why when i tighten the gibbs on my dovetail mill the backlash goes up?

All things related to the Centroid Acorn CNC Controller

Moderator: cnckeith

Post Reply
rk9268vc
Posts: 267
Joined: Fri Nov 13, 2020 4:12 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No

why when i tighten the gibbs on my dovetail mill the backlash goes up?

Post by rk9268vc »

Ive got a dovetail ways mill with double ballnut ballscrews. (PM-30MV)

When i tighten the gibbs, driving the wedge in, all of a sudden i get more backlash.
Backlash meaning that when i jog the mill forward 5 thou, when i jog reverse 5 thou it only goes one thou etc

If i tighten my gibbs a lot, i can get up to 4 thou backlash, but when i loosen the gibbs some (not all the way) i get half a thou of backlash.

what causes this? I understand that the servos have to work harder to move the table when the gibbs are tighter, but i dont understand getting more backlash when i tighten them. does the added resistance like stretch the ball nuts apart or something?

I added a one pump oiler, so i thought it would be a good idea to go tighten all my gibbs some, but now im getting this backlash ive never seen before, and wondering what gives? If i loosen the gibbs back, the backlash goes away.

I guess the next question is, it it better to have extremely tight gibbs and run a lot of backlash comp, or have looser gibbs and less backlash comp?

(my "loose" setting with half thou backlash, i can lean my whole body on the table and only move an indicator 1 thou)
suntravel
Posts: 1967
Joined: Thu Sep 23, 2021 3:49 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: 6433DB0446C1-08115074
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No
Location: Germany

Re: why when i tighten the gibbs on my dovetail mill the backlash goes up?

Post by suntravel »

That is because more force is needed to move the slides, some tension is build up in couplings, machineframe before it moves.

I tighten the gibs so the slides have no play and then adjust the backlash compensation.

Uwe
Muzzer
Posts: 728
Joined: Mon Feb 19, 2018 2:52 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: 38D269594F9C-0110180512
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No
Location: UK
Contact:

Re: why when i tighten the gibbs on my dovetail mill the backlash goes up?

Post by Muzzer »

Adjusting gibs is an art in itself. When you have no means of manually moving the axes and have 3 gibs per slide (on a machine with box ways), you have to think very carefully about what is going on. As Uwe says, if you overtighten, you end up with tension building in the compliant (springy) parts of the system before the axis finally moves. When it does so, you need to hope it does so consistently, rather than overshoot.

The instructions I have seen for gib adjustment seem to suggest you should gradually tighten the gibs until the backlash starts to increase. This is based on the notion that the force required to move the axis is constant until the gibs start to bite. And that this force doesn't vary significantly. The degree of table overhang affects it on my machine.

I'm not sold on the ability of backlash compensation to make up for backlash. Backlash (hysteresis) is a very nonlinear behaviour and its effects will depend on the type of cut you are making at the time. Depending whether the cutter is pushing or pulling the work isn't just a function of whether you are climb milling or conventional milling - it also depends on the depth of cut. You might argue that it only matters for finish cuts but that would assume that the tool never changes from push to pull or vice versa during a finish operation.

To put it all in context though, Swissi's probe app allows you to measure the backlash on your machine at any position. I found that quite a learning experience on my Shizuoka mill(!), where the actual measured values are much worse than you will see when playing with a DTI. Perhaps using his app would be a good place to start. That way, you can log the measured backlash at different positions and plot them against gib adjustment. But there comes a point where you realise you may be slapping lipstick on a pig!
rk9268vc
Posts: 267
Joined: Fri Nov 13, 2020 4:12 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No

Re: why when i tighten the gibbs on my dovetail mill the backlash goes up?

Post by rk9268vc »

Muzzer wrote: Mon Sep 05, 2022 5:19 am Adjusting gibs is an art in itself.
if you overtighten, you end up with tension building in the compliant (springy) parts of the system before the axis finally moves. When it does so, you need to hope it does so consistently, rather than overshoot.
My mill used lovejoy style couplers from the servos to the ballscrews, i wonder if the plastic spider is squishing slightly under the high force when tightened down. I oversized the servos for my mill, and with the power of the inclined plane, i had put a ridiculous amount of clamping force on the ways.
I might look into buying more servo specific shaft couplers. I use them a lot at work and they work pretty well.
https://www.mcmaster.com/2764K426/
Muzzer wrote: Mon Sep 05, 2022 5:19 am To put it all in context though, Swissi's probe app allows you to measure the backlash on your machine at any position. I found that quite a learning experience on my Shizuoka mill(!), where the actual measured values are much worse than you will see when playing with a DTI. Perhaps using his app would be a good place to start. That way, you can log the measured backlash at different positions and plot them against gib adjustment. But there comes a point where you realise you may be slapping lipstick on a pig!
I have the swissi probe app and have never seen this feature? Where do you find that?
rk9268vc
Posts: 267
Joined: Fri Nov 13, 2020 4:12 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: none
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No

Re: why when i tighten the gibbs on my dovetail mill the backlash goes up?

Post by rk9268vc »

update, i dont think it is the servo shaft coupler, even with tight gibbs and 4 thou backlash, and backlash comp turned off, when i jog in 1 tenth increments back and forth, the screw turns.

for now i just adjusted it to where i have around 1 thou backlash in x and y, and when i put my whole body weight into trying to move the table, it doesnt move more than 1 thou

ive got to get these parts done, then i can play with it some more.

I also notice that the backlash varies depending on how much i have the table overhanging.
I get less backlash in the center, especially on Y, when the table is centered. When at full X travel, i get a little more backlash in Y (like 2 tenths more).
I can also measure a slight change in Z, so i presume it is just the whole casting bending from the 300 lbs of table weight twisting it.
suntravel
Posts: 1967
Joined: Thu Sep 23, 2021 3:49 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: 6433DB0446C1-08115074
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No
Location: Germany

Re: why when i tighten the gibbs on my dovetail mill the backlash goes up?

Post by suntravel »

You also should move the vice from time to time, to get even wear of the slides over the whole length.

Most cheap asian machines need also scraping to get even play over the whole travel.

Uwe
Muzzer
Posts: 728
Joined: Mon Feb 19, 2018 2:52 pm
Acorn CNC Controller: Yes
Allin1DC CNC Controller: No
Oak CNC controller: No
CNC Control System Serial Number: 38D269594F9C-0110180512
DC3IOB: No
CNC12: Yes
CNC11: No
CPU10 or CPU7: No
Location: UK
Contact:

Re: why when i tighten the gibbs on my dovetail mill the backlash goes up?

Post by Muzzer »

rk9268vc wrote: Mon Sep 05, 2022 4:34 pm I also notice that the backlash varies depending on how much i have the table overhanging.
I get less backlash in the center, especially on Y, when the table is centered.
If it's an old machine (you may have said but I don't recall), there will be more wear on the ways where it was used most often. If you tighten up the gibs there, it may be very stiff elsewhere. Equally, with dovetail ways and big overhang, you will get a greater degree of preload due to the centre of gravity moving away from the saddle at the extremes.

You can also see the effect with (non-ballscrew) leadscrews when you eliminate the backlash around the centre position. When you then move to the ends of travel, you damage (stretch) the backlash adjustment screws
Post Reply