Giving the 1100 a "tune up"

[widget_maker]

Hi Guys,

I really haven't used my Tormach in the manner for which it was intended for a few years. Now retirement is upon me it's time.

I'd like to give the machine a once over, so far I've found a few things I am not excited about, but perhaps my expectations are not accurate.

Took off the way covers, and was surprised to see a lot of patches of varnish on the Y. The X I can't see, but the Z was as clean as a whistle. Checked the lube fittings, and all (that I could easily get to) seem to be putting out oil, the distribution is a lot heavier on some ways than others, for example the right Y gets a lot more oil tthan the left Y but both do get lubed.

I checked the machine for "level" again, looks like the shim I installed some years ago is still the correct one. However, I have pretty substantial Z non prependicularity. Moving the Z up along a 1-2-3 block gave me about 1 thou per inch. I do need to check another block however, could easily be the Chicom 1-2-3 I guess.

I looked for lost motion/backlash on the Z next. Bringing the head down, with a besttest in the spindle until it touches down, then reversing shows about a thou of slop. That seems pretty good to me, certainly for the junk I make. However, the stepping on the 0.1 thou range on pathpilot is bizarre. it will go tip-tip-tip-bop!, so you'll get a series of no or, tiny increments up in Z, then a large increment, upwards of a thou, then the same again, and again. Like stiction I guess.

Can you experts give me an opinion on this? Is this normal? Doesn't seem very good. I was hoping the Z gib was loose and tightening it would solve the Z axis "slope", but if anything doesn't this imply the Z gib might be tight?

thanks in advance.
ken

Similar Threads:

• New Machine Build- How can do ""manual stepper motor sizing for "rack and pinion"calculation""???
• Have to fine tune "Counts per Unit" - Just wondering why?
• Need Help!- VF-2 giving "low transmission oil pressure" warning
• Need Help!- How do you get the spindle to turn on with a PCNC 1100 for "maual" milling?
• Problem- X Axis "Goes Off Pattern", "Awry", "Skewed", "Travels"

[widget_maker]

OK, glad to have got the second post in - looks like the 1-2-3 block was the source of most of the issue. Z travel looks good when measured against a 6" precision square. X axis lost motion is about 1 thou when stepping on the path pilot 1 thou setting. If I step in one direction, then switch I always lose the first thou, but then it's good after that. Is 1 thou an acceptable number for a series 2 1100?
I still have questions about the 0.1 thou setting, seems beyond the 1100.

Lost motion (i hope I'm using the correct term) looks closer to 2 thou in the Y direction. By stepping Y+, then reversing, the first step is nothing, then the second step is around 0.2 thou.

[jwatte]

The 0.1 thou setting is still helpful once you take up the backlash slack.
Also, the software should have a setting for backlash compensation.
People have reported various degrees of success with tuning these parameters in this forum.

[popspipes]

Backlash compensation can be set in each axes in pathpilot, it can be set in the .ini file there was a post on it a couple years ago.
I use the .0001 setting on the jog shuttle, just a habit I guess.....

I found the backlash entry from the "re: renishaw probe" post:

[widget_maker]

Mike,
Thanks, I had n o idea that there was backlash compensation. I don't know if 1,1 and 2thou are decent numbers for a hobby level mill like the Tormach, but I suspect it's not terrible. Perhaps I will add the backlash compensation and be done for now?

A question for both of you: how linear are your .1thou steps on the jog? Mine are terrible.

[popspipes]

Mike,
Thanks, I had n o idea that there was backlash compensation. I don't know if 1,1 and 2thou are decent numbers for a hobby level mill like the Tormach, but I suspect it's not terrible. Perhaps I will add the backlash compensation and be done for now?

A question for both of you: how linear are your .1thou steps on the jog? Mine are terrible.

enter your backlash like it is shown in .000 -.0000 increments, I think this may be a linux thing rather than PP but it works regardless. I had trouble with it leaving witness marks on the part so I would up disabling it. I had .0028 backlash in the Y axis, I did shim the ballnut but havent put the backlash setting back in to see if it still leaves the witness mark.

I use the .1 setting exclusivly on the jog wheel, once the backlash is set its pretty much right on.

[TurboStep]

A question for both of you: how linear are your .1thou steps on the jog? Mine are terrible.

Quite some time ago I made a short video showing microstepping on my 1100. The .1 thou steps aren't exactly the same but each step can be clearly seen on the indicator and there's no real "tip-tip-tip-bop!". I suspect either your gibs are tight or you have debris between the way running surfaces.
https://www.cnczone.com/forums/torma...ml#post1672052

Did you tram you mill by adding shims under one corner of the base? If so, this could be the reason you have higher backlash in the Y (there are of course other possible causes). This method can leave the Y ways twisted and this would almost certainly increase backlash. I'd recommend leveling the Y axis using a machinists level as shown in the recent Tormach video.

While backlash correction sounds like the solution for all manner of problems - it's not! If your mill is set up correctly you don't need it. I leave backlash correction turned off except for very specific cases where I like to think I need it (but really don't) and I know that it will work for me. In (many?) other cases backlash correction can work against you and DOUBLE the effective backlash. I tried to explain this in the following, long-winded post:
https://www.cnczone.com/forums/torma...ml#post2118824
If you don't understand what's happening then it's best to leave it alone or risk pulling out what ever hair you still might have (this is becoming less of an issue for me )- this is the best compromise.
There are easier ways to enable/disable backlash compensation and I suspect this might have been one of contributing factors why Titan's first part on his Tormach 770 came out 3 1/2 thou oversize (search YouTube for "My New Tormach CNC Machine - Vlog #58").
@Mike: If my memory isn't failing me, the last time I remember you mentioning backlash correction on this forum you said you were no longer using it because it was causing problems, but I can't find this post. Are you still using it?
Step

[popspipes]

Quite some time ago I made a short video showing microstepping on my 1100. The .1 thou steps aren't exactly the same but each step can be clearly seen on the indicator and there's no real "tip-tip-tip-bop!". I suspect either your gibs are tight or you have debris between the way running surfaces.
https://www.cnczone.com/forums/torma...ml#post1672052

Did you tram you mill by adding shims under one corner of the base? If so, this could be the reason you have higher backlash in the Y (there are of course other possible causes). This method can leave the Y ways twisted and this would almost certainly increase backlash. I'd recommend leveling the Y axis using a machinists level as shown in the recent Tormach video.

While backlash correction sounds like the solution for all manner of problems - it's not! If your mill is set up correctly you don't need it. I leave backlash correction turned off except for very specific cases where I like to think I need it (but really don't) and I know that it will work for me. In (many?) other cases backlash correction can work against you and DOUBLE the effective backlash. I tried to explain this in the following, long-winded post:
https://www.cnczone.com/forums/torma...ml#post2118824
If you don't understand what's happening then it's best to leave it alone or risk pulling out what ever hair you still might have (this is becoming less of an issue for me )- this is the best compromise.
There are easier ways to enable/disable backlash compensation and I suspect this might have been one of contributing factors why Titan's first part on his Tormach 770 came out 3 1/2 thou oversize (search YouTube for "My New Tormach CNC Machine - Vlog #58").
@Mike: If my memory isn't failing me, the last time I remember you mentioning backlash correction on this forum you said you were no longer using it because it was causing problems, but I can't find this post. Are you still using it?
Step

I am not using it presently, I shimmed the Y ballnut and got the backlash down to about .0004. It did work but left a witness mark on the parts is why I stopped using it, after I got the backlash under control I never put it back in to see if the mark was still there.

[TurboStep]

I am not using it presently, I shimmed the Y ballnut and got the backlash down to about .0004. It did work but left a witness mark on the parts is why I stopped using it, after I got the backlash under control I never put it back in to see if the mark was still there.

I believe this essentially proves part of what I'm saying. The backlash correction didn't satisfactorily cover up your problem, but resolve your problem and there is no real need for backlash correction.
Step

[popspipes]

I believe this essentially proves part of what I'm saying. The backlash correction didn't satisfactorily cover up your problem, but resolve your problem and there is no real need for backlash correction.
Step

I agree Step, fix the primary problem first instead of applying a patch.
I still need to shim the X axis ballnut, but it isnt out that much and its working fine for the parts I make.

The ball nuts are easy enough to shim, the tricky part is cutting it to fit correctly, to repair it properly a new spacer needs to be made and ground to the exact thickness (larger) than the original.
I might mention the Y ballnut is a real "bear" to get to as well!

[widget_maker]

Yes, I checked level using my Starret 98-12 as per the Tormach video, looks like I got the shimming correct way back as no changes were needed.

If I could ask you guys to please answer me this:

Is x=1, y=2, z=1 (all thou) easily improved upon by gib and ballnut adjustment? Or should I just consider it par for the course?

[TurboStep]

Is x=1, y=2, z=1 (all thou) easily improved upon by gib and ballnut adjustment? Or should I just consider it par for the course?

Ballnut adjustment is not something for the light hearted. Angular contact bearing and gib adjustment are not too difficult. I'd certainly check out the Y axis gib, expecially if this axis does the "tip-tip-tip-bop!". How far you want to go depends on your needs and bear in mind you can do this at any time in the future.
Step