This is the open source thread for a cheap cnc design... benchtop size...

I did one of these last year, but it was not nearly robust enough... So I figured I'd bring it to the open source group and see what you guys think I did wrong, and then we can come up with a new, more robust cheapcnc design.

The only criteria for this design are that we want to use drill rod and bushings. Beyond that, anything is fair game!

Here's my original version of the cheapcnc machine...

Mike...

Hi Mike,
There have been several references to "drill rod" in US postings. Here in the UK the only reference I find is to Oil Drilling !!
Is this the raw material for making drills, or are you using very long drills as lead screws :confused:

Hey Greybeard,

Drill rod is blank shafting that can be used to make drills. It's basically untreated steel rod.

Look at this attachment from McMaster Carr...

Mike...

I think by going with the drill rod and bushings, you're setting yourself up with the same problems. The rod has to be supported somehow to eliminate the problems you had before. I'm a bit leary of recommending my method because I don't have it finished yet, but, I think it will work pretty good. Here's one using similar construction.
http://www.cnczone.com/forums/showthread.php?t=12056&page=1

It's cheaper, and it's easy to build. Just my 2 cents. :)

If you want to stick with the drill rod, that's fine.Make sure you can support it, though. Just set some basic rules and stick with them.

Does anyone have an idea how we can support a drill rod, and use split bushings, affordably?

I don't think a cheapcnc type machine can afford 4 of these on every axis. (see attachment)

Hi Greybeard,
Try "silver steel" for "drill rod" in the UK and Oz.
Cheers,
KW26

Do not use drill rod, use precision ground shafting.

Why?

-its cheap
- its hardened, DR isn't
- its round, DR isn't necessarily
- it has a better finish than drill rod
- and the top reason for using PGS, its tolerances are are zero over, drill rod is +-. DR often won't fit! Pressing an oilite into a reamed hole of perfect size produces a bearing ID that can't handle a half a thou over sized dia.

imo it would be very difficult to even agree on the objectives let alone a design. Think of how hard it would be to do in a workplace where there are pay cheques and boss to keep things in line. People will not spend their leisure time on something they disagree with. Want to see what can happen? Check out the thread on making a communal pcb mill at the parallax basic stamp site – there, the objective was predefined but what a mess!

As was posted, the idea of building a culture of open source designs is the best idea. Having a variety of designs will help people create what’s right for them and inspire many. Kudos’ to peeps like Phil and Alan and others for contributing to it – that sort of action will have the greatest effect imo. My view is that the culture should encourage people to share their designs and frown on amateurs trying to sell their designs. Me? I don’t have a design to share yet so try to contribute by answering machining Q’s :D

OOPS, bottom half of this was intended for open design thread <sheepish grin>

Do not use drill rod, use precision ground shafting.

I agree


As was posted, the idea of building a culture of open source designs is the best idea. Having a variety of designs will help people create what’s right for them and inspire many.
In a way I think that's what we are trying to do. If I can find any two people who agree on a type of cnc machine, then we'll draw it up, and let folks here work out the problems in the design. The end result hopefully is a set of plans, and even more hopefully, a machine that has been built to validate those plans.

Mike...

I believe either drill rod or gas pipe is acceptable for a machine as you have titled. ENCO www.use-enco.com has drill rod consistently on sale along with acme rod and nuts. Just make sure you size appropriately. Many machines have tried to use 3/4" DR for an X axis to support a moving gantry (my 1st attempt also) and 1/2" for Y and it just isn't enough. 1" drill rod is $14 there, 1.25" is $17. 7/8" is $12 and 3/4" for $8. The one thing about drill rod is it's easy to upgrade the machine to precision shafting of the same diameter it a person wants more precision. Same can be said for 1" gas pipe using roller bearings. my 2 cents

Just make sure you size appropriately. Many machines have tried to use 3/4" DR for an X axis to support a moving gantry (my 1st attempt also) and 1/2" for Y and it just isn't enough. 1" drill rod is $14 there, 1.25" is $17. 7/8" is $12 and 3/4" for $8.

So what size did you use on your x axis, and how long was the PGS (precision ground shafting)?

How about your Y axis? what size and how long?

And what kind of bushings did you use?

Thanks,

Mike...

Do not use drill rod, use precision ground shafting. Why? -its cheap??????????

I guess cheap as relative to what. I looked at mcmaster carr and the prices didn't seem that good as compared to drill rod. Maybe I was looking at the wrong stuff. Do you have recommended sources?

Herman

i think last time i bought it was $0.45 cdn per inch for the higher class .500" (there are two classes available), cut to lengh. i didn't mean to imply that it was cheaper than drill rod, only that as the right choice for the task at hand, it was not a lot of money. got mine from canadian belting, but any industrial supplier would do.

its not so much that its more precise (although it is) its more that the tolerance is only - whereas dr is +-

I tried 36" x 3/4" DR on the X and 18" x 1/2" on the Y, both two small. Since DR comes in 3' lengths I figured 36" straight out of the box was a good length for X and one biece cut in half was economical for the Y. Unsupported neither is big enough. I don't know if 1" would do 36" unsupported or not. Electronics is my education, not mecahnical.

Precision shafting in 1" for 36" is $48, if you use two per axis that is $96 vs $28 (for DR). If the design is flexable enough, those who want more precision could spend the extra money.

Hi Greybeard,
Try "silver steel" for "drill rod" in the UK and Oz.
Cheers,
KW26
Thanks for the info - all is now clear.
However I am dissapointed. I had visions of a cage of ball races spiralling around a 1" twist drill ;)

Unsupported small-siameter round rod is indeed a disappointment. "Cranky" John's designs using larger diameter pipe and roller-skate wheels are better.

Keep always in mind that the stiffness of a round rod goes as the CUBE of the diameter.

Also, tube is almost as stiff as solid rod.

Thanks for the info - all is now clear.
However I am dissapointed. I had visions of a cage of ball races spiralling around a 1" twist drill ;)
I believe Metal Supermarkets (http://www.metalsupermarkets.com/) have some UK locations, they sell it as drill rod, even in the UK.

Drill rod will be a high-carbon, high alloy steel, capable of being hardened through and through for cutting tools.

"Ground shafting" will as often as not be a lower-cost low-carbon steel that has been case-hardened.

Yeeas !
Greybeards got the ballscrews worked out .
Bear

P.S.will do what I can.

The engineering formulas found at http://www.engineersedge.com/Calulators_Online.shtml will help determine what shaft dimensions are suitable. Scroll down for beam deflection, and also for the section properties needed for the formulas.

I haven't run the numbers yet, but I'm hoping to use the two 5/8 x 18 printer shafts in my parts accumulation full length for the X axis of a rotated variant of Cranky's "Hawk". One thing to note is how much stiffer square tube is when rotated 45 degrees (called 90 degrees on the formula page). This will work nicely with a vee bearing system.

it is like anything else.
the entire machine will only be as good as its weakest link.
one can cut corners in some places , like axis movement speed
bit i think it must be determined what a machine is going to be intended for and what the cutting tollerences will be acceptable before one can select components.
i thought the hardest parts to find were going to be the shafts and bearings, but i found out real quick ( in my case anyway ) its the price of ball screws and nuts.
i am not going to cut a piece of steal or aluminum or whatever until i have all the components, so i know what i have to work with,
this will save me money in the long run. I have learned this from past experiences being it mechanical or electronic design.
i know what i want and what i'd like, but i always don't get what i want.

We have not come up with a good way to support the precision ground shafting yet... nor have we come up with a way to support the split bearings.

Any ideas? Anyone?

Mike...

support the precision ground shafting

looking at the above thumbnail of mikeschn's moving gantry pic:
(http://www.cnczone.com/forums/attachment.php?attachmentid=9641)

i imagine two easy ways:
1) in the moving gantry, use a high lubricity plastic sheet (UHMW or teflon) to support the gantry where it might touch the base if it were to sag... trading off ease of movement for no sag.
2) have span-following pillow blocks that attach to something like a spider divider/pantograph/adjustable divider/aka:some-kind-of-scissor-mechanism that moves the "span-supporting" pillow blocks (with oversize bearings) and distributes them equally across the open span using the same UHMW plastic/plastic sliding support for the moving pillow blocks.

i hope y'all understand what i'm saying otherwise i'll have to get out my crayons and color 300dpi scanner ;)

and if these are silly ideas, i'd like to learn why...

cheers,
marc.

1) in the moving gantry, use a high lubricity plastic sheet (UHMW or teflon) to support the gantry where it might touch the base if it were to sag... trading off ease of movement for no sag.


It's not just sag you have to worry about. It's side to side movement as well.

mikeschn,

What is the diameter of your shafting on theY axis? I used 32mm (1.25 inch ) seamless shafting for my JGRO Y axis.

Jason

I used 3/4" x 36" long shaft for my long axis. In retrospect it should have been a lot larger in diameter. How long is your 1.25" shaft, and do you have any flexing problems with it?

Mike...

We have not come up with a good way to support the precision ground shafting yet... nor have we come up with a way to support the split bearings.


I'm building a large steel CNC mill, using 2.3m 30mm shaft. since, my base frame will
bend a lot during welding, i need be be able to adjust each shaft support as well.
i'm going to drill, tap, and insert 6mm threaded studs, both down and sideways to
mount the shaft, and make it strait.
http://www.cnczone.com/gallery/data/500/medium/lineair_bearing.png
http://www.cnczone.com/gallery/data/500/lineair_bearing.png

now, for a simpler model, you could only support the shaft only in the up/down direction,
drill through the shaft (vertically). take a long M5 or M6 bolt, grind the sides of the bolt head
to make it narrow, drop it in from the top, put nut from other side. push threaded ends of
bolts through the base, and put a nut at above and below the base (for adjustment).

now you can have 2 (roller skate?) bearings on the top (at 45 deg), and add a third
bearing on the inside bottom (-45 deg) to keep things down.

Even just supporting it only once in the middle should help a lot.

update:
After looking around a bit, i think this is the perfect way to support the rods on a
wooden/MDF mill: (by spalm)
http://www.cnczone.com/forums/attachment.php?attachmentid=8976
http://www.cnczone.com/forums/showthread.php?t=12056&page=1
adjustments can be made by pushing shims in between the wood and the rod if needed.

jaap.

I think the idea jrgo had for the supported rails and bearing shuttle lends itself well to the cheapcnc concept. (attached)

Jrgo, do you want to develop that for us?

Mike...

Like Dave Kush's HBLB design...
http://www.buildyouridea.com/cnc/IG-88/phase_six/phase_six.html
http://www.buildyouridea.com/cnc/IG-88/phase_eight/phase_eight.html

Anoel,

That is an awsome looking machine. I wish I had casting facilities and welding facilities as well. I like Dave's acme tap too! ;)

Mike...

P.S. The cheapcnc machine is a 1 mill design, meaning that the parts are cut out on a previously built mill. That should open up lots of design possibilities for us. :)

mikeschn,

My 1.25" shafting is 45" long with no flexing problems. I supported each one with a 'L' shaped piece of wood to ensure no flexing. I will upgrade my 'X' axis to shafting rod as well as I notice too much flex with the black pipe when I twist the 'Z' axis with my hand.

Jason

Making an Acme tap was a great exercise, but you CAN buy them, you know.

Making an Acme tap was a great exercise, but you CAN buy them, you know.

But if you'r tapping Delrin, you'll get less backlash with a homemade one. And multi-start acme taps are a bit hard to find, if you can find them at all. And even a 1/2-10 will cost $30-$50.

But if you'r tapping Delrin, you'll get less backlash with a homemade one. And multi-start acme taps are a bit hard to find, if you can find them at all. And even a 1/2-10 will cost $30-$50.


Um, right about multi-start, I missed that. I make my own reamers for antique gunsmithing work, so I appreciate the work involved in making that tap. Just trying to help. There's a neat trick for making an anti-backlash nut, should you want one. Prior to tapping, you saw the nut block block about 85% thru, normal to the screw axis, and halfway along. Then you wedge the cut open a few thou whilst you run the tap thru. Remove the wedge, and the two threaded sections spring back and take up the slack. Best if you then clamp the entire nut against a flat surface when mounting, to keep the prelaod constant. Nothing new - been done at least since the 19th century.

uscra112,

A diagram or picture showing the making your anti-backlash nut would be great.

Jason

mikeschn,
I should have mentioned that my shafting is supported underneath which helps with any flexing. The shaft can flex outwards however if I twist the left and right sides of the gantry in opposite directions with some considerable force.
I solved this by using an idea I got from ger21 thread yesterday. I used 5/8 " threaded rod to hold the machine sides together. Result, No twisting possible.

Jason

As I am reading both the "cheapcnc" and "hardware store" for guidance, ahead of building my first machine, it occurred to me that the problem of levelling parts of the m/c will be a universal one, so I have posted here the following questions.

1. Do you adjust the shafts by trial and error using the carriage as a test for free running, or should you set the shafts up level and parallel then adjust the carriage bearings to the shafts ?

My m/c will be quite small(xyz = 30 x 12 x 3 inches), and my shafts (15mm ss pipe)are mounted, with bolts and epoxy, in a groove on the inner faces of two aluminium extrusions, about 10 inches apart. The whole bed is to be mounted on an off-cut of kitchen worktop.
I plan to use a pair of rods ground to the same length to deal with the shaft separation, but what about the levels ?

I could use a sheet of plate glass on the baseboard between the extrusions as a reference level , and with my one precision tool, a dial gauge, reference the heights of the shafts to the glass.
Then I started to worry about the flatness of the baseboard, and if bolting the extrusion down onto that would distort them.

In fact I'd even considered making the bed as a torsion box out of plate glass epoxyed together. Then I thought concrete, with the shafts epoxyed to raised webs. Then I went to bed. :confused:

I thought that I would start with a bubble level on the bench, and started to look at their accuracy. Could I build an accurate bubble level ? I've taken a cheap level to pieces meaning to remount it in a more solid, and adjustable housing, and then came face to face with the following question. How do bubble levels work ?
If you raise one end slightly, why doesn't the bubble travel to the higher end instead of stopping at some intermediate position ?
So the question is
2. How do you improve its accuracy - thinner tube or longer tube ?
Any thoughts/comments/links greatly appreciated.
Should I stop worrying, and take up sudoku ? :D

John

I think the idea jrgo had for the supported rails and bearing shuttle lends itself well to the cheapcnc concept. (attached)

Jrgo, do you want to develop that for us?

Mike...

I haven't heard from Jrgo lately. Is he hiding in front of his workstation, modeling up the CheapCNC concept?

Mike...

P.S. Jrgo, do you have a real photo of your shuttle?

If 36" drill rod @ 0.75" diameter (unsupported) is inadequate would the same sized Precision Steel Rod be any better. If its hardened steel will it tolerate more weight and longer runs w/o being compromised? What is the longest run likely to be acceptable with 3/4" Precision Steel Rod? It would be great if some of you engineer types (or others with experience) could generate a table showing the vaarious rods (DR, Presion Hardened Steel, Black Pipe, etc.) and show the maximum acceptable run and acceptable weight.
Thanks,
db113
Signal Mtn. Tennessee

You're in luck... two spreadsheets were posted in the hardware store design... it's not exactly what you were asking for, but it's a start!

http://www.cnczone.com/forums/showpost.php?p=99413&postcount=117

http://www.cnczone.com/forums/showpost.php?p=99426&postcount=119

Have fun!

Mike...

Ive been looking over all the various designs of these CNC machines, as Im sure everyone else has on this site. The question of flexion on various types of rods seems to be influenced greatly by the weight of the Gantry. Some of those Gantrys look pretty darn heavy and that weight would likely be a significant contributing influence in the rod flexion issue. I have seen a few designs that do not have a moving gantry but rather a fixed gantry and instead a moving table on the X axis. For me, an aknowledged novice to CNC, this design seems to be superior in that the weight of the gantry is never transferred to the X axis rods. What am I missing here? Are there reasons for having the gantry move rather than the X axis table?

also not only does the slide size need to be considered so does the motor and drive electronics and drive screw.
with the table fixed it doesn't matter how much the piece being worked on weighs.
3 lbs or 3000 lbs, as long as the table or floor can support it.
some machines have a u shaped frame and you can bolt the machine to the work piece, or roll the workpiece under the moving gantry.
with a moving gantry type the load doesn't very too much like it would with the various work pieces on a moving table.

Additionally, a moving table design has to be twice as big as the table...it takes up a lot of floor space for a given working envelope.

You're right, that it does create a very simple machine, though.

-- Chuck Knight

Mike's first posting - quote " this is the forum for a cheap cnc design, benchtop size"
One might think it problematic to consider a bench top machine that could handle a 3000lbs workpiece.
I assumed that most people looking for a cheap design might be the newbies like myself who were looking to build their first machine as cheaply as possible in order to gain experience, discover what the basic requirements were to move tool or workpiece, and when they could afford to invest cash and time, go on to design their own machine to suit their own needs.

Is this forum trying to be all things to all people, and design a universal machine ?

>> A moving table design has to be twice as big as the table...it takes up a lot of floor space...
>> Is this forum trying to be all things to all people, and design a universal machine ?

I can only say one thing, "Life is about tradeoffs". FIRST determine your goals, THEN have a SMALL group decide how to attack it. Too many supervisors and you'll never decide which way to go. Easier said than done? Ha Ha :)

Rance

Greybeard,

Yes, you are right. The cheap cnc was intended to either be a first machine, or a machine for someone who mostly just plays... like those guys in the RC hobbies.

Once you have built a first machine, and understand the principles, you can build another machine if you have the money and desire. The second machine will probably be custom fit to your set of requirements, in terms of size, cost, precision and durability. You might even be able to cut parts for your second machine on your first one.

The "Cheap CNC" design, of cheapcnc.com was a light weight machine, designed for either engraving or light cutting of wood. cheapcnc.com has gone out of business. I suspect his machines were too slow, (I should know, I had one), and not rigid enough.

The intent of this thread is to come up with an affordable replacement for the cheapcnc.com design that addresses the speed and the rididity issues. As such, you are not going to machine 3000# parts! But if you want to engrave, cut wood or aluminum parts for your RC hobby, this might be just the thing!!!

Mike...

Mike,

After my last post, I re-read a lot of the entire thread and composed the following before your last post. You've already answered at least one of my questions but I'll post this anyways.

First, I don't want to step on ANYONE's toes here, I'm just confused. :)

While trying to figure out whether I should follow this thread or whether it pertains to what I want to build for my first machine, I am getting so many mixed signals. The title says 'Cheap' CNC, but from what I can find (and that's spread way out) the design specs say 'benchtop size, robust, use drill rod and bushings. Beyond that, anything is fair game! Oh, and its a 1 mill design.'. There's talk of lots of money (in my opinion) which doesn't fit 'Cheap'. One user suggests a design around 5/8 x 18 printer shafts. Now that's more along my line of 'Cheap'. In fact I just harvested two of these and have two more to tear apart.

Possible questions that need an answer:
I would not expect that many folks have the tools or know-how to build a 'bearing shuttle'. This would have to be purchased. Are these available? Does the cost prohibit use on the 'Cheap' machine?
Should this be a 1 mill design?
What precision is expected?
What tools are the builder expected to have?
What knowledge is the builder expected to have?
Total expected cost?
Use of scavanged materials?
Expected working area for each axis?

Just some questions that I believe need to be answered UP-FRONT.

Again, I'm not trying to step on ANYONE's toes here. :) There is some good talk going on here about different aspects but how many of these apply to THIS CheapCNC machine's design specs?

Rance

Rance,

Here is the cheapcnc.com design as it existed before they went out of business.

http://www.cnczone.com/forums/showthread.php?t=1513


>Possible questions that need an answer:
>I would not expect that many folks have the tools or know-how to build a 'bearing >shuttle'. This would have to be purchased. Are these available?
No, bearing shuttles would have to built from scratch. I can’t tell you how easy or difficult that would be, since I haven’t looked at them in any detail.

>Does the cost prohibit use on the 'Cheap' machine?
That depends on what we could come up with. The cheapcnc.com design was a $1500 machine. I should think we would want to maintain that price ceiling.

>Should this be a 1 mill design?
That’s a good debate. Ger21 says we can cut one part, and use it as a template to cut the other parts. Building like that, it could be a 0 mill machine. But I built a cheapcnc.com clone and I used a mill. It sure made things easier.

>What precision is expected?
The cheapcnc.com design claimed a precision of +/-.0001. I’ve never tested that, but my guess is the precision was actually a lot less.

>What tools are the builder expected to have?
Normal wood working tools, like table saw, miter saw, drill press, router & router table, etc.

>What knowledge is the builder expected to have?
Normal wood working skills. If you can build a chest of drawers, or a classic cabinet, a cnc machine shouldn’t be a problem

>Total expected cost?
I’d like to stay under $1500


>Use of scavenged materials?
Absolutely, and also ebay materials.

>Expected working area for each axis?
I’d like to see something with the working area of the cheapcnc.com design. I don’t remember the exact dimensions right now, but it had to be at least 12 x 24” on the x and y.

>One user suggests a design around 5/8 x 18 printer shafts. Now that's more along >my line of 'Cheap'. In fact I just harvested two of these and have two more to tear >apart.
That is another approach. I am not pursuing anything like that at this time. It sounds like you wouldn’t have much working area. Maybe 12” x 12”? But if that’s all you need, and you’ve already got the shafts, then perhaps you want to adapt the jrgo design to your components?

There are many ways to get into CNC and no one solution is going to satisfy all the requirements. But we’ve been working on the hardware store design which I think will allow many people to get into their first CNC using regular woodworking tools and readily available components. I hope to be able to do this for a reasonable cost. I hope to be able to do this for roughly $200 in hardware, and $300 in electronic components. Time will tell how close I can get to those numbers. You can follow that thread here: http://www.cnczone.com/forums/showthread.php?t=12681

Mike…

Mike,

The $1500 limit tells me a LOT. Thanks for this and the other clarifications. :)

As to the 'bearing shuttle', maybe you can find some for purchase but the guy who casts his own did well. I think you could build one from plate steel too. Given the goal for your machine I wouldn't count these out, they are just outta my league (but that's ok). :)

Keep up the good work.

Rance

Rance,

What do you think of our $200 hardware store design?

Mike...

rance.
i agree, one must decide what the machine is going to be used for, set some size limits, and expectation goals and go from there.
i was told on here that you can make a machine to do 'a' good job, but a machine that will do all jobs good are far and few between.
i don't think there is such a thing as a universal machine

I've responded to Mike's question on the H/W Store thread since it is not related to this thread. I'll come by here and read from time to time but now that I know more of the goal, this machine really is not in my league.

Perhaps I got the wrong idea, but I assumed that the "cheapcnc design" would mean cheapest. The idea that a possible budget for it is $1500, compared with the "hardware store cnc" being $200 sems to stand everythjing on its head.
I think I'll go and lie down for a while to get my head round it. :confused:

Another option for rod in OZ is Bright mild steel.

2c worth