[typehack]

Hi, I've been lurking on this forum for a while and am at the point where I would like to attempt to start my first CNC build. What is bringing me up short is the confusion over the huge wealth of options and directions available.

I understand that each machine is built to it's intended purpose and with a specific budget in mind, but it is not always clear which choices are wisest for a specific path. While steppers might be perfect for one, servo's are the clear choice for another. likewise a plywood table is fine for one application and ludicrous for a different scenario.

I am working on a design currently and learning what I can but could really do with some specialist advice from the crowd-mind.

Myself and a good friend have a number of decades worth of mechanical and electrical experience, and would really like this first one to be somewhat useful at the very least.

With all that said here are the basic requirements:

=============================================
3 Axis
Minimum area 2ft by 3 ft with a 10" Z axis
~$3000 excluding software and computers
Capable of Milling aluminum with close tolerances.
=============================================

The conundrum as it stands is, at the price point I've stated what serves best. Where should we put our money?

Current areas where I could use any constructive advice include :

Servo vs Steppers. Which will we be able to continue to use as the machine is refined.

Ballscrews. For the application that I've mentioned what would be the ideal number, diameter and pitch we should design for? Is 30mm diameter too high? is 12mm too tiny?

Construction material. - What are the limitations of 80/20. This is more difficult to get in Canada (where we are based), is the simplicity of connection still worth it?

Guides/linear bearings, what kind of stats are we looking for in these?

Really the recommendations that could be pure gold are any hints or directions with regards to model numbers, suppliers, quality eBay vendors or manufacturers.
There is such a range of terribly manufactured cheap components out there masquerading as quality material that it has made me nervous about ordering anything from "Anonymous China Inc" This goes especially for eBay.

I understand this post is pretty general, but really anything on any of the points will help and I'm hoping people will be able to share what they are thinking for a similar machine and any stories of quality parts they may have encountered from a good source.

Thanks in Advance.

[ger21]

Is that Canadian or US dollars?

1) Your budget is too low for servos. I'd go with steppers.

2) Ballscrews. I'd recommend 2010 - 20mm diameter with 10mm pitch. Some people like a 5mm pitch on the Z axis, and you can get away with 16mm diameter on the Z.

3) 8020 Limitations? It canbe hard to mount extrusions, and it's not always perfectly straight. You can get it premachined for linear bearings from Misumi. It's also expensive.

4) I'd only recommend profile bearnigs like Hiwin or THK.

BST Automation is highly recommended for ballscrews and Hiwin bearings. BST AUTOMATION - Small Orders Online Store, Hot Selling carrier travel,carrier belt,carrier and more on Aliexpress.com | Alibaba Group

I'd recommend shortening the Z axis. Unless you need to machine very tall parts, you'll never use the 10" of travel, and the taller the Z, the weaker it is.

[wizard]

Hi, I've been lurking on this forum for a while and am at the point where I would like to attempt to start my first CNC build. What is bringing me up short is the confusion over the huge wealth of options and directions available.

Welcome aboard!

I understand that each machine is built to it's intended purpose and with a specific budget in mind, but it is not always clear which choices are wisest for a specific path. While steppers might be perfect for one, servo's are the clear choice for another. likewise a plywood table is fine for one application and ludicrous for a different scenario.

It is all about your expectations and intentions. Both of those is subject to change and more so people might not have the same idea as to what an ability means. For example people will claim their machines can mill steel and when you see what they are doing they may be using a cutter that is 1/16" in diameter. While technically that is machineing steel it isn't what many of us expect when we see the term machining steel.

To really nail down a design ahead of time you really need to get a handle on your expectations.

I am working on a design currently and learning what I can but could really do with some specialist advice from the crowd-mind.

Ideas and comments are worth what you pay for them! 😜

Myself and a good friend have a number of decades worth of mechanical and electrical experience, and would really like this first one to be somewhat useful at the very least.

This is a trick, I will say the more robust the machine the more likely it will work well in a wide array of uses. Building a robust machine though means more dollars up front. So you have to balance robustness against costs.

With all that said here are the basic requirements:

=============================================
3 Axis
Minimum area 2ft by 3 ft with a 10" Z axis

This isn't a bad size, it might be seen as the limit for a moving table design unless you have lots of space. I bring up moving table designs because I believe it is easier to build a stiff and accurate machine with a moving table design.

The 10" Z axis does cause a bit of concern as you start to have serious leverage issues which may require a more robust gantry.

~$3000 excluding software and computers

A machine that size is totally doable if you put a lot of manual work into it.

Capable of Milling aluminum with close tolerances.

This is where the discussion gets hairy because """close tolerances""" is not well defined. Close tolerances to one person could mean that your $3000 budget is woefully short of what is needed. When you start talking close tolerances you really need to start to understand the tolerances possible in the various classes of components

=============================================

The conundrum as it stands is, at the price point I've stated what serves best. Where should we put our money?

If a decent proportion of the machines usage is to be focused on machining Aluminum then I'd strongly suggest building a steel framed machine. Not a light weight steel frame either. The theory here is that a steel frame is harder to upgrade latter once there is an additional budget to work with. The is also relatively cheap. Most likely though you will need access to a machine shop and the ability to weld up a few assemblies.

Current areas where I could use any constructive advice include :

Servo vs Steppers. Which will we be able to continue to use as the machine is refined.

The budget demands steppers and probably acme leadscrews.

Ballscrews. For the application that I've mentioned what would be the ideal number, diameter and pitch we should design for? Is 30mm diameter too high? is 12mm too tiny?

Too high of a lead can be a problem if you want accuracy with steppers. You want to resolve more than the precision you are aiming for, often 4 times can be good enough. The problem is you don't want to eat up any speed potential in excessively small steps. Steppers imply very hard limits on how fast an axis can move while retaining resolution above the precision you are after.

Construction material. - What are the limitations of 80/20.

I hate it for critical machine structures. At work it is never used on critical parts of a machine. Frankly it doesn't even hold up as framing for machine guard doors. It is great for the lighter parts of a machine though. The trick in leveraging extrusions is to learn where using the stuff makes sense and where it should be avoided.

This is more difficult to get in Canada (where we are based), is the simplicity of connection still worth it?

Not for critical parts of a machine it isn't. It is extremely handy for stuff that isn't exposed to lots of stress.

Now I need to qualify this a bit, If you put together aluminum extrusions with plenty of gusseeting and corner reinforcements it can leave you with a strong serviceable structure though at a significant extra cost. Even then the morning arraignment for you linear rails levels a lot to be desired, especially if you are focuse on materials other that wood. A look at the cross sections should highlight the lack of support for the rails. There are of course ways around this but you just add more costs.

Guides/linear bearings, what kind of stats are we looking for in these?

Well if comes down to what the budget will afford you for the initial build. I really doubt that there will be budgetary room for profile rails but you should lean towards that direction.

Really the recommendations that could be pure gold are any hints or directions with regards to model numbers, suppliers, quality eBay vendors or manufacturers.

When going EBay it is basically a crap shoot, if you get lucky you can save $1000 of dollars. On the other hand it can be costly. For stepper control Gecko seems to get a universal approval but companies like Leadshine are making great progress.

There is such a range of terribly manufactured cheap components out there masquerading as quality material that it has made me nervous about ordering anything from "Anonymous China Inc" This goes especially for eBay.

Yep a big problem. China is a bit like what Japan use to be, it was very hard to weed out the quality manufactures from the not so good ones. You best bet is to pay a bit extra at a supplier that has done the weeding for you or directly with Chines from a vendor with a good reputation here in the forums. In other words deal with "pre-approved" vendors. You can however blaze trails for the rest of us.

I understand this post is pretty general, but really anything on any of the points will help and I'm hoping people will be able to share what they are thinking for a similar machine and any stories of quality parts they may have encountered from a good source.

Don't dismiss Amazon which has a large selection of mechanical components.

Thanks in Advance.

No problem. A couple of other things:

You seemed to focused on aluminum, depending upon what you intend to be doing a "router" type spindle might not be the best choice. You really want a spindle that puts the power into the speed ranges you expect to be using.

The other reason to consider steel is that you need (most likely) to use a coolant/lube when machining Aluminum. This of course macke wood out of the question. However machining fluids and swarf can and will get into every opening it can find on aluminum extrusions making clean up a problem.

[ger21]

Capable of Milling aluminum with close tolerances.

I missed this part. If aluminum is your primary goal, cut the Z axis travel in half, and build a welded steel frame with at least 1/4" wall tubing.

[typehack]

Thanks for the reply Gerry!
The amount was in USD. Sorry should have specified.
On your advice I will look at shortening up that Z. Any recommendations? Would 6" still sound like too much flex?
Also, I was working on an over-engineered moving gantry style, but it occurs to me that most milling machines use a moving table arrangement. Thoughts?

[typehack]

Seems I replied a little to quickly to your first comment Gerry and missed your subsequent post and Wizards. Thanks for the advice, and I will do just that .
I can live with 5" of travel.
To put this all in perspective, it will be used to machine closely interlocking armature pieces. Eventually looking at accuracies within a few thousandths of an inch. I'm planning for this first machine to be the first stepping stone towards something that can do this. Doesn't have to get all the way, but a significant percentage of that capability will allow a proof of concept.

Wizard, your reply was awesome, thanks! As for the fabrication from steel, this is doable. A friend has a small machine shop I'm sure he will not mind me stealing for a little bit.
Thanks to your comments I am now seriously looking at a complete redesign to transition to a moving table design. Though the footprint may be prohibitive, I'll spec it out and see.
With regards to this, what are peoples thoughts about picking up an older mill that fits the specs and adding control mechanisms and steppers to that?
I'm unsure how worn the screws can get with the older machines, but have the possibility in mind that there might be a few savings to be had. Is this something done often?
To be clear there is little to no overhead for me to weld a frame from scratch. Just thinking out loud.

If I was to go the welded steel frame route, what would you consider "not a light weight steel frame"? Should I be breaking out the 1/2 inch steel plate?
Steppers and Acme leadscrews it is.

For all the good reasons you mention, I am going to ditch the idea of using aluminum and just go with steel. As you say, we will be running coolant, and the very thought of the extrusion in that context has me realizing what a very dim idea it really is.

[ger21]

To put this all in perspective, it will be used to machine closely interlocking armature pieces. Eventually looking at accuracies within a few thousandths of an inch

How big are your parts?

It sounds like what you want to do is a job for a $7000-$10000 mill, not a router. Your budget would typically buy you a rather light duty 24"x36" router.

[rockcliff]

if cutting aluminum is full time , you definitely need to lean more towards a milling machine,

3' x 2' cut payload would be quite the beast of a machine, if your looking at used, or even new, you really need to look at the power requirements first, most of these industrial type mills will be 575vac, which is easy to get in an industrial shop, but very costly to generate 3 phase 575 with a phase convertor in a home style shop, plus they are tall, a mill this size most likely would not fit height wise in the average home shop, do your homework before you decide to go this route, there's a lot of potential issues that can come up,

[Tkamsker]

hi if you do a build look at the cnc step mills i have a 400 t and did cut a lot of aluminium with it till i got areal vmc mill so if you stay in that envelope fast not deep cuts are possible in high quality if you are able to build that quality will be shown over tine so go for it ,..


Gesendet von iPad mit Tapatalk

[wizard]

Seems I replied a little to quickly to your first comment Gerry and missed your subsequent post and Wizards. Thanks for the advice, and I will do just that .
I can live with 5" of travel.

There is a great thread in the stickies that goes into a bit of beam engineering for a gantry. Well worth the read before settling on the Z working envelope. You can design a machine for any sort of Z working space you want but it quickly becomes a massive machine.

Don't confuse Z clearance with Z axis travel either. Sometimes additional Z travel can be justified. This especially if you intend to use long drill bits, or other cutters. Before you jump on a Z axis value it is best to understand how you intend to use the machine.

.
To put this all in perspective, it will be used to machine closely interlocking armature pieces.

Armature as in a DC motor or as in a artists sculpture?

Eventually looking at accuracies within a few thousandths of an inch. I'm planning for this first machine to be the first stepping stone towards something that can do this. Doesn't have to get all the way, but a significant percentage of that capability will allow a proof of concept.

A few thou is certainly possible but it won't be easy.

Wizard, your reply was awesome, thanks! As for the fabrication from steel, this is doable. A friend has a small machine shop I'm sure he will not mind me stealing for a little bit.

A small machine shop with a big mill I hope!

Thanks to your comments I am now seriously looking at a complete redesign to transition to a moving table design. Though the footprint may be prohibitive, I'll spec it out and see.

The general problem with a moving table design is the space it takes up. However I kinda think it will be required for this project to give you the best hope of hitting the tolerances you want.

With regards to this, what are peoples thoughts about picking up an older mill that fits the specs and adding control mechanisms and steppers to that?

It depends upon the machine. Frankly a dead CNC machine that needs a controller upgrade is often a better deal.

I'm unsure how worn the screws can get with the older machines, but have the possibility in mind that there might be a few savings to be had. Is this something done often?

Yes it is done often. Sometimes by people that want a machine for their home shop and sometimes by companies wanting to improve productivity on a large piece of machinery. Companies are less likely to upgrade smaller machines if you can buy new off the shelf.

From the standpoint of wear it spends upon what you buy. Buy a machine that was in an R&D lab and you will see little wear. Buy a machine off a production floor and it might be worth scrap metal prices and not much more.

To be clear there is little to no overhead for me to weld a frame from scratch. Just thinking out loud.

You will need to decide if a router makes sense or if a mill would be better! Small mills can be cheap but a mill to cover the same space as your router specs might not be so cheap.

If I was to go the welded steel frame route, what would you consider "not a light weight steel frame"? Should I be breaking out the 1/2 inch steel plate?

A rough guess would be to go for square tubing of six inches in size, with 3/8" walls for the gantry.

As for plate you will need plate in more than one thickness. You will need internal gusseeting in the gantry beam where 1/4" thick plate may do. The Y Axis saddle might benefit from thicker material. External gussets and brackets may be in the 3/8" to 1/2" range.

Steppers and Acme leadscrews it is.

You will likely want to upgrade those in the future but this is a good place to keep an eye on the budget and at the same time get a working machine.

For all the good reasons you mention, I am going to ditch the idea of using aluminum and just go with steel. As you say, we will be running coolant, and the very thought of the extrusion in that context has me realizing what a very dim idea it really is.

Well do realize we are talking T-Slotted aluminum here. You can get aluminum in square and rectangular beams and sheets just like you can steel. It is however expensive to buy and weld. I just don't think it is worth the expense. Also aluminum doesn't hold threads well so you end up spending a lot of time dealing with that.

[alan_3301]

You have to say how big these armatures are, how many do you need to make, and how close are the tolerances?
Is this something cut from a solid block, with multiple vise setups, or is it cut out of sheet metal?
Then you can think about what kind of machine to build.
If this is a 2 x 3 foot chuck of aluminum, that needs half of it removed, you need a huge mill.
If this is a much smaller chunk of aluminum, at low volume, you could do it with a 2x3 router in a lot more time and poorer finish than a much smaller travel router or mill.
If this is a 2x3 piece of sheet metal, a basic router would do just fine.

[PatRyan]

Ger 21 above wasn't accurate when it comes to 80/20 extrusion - they are always straight. That is what you are paying for - tight tolerance and consistent clear anodizing color. Misumi is not the same as 80/20. They are a smaller manufacturer with less compatible parts, and less inventory in the US. 80/20 is made in the USA, and has the highest quality.
While 80/20 is not the cheapest extrusion you can buy - it will serve you well for years to come. The idea that the top brand is not straight is just nonsense.

[ger21]

Ger 21 above wasn't accurate when it comes to 80/20 extrusion - they are always straight.

Depends how you define straight.

8020 says it's straight to within .0125 per foot, and can have up to 0.25° of twist per foot. Not terribly straight in the world of machining.

[wizard]

I really have to disagree here.

Ger 21 above wasn't accurate when it comes to 80/20 extrusion - they are always straight.

Straight compare to what. 8020's specs are very clear here and not hidden. Beyond that if raw extrusions where that straight Mitsumi and other would not offer machining services to flatten surfaces for rail mounting.

That is what you are paying for - tight tolerance and consistent clear anodizing color.

Actually what you are paying for is a rapid assembly method, an adult erector set if you will. The tolerances are not machine tool quality!

Misumi is not the same as 80/20. They are a smaller manufacturer with less compatible parts, and less inventory in the US. 80/20 is made in the USA, and has the highest quality.

I don't know about the highest quality, we use a lot of 8020 at work and other manufactures as well and the quality is comparable across vendors. Aluminum T slotted extrusions have their place but it is a huge mistake in my mind to not consider alternatives to major parts of a machines structure.

While 80/20 is not the cheapest extrusion you can buy - it will serve you well for years to come. The idea that the top brand is not straight is just nonsense.

The comment is accurate, in the context of machine tool building extrusions are not straight, especially if you are considering them for axis linear rail mountings. There is nothing outlandish as 8020 documents their specs publicly. Now given that extrusions have tigher specs than you will see for most steel tubing suppliers. The fact remains that in either case you will require machining of the rail mounting surfaces or other approaches to flatten said surfaces. That being the case steel often makes good sense as it can be cost effective especially for a home build.

[PatRyan]

Although the specs may need not seem that exact, our experience is that 80/20 has an excellent QA process and so that spec would be a worse case scenario. As you know, company/product specs are easy to print in a brochure. But the good companies spend the money and time to check their manufacturing process to ensure they meet or exceed the published spec. or they re-melt it. That is part of the product cost that people forget about when comparing "foreign" extruded bars with lower prices that may not do this QA step. With many successful CNC frames in the field, I think you would see a lot of people jump in on this topic if their 80/20 t-slot extrusions did not work well. Yes you can probably find something less expensive that looks the same, or has the same spec - but you may or may not get what you want. Just saying specs aren't everything.


Straight compare to what. 8020's specs are very clear here and not hidden. Beyond that if raw extrusions where that straight Mitsumi and other would not offer machining services to flatten surfaces for rail mounting.

Actually what you are paying for is a rapid assembly method, an adult erector set if you will. The tolerances are not machine tool quality!

I don't know about the highest quality, we use a lot of 8020 at work and other manufactures as well and the quality is comparable across vendors. Aluminum T slotted extrusions have their place but it is a huge mistake in my mind to not consider alternatives to major parts of a machines structure.



The comment is accurate, in the context of machine tool building extrusions are not straight, especially if you are considering them for axis linear rail mountings. There is nothing outlandish as 8020 documents their specs publicly. Now given that extrusions have tigher specs than you will see for most steel tubing suppliers. The fact remains that in either case you will require machining of the rail mounting surfaces or other approaches to flatten said surfaces. That being the case steel often makes good sense as it can be cost effective especially for a home build.[/QUOTE]

[dmalicky]

I'm sure 8020 has very good QC and yes it can be successfully used in most DIY CNC routers. But I'd caution against alluding to Misumi as low quality because it is "foreign". Misumi is not cheap, and quality comes from many countries. Also:
- It's not realistic for a good extruded product to match the tolerances of a good machined product.
- 8020's (and most T-slot's) mounting surfaces are intentionally non-flat -- that's great for bolting 8020 to other 8020 or wide flat plates, for better joint rigidity. But it is not great when bolting a 20mm wide profile rail to it, as the rails may not sit in the same plane. This is part of why Misumi offers a machined-flat rail mounting surface.
- 8020 and any T-slot is not very rigid (see the Hardwoods thread for analysis), so it does not need to be straight and flat to 0.001" in order for the linear rails/bearings to not bind. I.e., the T-slot extrusion can bend to accommodate the requirements of the linear bearing. That is fine for a wood router, but a high rigidity machine will have more stringent requirements on mounting surfaces.

[PatRyan]

Didn't mean to allude that Misumi was low quality - our issue with them is limited products. But there are a lot of new extrusion companies coming into the US market through Alibaba that I would have concerns about.

By the way, 80/20 also offers custom machining for a more exact mounting surface if you need it.

[dmalicky]

Thank you for clarifying. 8020 does indeed have a wide product spectrum, and that's helpful for many apps. For larger CNC routers, though, Misumi has the best gantry option, with their GFS8 100200 (~4"x8") -- analysis shows it is about 2.5x stiffer than 8020's 3060 (3"x6").

I'd agree generic T-slot is risky.

[louieatienza]

Didn't mean to allude that Misumi was low quality - our issue with them is limited products. But there are a lot of new extrusion companies coming into the US market through Alibaba that I would have concerns about.

By the way, 80/20 also offers custom machining for a more exact mounting surface if you need it.

I think Misumi has a very extensive line. Finding product in their crazy format, however, is difficult. Misumi also offers custom machining, as well as surface milling on the GFS series, to accommodate mounting of linear components. Also, t-strips drilled to the same specs as linear rails, avoiding the use of single t-nuts and providing contiguous support.

[wizard]

This discussion about the various sources for aluminum extrusions is interesting but I still see Aluminum extrusions as being expensive and sub par for a gantry. Last week I was at the local steel suppliers "drops" section and steel was going for 90 cents a pound. Obviously with drops you may not find what you need immediately but eventually some nice tubing will show up.