiting.

When I tried to cut rings from 5 ply aircraft-grade everything was too tight to go together. The holes just too small to slip over the motor tube, the O.D. of the rings just too wide to fit in the BT80. I fiddled around in Fusion360, trying to dial in the dimensions using the "stock-to-leave" adjustments and tweaking the drawn dimensions of the model and got it to work.

Trying to figure out why I couldn't duplicate my beginner's luck, and why the cut dimensions seemed to be BOTH too large and too small, I noticed that F360 defaults to climb milling. The aircraft ply was just that much harder than the cheap stuff so that cutter (a 1/8" shanked x 1/16" cutter fitted into the router's 1/4" collet with a reducer bushing) was deflected out just couple of thou . Flipped the parameter to conventional milling, and I was able to get the cuts dead-on the drawn dimensions.

If hadn't ever used a stationary router table, I don't think I'd've thought to check the direction of the cut.

I don't use the reducer bushing anymore (extravagantly, I bought a second trim router and installed an Elaire 1/8" collet for when I want to use the smaller endmills), so it is a lot less sensitive to cutting direction.

The take-away, for me, was that there are a LOT of levers and knobs that can be adjusted to dial-in the precision for different applications.

 

[ttabbal]

If you want to do extreme DIY, there's the MPCNC. Need a 3D printer, but it's pretty much raw material after that. No idea what tolerance is like.

I've started building the MillRight. So far it's looking like a great machine. Note to builders of any machine though... Do NOT let a ballscrew nut unthread. That was annoying to fix and cost me some time today. Future screws are getting zip ties until the ends are captured.

 

[mpitfield]

Anyone have an opinion on the OpenBuilds Workbee 1010 (40x40) vs.the Shapeoko XXL? The Workbee is a lead screw machine vs. the Shapeoko which is belt driven.

Here is a 2 hour build video on the OpenBuilds.