Drawer slide design...

We've got several CAD users that are willing to work on some open source designs for us. I need someone to pursue the drawer slide design. Who would like to work on that?

jpot?
chronon1?
phillby?
someone else?

Mike...

I'll be concentrating on this thread, but CAD is one of the things I aim to learn from this project. Right now I have some pen on graph pad sketchs exploring my angled slide notion. Should I post cleaned-up scans here, or forward to whomsoever picks up the task here?

I'll be concentrating on this thread, but CAD is one of the things I aim to learn from this project. Right now I have some pen on graph pad sketchs exploring my angled slide notion. Should I post cleaned-up scans here, or forward to whomsoever picks up the task here?


Definitely post them here... we'll look at them and try to convert them into CAD.

Mike...

It cleaned up fairly well. Paint Shop Pro knocked out the red graph guide lines with just a few clicks. The two lines across the top which I left freehand are the shafting for the Y axis. I designed industrial equipment for close to 20 years using pencil and paper, and changed my direction as my employer was implementing CAD.

The main premise I'm trying to raise here is the notion that angling the slides should increase stability and rigidity. With a fixed table and moving gantry, the angle would go the other way. I expect to have access to a precision cabinet shop when I'm ready to start construction, but consistant angled wedges could be cut with a miter saw.

You'd be better off using 2 pairs of slides at 90° to each other. With your method you will still get a little play in the movement.

Originally posted by GER21
You'd be better off using 2 pairs of slides at 90° to each other. With your method you will still get a little play in the movement.

Believe me, if there'd been two pair in that dumpster I'd have grabbed them both.

We know that one pair of good slides, installed in the standard manner, does a minimally sufficient job. If my idea is valid, it would cut the investment, for anyone shopping from the plans, in half while still yielding some improvement. McMaster (never the lowest price, but cheaper for industry than chasing down the deal) wants $64 a pair for their closest equivalent.

I'm with Ger on this one.

Draw slides take loads in one plane only. Sit at your desk and half open a draw, lift it up and down and it shouldn't move, move it side to side and it will move a long way, wrack it diagonally and it will move about the same.

A gantry cnc will load its bearings in two dimensions and try to wrack the gantry in the same way as the draw test above. Add the leverage and distance of the cutter from the bearing and you've got problems.

I don't know what brand the slides are that were proposed for a diagonal mount but I've got some Home depot full extension high capacity slides here and trying it (before posting) they bind quite badly on a 45%.

Reading this post and the other recent posts about the new designs it strikes me there's little talk of the loads the machines have to be built for and how the different materials and bearings will need different design points in the construction of the machine.

For example given the draw slide option and the loads a cnc machine imposes the slides has to suport two planes of load and so in the ideal solution you'd need two sets one 90% to the other as Ger suggested.

Similarly the black iron pipe skate bearing arrangement has limits and challenges on the span a pipe of a given size can support, the need for support and the need for adjustment. It's not practical to significantly increase the dimensions of the JGRO router and expect it not to deflect big time.

In the same vein a lot of the gantry designs have simple uprights and back plates with no triangulation or support while the tube/rod is supported at the ends only by the thickness of MDF or the ends of bolt heads. They need way more support than that.

Some thought needs to go into the manufacture of the parts for alignment too. Since most of us ( except Ger apparently ! ) can't get our parts cut on a cnc machine first, then the design should avoid excessive number of joins or hole alignments, use techniques like cutting parallel dado's in one go on one board and then cutting the parts from that, bolting together enda and bdrilling all as one part to approximately line up holes, that sort of stuff.

It strikes me that there's a lot of enthusiasm ( which is great) to develop the machines but the engineering of each option has yet to catch up.

Andrew

Is this the kind of critter you are looking for??
don't have any cad [or any other, for that matter] drawing for, pretty simple to put together..
a good square, tape measure, and a saw..

It strikes me that there's a lot of enthusiasm ( which is great) to develop the machines but the engineering of each option has yet to catch up.

Andrew

Do we have any engineering types here? Or are we all machinists and cad users?

Mike...

Okay, let me just throw out a quick idea for the drawer slide design...

Both x and y have the drawer slides mounted at 90* to each other to keep things from shifting.

This is really going to be a simple thing. The table moves, and the gantry is fixed.

Is this the direction we should be taking?

Mike...

P.S. Are drawer slides in Europe metric? Can someone point me to a URL for metric drawer glides?

Okay, let me just throw out a quick idea for the drawer slide design...Is this the direction we should be taking?

Mike...


Mike,

Hi. My concern would be that the width of the table might work against you-see the attached pic. What about using the slides as they would normally be and use another method to limit movement in the unwanted directions? Something like delrin guides to contain the slop?

Lance

Okay, let me just throw out a quick idea for the drawer slide design...

....

Is this the direction we should be taking?

Mike...

P.S. Are drawer slides in Europe metric? Can someone point me to a URL for metric drawer glides?

As far as I know they are metric, but only on this side of the tunnel.:)

http://www.chambrelan.fr/en/pages/interface/iframe.htm

or www.chambrelan.com with intro.

Is it an idea to use more then two slides per axis ?
It's harder to line them out and there will be more friction but "accuracy" and the load will inprove.

Do we have any engineering types here? Or are we all machinists and cad users?

Mike...

There are a few, but then the engineering understanding necessary to be a machinist or professional Cad should be enough.

Your idea of using two slides at 90; draw slides support in one plane and allow movement in the other, so in your diagram the left hand slide would locate in the horizontal and allow some movement in the vertical. The right hand slide would locate in the vertical and allow movement in the horizontal. Together they could allow a rotation about the right hand slide. In other words the left hand side of the upper part could move up and down.

An alternative would be to put two slides at either end and another in the centre at 90%. Depending which way you orient them they will should locate the top part in both planes and have greatest capacity in the plane common to the two end slides...

Or put two at 90% at either end.

All three examples could work - it depends on what loads and their directions your machine produces, and your target precision might be.

Draw slides take loads in one plane only. Sit at your desk and half open a draw, lift it up and down and it shouldn't move, move it side to side and it will move a long way, wrack it diagonally and it will move about the same.

A gantry cnc will load its bearings in two dimensions and try to wrack the gantry in the same way as the draw test above. Add the leverage and distance of the cutter from the bearing and you've got problems.


Andrew,

I'm counting on you for a little bit of the engineering help. Is that okay with you?

Mike...

P.S. On the drawer slide design, is this what you were suggesting?

I'm new here and have been following the DIY posts with some interest since I'd like to put together a small CNC router to mill out PC boards at home. I liked the idea of using drawer slides but it seems to me that the table might tend to sag a bit at full extension even with full extension slides.

The simplest solution I can think of is to use 2 sets of full extension drawer slides on either side positioned so that when one is at full extension the other is and minimum extension in order to more evenly support the table. The only thing is that this approach might make it more difficult to mount.

The more I look at drawer slides, the more difficult it gets to come up with something that has any reasonable amount of precision.

I'm starting to think that the approach we should use is just build a cnc mill with 3 slides on the x axis and y axis, and accept the inherent play in the system. Like this pic...

How do we do the z on this system?

Mike...

Maybe add some roller bearings to help take up the play?

If you want something to work pretty well with drawer slides, You need to use 2 pair per axis. 1 pair at each end opening towards each other. The reason is, the farther you pull out a drawer slide, the weaker it becomes. By putting them opposite each other, one pair will be in the closed position when the other is fully open, so you should get fairly consistent operation that way.

A good place to get slides at decent prices is http://www.wwhardware.com
I'd recommend the Accuride 8400 series. Probably the best deal for the money. When you go anything heavier duty the price starts to go up fast.

Hey All,

I have been lurking around this forum for about a year or more, and have realy been interested in the drawer slide idea. I have a furniture building background so I have a lot of first hand knowledge of the guides.

I have seen you guys talking about the different brands out there and the differences among them. It have found that almost all of them are Accuride copies.

I found a Liberty brand guide at my local Home Depot (Dallas Texas). 16 inches, full extention, ball bearings, etc. $16 a pair aint nothin to wright home about, but it beats some of the other prices I have around here with a stick.

I am currently working on modifying one to work more like a pocket door slider, where it can travel past closed. This will keep the load bearing portion inside the base part of the guide, keeping its rigidity.

I am taking pictures of what I am doing and will post them when I have a sec.

All of the work will be done with a dremel, drill press (hand drill will work) and pop rivets. So the average Joe should be able to pull this one off with ease. I will make a tutorial if I get enough requests.

Keep Up The Constant Thinking,
Aaron

The more I look at drawer slides, the more difficult it gets to come up with something that has any reasonable amount of precision.

I'm starting to think that the approach we should use is just build a cnc mill with 3 slides on the x axis and y axis, and accept the inherent play in the system. Like this pic...

How do we do the z on this system?

Mike...

Mike,

this should work, but:
I want the drivenut in the exact center between the rails, if you put it too much on one side it will create twist and friction in the table. If the table is wide enough then the center rail and nut can be as close to the center as possible.
Ger21's solution is better but needs an extra rail.

I have rebuild my second machine using modified drawers (I think the same way Craft Worx does).
This is it:
http://users.skynet.be/sky50985/CNC/Xaxis1.dxf

I only have the 2 slides now and it looks stable, real tests will show if it's good enough.
Assembling like this is simple, just drilling holes and some bolts and screws.
The U shape will also hold the fixed sides for the Y
This is for a small machine mainly for pcb drilling and milling soft materials.
The drill/router will be 50 or 160W max.

The Z can also be done with drawer slides, disadvantage is that you need a rather long rail for stability resulting in a tall system, and again 4 pieces..........
I made my Z like this:
http://users.skynet.be/sky50985/CNC/Z_axis_concept.dxf

There are 2 things wich could be improved or changed for ease of construction.
1: I drilled holes in the guide bars, taped them and used screws to fix them between the top and bottom plate, can't do this precise so I had to use oversized holes for adjustment. I think it will work for me but for a 1HP router?
2: the bearing in the bottom has to absorb all up-downward forces. again I used a light construction and hope it will hold.

Another problem could be the nylon or teflon for the guides and nut.
I was lucky to find some leftovers but don't know if you can easily buy this.

Idea's for improving these?

Will try to post some pictures when it's reassembled, all 3 axis are finished, steppers are mounted and all seems fine.
I used standard M10 threaded rod and can run higher than 2500 mm/min (100IPM) with small steppers from a matrix printer (5V 0.8A).


PS: Do these DXF's create a problem or does anybody knows a good (0$) way to convert DXF to JPG?

You can export a DXF out in Autocad as a .BMP file. Open this in Windows Paint. Save out as a Jpeg

PS: Do these DXF's create a problem or does anybody knows a good (0$) way to convert DXF to JPG?
You can get a thirty day free trial of Paint Shop Pro 9 at the Jasc (http://www.jasc.com) website. It will open dxf files and you can convert to anything you like.

Evodyne

You can export a DXF out in Autocad as a .BMP file. Open this in Windows Paint. Save out as a Jpeg

Forgot to mention: it's autocad lite, under export there's only WMF and DWG.
I tried "Web publishing" but that creates a very poor JPG, used all settings just not good enough.

Paint Shop Pro, available for download as a limited time demo, can import .dwg, .dxf. and .wmf, and export wondrous .jpgs, .pngs, .bmps, and other more specialized raster formats. No guarantee that it won't choke on your particular dxf.

You may find the same capability in the graphics software which came with a scanner or digital camera.

If it's on your screen, just hit print screen go into photoeditor, paint.... and do a paste. Now you can save in .bmp, jpeg, png, gif............ and vwola you have a usable graphics file.

Thanks for all the suggestions on the DXF issue.
I was aware of some but just wondering if it was a problem, will look into it and try to find a good converter.

Back on topic:

What do you think about the Z axis?

Another issue with drawer slides:
Dust, how to prevent that "debris" from milling and drilling get's into them?

I am currently reassembling and aligning the machine and there's already some in the X.

One possibilty would be to use a vacuum cleaner and use a flexible tube mounted on the Y just above the table.

Another is to blow the dust away with a fan, this would require 2 upright angles on the front and rear of the moving table so that it can only get to the side and collect in the U profile (in my case anyway).

Attachment test.

Okay, that worked. Now a little about the picture. This was mid-build in my first router attempt. Panels are all MDF and the angle iron held the table very rigid. I wasn't as happy with the drawer slide setup on the other two axes and eventually removed them for different/better performance. I'm just throwing this picture out there to let you guys bounce more ideas around, though.

--Brian

Brian,

it looks like you used slides wich extend on both sides, I think these do not have enough ball-supported area.
What did you use as replacement?

What did you think about the precision on the x and y axis. Was it sloppy? Was there any more slop when you pulled the table out of the slides, (i.e. in the drawer open position?)

What would you do different?

Mike...

Update on my small machine:

The rework is almost finished and I have done some tests.
I started this to be a PCB drill only and if possible very light milling.
Drilling works perfect, light milling also but the light motors loose steps when going too fast or deep.
I will upgrade if I can find some stronger steppers but am on a very tight budget.
It's made out of 10 and 12mm Formica compact grade (at least that's what I think it is) and iron or aluminium profiles.
Work area is 300 x 160mm (12" x 6"), I know real small.
Y axis was originally with modified drawer slides but these had some play because the plate is only 60mm wide and had not enough ballcontact.
I changed it to 1/2" solid chromed printer bar with homemade nylon bushings.
The X has 2 modified Accuride drawers and I can't see any flexing or twisting.

A few pictures:
http://users.skynet.be/sky50985/CNC/full.jpg
http://users.skynet.be/sky50985/CNC/z_detail.jpg

Lucas, the slides I bought were not originally designed to extend on both sides, but I ground off a retaining tab and then they would extend on both sides. The goal was to keep the rolling ball section as close to the middle of the slide when the axis was in the middle to give it the most rigidity during the entire travel. I had to be careful after my modification, though, because overtravel in either direction would release the balls from their cages (oops!). Them little buggers are hard to find on the floor, too. . .

Mike, the Y-axis (left & right) and Z-axis (up & down) were sloppier than X. As you can see, I sandwiched all axes between two drawer slides. On the X-axis, I used angle iron to hold the slides, and I could squash the slides together toward each other. This made the table fairly rigid. On the other two axes, there was no adjustability. I suppose I could have shimmed them closer to each other, but I just never pursued it that far.

This being the ultra low-tech, low cost, use what you have, prototype, see if I can make anything work project, I used some 1 inch square 80/20 pieces I had sitting around (not enough to make a whole machine out of, though) to make some makeshift slides where one piece of aluminum extrusion would slide against another. The 1/4-20 threaded rod I used had plenty of mechanical advantage to drive this low tech arrangement around, and this is what I kept for quite a bit of my experimenting.

--Brian

Have you guys thought of using runners like these?

Vending machine Runners (http://shop.comdir.co.uk/Products.aspx?intGroupID=999&strProductID=fhgltcfgkw&strWebTemplate=C)

I have a pair i plan to use for my x axis when i finally get arround to building my machine.

having worked with lots of different styles of drawer runners when i worked for a company making bespoke kitchen, bathroom and bed room suites, i am truly impressed with the build quality of the runners, there is hardly any play in the bearings at all, and for a mass produced runner the build quality is excellent, compared to other "heavy duty floor runners" these are in a league of there own.

Also the range of runners offers wide range of travels 225.6 to 632mm of travel.

just a bit of fodd for thought...

nak.

How's your machine working out for you? Did you get those stronger steppers? If you were to rebuild it, what would you do differently?

Mike...

Update on my small machine:

The rework is almost finished and I have done some tests.
I started this to be a PCB drill only and if possible very light milling.
Drilling works perfect, light milling also but the light motors loose steps when going too fast or deep.
I will upgrade if I can find some stronger steppers but am on a very tight budget.
http://users.skynet.be/sky50985/CNC/full.jpg
http://users.skynet.be/sky50985/CNC/z_detail.jpg

Mike,

it's working fine for what it was designed for: a small machine for PCB dilling.

Improvement has not been done for several reason's:
The work area is too small for my milling needs.
It will never be able to do the heavy stuff like aluminium, the entire construction is too light using 10mm and 12mm HDF, 1/2 " guide rod's for the Y....
It would also need a larger router and there's no space.

A while back I decided to leave it as it is and start a bigger, much stronger and all metal machine.
I considered using modified drawer slides again as I am really pleased of the results and the easy construction.
Then I found a cheap bunch of skate bearings and aluminium plate and decided to take on the more difficult but "unlimited in size" build.

Sometimes I regret not having build an intermediate drawer slide machine that could be used to make some parts for the big one.

i was wondering around surpluscenter.com and found these drawer slides that look like a good deal.
http://surpluscenter.com/sort.asp?UID=2005112804082395&catname=misc&keyword=XD11
look at the last slide in the list 25-7/8" closed 52-1/4" open.
hope it helps

I was looking at some slides similar to the ones that "mdreitzusa" has the web site for and I was kinda wondering, the 26" slides have like a 2 step slide when u fully extend them. theres like 3 pieces to them, do you guys think that the transition from the first sliding action to the second part of sliding action will have any effect on the steppers jumping. Its not like there is a "pause" but you can feel the transition when u fully extend them.
Hope u can kinda understand what I am tryin to say...
Chris

do you guys think that the transition from the first sliding action to the second part of sliding action will have any effect on the steppers jumping. Its not like there is a "pause" but you can feel the transition when u fully extend them.
Hope u can kinda understand what I am tryin to say...
Chris

You can probably feel it when moving it with your hand, but when a screw is moving it, it shouldn't be noticable at all.

milford electronics uk sells a kit for a cnc with drawer slides , it has .1 accuracy..

what do you guys think of that design ?

Do you have a link to milford electronics uk ?

here's the link http://www.milinst.com/

I put together a little table about a year ago using scrap aluminum, and 24" full extension slides. I modified the slides so that their are only two pieces (keeps the part with all the little ball bearings in the metal cage), the bottom (widest) part, is the part with all the play. Then I removed and shortened the "carriage" careful to shorten it only enough so that it was supported by balls along full travel. Drill and countersink mounting holes in the bottom piece, and presto, a cool, tight, cheap slide. Using only two for each axis, and shortening them to the lenghts needed, I have about 11"x11"x3" of travel. Not gonna be building any doors with it, but could be used for casting molds, jewelery boxes, stuff like that. I still need to get motors, controller, and mount lead screws, as I had to put the toy away and focus on remodeling our house (6months, 7days a week, 8-10hours a day). Now I am busy building stuff for our new house. I am going to keep it around though, and one of these days I will be able to finish it. I will try to remember to take photos and post them.

Did you mount the rails horizontal or vertical?

It looks like you modified the rails like I did to obtain a carriage wich is always on it's entire lenght supported by the balls, this is not possible with all rails, in some the ball holder part isn't long enough.
Mine are mounted vertical and squeezed between 2 profiles, pushing the carriage into the long part of the rail, this reduces the play even more but slightly increases friction, not much but you can feel it.

I am currently finishing and testing a quickly build temporary "machine" with these rails also on all 3 axis, it flexes a little when pushing hard enough but this is probably due to the inferior materials used, laminated partical board and profiles wich are too thin etc..
This is not yet a problem as the spindle uses 1/8 bits and these flex a whole lot more.
Just need to find some decent spiral bits and see what it can do.

The x axis rails are mounted in a vertical fashion to the side of the aluminum casting that I used for the table. The y is a 90 deg mounting, with one rail on its side, and one on the flat on top of the gantry. The z are both flat. In my design, the y axis carriage is a piece of 1/4" aluminum bent to a 90, projecting past the rail, and the z carriage is the same. The z slides are between the 2 90's and the stepper will mount to the carriage itself, on top. The nut will be stationary on the y carriage. Kind of hard to explain. The z axis slides were modified to be only 9" long, and have a carriage 6 " long. To do this, the ball bearing cage had to be removed and trimmed. Be careful you don't lose your balls! I drew this in paint really fast for you.

any questions or replies? :violin:

One question I had was about it's performance, well no need to ask because it's not finished and tested yet.

You expained it very well, I understand the concept of the Z motor lifting itself together with the z carriage. I assume you did it like this because there was not enough room between the Y and Z carriage for the nut?
From your description it looks like you did a better job than me, you are using aluminium and I used left-over partical board, quite a difference.

When I first looked at your drawing I questioned the Y slide setup, thought it could be better: horizontal slide on the bottom and the vertical on the top.
In your setup the flex from the bottom (vertical slide) would be magnified towards the tool tip, if you would inverse the slides orientiation then the flex from the top vertical slide would not be magnified to the tool and stay approx. the same.
All this might not be true as the strenght of the "L" profile for the Y axis combined with the Z screw in the rear could prevent the Y from flexing.

Did you notice any flex from the drawer slides?
I would expect some from the flat mounted Z slides and the bottom Y one, also flat in the same plane, this would create some flex in the X direction for the router.

In my case the flex only became real visible when the router was mounted with a tool, then I could see the tool move over the table when pushing hard enough on the frame.
But it's no problem, I just milled a skate bearing holder out of UHMW and it's perfect, just set the parameters to match the machine's capabilities, it will take longer but the endresult is what counts, at least for hobby purposes or a temporary machine.

I'm convinced that a well build drawer slide based machine can produce real nice stuff, it's fairly easy and fast to build, definatly cheap and for a guy who has access to decent printer steppers and electronics it will almost cost nothing.

I could post some pictures of this one but it would be useless without pictures of some parts made on it, it's simply far too ugly to look at.
Maybe in a day or two, found a few more decent bits today and need a bit of time to see how it works out.

Yes that is why I made the Z to mount the stepper there. I am very good at seeing small differences in size or movement (or at least I was in the machine and fab shop), and I can't see much flex. A 1/8" bit spinning at 30,000 rpm would find it really fast though! I designed this machine to mount a laminate trimmer, and mostly cut signs, soap and candle molds (gotta keep the wife happy!), and more machine parts. I would like to put together a basic kit, that someone could mount to there own MDF, aluminum, or steel frame, and have all the parts aligned already, so that it could be more fun for most people. I think a lot of people end up getting agrivated with alignement in the machines they try to build, and it may lead to a lot of quiting. I am kind of on the other side of the coin...most people see their motors turn, and then never build the machine, I have built the machines, but never seen my motors turn! Several times I have had motors and drivers, but no machine, and by the time I had the machine close, I had to sell the motors or drivers to pay bills (stupid electricity and stuff), and so now here I am with a machine and no motors. It is one of the only things I kept from all of my neat stuff though, and will run it one of these days.

Here are those photos. The first one is the whole thing. Notice the scrap aluminum base. This is a nice cast piece, that came from my friends dismanteling. At one point I had 16 of these, but had to scrap em, took up too much space. As you can see, I mounted the carriage part of the z to the gantry carriage, and used the rails in the z carriage to hopefully give a little more strength there. Base will be covered with a sacrificial piece of MDF. I have some stainless acme bought in a previous life, but don't know if I will use it. Don't have a metal lathe to turn it down, don't like paying to have people build things for me. Probably just use allthread until I can buy a lathe.

I have been tossing the though of some kind of cover for the ways, and haven't thought of too much. Any ideas for really low cost stuff? Bout my only ideas are trying to find some kind of bellows on something else.z axis is a total stumper, as there is so little space in there, course because there is so little space, it may not be too much of a problem. Better hook up a vac!

Maybe this post could help in covering the ways

http://www.cnczone.com/forums/showthread.php?t=15195&highlight=origami

I hope it helps