PDA

View Full Version : CONVERSION: King Midas mill conversion



diy-john
04-03-2012, 10:18 PM
Subject of conversion in this thread is a mill from 1980's.

5426

There are some underlying issues that I would hope to address during the conversion:
- huge slack on Z-axis (covered here in detail: http://www.mycncuk.com/forums/showthread.php/4231-Z-axis-huge-slack-beginning-a-conversion )
- large slack on both X and Y -axis

Have read as much as possible, trying to learn from other machinists build-logs. As a result, here is a draft shopping list
- items in blue relate to axis/ballscrews.


ITEM
NO.
PART NAME
Length
QTY


1
Nema 23 stepper motor, 3Nm (425 Oz-in)
-
3


2
PM752 Microstepping Driver

3


3
Power supply 400W
-
1


4
Coupling
-
3


5
MBA motor mount
-
3


6
X, Y, Z-Axis Ballscrew Fixed Support Block, FK-type


3


7
X-Axis 1605 / C7 ballscrew 581mm + ballnut
581mm (520mm usable thread)
1


8
Y-Axis 1605 / C7 ballscrew 346mm + ballnut
346 mm (285mm usable thread)
1


9
Z-Axis 1605 / C7 ballscrew 386mm + ballnut
386mm (325mm usable thread)
1


10
Ball nut housing (X,Y,Z-axis)


3


11
X and Y-Axis Ballscrew Floating Support Block, FF-type


2


12
Z-Axis Ballscrew Floating Support Block, BF-type


1




In view of the mill, please can anyone spot here
1) Something clearly wrong ? (something one would regret later?)
- eg. size of motors

2) USB vs. parallel port
- that is something I keep on thinking. Have read about arguments for / against.
- Kind of would like to go USB, but I have read parallel port is the right thing to do.

3) Cables and sundries
- worried about little bits missing from the list. (trying to buy all in one go, to save on shipping fees)


For now I have left out from shopping list "pulleys" and "belts". Hope could have this machine running some day, then later add belts/pulleys. Not sure how this conversion is going to turn out, will it run or not.

Jonathan
05-03-2012, 01:08 AM
Motors and drivers are a good choice. What voltage is that 400W PSU?

I would substitute the couplings (4) for timing belts and pulleys. It makes it much easier to mount the motors as no faffing about getting them on concentric, reduces resonance and allows you to choose the ratio to get either better resolution or speed.

For (5),(6),(11),(12) make your own as you can use the milling machine to make them better than the commercial ones and much much cheaper.

USB / Parallel. If you use USB I doubt you'll regret it, but why not try the parallel port first as it costs virtually nothing to try?

For the stepper motor cables most people use 4-Core CY, 1.5mm^2 cable. Also add E-stop switch, limit switches and connectors for stepper motors.

diy-john
05-03-2012, 10:36 PM
Motors and drivers are a good choice. What voltage is that 400W PSU?


- The one I was looking at, is this. Is it any good?

Wattage 400 Watts
Input Voltage 240V AC
Output Voltage 36V DC adjustable +- 2.5V approximately
Output Current 11 Amps



I would substitute the couplings (4) for timing belts and pulleys. It makes it much easier to mount the motors as no faffing about getting them on concentric, reduces resonance and allows you to choose the ratio to get either better resolution or speed.


Think I just started to like belts and pulleys.

Not sure if one wants more speed, at 5mm pitch. Thought of going lower pitch, even 2mm, but most of screws available seem to be 5mm.
- Perhaps my preference is machine / cutter longevity and accuracy.



For (5),(6),(11),(12) make your own as you can use the milling machine to make them better than the commercial ones and much much cheaper.


In my calculations,perhaps half budget is for mounting hardware. On Ebay there seems to be great bargains, for example I like these mounts:
http://www.ebay.co.uk/itm/Aluminium-Motor-Mount-Nema-23-Stepper-Heavy-Duty-Industrial-but-very-light-/270927225819?pt=LH_DefaultDomain_3&hash=item3f148563db

Good an idea to make home-made mounts etc., though worry a bit about increased complexity.




why not try the parallel port first as it costs virtually nothing to try?


Good point, parallel it will be.



For the stepper motor cables most people use 4-Core CY, 1.5mm^2 cable. Also add E-stop switch, limit switches and connectors for stepper motors


Going now back to drawing board, I will soon come up with a more detailed shopping list. Thanks Jonathan for comments, they are very helpful.

Robin Hewitt
06-03-2012, 11:26 AM
The proposed set up looks good for a backlash of around .015 -> .025mm if you get it right. Plenty good enough to remake the machine after the screws bed in and .025 mm start to hack you off, which it probably will :smile:

A stepper does 400 half steps per rev. A 5:4 belt reduction on to a 5mm screw would give you 0.01 mm per half step, add one microstep for .005 mm and it becomes credible. You could reduce it 5:2 and avoid the springy microstep but that could seriously affect your rapid at 36 volts.

Jonathan
06-03-2012, 03:29 PM
- The one I was looking at, is this. Is it any good?

Wattage 400 Watts
Input Voltage 240V AC
Output Voltage 36V DC adjustable +- 2.5V approximately
Output Current 11 Amps

No good, why spend more on 70V drivers then use them on only 36V? The torque from a stepper motor is inversely proportional to speed and proportional to voltage, so if you get a 70V PSU you will get almost twice as much torque from the same motors. It's risking to operate right on the limit, so look for something just under 70V. One of the cheaper ways to do it is to use a toroidal transformer.


- Perhaps my preference is machine / cutter longevity and accuracy.

Cutter longevity is a vague one... if you feed a cutter too slow it can break just as quickly as from cutting too fast. You need to estimate the feedrates required for the materials you wish to cut taking into account the range of cutter diameters. Then make sure it will go a bit faster to be sure. However clearly you don't want to optimise the system for a high feedrate, and thus low resolution, if the vast majority of the time you'll be using much lower feedrates. You could select different size pulleys for each axis and swap them round / experiment to find the best compromise.


In my calculations,perhaps half budget is for mounting hardware. On Ebay there seems to be great bargains, for example I like these mounts:
http://www.ebay.co.uk/itm/Aluminium-Motor-Mount-Nema-23-Stepper-Heavy-Duty-Industrial-but-very-light-/270927225819?pt=LH_DefaultDomain_3&hash=item3f148563db

Good an idea to make home-made mounts etc., though worry a bit about increased complexity.


It should be decreased complexity if you make your own mounts for pulleys. This is what I started out with on my milling machine as it was easy to make without CNC and does the job. You can see them, briefly, in some of my very old videos:

http://www.youtube.com/watch?v=tasl9ua4R2s&feature=channel

diy-john
06-03-2012, 09:55 PM
The proposed set up looks good for a backlash of around .015 -> .025mm if you get it right. Plenty good enough to remake the machine after the screws bed in and .025 mm start to hack you off, which it probably will :smile:
...
You could reduce it 5:2


Robin, with this rate, the mill will be cutting a couple of atoms per slice - maybe. :smile:

So far, added to shopping list some 48 teeth (around 75mm diameter) pulleys, and smaller 24 teeth (around 35mm diameter). Could make it into 5:2, guess it will be a bit more precise then.




No good, why spend more on 70V drivers then use them on only 36V?
(http://www.youtube.com/watch?v=tasl9ua4R2s&feature=channel)

Good Jonathan you say that, didn't even come to think about that. Updated into shopping list. So far only found some 50V PSU's, but will keep on looking, to add to the list.






Shopping list colour codes:






RED=NOT NEEDED, going DIY






ORANGE=Maybe not needed (DIY?)






GRAY=check quantity






PALE BLUE=Ballscrews+ballnuts as units






YELLOW=New since last list











ITEM
NO.
PART NAME
QTY


1
Nema 23 stepper motor, 3Nm (425 Oz-in)
3


2
PM752 Microstepping Driver
3


3
Power supply 50V – 68V, at least 400W
1


4
Coupling
3


5
MBA motor mount
3


5
Aluminium motor mount
3


6
X, Y, Z-Axis Ballscrew Fixed Support Block, FK-type
3


7
X-Axis 1605 / C7 ballscrew 581mm + ballnut
1


8
Y-Axis 1605 / C7 ballscrew 346mm + ballnut
1


9
Z-Axis 1605 / C7 ballscrew 386mm + ballnut
1


10
Ball nut housing (X,Y,Z-axis)
3


11
X and Y-Axis Ballscrew Floating Support Block, FF-type
2


12
Z-Axis Ballscrew Floating Support Block, BF-type
1


13
Pulley for 12mm width belts. AT5 -type. 24 teeth. Total width circa 22mm. Total diameter circa 37mm. For X, Y, Z -axis
3


14
Pulley for 12mm width belts. AT5 -type. 48 teeth. Total diameter circa 75mm, for X ,Y Z -axis
3


15
Belt 12 mm 12AT5 length circa 450mm, for X, Y-axis
2


16
Belt 12 mm 12AT5 length circa 330mm, for Z-axis
1


17
E-stop switch
1


18
Limit switches
6


19
Connectors for stepper motors
3


20
4-Core CY, 1.5mm^2 cable








http://www.youtube.com/watch?v=tasl9ua4R2s&feature=channel

Nice video Jonathan. Makes it easier to understand how the pulleys are setup.
- 6, 11, 12 ... not quite sure how involved a procedure it is to make them DIY? (and how big a saving can be made, and can one create those without CNC? )

Think the shopping list starts to look quite good (ready to go?). Really appreciate, your help has been great. Thank you!

diy-john
08-03-2012, 01:32 PM
Here is a new draft of the ballscrews. Lengths change a bit, from going to pulleys/belts instead of directly driving the screws
.5442

Robin Hewitt
08-03-2012, 01:59 PM
A ball screw is best held in tension by Belleville washers to compensate for thermal expansion, one thrust race at either end.

If you lengthen the XY shafts at one end you can keep the handles.

Run the table to it's ful travel and ask yourself where the CNC parts are going to fit in the available space.

Experience tells me that being new to this you are very unlikely to believe this is necessary. You feel a compelling desire to see it cutting, don't want to go back to the drawing board now and spend extra money on machining screws. I only mention it so I can say, "Told you so" later on :naughty:

diy-john
08-03-2012, 10:18 PM
Yes Robin, back to drawing board it is. Really appreciate your advice, glad I didn't yet buy wrong length ballscrews. :smile:

Went to garage, moved X/Y -axis back and forth full travel. Need to lengthen the ballscrews, pulleys would now crash to the table.

Not only that, but realized that it might be possible to extend X/Y -axis travel a good couple of inches, with minor changes.
- Had not really thought, where one would fit the ballnut, and where screw would be mounted. (attach ballnut to upper part (the one sliding), or lower part vs. the other way around, that kind of things)
- Below is a sketch of the Y-axis movement (current vs. possibly future). Changes would be: 1) a longer ballscrew; 2) attach ballscrew to the table that does not move; 3) attach ballnut to the table that moves, at the centre.

5443



Belleville washers to compensate for thermal expansion, one thrust race at either end.

This is all new to me. Thanks will look for pics from google. Another thing I don't quite understand yet, is the construction of the "fixed mount" for the ballscrew. (still thinking whether to buy ready made, or DIY).

Robin Hewitt
08-03-2012, 11:26 PM
Yes Robin, back to drawing board it is.

Wow! If you are okay to take advice you can make one hell of a machine. I got it right on my third set of screws when the penny finally dropped.

Problem is, you need a CNC mill to make a pretty CNC mill which is why I was suggesting you cobble something together first. Hope you have a lathe, really do need a lathe.

I used pairs of Belleville washers to apply a preload of 500 lbf. One pair for each screw to hold the axial bearings in tension, one pair for each double ball nut.

That means there is no play in the system, apart from the column bending, so long as the load does not exceed 500 lbf. It was a lot of work but is quite delightful to use and all self adjusting.

The XY screws appeared much too long until it was all together, have a pic...

diy-john
11-03-2012, 11:25 AM
The XY screws appeared much too long until it was all together, have a pic...

Robin, that is one sweet looking machine. One can tell, a lot of planning has gone into it. Like the kind of contrast with the big, sturdy, used Warco mill, and then the precise CNC -parts. Very cool.

King Midas conversion last couple of days has been measuring + and thinking how/where conversion related components will be. From all your comments, I have realized how distant my thoughts have been from this project. Feel one has been watching this project sort of from cloud -level, but now there is need for much closer look. Exciting.

1) Y-axis ballscrew -related
1.1) Ballnut would be attached to "lower", "stationary" bed of the mill. (view from front of the mill)

5448

1.2) Y-axis ballnut would be attached to "lower", "stationary" bed of the mill. (side -view, from right hand -side of the mill)
- white box, height is 55mm, width is 18mm (for reference only)
5449

1.3) Overview / sketch of Y-axis ballscrew (side view, from LEFT -hand side of the mill)
- colours have no specific meaning, just different colours for different components
- length of ballscrew is now at 508mm. (note how it has grown, initially one was going to buy a 346mm screw...LOL)

5450

1.4) A challenge in Y-axis ballscrew is a 35mm vertical space here
- thought initially there is 40mm space, since the "cave" bottom looked like level with "outside" (see picture attached)
- need to make DIY "slim" ballscrew floating end-mount. Haven't found a slim-enough ready made mount.
- SKF 7200 BEP bearing outer diameter, for a 16mm diameter screw, is 30mm
- Would mount screw to "moving upper deck".

5451


I haven't got a lathe yet. Tool/workshop -wise I am in the wrong league. Looking to do premier league, but my skills/tools are only 5th division. But thanks to this forum and careful planning, maybe this game is possible..

Robin Hewitt
11-03-2012, 12:35 PM
You will need to get the ends of the 16mm screws turned down to 12mm and partially threaded 1mm pitch.

Watch out, 12mm single row, angular contact bearings don't have dust shields, you will want to bury them in grease.

You can't use double row bearings with shields because they aren't crushable.

This is how I built up my Y screw assembly...

I put on one bearing
the aluminium mount that bolts to the machine
then another bearing
then a pair of Belleville washers
then the pulley
then a stepped iron sleeve to carry the indicator dial and the handle
then a 12mm half nut

I nipped in flat ended grub screws through the iron sleeve and pulley, flats cut on the screw shaft.

I wound the 12mm nut in one turn to preload the Bellevilles and crush the bearings.

Then I added another 12mm half nut to lock the first one.

Then I tightened the grub screws.

There was no need for a bearing at the far end of the Y screw, it simply isn't long enough to whip.

diy-john
11-03-2012, 07:39 PM
Robin, tried to understand how things would go on the ball screw. Now, this is getting to the far edge of my knowledge, this is all new.

Made a sketch of how I have understood:


5459

So, starting from the right hand side, in the picture, requirements for machining screw:

- Machine 12mm length of screw to 10mm diameter (if want a bearing at the far end of the screw). Moving further from right to left, add



I put on one bearing <--- 1 bearing, shaft=12mm, OD=32mm, W=15.9mm, like this http://3.bp.blogspot.com/_3D8MX6Q_YGw/S60zwpUN31I/AAAAAAAAANE/gNUVeaBYLtU/s320/IMG_0326+sml.jpg so machine ballscrew width=15.9mm diameter=12mm
the aluminium mount that bolts to the machine <--- For this, machine 3mm length of ballscrew diameter=12mm. Bearings will go mostly inside this mount(?)
then another bearing <--- machine ballscrew width=15.9mm diameter=12mm
then a pair of Belleville washers <--- like this http://www.trakar.com/spring-training/disc-springs/disc.jpg machine 2mm length of ball screw to diameter = 12mm
then the pulley <--- machine total width of pulley (eg. 25mm) diameter = 12mm + flat spot
then a stepped iron sleeve to carry the indicator dial and the handle <--- machine total width of pulley (eg. 25mm) diameter = 12mm + flat spot
then a 12mm half nut <--- machine a thread M12 x 1mm to the end of the screw, perhaps for a length of 10mm


OR: machine the screw like this:
http://www.slidesandballscrews.com/images/machining ball screw-1100mm.jpg (http://www.slidesandballscrews.com/images/machining%20ball%20screw-1100mm.jpg)
- Except no.1 : that there would be a longer piece of screw machined to 12mm diameter (perhaps 35.8 mm length instead of 23mm in picture, to allow 2 x bearings (30.8 mm) + bearing housing (3mm) + belleville washers (2mm) = 35.8mm
- Except no.2 : Make a 12mm "flat spot of 50mm length" (to take pulley and handwheel)


For now, this project focus is trying to understand requirements: how / to what specification machine ballscrews.

Please comment!

Thanks!

Robin Hewitt
12-03-2012, 10:58 AM
Can you read an AutoCAD .dwg or .dxf? My drawing is a bit chaotic but it's al there if you look hard enough. I draw to get the dimensions then draw individual pieces for cutting.

I have lots of 12mm Belleville washers left, the minimum order was generous. I can send you half a dozen for the three screws if you PM an address, may take a while to find them though. I may have some 16mm I used for the ballnuts, spec is at www.bellevillesprings.com (http://www.bellevillesprings.com)

diy-john
12-03-2012, 09:42 PM
Robin, your offer is very generous, thank you very much. Will send you a PM.

Did not realize how critical these ballscrews are to the build, and how much thinking has to go to them.

The X-axis, which I have so far thought as being "the piece of cake" in this project, and hence did not really think about it much.

Today thought better still check available space for X-axis ball screw.

- X -axis ball screw would have to go through this hole:

5469

- The hole height is equal to 28.66mm, which sounds great when you think it is only a 16mm ballscrew to put in there.
5470 5471

- However it comes so close, because the ball nut height = 40mm, and the nut must clear the moving "top deck" (which slides over the ball-nut).
5473

Think there will be only 0.66mm of air underneath the screw, if one will go for 16mm screw. (and if the screw really is 16.00 mm). Challenge is, that one can not move the ballscrew higher, because then the ball nut would not clear the sliding "top deck".

A bit frustrating perhaps to just measure and measure, when one wants metal chips and sparks to fly. But guess you really don't want buy ballscrews that are only good to post to classifieds section... so slowly it goes. :)

Robin Hewitt
13-03-2012, 09:14 PM
Did not realize how critical these ballscrews are to the build, and how much thinking has to go to them

If you get the ballscrews right you never notice them. Get them wrong and you will be for ever adjusting and compensating.

I can export that drawing as a PDF if you think it might help, I have never exported anything as a PDF so I don't know.

C_Bubba
13-03-2012, 09:45 PM
Robin,
Try http://sourceforge.net/projects/pdfcreator/ which will give you a "printer" that will allow you to print from cad or any application and give a pdf file.

Gary
13-03-2012, 10:20 PM
In most conversions of machines of this size that i have been involved with the ballnut between the X and Y needed to be the FSI of DFI type that has two fixing holes either side of the flange and 2mm lower profile than the standard FSU type.
This will get over you space problem.

diy-john
14-03-2012, 10:28 PM
FSI of DFI type that has two fixing holes either side of the flange and 2mm lower profile than the standard FSU type.


Thanks Gary - Please do you know the "height" of the lower-profile ballnut, for a 16mm diameter screw (1605)?
- Have used 40mm in my calculations, thinking might have by coincidence already used a lower profile version(?)
- If one could get 2mm lower than 40mm, then it would be great yes.

With a bit of help from Robin, latest draft -version for X -screw is here:
5475

Total length of screw now is 871mm. (it has grown an extra 290mm from the original plan).

Hope that "custom" screw would not break the bank. (don't know how much extra cost that machining will add on top).
- Have prepared to go "conventional" cuttings, too, if it becomes really dear.
- One facet: want to get the screws as good as possible
- Another facet: being a newbie at this, would be very, very surprised if one gets it right first time round. Not sure how much point there is in paying top money for ballscrews, only to want something else next month. Of course trying very hard to get it right first time, but have to be realistic.

Slowly forward..

Jonathan
14-03-2012, 10:47 PM
Lots of drawings on Gary's site:

http://www.zappautomation.co.uk/ball-screws-c-1.html?osCsid=d100eafe88953dd6b02cf54c2f87cf0d

I can machine the ballscrew as you've drawn it for £15 plus postage.

Robin Hewitt
14-03-2012, 11:09 PM
latest draft -version for X -screw is here

The X screw is best held in tension, suggest you put a thread and one bearing at either end.

I have found 31pcs of the 12mm Belleville washers I used...
www.bellevillesprings.com (http://www.bellevillesprings.com) pn: D2812215

You are welcome to half a dozen, just shout.

Jonathan
14-03-2012, 11:26 PM
Was going to mention adding the second thread for tensioning but for some reason I thought that drawing was for the Y-screw...oops. I've got a thread on both ends of my router X-axis screws for the same reason.

Do you reckon the D2812215 Belleville washers would be suitable for preloading RM1204 ballnuts? The screw I have here is actually 11.60mm diameter so it's going to close... Apparently 10-15% of static thrust rating is a good compromise for preload, so 1KN. I probably need a lower spring constant for that as such a small deflection will be hard to measure.

Robin Hewitt
14-03-2012, 11:43 PM
Do you reckon the D2812215 Belleville washers would be suitable for preloading RM1204 ballnuts?


I used them with 1204's on my Roland mill at considerably lower pre-load than on the Warco. I don't go above 5mm tooling on that.

Works okay, but agree a lighter section washer would be easier to adjust and give you more take up.

Gary
15-03-2012, 08:16 AM
Looking at the drawing is seems over complicated.
If you want zero backlash and don't want any movement on the support bearing just go for a good quality FK12, the gten R1605 ballscrew and the R1605T3-FDID-P1 (http://www.zappautomation.co.uk/r1605t3fdidp1-ballnut-p-574.html?cPath=1_59_66)

This is the link to the datasheet for the FDI and FDU type ballnuts and it is 34mm between the flats, so it is actually 6mm difference.
http://www.zappautomation.co.uk%2Fpdf%2FBallscrews%2FFDU-FDI.pdf

The ballnut is already pre loaded for zero backlash and has been tested on the screw.

you will need ti install chinese fonts in acrobat.

Robin Hewitt
15-03-2012, 10:34 AM
The ballnut is already pre loaded for zero backlash and has been tested on the screw.


That looks interesting, is it shimmed or does it have springs? The data sheet is vague.

Gary
15-03-2012, 10:52 AM
No its a spacer between the two nuts.
I dont know of any commercial double ballnuts that use a spring.
All the manufactures i have worked with use a spacer to pre load the double ballnut, and there are reasons for this.








That looks interesting, is it shimmed or does it have springs? The data sheet is vague.

Robin Hewitt
15-03-2012, 11:18 AM
All the manufactures i have worked with use a spacer to pre load the double ballnut, and there are reasons for this.

Good point, I forgot to mention that John will need to leave room to fit an oil pipe to the nut. I use 4mm OD nylon tube and standard push fittings which are a bit sticky out and it will drip :beer:

diy-john
15-03-2012, 10:50 PM
I can machine the ballscrew as you've drawn it for £15 plus postage.

Wow, cool Jonathan. Can't go wrong with that. Thanks!



www.bellevillesprings.com (http://www.bellevillesprings.com/) pn: D2812215

You are welcome to half a dozen, just shout.

Robin, yes please, need those.



suggest you put a thread and one bearing at either end.


Made another draft of X-axis screw. Please is this any good now?
- added a thread to the other end. And also a similar "structure, a fixed mount" for the ballscrew. So there are now two such mounts, one at either end of the screw.
- there are two (2) screws in picture below, but it is the same one single screw.
-- The other one just highlights, what components would fit, and where. Like hand-wheel, pulley, shims, Belleville washers.

5483

Not sure about the dynamics, kind of figured out, that with this setup, the Belleville -washers + half nuts at either end of the screw ---> would tension the screw sort of around either end fixed mounts. Please correct if you think one has mis-understood the principle.




This is the link to the datasheet for the FDI and FDU type ballnuts and it is 34mm between the flats, so it is actually 6mm difference.
http://www.zappautomation.co.uk%2Fpdf%2FBallscrews%2FFDU-FDI.pdf (http://www.zappautomation.co.uk%2fpdf%2fballscrews%2ffdu-fdi.pdf/)



Gary, the 34mm height ballnut gets me off the hook. Thanks for pointing out these alternatives. Will speak to you about other components.


Thanks!

Jonathan
15-03-2012, 11:08 PM
Robin - thank's for confirming that about the washers. I'll order some 'lighter' ones (http://www.leespring.com/uk_product_spec.asp?partnum=5000351000&springType=W&subType=)


Wow, cool Jonathan. Can't go wrong with that. Thanks!

:) Discuss details when the design's ready.

To put the screw in tension you don't want to have pairs of angular contact bearings at both ends, as that fixes both. You need a pair of angular contact bearings at one end, and a standard deep groove bearing plus Belleville washers and thrust bearing at the other end to provide contact force for the tension. If both ends were fixed rigidly with angular contact bearings the tension in the screw would depend on the relative spacing of the bearing blocks and the length of the screw between them, so there's no way to set it.

By 'standard washer' I assume you mean an accurately machined washer with parallel faces. You don't need a washer between the ballscrew end and angular contact bearing as the bearing is most accurate resting on the machined shoulder on the screw.

Robin Hewitt
16-03-2012, 01:41 AM
That looks horrible. I'm obviously not explaining this at all well.

To put the screw in tension you just need one angular contact bearing at either end. The Bellevilles try to stretch the entire screw end to end. Adding more bearings just gives you an alignment nightmare. There is hardly any side loading on the screw to worry about, so don't worry. I used one bearing at either end of my overlength X screw and it was fine and dandy.

I used a pair of bearings at the handle end of my Y screw simply because I couldn't fit any bearings at all on the far end.

Not a problem for the X so I put one bearing at either end.

The whole she-bang should be sloppy, right up to the moment you tighten down on the Belleville's.

diy-john
16-03-2012, 10:52 AM
You are all doing a good job explaining the principles.
- It was me who assumed X and Y -axis ballscrews are put together in similar way.
- Assumed too, that idea is to make it as sturdy/rigid as possible.
- There are some gaps in my knowledge, filling the gaps with assumptions doesn't always work

But it will be good. Will create another drawing of the ballscrews -setup.

Thanks again for comments!

Robin Hewitt
16-03-2012, 12:13 PM
... and only one pair of Belleville washers, you don't want to balance the screw between two springs because that can move.

A few design tips.

I bolted the X screw pulley solid to the inner race. I sprung the sleeve that takes the handle at the other end. This reduced the bearing overhang I couldn't avoid on the Y.

Putting the handle on a sleeve means I can pop the handle off without detensioning the screw. Handy if I want to work on something that overhangs the end of the bed.

I mounted the X motor facing the unlikely way round and extended the screw to get clearance on it. This makes it a lot easier to mount, easier to change the belt and easy to pop a cover over to keep the crut and fingers out.

My X motor mount isn't actually bolted to the machine, it's free to turn. I was going to Loctite it in place but found the 500 lbf screw tension quite sufficient to hold it. I lean on it occasionally, if it ever moves I will know I have lost tension.

diy-john
18-03-2012, 03:26 PM
To put the screw in tension you just need one angular contact bearing at either end. The Bellevilles try to stretch the entire screw end to end.
I used one bearing at either end of my overlength X screw and it was fine and dandy.



Thanks again Robin

Please would this setup make the X-ballscrew dandy? (is this what you mean?)

5500


Sorry if this obvious to some of you. I tried to figure it out, that
- if you now tighten the screws at each end of the X-axis screw ---> the nut(s) will push the Belleville washer ---> the Belleville washer will push the bearing ---> the bearing will push the fixed bearing mount. But because the bearing mount is fixed, then tightening the nut(s) will pull (try to stretch) the screw from either end.

Please correct. (is this getting better or worse? Tell me)

Thanks!

Robin Hewitt
18-03-2012, 05:19 PM
Try this pic for the X screw in tension... :smile:

Right hand nut down tight.

Left hand nut tensions the screw.

Personally I'd put the pulley at the other end to reduce the overhang.

I should have drawn the handle fitting bigger than the nut. That way you can put the handle on and off without removing the nut.

Edit: I should also have put the Belleville washers on the other side of the pulley so the pulley grub screw doesn't have to slide. I can redraw it if you want.

diy-john
18-03-2012, 07:19 PM
Cool Robin :) - Thanks!

That picture is worth a thousand words. Updated X-axis draft accordingly.
- added Belleville -washers on the other side of the pulley, like you say
- the sleeve, on top of which the handwheel comes is a nice feature.

5503



Personally I'd put the pulley at the other end to reduce the overhang.


Now, have thought about this overhang. Might as well say now, what I have had in mind, but before reading further, make sure you sit tight and don't spill any coffee.. :lol:

The challenge is, that the X-axis movement is a mere 220mm, though the table is 500mm wide.

Thought of moving the
- handwheel
- pulley
- motor

way out of the end of the table, on the right hand side.

5504

5505


This "extension" would change X-axis movement from 220mm to around 400mm. Think there might be challenges ahead, with accuracy/slack, at the far end of the X-table movements. However, I have reasoned, that perhaps one could bring the slack down with careful adjustment. Just have been thinking, that there may be situations, I would rather live with a small slack at far end, than reposition the object on the table (because of lack of X-axis movement).

Of course people will wonder, why one didn't buy a larger mill in the first place. But what do you think, is this "extension" something one definitely should not do? Please comment :)
- extension would not be for attaching, or supporting pieces on table, but rather bring out the X-axis motor, pulley, bearing block.

Jonathan
18-03-2012, 09:09 PM
On my mill the original handle / bearing arrangement already clears the casting that the bed runs on. The extra travel that 'gains' has proven invaluable on numerous occasions. As you say it's bound to not be as rigid as using the standard travel, but I've never noticed it.

You could use some decent size aluminium bar for the extension and use it to house the bearings. Convenient as you can do it all on a lathe.

Is it convenient to add another (e.g. 6201) bearing at the driven end of the screw? I'd prefer to have the screw supported with two bearings at one end as otherwise I suspect the significant radial force due to the tension of the timing belt will tend to bend the screw about the single angular contact bearing. Might be negligible in reality as clearly it works for Robin without ... ?

Robin Hewitt
18-03-2012, 11:56 PM
Good, only one small mistake. You only leave a gap on the left hand side, the tension end. On the right you want the 12mm nut to grip the angular bearing inner ring against the step in the screw. That gives you something to tighten the nut against so it stays where you put it. Don't want that right end nut free to unscrew itself.

The gap on the left only needs to be long enough to stop the 16mm part of the screw ever touching the bearing. If it does touch then the Belleville washers stop pulling on the screw.

My handle fitted inside the saddle, if it didn't I would have extended it. My X motor didn't fit so I extended that end a long way.

The extension can be as long as you need, it doesn't have to be wide.
A wide extension could get in the way of the motor. Allow room for the motor, you don't want the motor hitting the saddle or getting in the way when you want to turn the handle.
The extension has to take a big axial load from the Bellevilles.
The radial load is the motor torque.
The side loading is ordinary workshop bumps and knocks.
It has to hold the bearing square, otherwise the screw will flex when you turn it.
It is comforting if the extension carries any motor heat in to the bed, so the screw doesn't warm up.

Swarfing
19-03-2012, 12:05 AM
A good tip here is to also if you are going to use the machine manually is to be able to disconnect the motors from any circuit. Otherwise they will act as little generators sending a nice bit of current back to your circuit boards :whistling:

Robin Hewitt
19-03-2012, 12:23 AM
they will act as little generators sending a nice bit of current back to your circuit boards :whistling:

Have you actually had a problem with that?

If you short out a stepper coil you get a definite resistance when you try to turn it.

I don't feel any resistance or cogging when I turn a motor with the driver OFF or FREE so I don't think there is any electrical path for it to push against.

OTOH I could be delusional, it happens at my age.

Swarfing
19-03-2012, 12:30 AM
Ezecnc Roger had this issue when he converted his viceroy, he did not suffer any damage luckily but did have a huge issue with resistance. Made it hard work to mill by hand. He was using large steppers but goes to show that the issue is there no matter what size steppers you use. It would be worth whipping out the DTM to check on any back current flow?

Jonathan
19-03-2012, 03:03 AM
The back-emf from stepper motors is significant - they're good little generators. The protection diodes in the stepper drivers will rectify the back-emf and if you turn it fast enough power up the driver which is clearly bad news. I found that below a certain speed they manually turn quite easily, but when that threshold is exceeded it gets a lot harder. Clearly it will vary significantly with the size of motor.

Easily solved - just unplug the motors.

diy-john
19-03-2012, 09:54 PM
You only leave a gap on the left hand side, the tension end. On the right you want the 12mm nut to grip the angular bearing inner ring against the step in the screw. That gives you something to tighten the nut against so it stays where you put it. Don't want that right end nut free to unscrew itself.

The gap on the left only needs to be long enough to stop the 16mm part of the screw ever touching the bearing. If it does touch then the Belleville washers stop pulling on the screw.



A tiny update to X-axis screw.
- Red texts are updates.

5516



It is comforting if the extension carries any motor heat in to the bed, so the screw doesn't warm up.


Will take a note of that. Again really good suggestions. Thanks!

Project has been moving at a snail's pace, but that is okay, have already avoided many pitfalls.

Jonathan
20-03-2012, 12:08 AM
What's the dimension I've marked with '???':

5518

Pretty minor, but I changed the end bearing surface from 7mm to 9mm since the bearing is 10mm wide and it's best not to have the bearing resting on too much thread.

Robin Hewitt
20-03-2012, 11:41 AM
What's the dimension I've marked with '???':

I'm confusing my milling machines, on this one I used full nuts rather than pairs of half nuts. I didn't want to stress a half nut by asking it for 500 lbf of tension.

10mm for the bearing, 10mm for the nut, say 1mm protruding, 21mm total.

Doesn't really matter how far the thread goes inside the bearing, the big axial load from the Belleville washers overwhelms all the other loadings.

I made all the bearings an easy sliding fit, inside and out. This made it very easy to assemble but it went rock solid the moment I tensioned it.

How do you plan to mount the motors? I suggest you cobble them on somehow with plates and studding then replace everything when you have CNC. Motor mounting is important now only because it determines the length of the screw.

Things to remember when designing motor mounting plates...

Putting the motor on so it faces away from the machine is best.
Putting the motor on facing the machine saves a couple of quid on the ballscrew length but makes everything else difficult.
Leave a tadge of motor movement to adjust the belt tension.
Cut the outline for a belt cover and it's mounting bolts, assume you have CNC for making the belt cover.
Somewhere to attach a flexy conduit over the motor wires saves bodging it later on.

Jonathan
20-03-2012, 03:03 PM
10mm for the bearing, 10mm for the nut, say 1mm protruding, 21mm total.

My point was not knowing those dimensions means the length of ballscrew directly between the bearings is unknown, which is a critical dimension.


Doesn't really matter how far the thread goes inside the bearing, the big axial load from the Belleville washers overwhelms all the other loadings.

I guess so long as the mounting surfaces are accurate, but there's no reason to have the thread far in the bearing surface.

diy-john
20-03-2012, 09:20 PM
5518

changed the end bearing surface from 7mm to 9mm

That is a sweet picture Jonathan. Good point that 9mm.



10mm for the bearing, 10mm for the nut, say 1mm protruding, 21mm total.


Final versions of X, Y, Z -axis screw are like this (changed according to Robin's / Jonathan's comments)
- X-axis:::
5527

- Y-axis:::
5528

- Z-axis:::
5529



Revised shopping list is here. A couple of notes about that
1) Power supply perhaps is slightly undersize. Thought to trial and see.
2) Need two (2) of the Nema 23 -motors WITHOUT the rear shaft
- Need one (1) --------------//----------------- WITH the rear shaft (Z-axis, will attach hand wheel there)
3) Items 12, 14 -- 19 not sure where to source from, have not found yet
4) Bearing housings / fittings ---> DIY ?



ITEM
NO.
Item




QTY


1
Nema 23 stepper motor, 3Nm (425 Oz-in)
Zappautomation
http://www.zappautomation.co.uk/sy60sth883008-nema-stepper-motor-p-24.html?cPath=9_159_42
2



Nema 23 stepper motor, 3Nm (425 Oz-in)
Zappautomation

1


2
PM752 Microstepping Driver
Zappautomation
http://www.zappautomation.co.uk/pm752-microstepping-driver-p-409.html?cPath=9_3_4
3


3
Power supply / PS705 - 75V DC at 0A & 68V DC at 5A
Zappautomation
http://www.zappautomation.co.uk/sps705-p-587.html
1


4
X-axis R 1605 ballscrew 853 mm (not machined)
Zappautomation
http://www.zappautomation.co.uk/r1605-ballscrew-p-93.html
1


5
Y-axis R- 1605 ballscrew 508 mm (not machined)
Zappautomation
http://www.zappautomation.co.uk/r1605-ballscrew-p-93.html
1


6
Z-axis R 1605 ballscrew 259 mm (not machined)
Zappautomation
http://www.zappautomation.co.uk/r1605-ballscrew-p-93.html
1


7
Ball nut R1605T3-FDID
Zappautomation
http://www.zappautomation.co.uk/product_info.php?products_id=574
3


8
Ball screws machining
Jonathan


3


9
Pulleys, larger around 70mm diameter, 21T5/44-0
Zappautomation
http://www.zappautomation.co.uk/synchroflex-timing-pulleys-10mm-p-190.html?cPath=14_72_153
3


10
Pulleys, smaller around 35mm diameter, 21T5/22-2
Zappautomation
http://www.zappautomation.co.uk/synchroflex-timing-pulleys-10mm-p-190.html?cPath=14_72_153
3


11
Belts, 3 for the pulleys (10mm)
Zappautomation
Ask Gary
3


12
Angular Ball Bearings 7201B -type
Local bearing shop
http://www.vxb.com/page/bearings/PROD/Kit1085
6


13
Belleville washer
Robin


6


14
E-stop switch
TBD (to be decided)






15
Limit switches
TBD






16
Connectors for stepper motors
TBD






17
4-Core CY, 1.5mm^2 cable
TBD






18
Sleeves for handwheels
TBD


2


19
Handwheels
TBD


3




Think this should be pretty good now. (until someone points out inconsistencies :smile:)

Jonathan
20-03-2012, 09:50 PM
£300 on ballnuts! Are you sure you can't get the cheap RM1605 ballnuts from china to fit as they're $25 each from eBay. Surely you can at least use the cheaper ones for Z?

Pulleys - the tooth profile of HTD pulleys is much better suited to this application than T5 and similar price, so get pulleys and belts from here:

http://www.bearingstation.co.uk/products/Pulleys/HTD_Pulleys/HTD_Pulley_5mm

Most (if not all?) the 3Nm motors have dual shafts, so unless the rear shaft is going to get in the way you can just get 3 of the motors you linked to at Zapp.

I've got lots of 4-core CY, 1.5mm^2 cable for the motors...yours for £1.20 (I think, will check) per meter.

Making the handwheels can be the second project on the machine after cutting the motor mounts! I could make the piece that attaches the handwheel to the screw (18), wont cost much.

The power supply you've linked to is probably sufficient, but more power would be better. It's cheaper to make your own using a transformer like 88-3839 (http://www.rapidonline.com/Electrical-Power/Toroidal-Transformer-500va-0-35v-0-35v-88-3839) from Rapid Electronics. Just -earch on google for making a power supply with a toroidal transformer. There's not much to it - you just need the transformer above, bridge rectifier and capacitors. Let me know if that's not enough information and I'll explain better.

Have you considered in the Z-axis screw mount how to temporarily disengage it from the spindle when changing tools so that the force from hammering on the draw-bolt to release the tool taper isn't transferred to the relatively delicate ballscrew/ballnut?

Cheapest e-stop switch I can find is here:

http://www.tlc-direct.co.uk/Main_Index/Distribution_and_Switchgear_Index/Emergency_Stop_Buttons/index.html

One day I'll get round to buying one...

Robin Hewitt
21-03-2012, 09:32 PM
Are these cheap chinese ballnuts doubles like the Zapp DF1605 at £44 ?

That would give about 10 microns backlash until it beds in, but with a bit of Belleville jiggery pokery you can get that down to around zero microns.

Transformers are okay until you get a mains surge and everything goes pop. Unlike back EMF from turning a motor, I have personal experience with this little snag :naughty:

diy-john
21-03-2012, 10:21 PM
Updates:

1) Placed an order with Zappautomation, via the online-store.
1.1) Replaced pulleys / belts with HTD -type, as suggested by Jonathan
1.2) Plan is, screws would be going to Jonathan for machining.
- not sure if the ballnuts could be detached from the screws. They are preloaded. Will talk to Gary at Zapp(automation).


£300 on ballnuts! Are you sure you can't get the cheap RM1605 ballnuts from china to fit as they're $25 each from eBay. Surely you can at least use the cheaper ones for Z?


Good point Jonathan. Thought about that, this time key criteria came to four (4) facets
- Overseas shipping costs - located in Finland, buying from different places will hike costs up. Shipping costs add zero value to the conversion, so minimize them.
- Need for the smaller dimensions of the FDI -type -ballnuts.(X/Y-axis tight space)
- Quality ... don't know much about ballnuts, but did rollerskating / skateboarding in my youth.. what a difference bearings can make :smile:

Perhaps biggest factor I have is "risk aversion". Still must admit, not sure if one will ever complete this thread. It is a big project for me, and if there are many set backs, project gets longer and longer, you never know. Does this thread end up in classifieds -section? Need to stick to safe choices.


a toroidal transformer.


Erm, would love to try that. But my skills..not very confident, have set a soft limit of "no mains" electrical appliance -jobs for me. :redface: Think one can manage stepper motor -cabling, and drivers, because they are sort of behind the power supply (not directly mains)


Have you considered in the Z-axis screw mount how to temporarily disengage it from the spindle when changing tools so that the force from hammering on the draw-bolt to release the tool taper isn't transferred to the relatively delicate ballscrew/ballnut?


Have not yet figured out how that could be best done. Some sort of a quick-release system would be cool. But not sure about design, and whether it would introduce slack to the Z-axis
- Something like this maybe, to quick-release?
5531
Really grateful for all comments and advice in this thread. They have been very helpful. Thanks!

Gary
21-03-2012, 10:30 PM
You can remove them, but i would just leave them on the screw and screw the nut out of the way when doing the machining.



Updates:

1) Placed an order with Zappautomation, via the online-store.
1.1) Replaced pulleys / belts with HTD -type, as suggested by Jonathan
1.2) Plan is, screws would be going to Jonathan for machining.
- not sure if the ballnuts could be detached from the screws. They are preloaded. Will talk to Gary at Zapp(automation).



Good point Jonathan. Thought about that, this time key criteria came to four (4) facets
- Overseas shipping costs - located in Finland, buying from different places will hike costs up. Shipping costs add zero value to the conversion, so minimize them.
- Need for the smaller dimensions of the FDI -type -ballnuts.(X/Y-axis tight space)
- Quality ... don't know much about ballnuts, but did rollerskating / skateboarding in my youth.. what a difference bearings can make :smile:

Perhaps biggest factor I have is "risk aversion". Still must admit, not sure if one will ever complete this thread. It is a big project for me, and if there are many set backs, project gets longer and longer, you never know. Does this thread end up in classifieds -section? Need to stick to safe choices.



Erm, would love to try that. But my skills..not very confident, have set a soft limit of "no mains" electrical appliance -jobs for me. :redface: Think one can manage stepper motor -cabling, and drivers, because they are sort of behind the power supply (not directly mains)


Have not yet figured out how that could be best done. Some sort of a quick-release system would be cool. But not sure about design, and whether it would introduce slack to the Z-axis
- Something like this maybe, to quick-release?
5531
Really grateful for all comments and advice in this thread. They have been very helpful. Thanks!

diy-john
31-03-2012, 02:00 PM
It has been quiet lately, while waiting bits to arrive. Status update:

- Jonathan now has the ballscrews.
-- Screws are R 1605 -type. They are un-machined, and only roughly cut to length
--- 900 mm screw will be cut to ---> 853 mm (X-axis)
--- 600 mm screw will be cut to ---> 508 mm (Y-axis)
--- 300 mm screw will be cut to ---> 259 mm (Z-axis)

Jonathan will have a look at the screws, and lets see what can be done to them.

Motors, drivers and other bits were dispatched on Wednesday 28.03. Now they are in Germany, latest status on EXPD -site:


30/03/2012

11:43
IN TRANSIT - DEPARTURE SCAN - LANGENHAGEN, GERMANY



Meanwhile, thought time will pass more quickly, if one would do something, instead of constantly checking where the parcel is.

1) Downloaded today Ubuntu / linuxcnc from:
http://www.linuxcnc.org/docs/2.5/html/common/Getting_EMC.html

2) Burned on a blank CD ubuntu-10.04-linuxcnc1-i386.iso (http://dsplabs.upt.ro/%7Ejuve/emc/get.php?file=ubuntu-10.04-linuxcnc1-i386.iso) from
http://dsplabs.upt.ro/~juve/emc/ (http://dsplabs.upt.ro/%7Ejuve/emc/)

2.1) For burning used my linux laptop: (commands given on command-prompt, screen output on violet)
- though could use Windows, or whatever software you have. One just likes linux, so that it will be.
cdrecord -scanbus
scsibus1:
1,0,0 100) 'MATSHITA' 'DVD-RAM UJ890AS ' '1.00' Removable CD-ROM
1,1,0 101) *
1,2,0 102) *
1,3,0 103) *
1,4,0 104) *
1,5,0 105) *
1,6,0 106) *
1,7,0 107) *
cdrecord -v dev=1,0,0 ubuntu-10.04-linuxcnc1-i386.iso

3) Found a dusty old PC, with a parallel port, it is an 800mhz pentium processor -pc
- vacuum cleaned the inside a bit (it was really dusty)
5596

3.1) Attached an old 160Gb maxtor -drive into it (to be used for the Ubuntu-linux installation)
- there are some DIY-cabling inside this old PC. She used to have a lot of hard-drives inside it, so the extra cabling is because of that. Not needed for this project.
5597

3.2) Going to boot the computer with this freshly burned Ubuntu (linuxcnc) -bootable CD:
5598

3.3) Almost ready to boot now
5599

3.4) On bootup newly added hardware shows
5600

3.5) Installation process begins and there are some easy-to-answer questions
5601 5602

3.6) Installation took perhaps 30-40 minutes. Once installation is finished, installation says to remove the CD from drive, and press enter for a reboot
3.7) On starting up the Ubuntu first time, noticed there are indeed CNC -related software there, EMC2 for example
5603


Okay this is how far we got now. Did not try the software more. Next steps would be:
- postman brings the Nema 23-motors, drivers etc.
- We buy some cable and sundries, try to connect them to the old PC.
- try to make the motors turn with the computer. (can't wait for this to happen :smile:)

Jonathan
31-03-2012, 02:27 PM
Just had a look at the drawings again...

You've currently drawn the Z-ballscrew with similar end machining to the X and Y screws which will put the screw in tension. On the Z-axis putting the ballscrew in tension is surely such a large force that it will pull the quill to one side. You also can't really combine a quick release with a tensioned screw as it would take too long to setup every time.

diy-john
31-03-2012, 04:33 PM
Jonathan, here is a sketch of how I have planned Z-axis::::


5604

So there is not going to be tension on quill at all. If you think I have mis-understood how it could work, please let me know.

Thanks!

diy-john
04-04-2012, 07:53 PM
Items arrived yesterday from Zappautomation.

5667 5668 5669

In box were:
3 x SY60STH88-3008 Nema 23 stepper motor (http://www.zappautomation.co.uk/sy60sth883008b-nema-stepper-motor-p-24.html)
3 x PM752 Microstepping Driver (http://www.zappautomation.co.uk/pm752-microstepping-driver-p-409.html)
1 x SPS705 (http://www.zappautomation.co.uk/sps705-p-587.html)
3 x R1605T3-FDID-P1 Ballnut (http://www.zappautomation.co.uk/r1605t3fdidp1-ballnut-p-574.html)
1 x HTD3M-9MM Timing Belts (http://www.zappautomation.co.uk/htd3m9mm-timing-belts-p-660.html)
1 x HTD3M-9MM Timing Belts (http://www.zappautomation.co.uk/htd3m9mm-timing-belts-p-660.html)
1 x HTD3M-9MM Timing Belts (http://www.zappautomation.co.uk/htd3m9mm-timing-belts-p-660.html)

Ballscrews, ballnuts, pulleys will come later. Jonathan will have a look at them, in view of machining them. Though I appreciate they are going to be quite tough to machine. (hardened..)


Thought the summer was here, and one would have the motors running in no time. But it is winter back again and the motors will not turn at all.

What I have done, is wire one of the drivers like this:
1) High voltage side (in picture, the upper set of wiring)
- GND ---> goes to power supply "GND"
- +VDC ---> goes to power supply "V+"

Then Nema 23 motor wires go into:
A+ ---> BLUE and RED/WHITE
A- ---> BLUE/WHITE and RED
B+ ---> GREEN and BLACK/WHITE
B- ---> BLACK and GREEN/WHITE

Then the Signal wires:
PUL+ ---> into PC parallel port PIN 3
DIR+ ---> into PC parallel port PIN 2

The "PA settings" (SW1 ... SW8) are:
- first four (4) are "ON"
- last four (4) are "OFF"

5670


Think the high voltage side wiring is correct. Signals not so sure about. Is two (2) wires enough, or does one need more connected?

I have sort of verified parallel port outputs some signals, by doing a similar test as here
http://www.youtube.com/watch?v=EtX3p4HFbtc
(http://www.youtube.com/watch?v=EtX3p4HFbtc)
...by running this little C-program on linux:
http://eagerfish.eu/lpt-programming-example-in-c-for-ubuntu-linux

Tried turning a single motor by running EMC2 (linuxcnc). But no luck.

Have tried googling, where to connect parallel port PINs on the stepper drivers. Have found some information, but not sure one has it right. Please can someone help?

Thanks for looking

Jonathan
04-04-2012, 08:19 PM
With only one wire per input the circuit is incomplete, so no current can flow... so no go.
Connect a wire to join up PUL- and DIR-, i.e. the negative terminals, and connect those to one or more of the ground pins on the parallel port. If it's not marked pin 25 (among others) is ground.

If you still get nothing then check the voltage between PUL+ and PUL- (then DIR+/-) and it should be 4-5V ideally when active.

diy-john
04-04-2012, 09:00 PM
WOW it runs now

- connected PUL- and DIR- to pin 25 on parallel port, like you say Jonathan.

The cables are so loose one does not dare to breath near the machine.

Going to try to buy some electrical connectors or something for the parallel port, tomorrow.
- I have tried to use an old printer cable, which I cut in half. But the wires are so small it is difficult to handle them.

Thanks so much, very happy now.

diy-john
06-04-2012, 06:31 PM
Status is now that all three (3) Nema 23 -motors turn. Driving them with EMC2 (linuxcnc)

To get here, what I have done yesterday and today is:

1) Bought some sundries from a local Electrical store:
5684 5685 5686 5687

2) Parallel printer cable - test each of the small 25 different wires with a multimeter.
- try to understand which colour cable links to which PIN on the parallel port
- here is the printer cable cut, from the "printer end" of the cable
- measured with "continuity mode" in multimeter, each of these 25 small wires.
- for all of you these are probably really elementary steps. But for me it was a bit of a surprise, that one would have
to get into this level of detail.
-- Had relied on some charts I had found from internet, on colour schemes of the parallel cable wires. But they did not seem to match my cable, so had to manually check each of the wires,
to which PIN they connect in parallel port.
- Found these "alligator jaws" handy
5679 5680

3) Here are the parallel port PIN's, cable colours, and where I have connected them on the Leadshine M752 drivers.


5681


4) A picture of how they are on the table now
- had bought 30 LED's for troubleshooting.
- found them helpful, though 30 is too many. I ended up using just one single
5682 5683

5) Now, what do you think?
- Could one now solder cables neatly in place?
- Have been thinking of using this kind of 9 PIN -cables for each stepper-driver
5688

-- So that the cables could be quickly attached / detached.
-- Would use these for the low-voltage, parallel cable wires.
--- Parallel cable (25 wires) ---> divide to 3 smaller cables (9 PIN) ---> each stepper driver

Motors seem to work okay, though they don't "home" itself on the EMC2. Not sure why that is. Haven't got a stop-switch, limit switches yet.

Please comment on whether one can now solder things, or should one wait until have limit switches and emergency stop-button?
- actually, any and all comments welcome

Thanks!

Jonathan
07-04-2012, 09:05 PM
- for all of you these are probably really elementary steps. But for me it was a bit of a surprise, that one would have
to get into this level of detail.

Most people buy a breakout board and connect the drivers to that which simplifies things as the pins are labelled. Breakout boards also add isolation, but since the drivers you are using are already optically isolated there's no point having additional isolation on the breakout board, so connecting it directly is fine. However if you connect the switches directly (with just one pullup resistor on each) clearly there is no opto-isolation, unless you DIY.

diy-john
07-04-2012, 09:12 PM
Just a brief update;
- decided not to solder cabling yet, as one might still want to make changes
- today have been thinking what kind of functionalities one might want from the mill, in terms of
-- homing
-- limit switches
-- zeroing

Fiddled with EMC2 (linuxcnc), added a similar function, as here
http://www.youtube.com/watch?v=DOBxt9MIbFo

...to auto measure workpiece surface level.

Of course haven't got the mill converted yet. Just simulated the real thing, with Z-axis motor running on table, and two wires in my hands (one cable attached to parallel port PIN 13, the other ground, think it was PIN 21).

diy-john
19-10-2012, 10:11 PM
Hello!

Well, it has been some 6 months since this project has been on hold.

Not much has happened since. Got screws machined - thanks to Jonathan. Then got tied up with work, had to put all hobbies aside, and let's leave it at that. Really quite embarrassed and sorry for being away for this long.

1) Plan now is to pick up the pieces, try to remember which planet we are on, and get the mill running. Not pretty. Not good, but get it running.
1.1) For that need some bearing mounts made. A local machining workshop will make some simple mounts, hope to get them in a couple of weeks
1.2) Need attach the ballscrews
1.3) Need attach the ballnuts

For 1.2) attaching the ballscrew on Z-axis, I have today started removing the Z-axis existing pinion / spring.
- For reference, discussed this Z-axis matter earlier here: http://www.mycncuk.com/forums/diy-machine-building/4231-z-axis-huge-slack-beginning-conversion.html

What I have done today is:
A) Started removing these levers
7155

B) So far quite easy, just use a HEX -key to open screws
7156

C) After about 5-minutes, got already most things away from right-hand-side of the mill

7157

D) Spring mechanism was on the left-hand-side
7158

E) Now the spring-mechanism was easy to remove, but this is as far as we got today
7159

F) On the other (right) hand side it looks now like this:
7160


Now, would appreciate some tips:
- please should one remove the rest of the gears?
-- Reason I would like to remove the gears,is that I am hoping to use the hole(s) for attaching a rigid structure, on either side of the quill, to mount Z-axis bearings + motor
-- Or do you think one would be happier having the hand-lever to lower the quill?
- If gear should be removed, how does it come out? Attached is a technical drawing of the mill components, if someone would take a care to give an opinion (on how to remove the stuck gear/shaft)?

7161

Thanks for reading

Jonathan
19-10-2012, 10:32 PM
Have you tried tightening the two nuts in picture C against each other, so that they lock on to the thread, then use the inside one to unscrew the threaded rod? If it's in a tapped hole that should get it out.

I wouldn't worry too much about the hand lever. I tend to remove the hand levers on my mill when I'm using it with CNC since they have a tendency to hit stuff (or me) if I don't!

diy-john
20-10-2012, 07:50 PM
A small update:

- the pinion gear eventually came out by punching the bearing out, slowly and carefully, from the opposite side
7180

- here is the gear/bearing that came out - this is all obsolete now
7181

- now the Z-axis looks mean and lean (LEFT: BEFORE, RIGHT AFTER)
71837182

- Now, I have bought a pile of these sort of building materials, and plenty of short 5mm bolts/nuts
-- They are made from galvanized iron, (or the more expensive ones, are stainless steel, very rigid)
-- Aim to attach bearing mounts to structures made from these building materials
-- This is cheap, but thought would make a running mock up with them.
-- Hope that way one could run the machine a bit, check dimensions again, check where to improve --> then make rigid, nice looking parts from aluminium.

Maybe this ugly duckling will turn to a beautiful swan some day.

7184 7185 7186 7187

Robin Hewitt
20-10-2012, 10:43 PM
The sensitive drill part is really useful when you come to hammer a collet out of the taper, it gives you something to hammer against.

Best if you can release the Z axis screw before you hammer so you aren't bashing against expensive ball nuts and bearings.

diy-john
23-10-2012, 08:50 PM
Good point Robin. Thanks!

Alternatively, perhaps one could use the lever on the left hand side of the quill, to lock it (during hammering)
- Not sure though is it strong enough / meant for that purpose.

Hoping to receive some bearing mounts, before weekend, so would have something to play with..

diy-john
28-10-2012, 09:22 PM
Status -update:

[x] received six (6) bearing mounts for the ballscrew -bearings
- a small local workshop made them for me.
- these are temporary bearing mounts, but very good at such.

7222

- Holes for bearing are 32mm in diameter. Used heat to expand the mounts, so that bearings would go in easier. Holes are quite tight and precise, think it is impossible to remove the bearings in one piece.

7223 7224

[x] continued with a temporary, crude build
- in particular Z-axis

- Z-axis now looks like this 7230 7225

- Z-axis lower bearing mount. Used an electric drill to fabricate some supports
7226 7227
7228 7229

- Purpose is to make a mock-up of the build, write down anything that needs changed, and then hopefully make it a nice looking mill.

[x] an un-necessary side-step was when I dropped balls from a ballnut. Thought first oh well, that was 100 quid down drain. But to one's surprise, managed to find all balls and with instructions found from internet, put it back together. For record, took out the remaining balls, and used a plastic straw from a WD40-bottle to put them back in, one by one. Used bearing grease (like Vaseline) to make the balls stick to the plastic straw / and also stick onto the inside-walls of the ballnut.

diy-john
16-12-2012, 09:46 AM
Latest is:
1) At garage, have a crude build ready to plug motors in
2) At home, have a pc with with LinuxCNC + motors running on table ( this has in fact been ready already for months)

Next steps are
- put 1) and 2) together... by Christmas time will take pc, monitor, motors, drivers, to garage and plug them in.
-- before that, need to study calibrating LinuxCNC X, Y, Z -axis

Lessons learned recently:
- making crude build takes more time than expected. Probably spent 4 days x 8 hrs = 32 hrs to make it, over a few week period.
-- need to buy small sundries here and there
-- compressed air die grinder similar to this is helpful. I find it safe to use (no sparks + quite difficult to cut yourself accidentally), and cuts material quite quickly, but not too fast. Makes a nice sound, too. Of course not a CNC -machinist / milling machine operator approach, but hey, one haven't got a mill running now.
7646
- have already noticed a few design errors by making a crude build, so it has been worth it.

- Longer term plan:
-- Once machine is running, make revised intermediate parts. Use possibly less thick Aluminium than one would have in final build. Cheaper. One expects there will be errors as I am a newbie in all this.
--- want these parts to be relatively rigid, so that it is possible to make good quality "final parts" later.
- Hope to submit more photos next weekend + a compulsory video

76477648

diy-john
31-12-2012, 11:37 PM
Well she runs now, albeit slow and ugly, though to my eyes it is pretty


http://www.youtube.com/watch?v=segC-xqdTxE

Is this it? CNC -virginity lost?

My smallest milling bit is a 10mm wide end mill, so the letters are fat beyond recognition.

7807

Machining marks are quite visible, and think all the axis give way when pushed hard.
- Think the cut aluminium chips are around 0.2mm thick.
- Have to check what sort of feed rates I used actually, and how fast the mill bit was turning. With my zero experience I just set them to something that didn't seem overly fast.

Happy New Year to all!

diy-john
03-02-2013, 07:50 PM
Things have been moving slowly, due to work etc. commitments.

I have now officially begun an intermediate -build of the CNC -conversion.
- Focus is making now parts more rigid, and more like the final, beautiful build.
-- One is using less thick (cheaper) materials here, than in what will be the final build.

Today I have been using the mill as a manual machine, with LinuxCNC, using DRO's and Power feeds.
- Trying to get used to G-code, use simple commands like G0 somewhere, G1 elsewhere... while cutting aluminium.
- Measuring with a digital vernier, cutting a tenth here and there.

Perhaps one had thought I am going to make a fancy drawing and export it into G-code. Send chips flying and ta-da, bolt everything in place, and have a ready CNC conversion.

But it has not been like in dreams. Need to take one step at a time. It was fun cutting aluminium today. :)

8137

diy-john
10-02-2013, 08:52 PM
Small update:
- have been getting more familiar with G-code, LinuxCNC, QCAD, dxf2gcode.
-- Also have Heekscad / HeeksCNC + Blender installed (http://www.mycncuk.com/forums/cad-cam-software/5567-ubuntu-10-04-linuxcnc-heekscad-heekscnc-installation-steps-what-i-did.html).
--- However so far,I have not transferred anything useful to work-piece from Heekscad/Blender.

- This weekend I started to make X-axis parts.
-- use QCAD to make DXF.
-- use dxf2gcode to make G-code.
--- each task I do in several smaller tasks.
---- I could not get dxf2gcode make one long piece of G-code, with climb-milling.
---- So instead, I have chosen to create individual smaller G-code -scripts, that I run in series.
----- Let mill return to home between each G-code script.
-- Getting used to mounting a work-piece.
--- Working with coordinates
--- Zeroing coordinates. Touchoff. In short: mount a work-piece, and practice how one can get cuttings exactly where one wants them.

-- On Friday bought a new 4mm and 6mm carbide end-mill. Smallest I had before this, was a 10mm end-mill.

It is slow progress, however rewarding. Would just love to go back to garage and start cutting again. At this stage, the machine is crude but it does what it is told to do.


8197

WandrinAndy
10-02-2013, 10:15 PM
It's looking good John! ...Thanks for a great thread.

diy-john
24-02-2013, 07:27 PM
This weekend have
- been milling parts for X-axis
- managed to snap first endmill (4mm carbide)

Next week:
- hope to receive new material for the build, 125 kg of 15mm thick plate 6082 -Alloy :biggrin:
- hope to send chips flying

Up until now I have had very little aluminium to work with, just a couple of used bits of 5mm plate (seen in photo)

8270

diy-john
03-03-2013, 05:58 PM
Today is an eve of first Anniversary of this build thread.

- Removed all crude -build structures from X-axis.
- Installed "intermediate build" components on X-axis

BEFORE:
8322 8323

AFTER:
8324 8325 8326


Some other pictures taken today:

- These are for motor mount. 5mm thick aluminium, what I have had lying around. These are mock -pieces.
-- Will use 15mm thick aluminium soon.
8327

- Once removed all the crude structures, had a chance to clean table.
- BEFORE:
8328

- AFTER cleaning with denatured alcohol
8329


It is still a crude build, but it is starting to make parts. Everything is coming more quickly now, that have taken first steps with G-code and CNC.

Got a smile on my face, when made mounting plate for motor. Keyd in coordinates, make mounting holes 47.14mm apart, as it says in technical data sheet. Guess what? The motor is a perfect fit for the plate :smile: CNC is amazing.

Swarfing
03-03-2013, 09:51 PM
The best bit of this is where you said "intermediate build", this means a lot as you understand the pitfalls but not afraid to plough on. You learn a hell of a lot this way and it was the route i took when first started out. I worked out was bad by building bad, but i understood that. Training would cost you an absolute fortune. Well done!:wink:

diy-john
04-03-2013, 08:30 PM
Yes, learning by doing. One year from start of thread now, and we are here.

Please how would you see a 'final X-axis build' like this?

1) Make sort of table-top, out of 15mm aluminium (or possibly 2 x 15mm = 30mm, even 3x15mm=45 mm possible?)
1.1) The top is level with current table top
1.2) Size of original table is 500mm x 130 mm
1.3) Size of new table top is extends current table top quite substantially (60mm at the mill-side of table, more at front)
- will not reduce travel, though.

8339

Reasons why I would want a larger table top:
- keep chips away from ball screw
- perhaps aid attaching a larger item on table (would drill + tap holes. Threaded holes covered with plastic plugs, except when in use)
- add rigidity to X-axis
- add rigidity to motor mount

2) On either side of table, would have 15mm thick aluminium plates, like in photo (except photo has 5mm thick aluminium)
3) The "boxed sections" at either end of table, would add 15mm thick aluminium blocks between original table end, and bearing mounts.
- purpose: add rigidity both longitudal and horizontal
8340

3.1) Motor mount: 2x15mm thick aluminium
- secure motor to larger 15mm thick table top (from above)

4) Bearing mounts on either end of table: 2 x 15mm thick aluminium

I see difficulties with the build in many ways. I am sure one would have been happier with a little bigger machine to start with. But this is what I have, and it will be a nice machine.

But do you see issues with this structure? Would one regret the larger table top?

Would appreciate comments :D

Thanks!

Swarfing
04-03-2013, 09:58 PM
John personally i would stay as you are for now and complete what you started. I will almost guarantee that by the time you do complete this your thoughts will be completely somewhere else....A new build!

In most cases the size of the parts you think you may want to make in the future end up being smaller. After i finish my mill build i shall be making my router smaller. The reason for this is because i am just not using the whole thing and it is space i could do with back in my workshop.

The bigger the part you want to make lends itself more to being done manually i always find. If you happen to come across a larger table for a few quid then just buy it. It will save a lot of heartache in the end.

diy-john
31-03-2013, 09:57 PM
Good point about larger items and most appropriate methods, and manual work. I will make this machine as good as I can. There will never be other one like it.

This week I have started to make the X-axis "final build". This is a slight revision of plan. I have realized time it takes to make the final build is going to be very long anyway, so not going to make Y and Z -axis any "intermediate build"... but next items are final.

On milling table there is a 708mm x 50mm new piece for X-axis. Had bought 15mm thick aluminium from Ebay for this purpose.

8595

Dialled in G-code to cut a through-pocket (15mm depth), but to one's surprise, end-mill did not penetrate all the way through. Have measured thickness with a digital vernier, and plate seems to be 16mm ( 16.2 mm ). Had to rewise drawings.

8596

I have cut the 708mm x 50mm parts with a cheap circular saw, bought for this purpose + blade made for aluminium. If you try this at home, please use a good respirator (aluminium dust) + eye protection. I did this part of work outdoors, due to mess it creates. Used piece of wood + clamps as a guide rail.

8597

Resulting parts are quite straight, no curvature or windings to mention. There may be a 0.5mm difference to the 50mm I wanted, but I will not try to machine them any further. Will make holes and fittings as needed for this build. This part will attach to the "front of the X-axis -table". The 16mm thick part will replace current 5mm parts.

8598

Swarfing
31-03-2013, 10:36 PM
Nice one John good to see your back at it.