MadModder
The Shop => CNC => Topic started by: Will_D on July 19, 2016, 05:00:39 PM
-
First of all thanks to the mods for creating this new area.
I have been maching metal as a hobby for many many years and am quite computer literate.
Now all this talk of CNC has got me interested in spending more money on the hobby.
I have a Sieg SX2 Plus from Arc fitted with 3 axis magnetic DROs and am getting tired of twiddling the X- axis on long cuts.
So some form of stepper motor control may be needed for X & Y and maybe Z.
First major question is:
1. What about backlash in the feed screws? Does CNC and steppers etc aleviate the problem or do I first need to get rid of the backlash? Looking at the price of some ball screws it would be cheaper to just buy a new Mill!
-
Backlash is a
bitc beast.
When machining by hand, you generally know which way the backlash is going and can compensate for it. Most software allows for a backlash compensation but all that does is turn the crank enough to take up the backlash when you change directions. You can't reliably cut circles and arcs with this because the cutting force will either push/pull the cutter into the backlash either before or after the compensation move happens.
That said, if you're taking light enough cuts that a bit of pressure on the gibs/locks can prevent the cutter from self-feeding, it can work ok. Just don't get too excited about complex curvy shapes if you don't eliminate physical backlash first. You can also use "CNC Assist" for some jobs, where you insert stops and comments like "LOCK THE Y AXIS" (do some machining with X) "UNLOCK THE Y AXIS" and so forth.
My own strategy will be to try to plan my machining cuts so that backlash adjustments aren't necessary while in a cut. Instead, I'd take the cutter away from the part, change directions and then do another cut. If this is unsatisfactory then I'll try anti-backlash nuts. If I'm not too old to be machining by that point, I may install ball screws or just scrap the entire works and get a real machine.
Thanks for suggesting this area.
-
Whilst there are many strategies for backlash compensation, its a bit like owning an old car. You can prop it up and keep it going for years, but when you eventually buy a new one you realise you should have just replaced it much sooner.
In CNC that means replacing leadscrews with decent ballscrews. There really just isn't an alternative.
-
I do agree that ball screws are the way to go for a CNC conversion, however they can be a bit of a menace if you also want the machine to be able to be used by hand.
The free movement that they provide meaning that the carriage can more freely be moved by cutting forces, so locking a stationary axis becomes even more important.
-
For what it's worth Will.
Some with these small machines make and use Delrin / Teflon nuts moulded to the actual lead screw. Here is one link that will give you the idea.
http://www.instructables.com/id/DIY-Anti-Backlash-Nuts-Super-Cheap-and-Simple/
Naturally, there was another rather well written build log but alas, no can find.
Possibly not for the purists among us but if finances are limited they seem to work for those who have used them. Ball screws are great though. I still turn handles too. Champagne taste on a beer budget is yet another of my difficulties in life.
If you don't want to go the CNC route there are others that just motorize the x axis using a car windshield wiper motor and voltage control.
John B
-
John B: Naturally, there was another rather well written build log but alas, no can find.
John,is this the one you refer to? http://bbs.homeshopmachinist.net/threads/43645-Making-Acetal-leadscrew-nuts-the-easy-way
I have made about 4 of these to date for acme threads I have screw cut on my Harrison.
The Evanut is really simple to make and has no backlash at all.......Oz
-
Excellent THREAD Oz, I particularly liked the idea of using a glow plug in a short length of leadscrew. My twist would be to put a thermocouple in with the glow plug to have accurate control of temperature. I recon Evanuts will be going in my (distant) future desktop CNC mill project.
Will, a good starting point might be a cheap stepper motor and driver controlled by the new BBC Μicro Bit. It would allow you to play with a smart power feed for the mill, be it X or Z axis.
-
Yes indeed,very good thread. Evan sure hit the sweet spot with that idea,the thread is still being usefully contributed to 5 years from the first post.......OZ.
-
One of the great truisms of DIY CNC is that everyone ends up using ball screws and linear bearings.
The interewebs are overflowing with "Cheap ballscrew alternative" and "Skate bearings just as good as ...." type articles. Ask those guys to show you 5 things they have made on those machines and the excuses start coming.
-
Chinese ball screws and nuts don't cost that much these days. Unless you are just experimenting on a shoestring budget, there is very little reason not to spend some money for the massive improvement they will bring.
Chinese and other far eastern products get slagged off by a lot of people (sometimes with some justification) but they have done more to improve the average homeworkshop than most people could dream of a few decades back. The few examples of Chinese made ballscrews and Hiwin rail knock offs I've seen and handled have been excellent for the price. I would have no hesitation in buying or recommending them.
-
Thanks Manx,
Yup, that was it. Strange how one finds such a thread, bookmarks it then looses the thing by some mysterious computer gremlin. :scratch:
In my case I think it was my junk removal software being a tad too efficient - like, all bookmarks disappeared as it seems to have been treated as a, "temp" file location.
John B
-
Where many seem to go wrong with producing a successful 'Evanut' is they omit the final stage of putting it in the freezer to shrink the acetal co-polymer and using an end portion of the threaded rod as a thread relieving tap.
By not doing this action they end up with a nut that is far too tight to be functional in many cases......OZ.
-
So after googling up some ball screws they seem well priced from China.
SFU 1204 seem they way to go, Tolerance is C7, and 12 as in 12mm od and 04 as in 4 mm pitch I assume??
Now what about the handle end. As already mention my SX2P has a fair amount of backlash at the handle (no thrust bearings). I posted a potential fix (http://madmodder.net/index.php/topic,10428.msg118625.html#msg118625 (http://madmodder.net/index.php/topic,10428.msg118625.html#msg118625))based on the ARC Myford cross-slide fix. If I just fit a stepper motor at the other end will this remove the backlash?
Also do I need to use those "bellows" type flexible couplings 'tween stepper and feed screw?
I already have 5 micron DROs fitted, can these be used as feedback for a CNC stepper controller?
-
So after googling up some ball screws they seem well priced from China.
SFU 1204 seem they way to go, Tolerance is C7, and 12 as in 12mm od and 04 as in 4 mm pitch I assume??
Now what about the handle end. As already mention my SX2P has a fair amount of backlash at the handle (no thrust bearings). I posted a potential fix (http://madmodder.net/index.php/topic,10428.msg118625.html#msg118625 (http://madmodder.net/index.php/topic,10428.msg118625.html#msg118625))based on the ARC Myford cross-slide fix. If I just fit a stepper motor at the other end will this remove the backlash?
Also do I need to use those "bellows" type flexible couplings 'tween stepper and feed screw?
I already have 5 micron DROs fitted, can these be used as feedback for a CNC stepper controller?
Correct 12mm diameter [ nominal ] and 4mm pitch
For the handle end you want a pair of angular contact bearings mounted back to back.
For couplings use Oldham couplings, they are backlash free and the only type f coupling that can handle mis-ailgnment in two planes without binding.
This is how I like to mount motors.
(http://www.stevenson-engineers.co.uk/files/x3xaxis.jpg)
Inside that block are the two bearings and the nut that retains them is the first half of the Oldham couplings. Don't mount the motor on pillars like many do, it's a machine tool not a bloody clock.
No you can't use your scales, none of the hobby controllers like Mach can accept feedback.
-
(http://www.stevenson-engineers.co.uk/files/x3xaxis.jpg)
Inside that block are the two bearings and the nut that retains them is the first half of the Oldham couplings. Don't mount the motor on pillars like many do, it's a machine tool not a bloody clock.
No you can't use your scales, none of the hobby controllers like Mach can accept feedback.
John... I am trying to picture how that is setup in mind. Having difficulty seeing it (my mind is already full). Can you describe a little more?
Eric
-
In my experience, the use of a Flexible Coupling is redundant. If all the Bearing Supports and Stepper Mounts are square and perpendicular there is no alignment issue to compensate for. If the Thrust Bearings are on the Stepper Motor end of the BallScrew, a little Axial play is built into the System at the support Bearing end to compensate for thermal expansion.
Copied from my reply to another Thread;
My method for lining every thing up;
Make sure the BallNut is secured in the intended centre of the BallScrew Axis.
When fabricating the End Plates which hold the Bearing, make the Mounting Holes a mm oversized. (eg 6mm Bolts, use 7mm Holes).
Bolt everything together so that the End Plates can move laterally.
Drive the Axis to the Motor end of it's travel and nip up the Mounting Bolts at that end.
Drive the Axis to the other end of it's travel and nip up the Mounting Bolts at that end as well.
Drive the Axis back and forth between limits and check that there is no binding anywhere.
If all is well, tighten the Bolts fully, and repeat for the other Axes.
I've converted two Mills and built a CNC Router from scratch, and have used this method with success on all three.
The Dual Motor Y Axis on my CNC Router and both the X & Y Axes on my SX3 are directly driven without Flexible Coupling.
-
In my experience, the use of a Flexible Coupling is redundant. If all the Bearing Supports and Stepper Mounts are square and perpendicular there is no alignment issue to compensate for. If the Thrust Bearings are on the Stepper Motor end of the BallScrew, a little Axial play is built into the System at the support Bearing end to compensate for thermal expansion.
Hi,
You say "there is no alignment issue to compensate for".
Everything is made to a tolerance. e.g. the stepper motor flanges may not be exactly perpendicular to the shaft and the bearings in the stepper probably aren't made to highest standards so will have some eccentricity. These errors may be very small but would have some detrimental effect. Using a solid coupling might seem to work OK but I feel sure that wear on parts would be accelerated and efficiency would suffer. I have used zero backlash couplings on my X & Y axes and timing belt drive on my Z axis.
On my machine I can push the table and turn the stepper motors (providing they are powered down). You may find that wouldn't work with a solid coupling.
I'm not saying that it won't work - it's just better to use a coupling in my opinion.
We had many machines at work (in the semiconductor industry) working to micron (or sub-micron) accuracy and, without exception, all the ballscrews were connected to the steppers or servos with couplings.
Cheers.
Phil.
-
@John: thanks for the pics. That looks like a SX3 hence the big steppers.
I would also some more info (pretty please: a drawing) of the thrust bearings etc.
Another question has raised its head: While googling Expensive pro quality UK ball screws one company was showing how to anneal the ends (with an induction coil R/F heater) before maching to suit!
Do the unfinished Chinese screws be that hard??
-
@John: thanks for the pics. That looks like a SX3 hence the big steppers.
I would also some more info (pretty please: a drawing) of the thrust bearings etc.
I used something similar to John's method, but I used a nut to hold the angular contact bearing on the ballscrew instead of using one half of the oldham coupling, see halfway down this post for a pic of the coupling in place - http://madmodder.net/index.php/topic,3535.msg40859.html#msg40859
Another question has raised its head: While googling Expensive pro quality UK ball screws one company was showing how to anneal the ends (with an induction coil R/F heater) before maching to suit!
Do the unfinished Chinese screws be that hard??
They're not that hard, and the hardening is not too deep either. I used a tipped tool and had no problem - http://madmodder.net/index.php/topic,3535.msg39297.html#msg39297
Hope this helps,
Tim
-
Many thanks for the replies -all are very helpfull
This is why the thread is called "The Basics"
I am really trying to 'drill down' to the fundamental knowledge we need!
-
SoI will need steppers/drivers/psu so from the UK I see a nice priced package:
http://www.ebay.ie/itm/3-Axis-CNC-Kit-425oz-in-Nema-23-Stepper-Motor-Driver-CNC-Mill-Router-Lathe-/121787136378?hash=item1c5b14397a (http://www.ebay.ie/itm/3-Axis-CNC-Kit-425oz-in-Nema-23-Stepper-Motor-Driver-CNC-Mill-Router-Lathe-/121787136378?hash=item1c5b14397a)
Will the 425 oz N23's be up to the SX2P?
I'm sure X & Y will be fine but what about Z? (Uses counter weights/gas struts may maker for zero'ish loading on Z
And of course there's that USB to 4 axis controller box thingy that got me going down this path!
WONGA WARNING!! DO NOT FOLLOW THIS LINK!!
http://www.banggood.com/50KHZ-CNC-4-Axis-Offline-Controller-Breakout-Board-Engraving-Machine-Control-System-Card-p-1021134.html (http://www.banggood.com/50KHZ-CNC-4-Axis-Offline-Controller-Breakout-Board-Engraving-Machine-Control-System-Card-p-1021134.html)
-
am getting tired of twiddling the X- axis on long cuts.
So some form of stepper motor control may be needed for X & Y and maybe Z.
Sounds like a power feed may be better suited rather than a full on cnc. That said, my first ever project many years ago, was a stepper motor based power feed. This got me into the basics of steppers, drivers and cnc on one axis, and admittedly, it required very little accuracy so backlash wasn't an issue. It certainly introduced me to cnc ...
-
Indeed Raynerd,
I am not (yet - or even ever) loooking at full CNC capability.
What I am looking towards in 2 axis control form either manual or very simple g-code (like what is generated by the G-Wizard Simulator Wizards)
If I get the basic mechanicals correct then the full 4 axis g-code will follow.
-
Here's how I did it on my X3
I used double shaft motors which let me keep the hand wheels, albeit with them sticking out a bit more.
The mounts were fabricated and then machined true.
The helical couplings are inside the tubes and you can see the access holes for the clamp bolts..
-
This is how I did my X & Y Axes.
First Photo is the X Axis. Solid Coupled Motor and twin Thrust Bearings at the other end. This setup relies on the Stepper Bearings for thermal compensation. No issues in 6 Years of operation, 65micron accuracy.
Second Photo is the X Axis from above.
Third Photo is the Y Axis. Solid Coupled Motor with twin Thrust Bearings. This LeadScrew is so short it didn't require a Bearing at the far end.
Fourth Photo is the Y Axis from above.
-
I'm sure X & Y will be fine but what about Z? (Uses counter weights/gas struts may maker for zero'ish loading on Z
I used the same size stepper for all 3 axis', but on the Z-axis I used a timing belt to reduce 2:1, figuring that I wouldn't need the same speed of movement as X or Y, and that the increased resolution and extra torque would be handy. I also got rid of my gas strut as well and it copes fine with no mechanical assistance.
Tim
-
I did the same with my Z Axis for the same reason. I rigged up a Sprocket above the Spindle Head and another at the back of the Column. This carries a Bicycle Chain running from the Spindle Head to a 16Kg counterweight at the back.
I'll try to get a Photo tomorrow. Nice & simple, and works perfectly.
-
Now what about the handle end. As already mention my SX2P has a fair amount of backlash at the handle (no thrust bearings). I posted a potential fix (http://madmodder.net/index.php/topic,10428.msg118625.html#msg118625 (http://madmodder.net/index.php/topic,10428.msg118625.html#msg118625))based on the ARC Myford cross-slide fix.
An update:
Fitted the 2 thrust bearings and the ordinary bearing and great result - no backlash at the handle end.
There is very little clearance under the tables - so fitting ballscrews maybe a nono!
There is only 32 mm on the Y axis and only 25 or so on the x-axis.
I have read up on the EvaNut and ordered some Delrin bar to have a go!
So just remembered that I have some 20 mm nylon offcuts (Ok Higher Mp but free) so can have a go while ePray and Parcel Motel do their thing
-
For low clearance applications, you might consider a spinning nut ballscrew installation. Not as pretty, but it works well and can be easier to keep swarf out of...
-
There is very little clearance under the tables - so fitting ballscrews maybe a nono!
There is only 32 mm on the Y axis and only 25 or so on the x-axis.
Rubbish, Read up on this thread
http://www.model-engineer.co.uk/forums/postings.asp?th=119249&p=2
Where only tonight I mocked up a SX2P to take ball screws.
-
That's a neat install
-
Counterweight System for my SX3 CNC Conversion as promised.
-
Excellent, multi story milling. Re purpose some of those old lift panels.
4: Setup and air cutting
3: Fine surfaces and finishing
2: pockets and slots
1: profiles and loose parts
B: Bed
£££: Below Bed
-
Rubbish, Read up on this thread
http://www.model-engineer.co.uk/forums/postings.asp?th=119249&p=2
Where only tonight I mocked up a SX2P to take ball screws.
Many thanks John for this link. What a co-incidence!
I will try the Delrin EvaNuts first and see how that goes. I may well do a few mods to the X2 and the ball screws.
I see you used 1605s would you recommend these over 1204s?
-
I used 1605's on my X2 cnc conversion, http://madmodder.net/index.php/topic,3535.msg40209.html#msg40209 shows the X axis ballscrew and nut fitting.
Tim
-
Rubbish, Read up on this thread
http://www.model-engineer.co.uk/forums/postings.asp?th=119249&p=2
Where only tonight I mocked up a SX2P to take ball screws.
Many thanks John for this link. What a co-incidence!
I will try the Delrin EvaNuts first and see how that goes. I may well do a few mods to the X2 and the ball screws.
I see you used 1605s would you recommend these over 1204s?
Will, I seem to remember there being debate on that forum about making the Evanuts from 'Delrin' which is Acetal homo-polymer.
There was discussion about the varying sectional density of Acetal homo-polymer bar causing a less reliable result than using the Acetal copolymer.
I used Acetal copolymer for mine,but I believe some of the ones seen being made in various videos may be using 'Delrin'
Do it in a well ventilated area,whichever plastic you use.....OZ
-
Counterweight System for my SX3 CNC Conversion as promised.
Why a counterweight ?
Done over fourty X3's and SX3's and none needed counterweights and any that had a gas strut fitted it got removed.
350 oz/in motor at 2:1 reduction on 5mm pitch screws
-
No great progress on the Eva nuts but a 500mm by 12 ball screw arrived by slow boat.
Am making progress in fitting a VFD motor and controller to the old girl (Myford ML7!!)
Now the nut is fitted to the shaft and I recall somewhere NOT to take it off the shaft less the balls fall out!
Can someone please explain a bit more as to how to safely remove the nut so I can machine the ends?
As a laugh at bede-time I love this comment from the Rio Equestrian commentator on Irish TV:
"Her ears are pricked back, so she's ready for riding!"
-
Although I've never done it, instinct tells me not to take the nuts off, but to fix them in place whilst machining! Some kind of split tube that bolts together with "U" bolts or something. Or maybe just "gaffer" tape? Anything but have to play with all those escape prone balls!
Good luck, Matthew.
-
Can you butt to ballscrew a plastic pipe that OD is thread ID and screw the ballscrew onto this plastic pipe?
The way we have used is:
1) to make sure this pallscrew is filled proper grease - grease will retain the balls.
2) use plastic pipe to substitute ballscrew, you will fumble at one point...
Assembly in reverse order. Plastic pipe OD should be "snug" fit into the ball screw and ID should slip over the future spigot.
Pekka
-
Thanks for replies - I think I will leave the nut in-situ and see what happens, with Pekka's idea as backup.
-
My only experience dismantling ball screw nuts was unintentional a few years back :bang:
I bought a job lot of ball nuts each on a short length of screw as a retainer, and fiddling with one of course it all came apart in my hands :bugeye: Fortunately it was over a tray and I retained the balls but it was quite a palaver re-assembling it !
Fiddly little things
-
So digging deeper and order hardware and RTFM 9all I can get)
So what are "Home Switches"?
Limit Switches I ubderstand, Kill Switch also.
Is it some kind of auto zeroing?
And one more:
How to calibrate steps versus mm travelled?
-
Limit switches kill the machine. Home switches set accurate home position. Basic homing routine is to move the carriage until the home switch is triggered, move back a small distance, then move towards the home switch at a slower speed. When the home switch is triggered this time, set the home position. This is just one homing strategy, there are several others.
-
To add confusion, many machine controllers will allow you to use the limit switches to home.
-
Thanks, it seems there is a lot to learn still.
Its not just about engineering the mechanicals is it?
-
A machine controller is a fairly simple thing. You can expect.some minor frustrations as you learn how it does.things. Hobby level Geode is so simple you can almost ignore it's existence. Most of the grief comes from electrical noise and idiosyncratic software.
-
Will, start out with something really basic like a desktop router. Forget limit and home switches, play with belt and lead screw designs. You can try out different arrangements based around the same basic frame, add bells and whistles as you like. Get a grip on the software side, see how much force your drive can produce (luggage scales) then move onto a mill or lathe as the fancy takes you. Even start with just one axis !!! Most of all, have fun playing.
-
Thanks, it seems there is a lot to learn still.
Its not just about engineering the mechanicals is it?
Funny, it's the mechanicals that usually turn around and bite me - not the electronics and software. I blame myself for this though, as my shop space and access limitations give me a legitimate excuse to avoid quality (big and heavy) equipment.
Wanna team up? :P
-
Good advise from Joules. Convert a dividing head or rotary table into a 4th axis.
At least this way you will still have something of use.
I never fit limit switches to my machines or homing. Homing wastes far too much time. just set work co-ordinates and machine co-ordinates to the same point and then zero on the corner of the work or vise and you always know where you are.
OK for the big boys who use the bottom paddock as a bed [ you listing Mawson ? ] and use G53, G54, G55 etc to machine 10 of the same item at one setting.
Us mere mortals are had pressed to get a fag packet central on the bed and under the cutter.
-
Tool Changer equipped machines John - you need home switches on such a machine so it knows where to grab the next tool :lol:
-
Yer, North 40 as a table, rub it in.......................................
-
I never fit limit switches to my machines or homing. Homing wastes far too much time. just set work co-ordinates and machine co-ordinates to the same point and then zero on the corner of the work or vise and you always know where you are.
I have to respectfully disagree with omitting limit switches. While my project is still ongoing, I've already made a few things under computer control and the limits have come in very handy. Initially when you're setting up drivers and other parameters, hitting the limit switches is pretty common. Even under moderate speeds it's nice to have the machine hit a switch and error out rather than hitting mechanical end of travel <cringe>. Perhaps if you're using under-powered motors it's ok because they'll just stall out, but when you ramp up the power or mechanical advantage you risk damage without working limit switches.
After finalizing your control/drive settings, the limit switches are very handy when you're jogging around and inadvertently go too far. With my current configuration (LimuxCNC), I can repeatedly bump into the soft limit (just shy of the actual switch and safely away from travel limit) and it just stops jogging that direction and lets me reverse. In order for this to work properly, you _do_ need to home the machine at least once at the start of your session using fixed switches. Better yet, the switches plus the index pulse from an encoder.
Oh, and I don't know about anyone else but I'm never in so much of a hurry that I can't watch a machine home itself a few times in a session. I like to know that the limits and motors are functioning correctly and this is a "quick" and easy way to do it.... assuming you have a system that can move faster than a snail. With my previous iteration (under-powered steppers and cheap driver with Mach3) I'd have to go get a snack while waiting for it to travel the length of the machine. Now with decent servos, it's only a few seconds.
-----
But I digress. Habitually.
Back to Will's questions, one that wasn't addressed was the setting of distance traveled relative to steps. The first thing to do is the math that involves the angle of rotation per step (commonly 1.8 degrees, or 200 steps per rotation), fractional driver steps (full, 1/2, 1/4, 1/8th, etc), the gear ratio between motor and axis (if not driven directly), and the pitch of the leadscrew. You can work it out logically or do a search for some calculations you just plug the numbers into. This should get you _very_ close.
After that you'll want to tweak it as close as you want, within the mechanical limitations of the system. You can do this with a 1" travel indicator, which is ok for hobby use but does have some "significant" error in full travel. Or using a DTI and gauge blocks - a more accurate method, or if all you have is a set of digital calipers and you can figure out a clamping arrangement this can be pretty good too.
The idea is to determine the difference between what the machine "thinks" it's moving, and what it's actually moving. Then you adjust one of your parameters by the percent error. If it gets worse you're doing likely it backwards. Keep doing iterations until you get as close as your hardware lets you. Do it for each controlled axis. For some really long axes it may be worthwhile to set up a "leadscrew map" to account for differences in pitch along the length, but for most machines this probably won't gain you much.
-
If you keep belting the end of the table, pay more attention.
Don't be lulled into a false sense of security, seen it happen all to often. Fit a 4th axis at one end of the table and that limit switch is now null and void because it's behind it. now if that switch is also the homing switch and you aren't paying attention........................................ :palm:
-
For starting out, you want underpowered machines, you also don't want to take your eyes off it whilst it's moving, no limit switch will save over confidence, you don't have a Z limit switch. If you go for the controller I am currently using, it asks you for steps per mm which in my case is 533.333 YES, it handles to 3 decimal places and will have rounding errors in the final moves, but as it does 32bit maths I won't loose sleep over it.
Guy's TOOLCHANGERS !!! Will is asking for basics here.
-
If you keep belting the end of the table, pay more attention.
Don't be lulled into a false sense of security, seen it happen all to often. Fit a 4th axis at one end of the table and that limit switch is now null and void because it's behind it. now if that switch is also the homing switch and you aren't paying attention........................................ :palm:
I'm not convinced you got the point. To summarize: the significant effort required to select, mount, configure and test all of my limit switches was well worth it.
It's easy to be lulled into the trap of omission to save time and money, but if you want to move to the next step, they're essential.
Kinda like ballscrews...... for which I'm sure I'll eventually give into the purchase thereof.
-
No seriously do a rotab or dividing head.
It learns you do do the mechanical of connecting a stepper to the axis of the table and making sure it all revolves free.
Then you move on to the electrics with a simple breakout board like this.
http://www.ebay.co.uk/itm/361459921691
A power supply robbed from a computer, either laptop or desktop and a driver like this.
http://www.ebay.co.uk/itm/TB6560-3A-/301199977453
A simple stepper motor size depending on table but 180 oz/in for a 4" table is fine.
UK based one here.
http://www.ebay.co.uk/itm/Nema-23-Stepper-Motor-1-26Nm-2-8A-4-wires-6-35mm-Shaft-DIY-CNC-Robot-3D-Printer-/121651189123
then it's just a simple thing to wire up and that has taught you how to build a single CNC axis.
And instead of wasting time building a crap wooden router that will only suffice as part for the MKll this will last you all your life.
Every aspect is a learning curve and a keeper, nothing will be wasted.
Plenty of people on here to help you
-
I'm not convinced you got the point. To summarize: the significant effort required to select, mount, configure and test all of my limit switches was well worth it.
It's easy to be lulled into the trap of omission to save time and money, but if you want to move to the next step, they're essential.
Kinda like ballscrews...... for which I'm sure I'll eventually give into the purchase thereof.
.
You put limit switches before ball screws ???????????????
-
My CNC has limit switches on all three axes and a dedicated home switch only in z. (Limit switches can be just that or can be a home switch.)
When I switch on I always do a "home all axes". This first moves the z axis down to my home position (to get to the bottom limit would take a lot of time - but would be the safest option) before homing x & y to the limit switches. It takes less than half a minute.
Mach3 is configured to home in the same z, y & z directions every time. I mount my 4th axis on the LH end of the table. The x-axis always homes to the left away from the spindle.
In Mach3 you can configure soft limits. When you home all axes it sets the limits accordingly and doesn't forget them even if you zero on the corner of a job. In theory you will never hit the limit switches whilst machining.
If you are machining a job which is towards the limit of the capacity of your machine Mach3 checks the GCode for the job and gives a soft limits warning if executing the GCode will exceed these limits. Better than machining an item and only discovering you have a problem when everything stops as you hit a limit switch (or even worse - if you haven't got a limit switch the control won't know and will try to continue machining - wrecking the job.)
Just my 2p's worth.
Phil.
-
You put limit switches before ball screws ???????????????
Indeed.
On my vehicle, I'd fix a brake problem before I became concerned over a small amount of play in a ball joint or tie rod end. Priorities.
Backlash can be dealt with to an extent in your machining strategy, though it does severely limit the type of cut that can be made based on the direction of travel and where those direction changes occur. Want to drill an accurate hole pattern? Just come in from the same direction (in all axes) after every move. The fundamentals of machining still apply here.
While bordering on being moderately insulting, you're also demonstrating excessive faith in the human operator and control software (written by humans). Do you omit the E-stop switches as well? If you do include at least one E-stop in your retrofits, I bet it would get a lot less use if you included limit switches as well.
Nothing is a panacea but limit switches will remain on my list of CNC essentials.
-
I think the consideration is different depending on the power of the drive systems. With a a few kW or so of servo power behind the axis drives of my big CNC lathe or the Beaver Partsmaster CNC Mill it would (in my opinion) be irresponsible not to have limit switches as over travel can cause considerable expensive damage if things go awry.
However if the system is using relatively lower power systems where over travel just stalls the drives without mechanical damage it's another kettle of fish.
When I cocked up and drove the CNC Mill Z axis into the work once it sheared off bolts holding the soft jaws to the vice the sheer strength of the bolts I calculated as 9 tons :bugeye: No limit switch would have saved that situation, but my point is that the forces can be very considerable.
http://madmodder.net/index.php/topic,9305.msg102982.html#msg102982
-
I'll add my voice to the limit switch thing. I wouldn't fit them to anything with NEMA 23 or smaller steppers. Our full sheet router has them now, and you can still see the sheared off mechanical stop bolts from when it didn't!
Homing is only useful if you use fixtures to produce parts.
-
And the title of the thread is ? The basics.
-
To rephrase:
Basically, you don't need limit/home switches on a basic machine.
-
Sorry PK, wasn't replying to you, only the post in general.
-
Have managed to fit a stepper to the x-axis and am getting into the DDCNC controller and its parameters (There are a lot).
Have set up soft limits, backlash compensation, and am starting to speak "baby g-code"
Question about Z axis tool sensor:
Helps if you have one and know how it works!
Quick google finds the cheapest at about €15 and the most expensive is €150 plus.
Am I right in assuming that its just an insulated (from the table) block of metal with an exact, known height of say 30 mm that when touched by the tool closes an electrical circuit of the controller (+12v to Probe Gnd)?
Just checked that Tungsten Carbide conducts but what about exotic ceramics (not that I can afford any)
I will be simulating this and the hard limits/home limits by a simple switch (NO/NC) tomorrow.
-
The Z height sensor doesn't NEED to be a precisely known height as it is used comparatively. Set your longest tool using it to zero off set, then all your other tools are measured and their 'shortness' is set as an offset in your tool table.
Touch off your workpiece top surface with the longest tool and this becomes Z Zero, and all the other tools will be correct when their offset value is read from the tool table.
Of course if the height of the sensor IS known precisely then it can be used on the workpiece instead of touching off Tool Zero
-
.....Question about Z axis tool sensor ...... Am I right in assuming that its just an insulated (from the table) block of metal with an exact, known height of say 30 mm that when touched by the tool closes an electrical circuit of the controller (+12v to Probe Gnd)?.........
Will,
The essence is that when the tool touches the sensor, a circuit is made (or broken) and the simplest form is, as you describe, a conductive plate or block insulated from the rest of the machine. However, a solid block is unforgiving, and you can damage the tool tip or block unless you sneak up via a Rizla paper or similar. Better sensors have a platform with some 'give' - the contact pad can be driven down (against a spring) if you over-travel but once you back off the platform rises back to its datum until it contacts a fixed lip (and, just like referencing a machine's axis, its usually better to take the position when you back off as that eliminates any over-travel). The platform can be just mounted on a plunger, but there is a small danger of binding especially if the contact is off the centre of the platform or button, so even better is the arrangement like the three-legged (but not Manx!) innards of a Renishaw probe.
Dave
-
Here's the first one I made:
Spent hours getting that sliding fit perfect..
All subsequent units have just been blocks of conductive things. The one that came with our 1325 router was just a 10mm thick disk of brass.
Others have used un-etched copper clad PCB material, and I've seen a commercial unit that did the same.
If you are bringing Z down at 30mm/s and it can accelerate at 3000mm/s/s then it's going to take about 1mS to stop. Average speed (ignoring fancy acceleration curves) will be 15mm/s, so it will overshoot 0.025mm..Now that might be a lot on a big machine, but small machines flex an order of magnitude or two more than that...
PK
-
Will - what size motor have you used? nobody answered your original query about the 425ozin ebay advert.
John gave us
1)180 ozin for a 4in rotab.
so advice please on
2) 6in rotab
3) dividing heads of various sizes
These would always be direct drive but a mill could use 2:1 or 3:1 gearing with effect on speed too so next question is for the basic intro
4) what step rate is sensible for the stepper pulses and
5) what traverse speed in/min is adequate for a hobby mill
6) what speed for Z axis if different from above
having established speed, and assuming direct drive
7) what size stepper for each axis on say X2 and X3 mills.
John again gave us I think 325 ozin for a Z axis on a ?? mill moving the whole head without counterbalance what about
8) quill only and
9) raising knee
as opposite ends of the spectrum.
-
.....Question about Z axis tool sensor ...... Am I right in assuming that its just an insulated (from the table) block of metal with an exact, known height of say 30 mm that when touched by the tool closes an electrical circuit of the controller (+12v to Probe Gnd)?.........
Will,
The essence is that when the tool touches the sensor, a circuit is made (or broken) and the simplest form is, as you describe, a conductive plate or block insulated from the rest of the machine. However, a solid block is unforgiving, and you can damage the tool tip or block unless you sneak up via a Rizla paper or similar. Better sensors have a platform with some 'give' - the contact pad can be driven down (against a spring) if you over-travel but once you back off the platform rises back to its datum until it contacts a fixed lip (and, just like referencing a machine's axis, its usually better to take the position when you back off as that eliminates any over-travel). The platform can be just mounted on a plunger, but there is a small danger of binding especially if the contact is off the centre of the platform or button, so even better is the arrangement like the three-legged (but not Manx!) innards of a Renishaw probe.
Dave
I think relying on continuity through the cutting tool to a conductive block (even with an over-travel allowance) is flawed from the start.
With a contact system, in the event of e.g. a wire breaking, the tool will continue driving until something breaks or stalls.
A system which breaks a circuit on the tool touching down (but ideally still with over-travel allowance unless you're prepared to run the tool down at a ridiculously slow speed) is inherently safer in that if a wire breaks or a plug falls out the control thinks the tool has already touched down and the tool won't be driven down at all.
Hope this is of help.
Phil.
-
A system which breaks a circuit on the tool touching down (but ideally still with over-travel allowance unless you're prepared to run the tool down at a ridiculously slow speed) is inherently safer in that if a wire breaks or a plug falls out the control thinks the tool has already touched down and the tool won't be driven down at all.
But you still have wires, if they break then you're diving for the stop. Yes the over travel helps here. But I've done a LOT of tool height setting with both systems; if I broke 3 tools tomorrow because of a dodgy wire, I'd still reckon I was way ahead of a non automatic tool setting machine.
Oh, and:
As a reliability engineer I offer that a single wire to a conductive block has half the failure modes of two wires going to a switch.
:smart:
PK
-
A system which breaks a circuit on the tool touching down (but ideally still with over-travel allowance unless you're prepared to run the tool down at a ridiculously slow speed) is inherently safer in that if a wire breaks or a plug falls out the control thinks the tool has already touched down and the tool won't be driven down at all.
But you still have wires, if they break then you're diving for the stop. Yes the over travel helps here. But I've done a LOT of tool height setting with both systems; if I broke 3 tools tomorrow because of a dodgy wire, I'd still reckon I was way ahead of a non automatic tool setting machine.
Oh, and:
As a reliability engineer I offer that a single wire to a conductive block has half the failure modes of two wires going to a switch.
:smart:
PK
Sorry PK,
My system is fail-safe and there's no diving for an E-Stop button
Yes I still have wires (and 2 at that) which connect to the control via a jack plug. If the plug gets pulled out or one or both wires breaks then Mach3 gives an error message (from memory "Tool Already Grounded") and the Z-axis doesn't move. Try breaking a wire connected to a piece of pcb material used as a sensing pad and the control will keep driving the tool towards the table until .....
My system also works with e.g. a diamond drag engraving tool which is non-conductive.
Phil.
-
.....
Oh, and:
As a reliability engineer I offer that a single wire to a conductive block has half the failure modes of two wires going to a switch.
:smart:
PK
How many well defined nodes you have on return leg? Your failure analysis is only half done :poke:
Pekka
-
Thanks for the replys and warnings about the Probe Overshoot! As I am used to manually touching off the tool didn't realise that the Probe function in the controller will drive the tool into the tool height sensor and so some flexability is required. If and when I motorise the Z axis then I am forewarned.
Am also using a simple switch to explore the Home function of the controller and the Hard limits.
Just waiting for the Y axis ball screw to arrive as I think I have found a simple (ie no milling of the machine) way to fix it in place.
Will
-
so some flexability is required.
To reiterate. It isn't. There may be better ways, and they are all more work....
-
Thanks for the replys and warnings about the Probe Overshoot! As I am used to manually touching off the tool didn't realise that the Probe function in the controller will drive the tool into the tool height sensor and so some flexability is required.
I just use the "Painfully Slow" setting in the touch off routine :coffee:
-
Well its been a while but eventually I have all 3 axes under control.
The Fusion 360 knowledge is slowly growing and I am starting to generate g-code and cut air!
A really great support site is https://www.nyccnc.com There are some great videos here and a ton of good info. What I like is that they take things slowly, zoom in to the pop-up windows so you can read the values and don't assume you know everything (like the Fusion videos).
Only problem is that they tend to use Tormach CNCF machines so there recomended feeds and speeds are way too high for my little SX2P. Also everything is imperial.
Does any one have a link to a spread sheet that will allow reasonable calculations of feeds and speeds for a machine like the SX2P and for Aluminium/304 SS/ and FC steel?
-
So now I am getting very confused!
I have run F360 to generate a simple "spot drill 3 holes" program. This outputs a lot of G-code and some of it involves the machine home position.
As I understand the Machine home co-ordinates in the DDCSV controller are the smaller co-ordinates to the left of the Large co-ordinates which represent current workspace. No mention of this in the old manuals!
The machine co-ordinates are manipulated by using the red MACH co-ordinate system .
This system defines the soft limit sets.
X and Y are understandable in both co-ordinate systems (using G54 as default) but Z is totalling confuding me.
Do I set my G54 Z home (zero) position to the tool tip on the workpiece?
I am hoping to find a video entitled "first cuts with F360 gcode" and how to set up the machine.
One other question: I can fit hard limit switches to X+, X-, Y+, Y- and Z+. Does any one use a limit switch for Z-?
Having spent a lot of time and money on this project I am not sure if this controller is up to the job as a Mill controller. It seems to designed as a CNC router.
See the attached file
-
Can you export your Fusion 360 file (as Archive *.f3d) and attach it to the thread?
Do I set my G54 Z home (zero) position to the tool tip on the workpiece?
It depends on where you set the Work Coordinate System in Fusion 360. Generally setting it to the top of the workpiece is usually easiest.