weird backlash or should i say fowardlash

[ataxy]

ok as the title say i am experiencing a weird fowardlash on the z axis
example:going z down -0.001 per step here is what the dti will say
-.001,-.002,-.003,-.004 now going back up you would expect a bit of backlash so it would go -.004,-.004,-.003,-.002,-.001 and back to 0.000 but instead i get this -.004,-.0045,-.0035,-.0025,-.0015 and 0.0005 so instead of missing a step befor catching the backlash and resuming i get a slight move in the previous direction of 0.0005 before it resume in the right direction

the only thing for now that i can suspect is the z axis dovetail binding a bit due to poor fit of the dovetail and gibbs of the z axis any other ideas, i have checked the screw in the back and aside for the play between the screw and nut (backlash) the rest is tight and fit. i wish i would get news from syil about my gibbs and other part like bellow they promissed me in .....january

[Ed from NY]

Sorry for this question, but sometimes we do forget... did you turn off backlash comp in Mach3?

Another thought that comes to mind is to do a more gentle ramp on the motor acceleration and see what gives.

What is your regular mechanical backlash?

You can also repeat it in the opposite direction and that might give you an indication if it is related to the weight of the head or the strut.

[ataxy]

Sorry for this question, but sometimes we do forget... did you turn off backlash comp in Mach3?

Another thought that comes to mind is to do a more gentle ramp on the motor acceleration and see what gives.

What is your regular mechanical backlash?

You can also repeat it in the opposite direction and that might give you an indication if it is related to the weight of the head or the strut.

yes it does it both way as for backlash it is on but only for x and y, z is set to 0 or no backlash compensation.

[neilw20]

What you have observed is correct, although not what one would expect.
I have even seen this same effect on big 10 ton OKK mill.

With the SX3, starting with the head stationary, if you begin to move down, the ball screw (behind the slide-way) pulls down on the head.
The head does not initially move down, but as the screw applies more force to overcome the friction in the slides the front of the head kicks up pivoting around the slide-way. Until it breaks the friction and starts to move, the spindle nose continues in an UPWARDS, OUTWARDS and TO THE RIGHT. Once it starts to move all is well but the spindle nose lags behind the command position of the screw due to the friction, and compounded (quite forcibly) by the gas strut, which is 50% too strong for the weight of the head (at least until the gas leaks out because the strut is upside down and it's seal is not covered in oil as intended).
If you look from the front, as it goes down, the spindle also kicks to the right also because the strut is on one side, putting the friction off the center line. It needs a strut each side charged to 20Kg. Not one charged to 60!

Now once you are down to position, the spindle has just not got to where you want it.

Once you figure out how much this effect is depending on your gibs and the strut, for anb accurate finish pass when I move down, I overshoot, by a small amount then back off a bit to unwind all the spring loaded effects.

I am repeating to within 0.0002" (in Z) all day once the machine is warm, but have to compensate for this flexing action. Had me tricked for a while until I remembered how a big OKK I once programmed had the same feature.

The reverse happens when you go up. The nose goes down, towards the column, and to the left, until the friction is overcome. By overshooting going down, than backing off just the right amount the spindle ends up where you want it. Then when you move up, there is already some tension allowing the first move to up so the cutter does not dig in before retraction.

With various X,Y movements and vibration, the head can settle a bit, then it will always dig in a bit before moving up. I try to make no Z moves once the head is settled for the finish cut, and hopefully finish off the job or over a hole before I do an upward move. Downward approaches over a hole or other space will help.

Using drill smaller than about 0.3mm (0.012") unless long enough to allow for the misalignment can be 1/4 to 1/2 a drill diameter off position.!!!
Because I have not destroyed the quill action, I still drill 0.2mm holes manually. Careful coordinated movement of all 3 axis is required to use small drills.

2 weaker struts will help, but unless you use frictionless linear ways on the Z just compensate for the error. What did expect for your $$$'s

[ataxy]

Well i was expecting to work as good as a 1 000 000$ machine!!! ;P no just kidding.
Thx for the info tweaking the file to compensate on the move might get hard as most part i do uses 3 axis and not 2.5 axis and some of the g-code used are sometime way over 1000 lines i will try changing the strut for two 20kg ones as this will not solve the problem but might solve part of it i also am thinking about making my own gibbs as syil are to slow on sending part they said they would, sorry about that but i am pretty disatisfyed with there aftersale service, not syil canada but syil china, as i have nothing to say against syil canada support they deserve there A+, anyway thx again fo the help

As for what i was expecting, well of course i was not expecting top notch quality but either was i expecting such lack of precision in the fit of the part, for example in order to reduce the play in the head i had to make a new stronger screw for the z gibbs as i had to push the z gibbs so far into the way that it is protruding about 1.25'' under the z saddle and to my knowledge this is not an acceptable fit for a 6k machine, i mean my laundry washing machine still as better fit then that after 20 years of usage.

[syil]

ok as the title say i am experiencing a weird fowardlash on the z axis
example:going z down -0.001 per step here is what the dti will say
-.001,-.002,-.003,-.004 now going back up you would expect a bit of backlash so it would go -.004,-.004,-.003,-.002,-.001 and back to 0.000 but instead i get this -.004,-.0045,-.0035,-.0025,-.0015 and 0.0005 so instead of missing a step befor catching the backlash and resuming i get a slight move in the previous direction of 0.0005 before it resume in the right direction

the only thing for now that i can suspect is the z axis dovetail binding a bit due to poor fit of the dovetail and gibbs of the z axis any other ideas, i have checked the screw in the back and aside for the play between the screw and nut (backlash) the rest is tight and fit. i wish i would get news from syil about my gibbs and other part like bellow they promissed me in .....january

i think you have been get the new gib and 4th axis long time ago,
when keith at china we also talk about exchange the gib questions.
we need konw customer machine gib space size,then can sent correct one.how to measure it keith konw that.
you also can email to me.
xushuo

[neilw20]

Without making a new tower, where the ball screw can react directly in line with the friction, and all the forces are in the same plane, bending will occur.
Have a look at an example where all the forces are much better reacted. Although the balance strut is off set, at least it is on the center line.

http://www.toolmach.com.au/product.p...age=1&featured

[Ed from NY]

What are your thoughts on removing the strut all together?

I have been running mine without one, but I'm not sure what other problems I may be causing.

[neilw20]

1. With no strut, head likely to wander down when no power applied.
Usually doesn't, but once it starts to move it really gets moving. Try it.

2. Maximum RAPID upward speed is reduced due to extra load on stepper.
That means down is the same. Limits rapids when doing lots of peck drilling.

At least it won't deflect stupidly in the X direction. Only Y and still the reversing backlash feature. Did you look at our CncPro1000 ?

Neat way someone suggested for struts is attach the lower end of the skinny part at BENCH level, way below where it is now. That allows larger travel struts, without the need for an extension at upper attachment point like my early attempt at increasing Z travel.

Using 2 struts allows the skinny 14 or 18mm diameter ones to be used. More elegant? and should be able to make space for access to the head rotate lock pin.

[Ed from NY]

Yes. The saddle is much larger on that machine though. Having the strut starting lower would be good. The reason why I removed mine is that it did not allow me to go low enough on the table.

Still worrying about my regular backlash. I think I'll get a few new parts like the bearing shims, take it apart and put it back together again replacing all screws. I used to have an ultrasonic cleaner, but that went the ebay way and I wish I had it now.

Swarf and coolant management are also something I want to get a handle on before starting messing with the column. They get everywhere. I took half a handful from the back where the Z axis is. A metal bellows would be nice.

So you think the reverse backlash is torsion ah?

[neilw20]

I said stiction or words to that effect, but it actually tilts the head, and while gravity points down and the ballscrew is behind it all, it will tilt, and get full of swarf. I made a nice swarf cover over the Z motor. It's bearing was getting full of cr4p. My coupling was Chinese tight too. I made some X and Y helical couplings, cos 1 broke, but it looked too hard to change the Z one, so I broke the cutter and can't make one anyway now. That's my excuse ?
One day, when I have nothing to do, and a spare machine to drill with I must add a lube system. Lots of hole drilling and messing around. Cos mine get dirty, they've bedded in nicely. Well I do oil them and all my dust stuff just sticks everywhere. I create 2 handfulls of fiberglass/stainless splinters every hour. Clean up workshop? Start the weed blower, take a big breath and it's all done in 15 seconds. Broom handle got used for a jig. (hehe)

[ataxy]

i think you have been get the new gib and 4th axis long time ago,
when keith at china we also talk about exchange the gib questions.
we need konw customer machine gib space size,then can sent correct one.how to measure it keith konw that.
you also can email to me.
xushuo

i did get the fourth axis but none of the gibbs so far and you told me you would send me way cover with the gibbs and fourth axis wich i did not get also. I did get a message from keith asking me to call him in regard of a part i asked him so ill get in toutch with him and see what are going to be the next step

in regard to mesurement of the gibbs... wouldnt it have been easyer to do it befor sending the machine to the dealer..!?!?