Siff' Build Log

Hi guys, below are some images and renders of my first proposed CNC machine. I intend to do a small amount of aluminium and wax routing, but mostly I intend to use this to aid in the manufacture of speaker cabinets and PCB milling.

I've just posted an RFQ over in the other forum asking for quotations on getting some of the parts made up.

Cutting area is 400mm x 400mm and the parts I intend to use include:

Ballscrews all around
Kress 1050-1 FME
Valuframe extrusion
Gecko 540 stepper driver

For the stepper motors i'm not sure, but I was looking at these: http://cgi.ebay.co.uk/3-Stepper-Moto...7#ht_535wt_911. If I understand correctly, even though the motors require 4.2A and the gecko only provides 3.5A, this would mean that the motors just run at reduced torque, and wouldn't damage them, or the controller. Please correct me if i'm wrong.

The only thing I don't like about the design is how far the Y axis stepper motor sticks out. I was originally looking at setting the motor in the upright support and using a timing belt/pulley combo, but i'm not sure i'm willing to deal with the extra cost and potential increased backlash for the sake of aesthetics.

http://farm7.static.flickr.com/6136/...ac2ddfb0_b.jpg
7 by SgtSiff, on Flickr
http://farm7.static.flickr.com/6026/...b2497716_b.jpg
4 by SgtSiff, on Flickr
http://farm7.static.flickr.com/6001/...1cf1b6a6_b.jpg
3 by SgtSiff, on Flickr
http://farm7.static.flickr.com/6010/...b7fafaa7_b.jpg
2 by SgtSiff, on Flickr
http://farm7.static.flickr.com/6147/...ddbabd60_b.jpg
1 by SgtSiff, on Flickr
http://farm7.static.flickr.com/6150/...7f168466_b.jpg
5 by SgtSiff, on Flickr

Thanks for looking.

Looks good but I am sure you will regret using unsupported rails. Changing to supported rails will give a much stronger and more accurate machine.

Unsupported will be fine as long as you make them thicker. I would suggest turning the Y frame around 180 degrease as well to centre the weight more and you could then shorten the bed/ rails. You could always mount the Y motor around the back and use a small belt 1:1 to power it as well to stop it sticking out?

Oh dear. Suggest you redesign it a bit and do everything possible to get rid of the overhangs. You really want the shortest possible connection between the router bit and the workpiece to minimise flex. Do you really need that much travel on the Z?

Flex is easy to overcome, you simply reduce the cut depth/feed. If you want to cut deep and hard fix it now before it's all too late :naughty:

Quote:

---

Originally Posted by everyreasonto

If I understand correctly, even though the motors require 4.2A and the gecko only provides 3.5A, this would mean that the motors just run at reduced torque, and wouldn't damage them, or the controller. Please correct me if i'm wrong.

---

That is correct. You could always run them in bipolar series ... then you only need 2.1A. That will reduce the torque slightly at higher speed than with parallel, but I'm pretty sure it will be less of a loss than using only 83% of the rated current. The power dissipation of the motors will be about 70% of rated running on 3.5A. Why gecko? Something like PM752 or PM542 on from eBay or Zapp should be cheaper and will let you run at the full current and (with PM752) higher voltage.

I agree unsupported rail is a bad plan, especially on the X-axis as the end fixity is poor. Even 30mm unsupported rail will flex more than 12mm supported. Granted the load ratings of the bearings on the 30mm rail would be higher but they're plenty more than required anyway. I would use 16mm, or more if you can afford it. 25mm supported rails are good as they have an additional row of balls ... though on this size of machine you don't need that size. Have you looked on eBay for them? They're cheap from linearmotionbearings2008.

You might want to bolt the Y axis plate to the back of the gantry, not the front, as that will get the centre of mass closer to being half way between the pairs of X-axis bearings.

There is a lot of overhang, as Robin said do you really need that much Z travel? If so you're best to eliminate the gantry sides altogether, like I did here:

Attachment 4252

Plate is also the worst profile for the gantry sides since it bends easily in the direction parallel to the Y-axis, however you've compensated for that by having a plate parallel to the Y axis, so the only thing it will lack is torsional stiffness. You could use something lighter if you changed the cross section - maybe extrusion? That's expensive though. Steel box section is cheap and the weight isn't really an issue since the gantry is fixed. Just make it as strong as you can.

Also if you want the machine to be fast use timing belts on the stepper motors/screws to get the motor operating in the right region. Since you're machine is quite light you will almost certainly want the pulley on the stepper motor to be bigger than on the screw. What pitch ballscrews are you using, probably 5mm? I have not noticed increased backlash from using timing belts/pulleys. They do stretch very slightly, but I don't think that's a big issue.

Nice drawing ... if you do decide to use supported rails I can send you the drawings I did of them to save you a bit of time if you like...

Thanks for the input guys, I really appreciate it. I hadn't seen the seller linearmotionbearings before.. boy is he cheap! I'm really liking the look of the machine you posted Jonathan but wouldn't I need two ballscrews and two steppers on the X axis?. looks like I could get a larger machine for the same budget, using the supported rails that everyone seems to be suggesting from linearmotionbearings. The rails I put in the original plan were 16mm so I guess I overestimated their strength.

Budget is my only constraint, excluding spindle, I have about £1000 to spend.

Taking advice on the stepper drivers, I've found some from a seller called cnc4you. I also intend to get my steppers and ballscrews from there.

EDIT: Looks like i'm going to go for this set: http://www.ebay.co.uk/itm/3-SBR20-se...#ht_1004wt_948. Would work out around £450 shipped (inc tax) and they are all 20mm rails.
It would fit very nicely to the 60x30 profile from hepcomotion.

I'm going to make a point of avoiding Valuframe extrusion because they said that delivery would be £33 + VAT if they were sending to a private address, robbing gits!

Quote:

---

Originally Posted by everyreasonto

I hadn't seen the seller linearmotionbearings before.. boy is he cheap!

---

When you've decided what lengths of ballscrew and rail you want send him a message with the sizes and he will list them on eBay.

Quote:

---

Originally Posted by everyreasonto

I'm really liking the look of the machine you posted Jonathan but wouldn't I need two ballscrews and two steppers on the X axis?.

---

Yes you would - it's the best way to do it anyway. You could link the two ballscrews with a timing belt and just use one stepper motor. Another way is to use cables beneath the gantry (like on a drawing board ruler) to link the two sides...

There's more about it in my build log, probably just read the last few pages!
http://www.mycncuk.com/forums/showth...outer-building...
Also here for more pictures, though this isn't neccecary for the size of your machine.
http://www.mycncuk.com/forums/showth...t-design-ideas

Quote:

---

Originally Posted by everyreasonto

Budget is my only constraint, excluding spindle, I have about £1000 to spend.

Taking advice on the stepper drivers, I've found some from a seller called cnc4you. I also intend to get my steppers and ballscrews from there.

---

£1000 not including spindle should be plenty...less than I spent.
Get your ballscrews from linearmotionbearings. They will be exactly the same and end machining is very very cheap. Look on eBay for stepper drivers - there's some 80V ones that look tempting, or if not this is what I got because at the time they were not on eBay cheaper:

http://www.zappautomation.co.uk/pm75...71a69c8def8e2f

I also used the 3nm motors...

I think steel box section is much more cost effective than using extrusion. Definitely for the stationary parts where weight isn't an issue.

There's a pdf with all the drawings for the ballscrews/blocks etc in one place here ...
http://www.file-vault.us/pdfs/ball_s...umentation.pdf

My router was built with rails, bearings and balls crews from linearmotionbearings. The cutting area is 500 x 600mm and the whole build came in at about £1000. I did what Jonathan has suggested and built it from mild steel box section.

Here is a link to the build thread, it may give you some ideas http://www.mycncuk.com/forums/showth...-framed-router

Good luck with the build.

http://img62.imageshack.us/img62/6769/tempge.jpg

Hi guys, this is my latest design. It is based off of Jonathan's design and uses:

15mm plate
20mm supported rails
twin x axis ballscrew
Aluminium extrusion for x axis rail support and Y axis.
Will be welded to an aluminium frame when complete

Let me know if you guys think this is an improvement. The Z axis isn't completely finished yet..

Quote:

---

Originally Posted by everyreasonto

Will be welded to an aluminium frame when complete

---

Why aluminium? Will be interested to see it welded ...

Quote:

---

Originally Posted by everyreasonto

The Z axis isn't completely finished yet..

---

Good, as it wont work like that!

Y-axis: You need to have the 4 plates that make up the 'box' round the linear bearings configured such that the plates with the bearings on can be adjusted to get the bearings precisely parallel. Look how I did it ... just don't copy my mistake - making two of the plates 10mm too short! At the moment it relies on the two plates in the YZ plane being machined very very accurately which just isn't going to happen.

X-axis: End supports for the ballscrews look a bit bendy? Remember that all the cutting force in the X direction will be resisted by those plates, so they want to be strong to stop the tool deflecting much. That's why I made mine the odd looking shape out of bar. The easiest solution I can think of is to replace them with box section then put a smaller plate/disc on that to mount the bearing. Have the holes slotted on both so you can align the ballscrews.
If you rotate the ballnuts by 90° you can get the ballscrews slightly closer to the tool...probably not significant.

The way your Y-axis box section is mounted to the X-axis is a bit bendy as the vertical plates only have a line contact with the X-axis bearings. That's why I used box section...

You could get the gantry even further down? Debatable if it's worthwhile.

Other than that looks good (but obviously I'm going to say that as you copied mine :smile:)

Using aluminium because I can get it cheaper than steel and my rails are sitting on ally extrusion. That would mean I could then weld the ally extrusion to the base.

I'll add something in to make the bearing adjustable.. something I never thought about, so thanks!

I've reinforced the joint between the x and y axis, but I believe the bearing plates will be strong enough. I know some builds don't even use bearings on the 'floating' side and they will be made from 15mm ally.

Wish I could lower gantry but i'm trying to avoid using timing belts/pulleys where I can, so I had to raise it so the motor would clear the ballscrew :sad:

Also, quick question, where do i get pulleys from with a 10mm bore to fit the ballscrews? And can I get the belts in any length? Should I add a tensioner further down the line?

Finally, once I have bolted together all of the y axis and confirmed it is square/smooth etc, If I weld along the seems (lack of a better word) will it cause the ally to distort at all?

Cheers, Joe :naughty::naughty::naughty: (looking forward to getting it built)

Quote:

---

Originally Posted by everyreasonto

I believe the bearing plates will be strong enough. I know some builds don't even use bearings on the 'floating' side and they will be made from 15mm ally.

---

That sounds reasonable as long as they are mounted securely.

Quote:

---

Originally Posted by everyreasonto

quick question, where do i get pulleys from with a 10mm bore to fit the ballscrews? And can I get the belts in any length? Should I add a tensioner further down the line?

---

I think this is the cheapest place (HTD is the best):
http://bearingstation.co.uk/products...HTD_Pulley_5mm

You could drill it out to 10mm, but that's very hit and miss unless you have a set of drills in 0.1mm increments and even then I'd bore it. Or use a reamer, but then it's still not likely to be concentric.
Get them delivered to me and it's only about a 10 minute job for me to bore them to 10mm accurately. :wink:

There's various length belts at that site ... you'll soon find the formula to work out the belt length for a given centre distance (it's pretty simple). If not I can send you a spreadsheet with it. I just made it so that the 4 holes that mount the motor are slotted so that you can accommodate different length belts and different pulleys.

Quote:

---

Originally Posted by everyreasonto

Finally, once I have bolted together all of the y axis and confirmed it is square/smooth etc, If I weld along the seems (lack of a better word) will it cause the ally to distort at all?

---

It defiantly would with steel, ... this implies it's worse with aluminium:
http://www.aws.org/wj/2003/02/028/

You need to design it such that there is adjustment designed in to the relevant parts such that it does not matter if it distorts. Obviously clamp it well when you weld it... I did that with mine and didn't really notice any distortion.

Ill probably do that then when I order if thats ok. Can you send me the spreadsheet please? I really can't be bothered to work it out, Just got back from work and i'm knackered!

Anyway, I've been buying today. I haven't ordered anything structural yet because i'm still working on it but I have bought:

Power supply, 3 x 3.1NM steppers, 3 x 4.2A drivers, 3 x couplings = £264
Breakout board = £13.99
4 x ballscrews, 6 lengths of rail, 12 x bearings, er11 collet set = £350 (exc import)
1.5kw spindle and inverter = £201 (exc import)

Still need to order belts, pulleys, extrusion, ally and some milling bits. Can't wait!

Oh, I also picked up a dedicated PC today for the workshop... 2.6ghz P4, 512mb ram, 40gb HDD, old ati graphics card and 15" monitor for £30!! Not bad eh?

Jonathon, what are the dimensions of the spindle mount that you made?

Regards

Quote:

---

Originally Posted by everyreasonto

Ill probably do that then when I order if thats ok. Can you send me the spreadsheet please? I really can't be bothered to work it out, Just got back from work and i'm knackered!

---

Found this, http://www.sdp-si.com/cd/default.htm ... simples.

Quote:

---

Originally Posted by everyreasonto

Jonathon, what are the dimensions of the spindle mount that you made?

---

This:

Attachment 4288

Or:

Attachment 4289

(but with more rounded corners...)

Or whatever you need really...the 150mm wide ones are 1/2" thick and the 125mm wide are 5/8" thick.

Looking so much better.

I agree with Jonathan, the X axis screw mounts look a tad flimsy. You need it rigid at the end which bears the thrust bearings.

Presume you are bolting it down in all 4 corners to something solid.

If you rearranged it a bit you could cover the X axis screws with aluminium angle and shield them from the worst of the swarf coming off the tool.

Anything you weld will distort. Okay to weld so long as you can machine it back to square after it has cooled down.

Very refreshing to see someone actually taking advice rather than rushing in to it and having to make a Mk 2 very soon after the Mk 1. Keep drawing, you won't regret it :smile:

Update time!

Rails etc haven't arrived yet so I've been busy getting my garage ready. It was empty before, so I've built a workbench and put it on it's own RCBO protected ring. I also got my workshop PC set up:

http://farm7.static.flickr.com/6009/...549f8e3d_z.jpg

My steppers/drivers have arrived so I began creating my box of electrickery! :dance: These are all terrible photos by the way (taken on camera phone with a sprinkling of couldn't be arsed) I don't have any progress photos just the almost finished box:

http://farm7.static.flickr.com/6005/...5acffed4_z.jpg
http://farm7.static.flickr.com/6127/...8f75958d_z.jpg
My homemade 24v/5v power supply for the cooling fans and breakout board:
http://farm7.static.flickr.com/6030/...875d1cf5_z.jpg
This shot shows the almost finished box except for the cooling fans because they haven't arrived yet. The stepper drivers have been bolted to the case and are covered in thermal paste where they make contact. You can also see my line laser on this shot, which is going to be for my 3D scanner project. Fans + Filters will be attached on the outside.
http://farm7.static.flickr.com/6030/...9d405bf6_z.jpg
http://farm7.static.flickr.com/6023/...f6907899_z.jpg

I'm still waiting on the breakout board and E-stop switch, which will be fitted in/on this case.

I've used glands and filters on all openings to keep dust out. The only possible place for dust to get in is the brush plate at the bottom, it will keep 99.9% of the dust out though so i'm happy.

I've also got a homebrew temperature sensor gizmo, that will let me know if the case is getting too hot. I doubt this will happen because, erm, well, it just won't! :)

Also, I've only got a baby pillar drill so I couldn't get the case under to drill those 60mm holes... they were a pain in the Harris to cut by hand and my baby walty almost burned out!

Thankya, goodnight, much lav!

Hi guys, just wondering if I could get some help. I can't seem to get my steppers to turn via mach 3. I know the motors are wired up correctly becuase I tested by shorting enable and the simulating pulses with a link wire and they turned just fine.

I've set up the correct port and pins in mach 3 (or so I think!), but it is as if it can't see the breakout board? The emergency stop button doesn't do anything and I don't know where to start looking... I don't know if I was doing it right, but I resetted the software on mach3, went to diagnostics and it showerd all 3 enables were on. I tested between the gnd pin a pin 1 on the parallel port and got nothing? I am howerver getting around 4v on on pins 12, 11 and 10.

Cheeers

What breakout board do you have? Does it have a charge pump and if so is it configured? You don't need to do anything with the enable inputs.

Hi, it's this breakout board:
http://www.zappautomation.co.uk/zp5aint-p-499.html

It's being run from a regulated 5v supply, I don't know if it has a charge pump
Didn't have time to recheck all of my wiring but i'm going to have a fiddle tonight.

Some good news though, my spindle arrived today! Seems excellent quality and I can't wait to get it fired up after i've configured it. Does anyone put anti freeze in their water to aid in cooling/ add corrosion resistance?

Cheers, Joe

Doesn't look like it has a charge pump...

http://www.slidesandballscrews.com/p...d%20manual.pdf

You should put something in the water to stop corrosion / algae etc. I've not yet but will do soon.

I've taken a break from getting the drivers to work to re-think (again!) my cnc layout, I came up with the design below. What do you guys think? The Y axis is more secure and lower down, the end plates are more substantial and will have 4 x m8 fixings each and the aluminium rail is 60 x 60 all around. I haven't added in the Y axis ballscrew yet but I think i'll put it along the top. I've also used one solid piece for the back-plate which will hopefully add a bit more strength. It needs it.. All my rails arrived yesterday and I didn't realise how heavy they all were combined :confused:

http://img825.imageshack.us/img825/5533/tempaq.jpg

Looks much better to me.

Putting the Y-axis ballscrew along the top is not a good place since it's so far from the cutting force. The Y-axis bearings obviously don't provide any support parallel to the Y-axis, so they would be pushed slightly in that direction. If the screw is lower down it's less of a problem. The best place would be to put it on the opposite side, but that's a bit in the way...

Not sure about the way you've mounted the X-axis linear bearings. Though the forces might by more even, it's much much harder to get them parallel. I'd keep the orientation you had previously. I think it's best to have them that way anyway as the forces are not even - you've got the weight of the gantry loading them in one direction.

Ok thanks, think I'll stick with the previous design then, because that solves most of the problems.

http://img829.imageshack.us/img829/4972/cadfinal.jpg

Hi guys, whilst i'm waiting for my extrusion to come, I've been faffing with my design and one think that has struck me is how off balance the z axis spindle is in relation to the linear bearings/center of gravity. Maybe i'm underestimating the strength of the rails/aluminium but would the design be stronger if I were to add some type of counterbalance or am I looking into this too much?

:confused:

That looks stronger than previously :)

Put the rails on the Z-axis and bearing blocks on the Y axis box as that's stronger. With how you've drawn it the Z plate will bend wherever the Z-axis is. Third time I've posted that this week...wait you had it like that in post #21, why the change?

The centre of mass is not ideal, but it depends how you look at it. Ideally you would have the contact force at each bearing equal. If you calculate the forces and compare to the ratings of the bearings (http://www.mycncuk.com/forums/showth...near-bearings-... I should find the spreadsheet I did for that) then you will find its not significant. If you're drilling something the forces will change and actually get more balanced.

You could compensate by mounting the gantry further back between the bearings so that the spindle is closer to them. Ideally work out the centre of mass of the gantry assembly, minus X bearings, and mount it such that that point is directly between the two bearings. That's what I did, but couldn't get it that far back so ended up about 30mm off the centre of mass.

Ignore the z axis, I've just been faffing (a product of impatience) and this is by no means my final design.

Cheers, if you have that spreadsheet still, id appreciate it! 'If you're drilling something the forces will change and actually get more balanced' is something I never considered. I imagine you could go on forever in the pursuit of perfection but i'm happy in the knowledge that it will probably be negligible :tongue:

Cheers, Joe

Just an quick update to let you guys know that this project is still alive...

I have been mega busy with work lately. I work in the solar PV industry, and as some of you may know, they have slashed the FiT so the last month has been hectic getting installs done.

Anyway, Not much progress has been made other than I have just bought a new stepper and driver so I can have dual steppers on the X axis.

I'm off for three weeks soon, so I'm going to get stuck back into it!!