[Cret]

Hi guys

Got most of my bits together at last at least for the X & Y setup, and I want to hook up the wiring and try out a few basic manual input tests with Mach 3 to check if things move how they should etc.

But I am confused about the correct way to wire together and then set up in Mach 3 for this kit because I've seen conflicting and contradictory info, so I'm not sure what's correct and what's not.

Basically, I've seen the wiring for these DM542A drivers done two ways:
1) with the + step & dir signal coming from the output terminals on the BOB, and then the ground shared between them
2) with the + step & dir feed coming direct from the +5v input to the BOB and the - inputs coming from the output terminals on the BOB.

Then besides that, I'm confused about whether in Mach3 the settings should be "Active Low" or not (I've seen a few people say it needed to be for these drivers, but not seen the associated wiring re' 1 & 2 above to know the overall setup), and have also seen comments about something needing to be set as inverted in Mach3 depending which way you wire things. Or something!

Sorry that's probably as vague & confusing for you as it is for me.
I'm paranoid about blowing something up by making a bad decision but with what appears to be conflicting info I'm unsure how exactly to proceed.

Would really appreciate any dummy proof help here!

Thanks

[cgo]

The DM542A can be hooked up either way, it has optical couplers inside. If you hook them up wrong, it will just not run the driver or hurt them. (as long as you dont exceed there voltage limits, typically 5 volts)

The bob is the part you would need to look at for wiring to drivers, I did a quick google search and it shows you bob wired as common negative, which means in mach3 theres no special settings from default, just the normal IO, limit and motor tune stuff. ( your option 1)

[Cret]

Thanks CGO

I've wired it common ground like you said.

So for now I've set it up with the two negatives (step/pulse and Dir) to Gnd, the step/pulse+ to pin 4 and dir+ to pin 5.
They're set as these pins in Mach3 but I haven't changed anything else yet after installing it.

I need to look more at the install & config manual but it's wired up, so is the motor, and the PSU, and 5v+ coming from a USB cable.

In theory therefore it's wired up how it needs to be in order to run, so it's just down to whatever software settings are needed, and then going to whichever screen is needed for choosing an axis to move, then use the cursors. * Edit* I think this is via the "motor tuning and setup" screen by selecting the axis required then using the cursors?

Gonna have a skim through the manual now to see how I actually do that, then see if I can make a motor spin!!

BTW I don't actually have an E stop yet (It's en route) so as I'm just wiring up to a motor on the bench I presume I need to disable that in the software.

[Cret]

Ace! I got some success with it!!

I could not get it to work with common ground for the drivers, so I tried the other way.

What I had to do was wire the driver so the step and dir + get permanent 5v+ from the USB power feed, and the - terminals are from the BOB output pins (4 & 5 for Y axis).
Then the motor outputs are set as 'Active Low' in Mach3. I notice when I use the cursors to power the motor that the signal is around 0.15v.

And that's all it needed to get the motors to turn! I haven't worked out the correct number of steps or any of that yet. I just wanted to get a response and movement from the stepper that's hooked up, so I'm VERY happy to achieved that.

Next is to swap out that stepper to the one physically connected to the X2, and see if it will move things.....

Appreciate the help thus far, thanks!

[Cret]

More success - sort of.
I hooked up the stepper for the Y axis and nervously ran a command to move it. It moved, but was incredibly slow. Presumably because of the steps etc not being set right. So I increased them from 2000 to 8000 so it would quickly enough to visibly see it move!

What happens when I send manual commands to it now is that it moves both ways without any horrible noises but as the saddle moves along the base, it appears to speed up and slow down rhythmically!!

The stepper is turning at a steady rate, so I don't understand why the saddle would constantly kind of oscillate its speed as it moves? In one or two places it more or less stopped moving briefly despite the ballscrew turning still.
Could that be that the ball nut is pinched up in the wrong position perhaps, meaning it's tight sometimes?

Without being able to turn it all by hand, I'm not sure how to check this or get a feel for how well it's aligned etc? With hand wheels you can turn them and feel if it's free or not, but I can't since I mistakenly got single output steppers and cannot attach the handles as a result.

Any suggestions?

This video should hopefully demonstrate what I mean if you look carefully at how the saddle moves, almost like it 'pulses':
http://s26.photobucket.com/albums/c1...ps904fd233.mp4

[cgo]

Good to hear its running.
Yea can only be the two ways to connect it.

Under ports and pins/ input signals when you get to that stage, this may help.
I use a normally closed estop so my setting for that is active low checked, same for normally closed end stops they are set "active low" checked, also the option to "emulate" will bypass the signal, good for testing stuff.

5 mm ball screw ? I use 1/4 step which gives a decent resolution, probably better than the machine could do.
Also use inches in my setups so my numbers probably wouldnt help you, guessing your using metric ?

If you look at the mach main window the default feed rate is very low to start, until you load some gcode like there roadrunner or manually over ride it.

In the video it does look like its binding, loosen the ball nut so it can move a little, then move the saddle as close to the stepper as you can get and re-tighten the ball nut while its close to the bearing.
If thats not it then maybe the dove tails need set, usually a new machine will need that checked by removing the end bearing mounts and manually sliding the table or saddle without the screws hooked up.

A way to test if its binding, you could remove the stepper and turn the coupler by hand.

[Cret]

Thanks for the tips!

I'm honestly not sure if I want to have it running metric or imperial to be honest as my brain works with both depending what I'm doing, and without really any consistency!
I may well set it all up metric thought. Just not certain.

But anyway, this is the ballscrew I have for Y:
1 anti backlash ballscrew RM1605-230mm-C7 end-machined | eBay

Yes though - pretty sure it's binding like you say as I scratched my head a bit and measured some stuff.
I figured if I measured the height between the top surface of the stand and the groove in the ballscrew at each end of the screw then I'd know if it was level with the saddle or not.
Ie measuring like this at each end (ignore the number shown on it actually - was just to demonstrate):


It wasn't level!
I looked closer at how the ballnut was positioned, and could see the top of the ballnut body was snug against the underside of the saddle and it occurred to me perhaps the saddle is forcing the ballnut/screw down, preventing it being level & causing binding. I think I need to dremel/grind out a slight concave on the underside of the saddle so the ballnut can sit maybe 1mm higher or so, which will then allow the screw to sit level and parallel to the saddle correctly & therefore cease binding. I hope.

Makes sense in my head anyway!

[Cret]

This is the bit I need to remove material from I think:



Sadly though it's late now and that will be noisy so it'll have to be tomorrow after work. Don't want to upset the neighbours too much!

Can't wait though. It's proper exciting now that after years of wondering about CNC that I'm a good step closer to making it happen.

[cgo]

That looks like the same bit of metal as the grizzly g0704 machine needs removed !

In mach you only setup the machine in inches or metric one time in config, but when your using the program you can then switch to either for machining by using either gcode commands (g20=inch g21=mm) or I believe theres a graphical button on one of the screens for that. I use linuxcnc so I never bothered learning mach, although I do have it installed on the windows side of my cnc pc.

I should have also said you can type in on tab 2 screen of mach, f for feedrate, like f20 or f100 to set max feedrate, on the inches side of mach it equals inches per minute.

[Cret]

All useful info thanks mate

Now that I got the motor to turn, I want to focus again on trying to make sure the axis move smoothly and the hardware fits together freely and operates correctly.

I removed metal from the saddle where marked, then put it all together but it still binds. BUT, when I undo the clamping grub screw that locks the ballnut in place, it moves very freely & smoothly, which says to me that it's still just an alignment issue and not a fault with the ballnut/screw etc. I will work on that, but I'm not worried since I know basically it's ok and just needs tweaking to get right, or potentially a bearing support or the like.

Right now I have a more ominous issue though I think.

Since the Y axis sort of works I thought I'd take a look at X to see what's going to be required. And it worries me.

Placing the X ballscrew/nut on top of the saddle it was very obvious I'd need some grinder action to get it remotely close. As per the area where I've begun grinding it here:




No by the way, that's not a rogue bit of clumsy grinding on the dovetail, it's some lithium grease!!

I've ground a fair bit away already (as per above pic) and it still sits nowhere near low enough for the table to clear it, as you can see in the next photo.
This is with the X ballscrew/nut in situ with the ballnut sitting in the ground out original recess. I have then pushed the table dovetails into the saddle dovetails so you can see how it sits in relation to the top of the ballnut, here:




So what's the best course of action to fit it? Can I grind a load off the top of the ball nut without ruining it somehow? I'm wary about trying it without knowing what other people have done to fit these on similar machines?

This is the shape of the ballnut I have here:




Has a bunch of holes through it, but part way up the side it has a hole/nipple for oiling the inside. My concern is that if I grind away the top, I might eat into the internal space which is for oil to be inside, and potentially trash it if that happens.

Appreciate any input on what to do here guys, thanks. I can grind more off the top of the saddle but it'll mean that I then have to cut a full length groove along it for the screw itself to sit lower where there is currently metal, and I'm not sure if that's a bit too aggressive in terms of what I remove. But then if I don't do this, I'm not sure how I'll be able to fit the table without taking quite a bit off the top of the nut?

Thanks guys!

[cgo]

It should be safe to cut the flat edge or any part of that mount down to the main part diameter but not deeper (if that makes sense), maybe a good hack saw blade will cut it ? Not sure I would want to grind on that and risk getting abrasive into the ball part. This forum probably has a few builds for that mill and maybe with some pics.

I have some blueprints for other mills where they cut away a good deal of the mount only leaving two screw holes to attach it, its too bad these companies dont make things a little easier to convert from the start.

Ive seen the X axis ball nut actually mounted on the outside of the saddle on a mill, they made a longer bearing mount on one side of the table to make up for the offset, that would also need a longer screw to match.
I forget what mill it was on but looks like they maybe ran into the same trouble your having.

[Cret]

Thanks - I just wasnt sure if grinding that part down could accidentally eat into an oilway, but did wonder if those would just be contained within the diameter of the main 'tube' section.
I think on closer inspection it must be the case that the outer is just solid metal since the holes round the side leave nowhere for a passage to reach that top bit from what I can see. Hope that makes sense!

I think this must be the other thread you mention as I was just reading it myself with interest:
http://www.cnczone.com/forums/bencht...onversion.html

[Cret]

I've managed to grind more out from the saddle as I'm using a bench grinder to do so and it's crude and rough at best (there's no-one I can take it to really to machine it out for me), so the nut sits lower now.
I'm hopeful that I can just remove a smaller amount therefore from the top of the nut as a result. I'm not sure if a hacksaw will go through this metal but I'll try before taking it to the grinder. I'll be wrapping it up carefully first to reduce the odds of metal dust getting where it shouldn't be.

First though I think I'll look for a diagram online of the ball nut to see if it shows internal oilway etc. I'm pretty sure on reflection and after your comments that it'd be fine to cut it down a bit in the correct place, but want to double check to avoid an upsetting mistake!!

[Cret]

Right, based on the diagram here for this ballnut, and what appears to be the inner oilway round it, I figure I can take off a minimum of 3.5mm down from both flat sides.
It's probably more, but I don't really want to risk it.

Hopefully between the reduced saddle height beneath the nut, and a bit off either side of the wide part of the nut then I might be in business. And if that's not enough, I'll consider options at that point.

One question I've not really pondered until now, is it possible to remove the ball nut from the screw without it all falling apart? At the moment I'm having healthy paranoia and assuming the moment I remove the nut off the end of the screw that it's all over, and it'll never go back together, but it'd be a plus if that's an incorrect assumption.

[cgo]

You can take them apart , and they do sort of fall apart if theres nothing to hold the balls in. if you look on the side theres a plug, thats how its assembled and also if it did come apart that would be how to repair it.

1) Several things you can do, one is if you had a short piece of the screw you could thread it in place of the longer one.
2) You can use a round piece of plastic tubing to slide inside as you remove the screw to hold the balls in place, probably wouldnt take much abuse unless you packed the center of it with something while its apart.
3) Last would be to unscrew it over something to catch all the balls, then rebuild it, thats where you would remove the plug and do whats called repacking, basically feeding the balls through the hole under that plug.
Doesnt sound like fun but at least its not a total loss if you end up needing to un-assemble it. I know theres a few youtube videos of this proccess.

[Cret]

That's great to know thanks, appreciate the info. Doesn't sound too bad but a bit fiddly. Rather keep it intact if I can though I think, all the same.

At least I know if I end up having to do that, cheers.

Well I removed more off both sides of the nut, and from the saddle so it's pretty close height wise to fitting under the table.
Of course what I then found is the top side of the nut is also way too wide to fit the profile on the underside of the table. Shouldn't take too much to take that down a bit though.

I've had to grind it as hacksaw barely made a scratch. Ideal thing would have been to mill the nut down but I don't have any friends who own a mill.
It's a nasty way of doing it but I taped up the whole thing first and used blu tac to keep the dust out of the join from nut to screw. I also cooled it frequently as I worked to try and prevent heat problems to anything inside the nut since there's plastic in there.

If it all works out ok and functions, then it guess one project could potentially be to get another nut and machine it into a nice neat fit, but then if this ends up working is there much benefit to doing so? That's a way off yet!

[antichip]

I ended up removing all of the one side of the nut so it has the 2 flat edges and one lobe. I even went as far as removing much of the flats past the hole that holds in the "wiper" the actual threaded section is in there pretty far and I still went down to near the threads. I do not think I took any pics though. Everything on the x on my machine is off center and I ground out a good portion from the underside of the table for clearance of the x nut even after removing all of the " flange" . I would bet I also removed more material from where the nut sits so that the screw is a little closer to the cross slide base. To secure it I just made a "shim" that was snug to go between the nut and the cross slide face of the slot. The shim is a U shaped so as to have as much surface area as possible. It was made from a piece of titanium flat stock as it was already perfect thickness. I started like 6 grass fires grinding out the shape. LOL.

[Cret]

Thanks for the description antichip. I'm kind of struggling to picture fully what you're saying about the nut re' which bits you removed and how much.
It'd be useful to understand so I know what I can get away with!

So you've effectively removed 1 lobe from the nut, but how far down? I.e. Is that down to match the diameter of the main tube part of the nut, or not that much?

Sorry if I'm being a bit slow!

I'll be home shortly and doing more abuse with the bench grinder.
I'll try the dremel too with some stones to do more specific bits on the saddle, although I expect that to be proper slow!

[Cret]

Is this what you mean about removing one of the lobes, i.e. the red areas were removed?

Attachment 239532

And if so, doing this didn't expose the oil ways?

What I've removed from mine is more like this currently:

Attachment 239534

So I've reduced the diameter across the 2 flats from 40mm to about 37mm, plus ground out space from the saddle. It's much closer now but not quite enough yet.
I'm nervous about trashing the nut by removing too much though.

[Cret]

Ok so the x ballscrew & nut fits now, but I'm not proud of what I had to do to the ball to get it all to fit!



It's nice and low now:



And another:



Tons of tweaking to be done yet. I've not tried at all to see if the end plate aligns with the new screw. Expect it'll be miles out as that would just be far just too easy for my luck if it fitted!