[GPT]

Hi All

Been reading for a while and I'm close to starting a build.

[] I'm looking at constructing primarily from 25 mm (1.6g) RHS.

[] Planning on making the linear rails from 3mm angle and using skate bearings (or similar).

[] Looking at basic threaded rod with some kind of anti backlash device (I've seen a number of designs on the forums (trying for extra cheap here)

What I wanted to know was:

[a] built carefully, am I likely to get 0.1 mm accuracy with this arrangement? I don't think that's stellar accuracy but all I really need.

[b] I only need a working area of 30 cm square but was wondering whether scaling up would reduce accuracy much. I'm thinking either 60 cm by 120 cm or 120 cm by 120 cm?

[c] Is the ACME lead screw worth the extra cost i.e. if you can remove the backlash is the threaded rod OK or is the ACME rod much better?

Thanks

Gerard

[louieatienza]

Hi All

Been reading for a while and I'm close to starting a build.

[] I'm looking at constructing primarily from 25 mm (1.6g) RHS.

[] Planning on making the linear rails from 3mm angle and using skate bearings (or similar).

[] Looking at basic threaded rod with some kind of anti backlash device (I've seen a number of designs on the forums (trying for extra cheap here)

What I wanted to know was:

[a] built carefully, am I likely to get 0.1 mm accuracy with this arrangement? I don't think that's stellar accuracy but all I really need.

[b] I only need a working area of 30 cm square but was wondering whether scaling up would reduce accuracy much. I'm thinking either 60 cm by 120 cm or 120 cm by 120 cm?

[c] Is the ACME lead screw worth the extra cost i.e. if you can remove the backlash is the threaded rod OK or is the ACME rod much better?

Thanks

Gerard

I would say get the ACME screw if you can. You can use regular threaded rod, but their pitch is very fine, so your machine will be very slow, maybe not as accurate or finished as ACME, and prone to wear faster.

Even if you use 1/2", 5-start ACME your resolution with a stepper, direct-driven, will be .0025", and you'll get better performance out of your steppers. To me, repeatability is just as, if not, more important, and making sure you minimize backlash as much as possible is important... Also adjustability to get your bearing systems tight as possible without binding. And ridgidity; making your structure ridgid enough to handle what you are cutting. Fortunately, since you are making a small machine, that makes it much easier...

[GPT]

Thanks for the reply.

From what I've read, everyone seems to come round to the ACME screws.

Since the first post I've found a cheaper supplier for the ACME lead screws. I guess it will be one of the last things I buy so if my machine looks up to it I suspect I'll bear the cost. And have more time to find the cheapest supplier

Do you think the steel angle / skate bearings will work OK? I've seen lots DIY designs that use that option.

I'm was planning on using threaded rod through holes in the RHS frame to hold the pieces true while I do the welding. Does that seem like a reasonable idea?

Again, thanks for your response.

Gerard

[Harpye]

Hi gpt - I don't want to disencurage you but 0.1 Tolerance on what???

A) Position ??? a standard rolled ACME 7e is up to 0.7mm accumulated modul tolerances on 30 cm length

B) repetition on no load situ - if you build a stiff frame the reverse play will probably take this tolerance

C) repetition on Load situ .... the Skaterbearing construction will probably not have the stiffness to keep 0.3mm on positioning

Regarding the size - think about the increasing flexibility in relation to the length .... with a 1,2mm long 20mm round shaftrail (not suported)
the ganty weight of 15 KG will cause the shaft hanging through in the middle more than 0.1mm
and a 1.2m long Aluminium bar has probably more than 0.1mm expansion on just 10K raise of ambient temperature!!!

I have a fully IT5 setup but would not claim to keep 0.1mm over all the ranges of temperature distance and so on .... but for the smaller parts ( distance) it is accurate...

[GPT]

Hi Harpye

I don't think there is any possibility of discouraging me

I guess what I mean is 0.1 mm with regards repeatability of positioning.

So, if I start cutting a 30 cm circle in several passes I'd hope that by the time I got back to the start of the cut I'd be within 0.1 mm of where I started.

I'm guessing if the software accurately controls the steps, that ought not be too difficult.

Knowing nothing at this point I'm not even sure if it's feasible that you can step the motors that accurately.

As to accuracy of lead screws, if I'm out by 0.7 over a 30 cm circle,I won't cry.

If I can't reposition the tool to 0.1 of where I started, I might

Thanks for taking an interest.

Gerard

[Harpye]

THe 0.1mm Are dependant on where you start in your circle which probably will not be a circle on the machine but somway a rounded square - you don't have playless / preloaded or at least tolerance reduced tread / nut combinations e.g. at mach you can try to implement the play compensation but this just works half way correctly on non circular objects

CIrcular objects will NOT be circular! To have just 0.1mm play on a spindle nut combination over time , load and wearout is an illusion ... I would assume and calculate with 0.3mm play which ( all other issues about tolerances are ignored) will cause a 0.3x0.3mm large field of missmatching start and end point ( as a compromise this can -as mentioned- be compensated by play comp. function ... additional there is a tolerance created by the BB system ( what will you use ... diagonals... good ... radial BB . NOT GOOD ...use motor bearing for spindle is a NO GO! THere is another tolerance by ortho of the axis ... etc....

Make a calculation on your machine statics and check out the tolerances scenario before going with the wrong material into the project...

Hj

[louieatienza]

GPT, I run 1/2"-8, 8 start ACME with anti-backlash leadnuts on all 3 axes. I've been able to cut very fine detail, and repeatability is excellent! If your cutting forces don't exceed the limits of your anti-backlash leadnuts, and your steppers are sized right, then your repeatability will be dependant on any backlash in your leadscrew bearings, couplings, belts, as well as the ridgidity of their mounting and the machine itself.

After upgrading my motor couplers, motor mounts, leadscrew bearings, I was able to remove as much play as I could from the drivetrain. Now it's just a matter of running my machine within it's limits to ensure it can repeat! And I do believe that it's more important than accuracy.

As for skate bearings/angle iron, my first machine used a skate bearing assembly on aluminum tube, and repeatability was pretty good! It may take a bit of fudging to get it set right, and as for my first machine (Solsylva plans) it relied on the flex of the bearing mounts to "preload" them. On industrial slides, they're made accurate enough to havve the preload "built in" or adjustable. The trick is to make your bearing configuration in a way that it can be adjusted!

[GPT]

Hi louieatienza

Thanks for stopping by.

It is repeatability that's seems to be more of an issue than pure 'accuracy', as such.

I've started looking at motors and controllers. I'm going to used scavenged steppers and home built controllers to start with, to get an idea of how the circuits work, before probably going with some commercial kit / stepper arrangement.

Can I ask what sort of (size / materials) machine you use, and what stepper / controller / software combination?

I plan to start building after Easter. Now undecided whether to go Steel RHS or plywood/MDF for the table and gantry. I can weld and steel is cheap, but as I'm looking to mostly do small stuff, maybe a lighter machine would be better?

Thanks

Gerard

[louieatienza]

GPT, I made my current machine out of 1" paper-based phenolic left over from a job. It's very stiff, heavy, but can be cut with woodworking tools. It doesn't smell pleasant. But it can be tapped and holds threads well. 1" is not easy to find, though you could find 1/2" and 3/4", as it is used mainly for bathroom stall and shower partitions... The cloth and glass-based stuff is way more expensive, but probably more durable.

I used ROLLON 43 series linear slides for the x and y axis. I just scored two more rails for $20 each on eBay! They're super strong and cheap, though getting the preload tuned on them takes a bit of tweaking.

Leadscrews are the aforementioned 1/2"-8, 8 start from McMaster Carr. I used dumpstercnc anti-backlash leadnuts. I used 1/2" bearings from vxb.com for the leadscrews, in bearing blocks made out of HDPE (better known as chopping board!) I put a slight preload on the bearings...

The x axis is run by 2 leadscrews, which are run by a timing belt and pulleys. The y and z axes are direct-driven. To make things easy, I used pulleys with a 1/2" bore, and 1/2" shaft collars, so I didn't have to turn journals on the leadscrews.

I used the 3-axis 425 in-oz xylotex kit. I use Mach3 for control. My cam software is Vectric Cut2d and VCarve Pro 5.5, and OneCNC XR4.

I don't think you have to worry too much about weight with a small machine. Mass and stiffness help, but only if your steppers can move them. The steppers are a lot stronger than you might imagine, and the leadscrews/pulleys can multiply that torque!

You can see my build here http://www.cnczone.com/forums/diy-cn...d_another.html and my machine in action here YouTube - AtienzaLouie's Channel

[GPT]

Hi louieatienza

Had a look at your build thread an videos. Cool! Particularly liked the BMX sprocket in aluminium.

I had a quick look at local prices for phenolic board and it's not cheap. I also checked out HDPE and cutting boards seem to be cheaper than uncut sheets (go figure).

The more I look the more I think I just have to jump in and build something cheap as a learning exercise.

So, the plan is:

[] mdf or plywood for a torsion box style table and gantry
[] steel angle and skate bearings for linear guides
[] threaded rod and some home built backlash solution
[] home made controllers and scavenged motors

Then, once that's all working, I'll think about the machine I really want to build

I will probably start out with Mach 3.0 though. All the video tutorials I've seen make it look like a user friendly product.

Gerard

[louieatienza]

Hi louieatienza

Had a look at your build thread an videos. Cool! Particularly liked the BMX sprocket in aluminium.

I had a quick look at local prices for phenolic board and it's not cheap. I also checked out HDPE and cutting boards seem to be cheaper than uncut sheets (go figure).

The more I look the more I think I just have to jump in and build something cheap as a learning exercise.

So, the plan is:

[] mdf or plywood for a torsion box style table and gantry
[] steel angle and skate bearings for linear guides
[] threaded rod and some home built backlash solution
[] home made controllers and scavenged motors

Then, once that's all working, I'll think about the machine I really want to build

I will probably start out with Mach 3.0 though. All the video tutorials I've seen make it look like a user friendly product.

Gerard

Phenolic is definitely not cheap! Though if you scrounge you can find some good pieces (check out commercial bathroom remodels) that would make a great work table, or bearing blocks... Cheaper, and easier to work with, than aluminum!

Mach3 is pretty awesome piece of software! You can even try it out free, it will output 500 lines of g-code, good to at least watch your machine move and do simple cuts.

When configured well the skate bearing/angle iron/pipe systems are remarkably robust and accurate for the cost!

[Harpye]

I built mine from fine plained Aluminium plate and pay about 10 Euro / KG including VAT

simple aluminum not plained is about 60 to 70% of the price...

dont know about phenolic board ( do you mean pertinax) but dependent on what you want to achieve in handling and regarding coolant...

I have seen guys dooing woodworking machines from this material ... pretty stiff and rigid, but not directly usable in combination with high humidities


Its a kind of multiplex - in Germany it is called Siebdruckplatte

http://www.schweiger-weilheim.de/images/ximg2335.jpg

[louieatienza]

I built mine from fine plained Aluminium plate and pay about 10 Euro / KG including VAT

simple aluminum not plained is about 60 to 70% of the price...

dont know about phenolic board ( do you mean pertinax) but dependent on what you want to achieve in handling and regarding coolant...

I have seen guys dooing woodworking machines from this material ... pretty stiff and rigid, but not directly usable in combination with high humidities


Its a kind of multiplex - in Germany it is called Siebdruckplatte

http://www.schweiger-weilheim.de/images/ximg2335.jpg

Harpye, they use stuff similar to that here, for concrete forms. Withstands multiple use, and yes, it's super-strong!

There are many varieties of phenolic, all are layered material bonded with resin. Paper is most common (think Formica backing and bathroom partitions and lab tables) to cloth of cotton even fiberglass. PCB boards are considered phenolics. Garolite, Bakelite, Micarta some brand names...

I paid about $550 (including $100 fretight) for 24 square feet of 1-1/4" thick Garolite in surplus. That would easily be about $3000 at least if I used aluminum!

[Harpye]

For me it is rather easy to use aluminum as I try to get a very light and small machine.... its about 250 € for Alu

[louieatienza]

For me it is rather easy to use aluminum as I try to get a very light and small machine.... its about 250 € for Alu

True, but your mill is rather small! Even 250 euro seems pricey! The phenolic I got ended up being $20/sqft before shipping, so at a size machine like yours would be pretty inexpensive. IF I could, I'd have built my machine from aluminum or steel as well, but after pricing it out, I needed a seat! Plus evevryone here seems to be building from steel and aluminum these days, and I've not seen another phenilic router here, so it's fun to be different! I have some more phenolic that I will use to build another smaller machine and hope to get it going soon...

[Harpye]

Did you consider ceramic / epoxy concrete to predefine the shapes / bed the rails??

[louieatienza]

Did you consider ceramic / epoxy concrete to predefine the shapes / bed the rails??

I actually DID think of using epoxy/granite. But then I'd have to make forms, so it didn't seem the mose expedient way. I would like to use a slab of granite for a work surface as it's super flat and hard, and inexpensive, but I have to figure out how to hold down material.

GPT, didn't mean to get a little off topic here... MDF is a very strong material with good vibration damping qualities; you just have to design your machine to work around it's limitations. Mainly, it's tendancy to delaminate with fasteners installed through its edge. A lot of guys here use barrel nuts or threaded bushings to add strength to these parts. It is also a good idea to coat or paint your machine as well. Baltic birch would be another good material. The chopping board stuff makes great bearing blocks, and I even used it for my motor mounts!

Look through the threads for the Grumblau design. It's an MDF machine, and a beautiful one at that! With a spinning nut on the y axis!

[GPT]

Hi

I had a look at the Grunblau machine. It is a very simple yet elegant design.

Looking at it makes me think I should discard my light torsion box design and just make a smaller copy of that.

I was wondering about the angle aluminium bearing design the Grunblau machine uses (which I've seen on a lot of other machines).

It seems to me that supporting the weight of the gantry with the bearings at an angle of 45 degrees to the load is not ideal? The axle bolts are parallel to the aluminum track surface but I'd imagine some shear force still acting on the bearings?

Gerard

[louieatienza]

Hi

I had a look at the Grunblau machine. It is a very simple yet elegant design.

Looking at it makes me think I should discard my light torsion box design and just make a smaller copy of that.

I was wondering about the angle aluminium bearing design the Grunblau machine uses (which I've seen on a lot of other machines).

It seems to me that supporting the weight of the gantry with the bearings at an angle of 45 degrees to the load is not ideal? The axle bolts are parallel to the aluminum track surface but I'd imagine some shear force still acting on the bearings?

Gerard

I honestly don't believe it's a huge issue, and you might be overthinking this. Believe me, with the power these steppers have, you'll bemd your machine before the bearings fail! The bearing bolts are actually parallel to the tangent of the point of contact of the rail and bearing., and perpendicular to the rail direction. Thousands of DIY CNC routers were and are made this way (even my first, which I also cut aluminum with) and are still in service, look at the Joe'sCNC and JGRO forums!

You could do a CNC Router Parts or Momus style bearing, but now you're making things cost more, so you have to decide how far yo're willing to go for the fist build...

[markskells]

Louieatienza- I was inspired by one of your videos to rebuild my 2x3 foot plywood-constructed Router ( ALA Gatton's Sidewinder) using mostly 90x90 aluminum extrusion and Phenolic Board for the Gantry plates and for almost any flat part. I have test fitted and I'm confident I can complete the build in a few good Saturdays. If the Phenolic doesn't work out, I'll recut the gantry plates in aluminum, bit I really don't thin I'll have to. I'm curious as to your opinion as to why phenolic board is not used more. The literature has few citations and you really have to dig deep to learn anything about what seems to be a terrific material. Regards, -Mark