View Full Version : A lucky find, a good machine?

09-11-2005, 01:13 PM
I am so very happy! I happened upon some linear motion stuff for free! Some was even new in box!
What I have found:
-2 times 120 cm (48") IKO LWHD25 rails with 2 bearing blocks per rail. These were new in box.
-4 times 110 cm (45") NSK LGY25 rails with 2 bearings each. These are mounted 5" apart on a 1/2" thick metal plate. I have two plates with two rails each. One of the bearing blocks is missing it's balls though.
-1 times 94 cm (37") 20mm NSK2009z ballscrew with front and end bearing with a 150oz/in SloSyn stepper attached.
-1 times 123 cm (48") 20 mm NSK2009z ballscrew with front and end bearing and a huge whopping 450oz/in stepper attached.
-2 times a 55 cm (21") 15mm rail with one THK RSR15VM bearing each.

I will post pictures in a few moments. I was thinking of using one of the plates of with the 2 NSK rails with the 37" ballscrew module as a complete Y-axis. The new IKO rails with the longer ballscrew will be the X-axis. I am unsure what to do with the z-axis though.

It is highly unlikely that I will happen upon more good stuff any time soon, and building the Z-axis with crappy parts would be a waste of this lovely stuff. Can anyone help me sort out my design?

Kind regards from a very happy Tim!

09-11-2005, 01:40 PM
Hi, here are the images! I have a crappy digital camera, and I hope this is posting is not against the rules.


09-11-2005, 01:44 PM
Those appear to be precision ground BS big $$$, not the cheaper rolled version. :cool:

09-11-2005, 01:53 PM
Yes, I am very happy with my find, as you can imagine!
Is there anyway to order new balls for the bearing block that has lost its balls? How much would that cost?

At the moment I do not have a way to make a good Z-axis. I could, in theory mount one of the two plates with the NSK bearings on the Y-axis and have almost 30" of Z-travel. I wonder if that is a good idea though. Perhaps if I built it like the one at rainnea graphics:
The x-axis rails up high with the large z-axis dropping down.

Another question I have is, will it be okay to drive only one side of the x-axis, not having the ballscrew in the middle of the moving way? I know this can cause binding problems in MDF/skatebearing/allthread machines, but is it really an issue with these high quality components? If it really is an issue I have to scrap that rainnea idea.

09-11-2005, 08:34 PM
The only potential problem I see with this design is the long reach between the router tip and the Y axis. This could allow lots of unwanted flex.
On my first machine I had a 400mm reach but was only cutting 20mm sheet. If you are cutting only sheet material ( unlike what Rainnea does) make your reach as short as possible.

09-11-2005, 09:21 PM
Have you any more pics of this rainnea machine

09-12-2005, 03:26 AM
Yes, more pictures are here:

I would like to do some 3d work as well, not only sheets. Would these large 25mm NSK rails flex so overly much in their bearings?

And what about driving it from one side only, would that cause problems with such high quality materials?

09-12-2005, 06:54 AM
You need to establish what would be the max size you will be doing, I would try to avoid long z axis travel as it is very difficult to get rigid enough to prevent flexing without spending big bucks.

I would be surprised if the 25mm rails would flex much, You need to decide what are you going to build your machine out of MDF, Alum, or alum profile section.

With these rails I would tend to bolt them to a length of the profile section it can be bought fairly reasonable if you look around.

You can also bolt them streight onto a mdf made machine, have a look at Balsamans machine on the forumn he first built one from mdf using rails like yours then a second machine all made from profile section and aluminium.

You can drive on one side no problem, you just need to have everything spot on.
Need to decide what type of machine you are going to build, fixed gantry or moving gantry, you don't need to have an axis overhead like the rainnea machine it still can be moving gantry with the axis driven in the middle underneath the cutting area.

09-16-2005, 05:43 AM
Right, I have decided that a metre long z-axis weighing 100 pounds isn't going to help me.

I have been looking around for specs on the bearings by NSK, the used ones and came up blank. I want to know if it is allright to mount them vertically like the way balsaman did on his first router. They are most likely far better suited to forces that push it "into the rail" but I want to know if it is okay to use them upright anyway.

Kind regards Tim

09-16-2005, 12:30 PM
Flex ...kinda depends on the material and how agressive the cut. I would think most of the NSK stuff is good for any mounting in any direction. Part Number??

09-16-2005, 12:41 PM
I think it's ok to use them vertically if you make sure there's not to much weight in front of them (if you mount them with the upside of the trains facing the front).
Or use something as counterweight.

Btw where in Holland can you find this stuff ?


09-17-2005, 01:02 PM
Flex ...kinda depends on the material and how agressive the cut. I would think most of the NSK stuff is good for any mounting in any direction. Part Number??

NSKLGY25, the complete list of stuff is in the start of this thread.
Jerber: I have no idea where to get more of this stuff. I got lucky. You might want to try PCB companies. The PCB hole drillers (like Posaluxes or the like) use this kind of stuff. Perhaps they are scrapping one?

I managed to revive the bearing block which lost it's balls, bought 100 new balls at a local bearing shop.
And my hobbycnc driver kit has arrived. Let's hope it all works soon!

Any more hints/comments about the construction of this machine?

09-17-2005, 05:06 PM
Free?? Some guys have all the luck! You should play lotto all the time.

09-18-2005, 07:21 AM
Actually some of my luck has run out. It seems that one of the linear bearing blocks is broken beyond repair. The ball bearing guideway cages have snapped right off so it won't hold any balls while moving on a rail. :( So I now have 4 identical rails but only 7 blocks. Ah well, can't complain for the price really.

Anyway, on with the story. What kind of base would be suitable for this machine? In a few hours I will upload what plans I have drawn up for this device.

09-18-2005, 10:12 AM
I hope you won't mind me talking to myself, but I am very happy. I just put together the hobbycnc kit, and it works a treat! I was amazed at how blisteringly fast rapids the ballscrew with the 150 oz/in (which is also 5A but limited at 2A on the hobbycnc board) motor could do. It would do like 350IPM or so. I like it :)

I will try to upload some pictures and maybe a movie when I find the time between all the happy jigs and jumps of joy.

09-18-2005, 12:47 PM
Another Update:
I have decided and I will be building a balsaman clone. Only mine will be a bit bigger: 120*100 cm, governed by the size of my bearings and ballscrews.

The baseplate will be 22mm MDF or perhaps 2 sheets glued on top of eachother to give a rock solid base plate. (Not to mention heavy)

So my machine will have about 70*70*7 of travel. Big enough to cut most things I'd reckon. It should have a fair turn of speed as well!

I hope someone is reading this and you don't mind my monologues.

09-19-2005, 07:27 AM
what voltage supply are you using

09-19-2005, 07:43 AM
I was using a half wave rectified 16V supply without smoothing cap. I am going to use 36V, perhaps from an ATX stack. It should have a lot more power then.

09-19-2005, 08:11 AM
You seem to be getting exceptionally fast speeds with only 16v must be the ballscrews. Dont know if I would go the ATX stack route, I know my atx supply can sometimes keep switching itself off after you have switched it on, although it has never done it when in the middle of a cut.

I like the idea of a transformer, rectifier and smoothing cap, nice and simple.

70" x 70" is fairly big for a MDF machine unless you are referring to balsamans second machine which is made from alum profile

09-19-2005, 08:20 AM
Oh sorry, I have forgotten about the imperial/metric change.

The machine will be 120 cm * 100 cm. The work envelope will be 70*70cm.
Yes, I was also very amazed at the speed. It would go from one end to the other in somewhere around 7-8 seconds. Works out to 300+IPM. Although it did not have much torque at that speed so it is most likely not a good speed to use in a real machine. Perhaps it will have more power with a higher voltage and the current maxed. It will benefit from a gecko anyway as it is 5A.

I will draw up what I am trying to design and post it later

09-19-2005, 08:31 AM
100IPM would be classed as a very good machine, there are comercial ones at $5000 that only rapid at 150IPM and cut at 60IPM

09-19-2005, 08:46 AM
Don't forget, it was a precision ground ballscrew and a big NEMA34 stepper connected to a good chopper drive, and it had no load to push it back except the friction in the bearing as it turned. No wonder it went fast.

Right, I have done the virtual equivalent of doodling on a napkin: I have drawn something in paint in my lunch break to try to explain what I am trying to do.

The drawing shows a cutaway view from the side. I will be making a 1 metre long box of 1" MDF with the metal plate with the bearings attached.
The baseplate will be one or two 1" MDF plates.

Anyway: here it is. Excuse the ****ty quality http://timkoene.4hv.org/design.jpg

09-25-2005, 12:07 PM
All right: Disregard the previous post.

Only the IKO bearings are new enough or good enough for 4 way loading. So they are the only candidate for the gantry. No matter, it only changes the orientation of the machine. The machine will be around 1600mm wide and will need a space of 1700mm in the X-way to accomodate a full sized panel.

My X-axis will be in the style of a bridgeport. I will have 6 block mounted to the baseplate with the rails moving over them. The rails will be left on the identical thick metal plates they are on already. 2 rails per side, 3 blocks per side (One block is faulty and I like symmetry). The X-axis ballscrew and motor will be on the underside of the table, with the nut fixed to the baseplate. This way the table will guard itself against dust. The downside is that the moving table will be 50kgs. Very good for stability, bad for huge speeds. (It should still do 100IPM I guess) My gantry Y-axis will be the fast one, with the new IKO bearings and the 450oz/in high speed NEMA34 stepper.

The gantry is still fixed, and made as heavy as I can to ensure stability and stop vibrations.

I hope to have a copy of solidworks next week to show you some nice models.

Any comments?

10-16-2005, 12:15 PM
Hello everyone,

I haven't abandoned this thread ;) I am having real troubles with making the mount for the ballscrew nut to couple it to the moving table. I tried to make it by making a suitable shape out of 2*22mm MDF, which worked out alright. But when securing this mounting block to the table the mdf cracked. Everytime I used enough force to secure the nut mount firmly to the table (I used 4 allthread pieces going through the length of the glued-together pieces), the wood cracked. I tried 3 different shapes and orientations, all sucked.

What can I use? I have no blocks of metal, nor the tools to machine metal to any precision. Please help me :)

Jason Marsha
10-16-2005, 01:34 PM
Post some pics of the actual part you are having trouble with.


10-16-2005, 01:50 PM
TimKoene....you'll need to select the piece of wood carefully....when wood cracks like that it is either dry or the grain is running in the wrong direction.....if you can't find a piece of wood that is quartersawn, you might make the mounts from some type of plastic...lexan, delrin, or if you have some phenolic board....you would also epoxy some PC board material FR4 into the size you need and use that like you would the wood mount.....

10-16-2005, 02:02 PM
Hmm, I only have access to wood for the moment. Wood and angle iron/alu. Here in holland acquiring blocks of plastic is very difficult.

I have attached a poor drawing of the wooden jig I want to use to mount the nut. The nut attaches to the face of this plate with the 4 mounting holes. Then the mounting bracket is attached with allthread through the length of the wood. I made this of two identical pieces of 22mm (~little less than one inch thick)MDF. The tightening of the nuts pulling the bracket to the table will make cracks appear in the side of the bracket rendering it useless.

Any ideas?

10-16-2005, 05:37 PM
Okay, maybe where or how it mounts to the mating surface is not flat.

Here's a solution, ...make the bracket but radius the inner corners, also another solutions that causes some headaches during assembly....make the bracket square or rectangular then bore out a hole in the center where the ballscrew nut will mount.....it's like having a rectangular/square donut.

10-16-2005, 06:18 PM
You could make the plates wider and bolt the MDF together before you tighten the allthread so that it can't split anymore. Also using a big washer on the allthread will spread the force.
Angle iron would be better but it looks like you won't be able to find a large enough size in normal shops to mount the nut directly, you could however use 40x40x3mm angle (readily available in hardware shops) and use 2 pieces of these to fix the MDF to the table, one on the front another on the rear.

12-04-2005, 02:02 PM
Hell no this thread ain't dead :D
I finally got a chance to make some good couplings between nut and table. Made them out of 1" thick PVC block I found. Works a charm!
I have been able to complete my first axis! The Y-axis is the one on the gantry and I managed to make it magically move all by itself. Does anyone have video encoding software, because the only video I have is 20mb for a 1 minute sequence. Crappy camera codec.

you can find it here: http://timkoene.4hv.org/test/y-axis.avi