Machsupport Forum

Mach Discussion => General Mach Discussion => Topic started by: Bob La Londe on July 28, 2012, 01:15:31 PM

Title: Backlash compensation
Post by: Bob La Londe on July 28, 2012, 01:15:31 PM
Over time I have noticed if I leave the Taig fairly sloppy it tends to stay close to the same amount of sloppy for much longer than if I adjust and tighten it up to reduce mechanical backlash.  One of Hoss's videos inspired me to try leaving it that way and instead using backlash compensation.  At his settings the machine would lose position and walk across the face of the work piece fairly quickly.   I used very conservative settings compared to what He had used and it seemed to work ok.  I cut a 1" register pocket with a .250" end mill and I was modestly satisfied with the results.  It was within .003" total tolerance of round.  Then I tried to cut some parts.  With a 1/4" ball end mill the 3D pocket I was cutting seemed to rough out very nicely, but when I tried to do a "finish" pass with tighter resolution and percentage of step over it seemed to drift a little on each pass.  No machine settings where changed between passes. 

I was cutting in constant velocity mode, but with a very small rounding number. 
Title: Re: Backlash compensation
Post by: Hal on July 28, 2012, 01:48:31 PM
Yep backlash issues.

The heavy cuts force the axis to one side. The lighter cuts allow the float of the axis to show up. You need to get machanial backlash out of the mill as possible. This may include the lead screw and nut. Check the gibs. Slop there could show up as backlash in the other axis. Sloppy will give you greater speed.  Any flex in the Z axis?

Try a circle cut. Is it round. Is it round?
Title: Re: Backlash compensation
Post by: Bob La Londe on July 28, 2012, 02:12:03 PM
Over time I have noticed if I leave the Taig fairly sloppy it tends to stay close to the same amount of sloppy for much longer than if I adjust and tighten it up to reduce mechanical backlash.  One of Hoss's videos inspired me to try leaving it that way and instead using backlash compensation.  At his settings the machine would lose position and walk across the face of the work piece fairly quickly.   I used very conservative settings compared to what He had used and it seemed to work ok.  I cut a 1" register pocket with a .250" end mill and I was modestly satisfied with the results.  It was within .003" total tolerance of round.  Then I tried to cut some parts.  With a 1/4" ball end mill the 3D pocket I was cutting seemed to rough out very nicely, but when I tried to do a "finish" pass with tighter resolution and percentage of step over it seemed to drift a little on each pass.  No machine settings where changed between passes.  

I was cutting in constant velocity mode, but with a very small rounding number.  

I can adjust to about to about .002 backlash per axis.  Unfortunately if I do that the machine exhibits excessive wear and loss of consistency within a few hours of machine time.  If I leave the backlash at about .005 to .01 excessive wear or change in backlash takes a about 20 hours of machine time.  I was hoping backlash compensation would work so that I coule leave the machine in that sloppier range and be able to actually complete a longer job between machine adjustments.  Your reply tends to indicate to me that backlash compensation CAN NOT BE USED for this purpose. 

Yes I do know how to adjust the machine. 

Perhaps another member might be able to suggest actual settings for backlash compensation that might work, but thank you for your feedback. 
Title: Re: Backlash compensation
Post by: Jeff_Birt on July 28, 2012, 02:42:48 PM
Quote
Unfortunately if I do that the machine exhibits excessive wear and loss of consistency within a few hours of machine time.  If I leave the backlash at about .005 to .01 excessive wear or change in backlash takes a about 20 hours of machine time. 


Then you are doing 'something' wrong. I set my Taigs so the backlash is under 0.001" and it will stay that way for hundreds of hours. I do have a tendency of moving my fixtures and vise around so that I use the entire length of the screw, not just a few inches in the center. How often are you lubricating things and what type of lubrication do you use?

You are never going to get 0.005"~0.010" of backlash out with open loop software compensation it will just not work as the table is pushed around by the cutting forces.
Title: Re: Backlash compensation
Post by: Bob La Londe on July 28, 2012, 03:11:10 PM
Yes, Jeff.  Whatever you say.  I suspect the answer to this is going to be like your answer to 250IPM rapids for a Taig, but I am willing to play.  

How much cutting force can it stand to stay within .001" for hundreds of hours?
How often do you lubricate the machine to stay within .001" for hundreds of hours?
What kind of lubricant do you use to stay within .001" for hundreds of hours?  
What rapid speed, cutting speed, and acceleration do you set your machine at to stay within .001" for hundreds of hours?  

What answer for all of the above questions at the same time allows you to stay within .001 for hundreds of hours?  


And bear in mind I am not running that overweight chuck of lead Taig calls a spindle.  I am running a spindle that turns 28000 RPM currently and making cuts in aluminum that a 28000 RPM spindle is capable of.  Actually conservative cuts at that speed, but far above what the stock spindle is capable of.  

If I could get the machine to stay within .001" for hundreds of hours (and not babying it at 10% of its "capacity") I would not even need backlash compensation.  

Bear in mind that the roughing pass with much more aggressive (more cutting load) cuts seemed to cut beautifully, but the finish pass with very light cutting loads was the one that wandered out into left field.  

In the long run if I continue to run this machine its going to get ball screws on all axis and probably a totally redesigned Z.  The nuts, and screws have been recently replaced and were broken in from limit to limit for about 2 hours with grease (as recommended by Taig - the grease not the breakin).  My typical job uses from 40-100% of the X and Y working envelope.  Often parts use nearly 100% of both for the over all part, and about 60-90% for intensive machining.  The gibbs are fine, and all the axis bearings were replaced when the screws were replaced.  Since the axis bearings are just cheap Chinese skate bearings I am really impressed that you can keep your machine running at high rapids, and adjusted to under .001" for hundreds of hours just because of the bearings.  


Yes, Jeff.  Whatever you say.  I suspect the answer to this is going to be like your answer to 250IPM rapids for a Taig, but I am willing to play. 

How much cutting force can it stand to stay within .001" for hundreds of hours?
How often do you lubricate the machine to stay within .001" for hundreds of hours?
What kind of lubricant do you use to stay within .001" for hundreds of hours? 
What rapid speed, cutting speed, and acceleration do you set your machine at to stay within .001" for hundreds of hours? 

What answer for all of the above questions at the same time allows you to stay within .001 for hundreds of hours? 


And bear in mind I am not running that overweight chuck of lead Taig calls a spindle.  I am running a spindle that turns 28000 RPM currently and making cuts in aluminum that a 28000 RPM spindle is capable of.  Actually conservative cuts at that speed, but far above what the stock spindle is capable of. 

If I could get the machine to stay within .001" for hundreds of hours (and not babying it at 10% of its "capacity") I would not even need backlash compensation. 

Bear in mind that the roughing pass with much more aggressive (more cutting load) cuts seemed to cut beautifully, but the finish pass with very light cutting loads was the one that wandered out into left field. 

In the long run if I continue to run this machine its going to get ball screws on all axis and probably a totally redesigned Z.  The nuts, and screws have been recently replaced and were broken in from limit to limit for about 2 hours with grease (as recommended by Taig - the grease not the breakin).  My typical job uses from 40-100% of the X and Y working envelope.  Often parts use nearly 100% of both for the over all part, and about 60-90% for intensive machining.  The gibbs are fine, and all the axis bearings were replaced when the screws were replaced.  Since the axis bearings are just cheap Chinese skate bearings I am really impressed that you can keep your machine running at high rapids, and adjusted to under .001" for hundreds of hours just because of the bearings. 


Yes, Jeff.  Whatever you say.  I suspect the answer to this is going to be like your answer to 250IPM rapids for a Taig, but I am willing to play. 

How much cutting force can it stand to stay within .001" for hundreds of hours?
How often do you lubricate the machine to stay within .001" for hundreds of hours?
What kind of lubricant do you use to stay within .001" for hundreds of hours? 
What rapid speed, cutting speed, and acceleration do you set your machine at to stay within .001" for hundreds of hours? 

What answer for all of the above questions at the same time allows you to stay within .001 for hundreds of hours? 


And bear in mind I am not running that overweight chuck of lead Taig calls a spindle.  I am running a spindle that turns 28000 RPM currently and making cuts in aluminum that a 28000 RPM spindle is capable of.  Actually conservative cuts at that speed, but far above what the stock spindle is capable of. 

If I could get the machine to stay within .001" for hundreds of hours (and not babying it at 10% of its "capacity") I would not even need backlash compensation. 

Bear in mind that the roughing pass with much more aggressive (more cutting load) cuts seemed to cut beautifully, but the finish pass with very light cutting loads was the one that wandered out into left field. 

In the long run if I continue to run this machine its going to get ball screws on all axis and probably a totally redesigned Z.  The nuts, and screws have been recently replaced and were broken in from limit to limit for about 2 hours with grease (as recommended by Taig - the grease not the breakin).  My typical job uses from 40-100% of the X and Y working envelope.  Often parts use nearly 100% of both for the over all part, and about 60-90% for intensive machining.  The gibbs are fine, and all the axis bearings were replaced when the screws were replaced.  Since the axis bearings are just cheap Chinese skate bearings I am really impressed that you can keep your machine running at high rapids, and adjusted to under .001" for hundreds of hours just because of the bearings. 
Yes, Jeff.  Whatever you say.  I suspect the answer to this is going to be like your answer to 250IPM rapids for a Taig, but I am willing to play. 

How much cutting force can it stand to stay within .001" for hundreds of hours?
How often do you lubricate the machine to stay within .001" for hundreds of hours?
What kind of lubricant do you use to stay within .001" for hundreds of hours? 
What rapid speed, cutting speed, and acceleration do you set your machine at to stay within .001" for hundreds of hours? 

What answer for all of the above questions at the same time allows you to stay within .001 for hundreds of hours? 


And bear in mind I am not running that overweight chuck of lead Taig calls a spindle.  I am running a spindle that turns 28000 RPM currently and making cuts in aluminum that a 28000 RPM spindle is capable of.  Actually conservative cuts at that speed, but far above what the stock spindle is capable of. 

If I could get the machine to stay within .001" for hundreds of hours (and not babying it at 10% of its "capacity") I would not even need backlash compensation. 

Bear in mind that the roughing pass with much more aggressive (more cutting load) cuts seemed to cut beautifully, but the finish pass with very light cutting loads was the one that wandered out into left field. 

In the long run if I continue to run this machine its going to get ball screws on all axis and probably a totally redesigned Z.  The nuts, and screws have been recently replaced and were broken in from limit to limit for about 2 hours with grease (as recommended by Taig - the grease not the breakin).  My typical job uses from 40-100% of the X and Y working envelope.  Often parts use nearly 100% of both for the over all part, and about 60-90% for intensive machining.  The gibbs are fine, and all the axis bearings were replaced when the screws were replaced.  Since the axis bearings are just cheap Chinese skate bearings I am really impressed that you can keep your machine running at high rapids, and adjusted to under .001" for hundreds of hours just because of the bearings. 


Yes, Jeff.  Whatever you say.  I suspect the answer to this is going to be like your answer to 250IPM rapids for a Taig, but I am willing to play. 

How much cutting force can it stand to stay within .001" for hundreds of hours?
How often do you lubricate the machine to stay within .001" for hundreds of hours?
What kind of lubricant do you use to stay within .001" for hundreds of hours? 
What rapid speed, cutting speed, and acceleration do you set your machine at to stay within .001" for hundreds of hours? 

What answer for all of the above questions at the same time allows you to stay within .001 for hundreds of hours? 


And bear in mind I am not running that overweight chuck of lead Taig calls a spindle.  I am running a spindle that turns 28000 RPM currently and making cuts in aluminum that a 28000 RPM spindle is capable of.  Actually conservative cuts at that speed, but far above what the stock spindle is capable of. 

If I could get the machine to stay within .001" for hundreds of hours (and not babying it at 10% of its "capacity") I would not even need backlash compensation. 

Bear in mind that the roughing pass with much more aggressive (more cutting load) cuts seemed to cut beautifully, but the finish pass with very light cutting loads was the one that wandered out into left field. 

In the long run if I continue to run this machine its going to get ball screws on all axis and probably a totally redesigned Z.  The nuts, and screws have been recently replaced and were broken in from limit to limit for about 2 hours with grease (as recommended by Taig - the grease not the breakin).  My typical job uses from 40-100% of the X and Y working envelope.  Often parts use nearly 100% of both for the over all part, and about 60-90% for intensive machining.  The gibbs are fine, and all the axis bearings were replaced when the screws were replaced.  Since the axis bearings are just cheap Chinese skate bearings I am really impressed that you can keep your machine running at high rapids, and adjusted to under .001" for hundreds of hours just because of the bearings. 

Yes, Jeff.  Whatever you say.  I suspect the answer to this is going to be like your answer to 250IPM rapids for a Taig, but I am willing to play. 

How much cutting force can it stand to stay within .001" for hundreds of hours?
How often do you lubricate the machine to stay within .001" for hundreds of hours?
What kind of lubricant do you use to stay within .001" for hundreds of hours? 
What rapid speed, cutting speed, and acceleration do you set your machine at to stay within .001" for hundreds of hours? 

What answer for all of the above questions at the same time allows you to stay within .001 for hundreds of hours? 


And bear in mind I am not running that overweight chuck of lead Taig calls a spindle.  I am running a spindle that turns 28000 RPM currently and making cuts in aluminum that a 28000 RPM spindle is capable of.  Actually conservative cuts at that speed, but far above what the stock spindle is capable of. 

If I could get the machine to stay within .001" for hundreds of hours (and not babying it at 10% of its "capacity") I would not even need backlash compensation. 

Bear in mind that the roughing pass with much more aggressive (more cutting load) cuts seemed to cut beautifully, but the finish pass with very light cutting loads was the one that wandered out into left field. 

In the long run if I continue to run this machine its going to get ball screws on all axis and probably a totally redesigned Z.  The nuts, and screws have been recently replaced and were broken in from limit to limit for about 2 hours with grease (as recommended by Taig - the grease not the breakin).  My typical job uses from 40-100% of the X and Y working envelope.  Often parts use nearly 100% of both for the over all part, and about 60-90% for intensive machining.  The gibbs are fine, and all the axis bearings were replaced when the screws were replaced.  Since the axis bearings are just cheap Chinese skate bearings I am really impressed that you can keep your machine running at high rapids, and adjusted to under .001" for hundreds of hours just because of the bearings. 

As to backlash....  it wasn't just a matter of eliminating backlash by software, but the fact that the cut drifted and continued to drift on every cut pass.  The net result would have been satisfactory if turning backlash compensation on had not caused the machine to drift out of position.  Every single cut pass was off from the previous one by a small amount.  The cumulative loss of position was pretty bad.  With backlash compensation turned off this does not happen. 




Title: Re: Backlash compensation
Post by: Bob La Londe on July 28, 2012, 03:33:29 PM
Oops.  Not sure how that happened, but here:  


Yes, Jeff.  Whatever you say.  I suspect the answer to this is going to be like your answer to 250IPM rapids for a Taig, but I am willing to play.  

How much cutting force can it stand to stay within .001" for hundreds of hours?
How often do you lubricate the machine to stay within .001" for hundreds of hours?
What kind of lubricant do you use to stay within .001" for hundreds of hours?  
What rapid speed, cutting speed, and acceleration do you set your machine at to stay within .001" for hundreds of hours?  

What answer for all of the above questions at the same time allows you to stay within .001 for hundreds of hours?  


And bear in mind I am not running that overweight chuck of lead Taig calls a spindle.  I am running a spindle that turns 28000 RPM currently and making cuts in aluminum that a 28000 RPM spindle is capable of.  Actually conservative cuts at that speed, but far above what the stock spindle is capable of.  

If I could get the machine to stay within .001" for hundreds of hours (and not babying it at 10% of its "capacity") I would not even need backlash compensation.  

Bear in mind that the roughing pass with much more aggressive (more cutting load) cuts seemed to cut beautifully, but the finish pass with very light cutting loads was the one that wandered out into left field.  

In the long run if I continue to run this machine its going to get ball screws on all axis and probably a totally redesigned Z.  The nuts, and screws have been recently replaced and were broken in from limit to limit for about 2 hours with grease (as recommended by Taig - the grease not the breakin).  My typical job uses from 40-100% of the X and Y working envelope.  Often parts use nearly 100% of both for the over all part, and about 60-90% for intensive machining.  The gibbs are fine, and all the axis bearings were replaced when the screws were replaced.  Since the axis bearings are just cheap Chinese skate bearings I am really impressed that you can keep your machine running at high rapids, and adjusted to under .001" for hundreds of hours just because of the bearings.  

As to backlash....  it wasn't just a matter of eliminating backlash by software, but the fact that the cut drifted and continued to drift on every cut pass.  The net result would have been satisfactory if turning backlash compensation on had not caused the machine to drift out of position.  Every single cut pass was off from the previous one by a small amount.  The cumulative loss of position was pretty bad.  With backlash compensation turned off this does not happen.  
Title: Re: Backlash compensation
Post by: Bob La Londe on July 28, 2012, 03:36:54 PM
P.S.  When I say hours I mean hours of run time cutting parts.  Currently I think I have over 20,000 hours of run time on this Taig.  
Title: Re: Backlash compensation
Post by: Jeff_Birt on July 28, 2012, 05:14:14 PM
Bob there is no need to be snide. I'm trying to help and by saying "you doing 'something' wrong" I'm meaning to imply that your results are not typical. (I was not trying to be rude.)

I do have to doubt your 20,000 hour estimate though, that would be 833 days of use 24 hours a day. So unless you have run your machine for more than 2.5 years non stop, or five years twelve hours a day then your estimate is a bit high. I think though that your point is that you run your machine a lot and judging my all the neat molds I have seen pictures of I would guess you do run it many hours.

I'm also not sure what your getting on about with your 250 IPM quip, again rudeness is not required. As is clearly stated in the video and in what I have written about it is that the point was to show big stepper motors are not required to obtain high speeds. As I said in the video it is not recommended.

For the last six months I've had my machine set for 85IPM rapids and generally cut anywhere from 4~30IPM in aluminum. I work mostly with aluminum but do plastics, steels and PCBs on occasion. Acceleration is about 10 i/s/s. Before that, for the last three years or so, it has been set to rapid around 50 IPM. I raised it to 85IPM when helping a customer tune his machine for best performance when doing lithopanes. With this machine I lube the leadscrews and ways about every 8 hours with automatic transmission fluid. On occasion I'll run an exercise program to run each axis from limit to limit to make sure the lubrication is well distributed.

I have another Taig with the MM2000 controller that is not capable of the high rapids but its cuttings speeds are similar. It is set up with flood cooling so it gets lubricated about twice as often. Both machines use the stock spindle motor and spindle. Both machines use bits from teeny-tiny sizes to 3/8"

As I recall Bob you do a lot of 3D profile work which can take many hours of continuous running. If your running these long sessions you would want to stop every now and again to lube things up or fit your machine with an automatic oiler. This is what I was getting at asking about how often you lubricate the machine and with what sort of lubrication. Lack of lubrication will wear things out in a hurry and this would be my first guess as to a cause. Cutting a small aluminum mold could take 20-30 hours and if your not stopping to relube every 4-5 hours you will wear things out quickly.

Large stepper motors can put more strain on drive components. Some folks have problems on some small machines breaking stepper couplers that were designed for 150oz-motor because they are using much larger motors. If your spindle motor is much more powerful than the stock Taig unit and your combining that with large stepper motors to get fast cutting rates with a large cutter than you could just be applying more force to the leadscrew/nut than they were designed to handle. This may be a contributing factor but I'm still guessing that it is lack of lubrication.

You could fit an automatic oiler to your machine but it is sounding like you might be growing out of it. A small Haas or similar machine might be better suited to the type of work you are doing and your need for a stronger spindle.

Good luck...
Title: Re: Backlash compensation
Post by: Bob La Londe on July 28, 2012, 05:43:21 PM
20,000 is a guess, but at one point I had the log in Mach showing over 13,000 hours.  That was the second computer on the machine.  Its now on its third computer.  I have had the machine for about 4+ years.  I'ld have to go back into my accounting software and see when I purchased the machine.  I have definitely outgrown the machine, but I just can't afford a Haas.  I am considering a Syil X5 Speedmaster (with the optional 35K spindle) though.  I would have to give up a little X axis, but the design is definitely a lot more robust than a Taig.  I did talk with some of the guys at Syil (China) and I might be able to tweak the machine by running dual limits or disabling limits for certain jobs to maintain the 12" X limit I really need.  The Syils now have auto oiling built in to all mills so that's a plus. 

I was also considered a Tormach 770, but I would have to back off on production (10K max spindle) or plan on buying a couple of their speeders to go with it.  Their lack of concentricty is not an issue since I probably wouldn't do both low and high speed machining on the same part.

Anyway, the practical answer is probably that I just can't get any more out of the Taig, and I have probably gotten a lot more out of it than most folks. 

In this case though, I just want to know why it started losing position incrementally when backlash compensation was turned on (with some pretty conservative settings).  Also why it did not lose position with the first very aggressive roughing MOP.  Full depth (less roughing clearance) in one depth increment with a 1/4" ball mill.

I run flood transmission fluid for coolant and cutting oil on aluminum by the way.  Even if it washes away my grease its a lubricant in its own right. 
Title: Re: Backlash compensation
Post by: Bob La Londe on July 28, 2012, 05:55:46 PM
The radical change just blew me away...
Title: Re: Backlash compensation
Post by: Jeff_Birt on July 28, 2012, 06:33:10 PM
With software backlash comp the controller is guessing about how to compensate for the backlash. Let's same you home your machine, it runs toward the home swtich, hits it and then backs off. At this point Mach knows which side of the backlash will be on (toward the home switch.) Anytime the axis reverses Mach will throw in some extra steps on that axis to take up the backlash. There are two big problems with this approach:

1) The amount of backlash you program into Mach must be exact. If you machine has more or less backlash than you program in Mach will be adding the wrong number of steps at each axis reversal.
2) The forces of the cutting process tend to push the table around so there is no way to tell where an axis really is so the backlash compensation will likely make things worse. If you have every been climb milling on an old manual mill you have likely seen the table jump like this.

I had a chance to meet the guys from Tormach at the CNC Workshop. Nice guys and nice machines. With a larger, more rigid machine that that you can take much larger cuts than you can on a Taig. Unless your using only really small bits I wouldn't guess that the 10K spindle speed would be limiting.
Title: Re: Backlash compensation
Post by: Bob La Londe on July 28, 2012, 07:06:57 PM
It may be "guessing", but its guessing by the amount you tell it.  As you can see by the example posted it wasn't just "guessing".  It was losing position by a little bit on each pass.  It was either exceeding the limits programmed (most likely acceleration) and losing steps or it was adding consecutively more steps in one diraction than in the other.  That just doesn't make sense.  Its ok to say you don't know too.  

My curiosity is peaked though.  I intended to try the same thing in axact stop although its much slower, and again with backlash speed set at a lower percentage of machine speed. 


I suspect either backlash compensation doesn't play well with CV mode or that somehow it doesn't use the standard accleration set when adjusting the screw for backlash. 

I would note the Y is the axis with the worst backlash, and always has been the one to go bad first and fastest under use.

Title: Re: Backlash compensation
Post by: RICH on July 30, 2012, 12:12:31 AM
Bob,
All I can say is that if one wants to do accurate 3d work then minimise / eliminate backlash and have the right components for the machine.
I found doing 3D with backlash is like playing pool with a noodle for a cue stick........... ???
Same goes for doing accurate lathe  / threading work.

Carry on folks...... ;)
RICH
Title: Re: Backlash compensation
Post by: Bob La Londe on July 30, 2012, 09:20:53 PM
Rich, I know there is no substitute for a tight accurate machine.  It still doesn't explain the radical failure and progressive loss of position illustrated. Especially when it cut the previous pass perfectly.  No changes to machine settings between passes, and yes cranking down the machine as best I can without using backlash compensation IS the way I have been doing 3D work.  Given the nature of the feedback I have received though I am going to have to conclude that BC as it is now is a waste of time.  A sloppy or rough cut would be expected, but what was shown is a progressive loss of position.  Backlash was measured with a dial indicator before experimenting with the compensation feature.  

Jeff, I don't know how you are getting sub .001" backlash with stock bearings and plastic pin couplers.  I finally broke down and ordered some angular contact bearings to replace the skate bearings.  I already converted the Z to helical coupler and will convert the X&Y when I convert the machine to ball screw.  (I have screws on the way.)

Tormach... push comes to shove I use a lot of .125 and smaller cutters for fine detail work.  There is a huge difference in material removal rates between 10K and 30K.  Even a 1/4" cutter likes speeds upto 24K.  


Title: Re: Backlash compensation
Post by: Jeff_Birt on July 31, 2012, 12:20:04 PM
Bob, I have no idea why you would not believe me about what backlash I'm seeing. I'm telling you what I see on multiple machines which is also what Taig says it will so. It sounds like your just pushing things beyond their mechanical design.
Title: Re: Backlash compensation
Post by: Bob La Londe on July 31, 2012, 01:24:20 PM
Bob, I have no idea why you would not believe me about what backlash I'm seeing. I'm telling you what I see on multiple machines which is also what Taig says it will so. It sounds like your just pushing things beyond their mechanical design.

Obviously because I am a total idiot and incapable of doing simple experiments like cranking down everything until it stalls and then backing it off enough to run even to do a simple no cutting load test of backlash under ideal circumstances.  

I did exactly that yesterday.  I started by turning backlash comp off obviously.  No load.  Cranked down nuts until it stalled, then backed them off.  Did the same thing with gibbs.   Mobil way oil on the saddles and spindle oil on the screws to give it the best chance of good results.  Found there was a some backlash in the bearings so I replaced them with exactly the same part number bearing (I had spares on hand) and found there was still some back lash in the bearings.  Sorry, cheap skate bearings are going to vary, but I knew that.  Then I did some rapid change tests and found with everything tightened just so and well lubricated there was still about .0005 to .0007 just in the plastic pin couplers.  Less if I loosened up the table, but then what's the point.  LOL.  

I also did some cycle tests to make sure of repeatability.  The machine was not losing steps.  With a hundred cycles on each axis repeatability was sub .001" while I was testing.  Tests were run travel limit to travel limit on X & Y.  

I took your word that is "possible", but I still doubted the hundreds of hours claim.  With relatively new (less than a couple hundred hours) screws and nuts  and new bearings it checked at .0015 on X and .0025 on Y apx with everything as tight as I can make it without losing steps.  Checked with two different dial indicators.  One was a Starret.  One was a Shars import.  

Now maybe I have a lemon, or maybe I am really am an idiot, but after spending all afternoon experimenting (not the first time) I do have some difficulty with the the sub .001 for hundreds of hours without adjustment claim even with light cutting loads.  I'ld be hard pressed to believe it is possible on this machine atleast even with virtually no cutting loads.  I can probably get it under .001 with ACBs and helical flex couplers, but I would still be doubtful about being able to keep it there for hundreds of hours without adjusting it.  I could be wrong though.  I hope so.  The worst of the backlash does appear to be in the bearings.  Since I couldn't find ACBs that small with grease seals I'll have no choice but to set up some kind of oil mechanism on them.  

I would note that the Z axis way plate on the machine was never straight from the factory, and I had to shim it in order to get it close.  Maybe I did get a Monday morning Taig, but the the design just doesn't impress me that its capable of hundreds of hours of actual cutting with sub .001 backlash with stock components without having to readjust it.  

When I got your "it must be you" comment I decided to test it to the best of my ability.  You have setup a bunch of these afterall, and I have setup one.  

BUT NOBODY WHO HAS RESPONDED TO THIS HAS YET TO GIVE A GOOD ANSWER ABOUT WHY THE BACKLASH COMPENSATION CAUSED THE RESULTS I GOT.  Which was the original point of the thread.  I do have an idea about what could cause it if it were machining in mixed milling mode with lots of short (high resolution) reversals on an axis, but this was not the case.  The cut shown was done in conventional mode with each cut pass only having short move reversals on the Z which amazingly only dialed in at .002 backlash before the cut.  I have not checked it after because I have been playing with the X & Y for this round of testing.  Sure a sloppy machine may jerk and move depending on varying cutting loads, but a consistent loss of position per cut pass in one direction doesn't make sense.  

Added:
I might add in order to isolate causes of backlash I turned the screws by hand for some of my tests. 



Title: Re: Backlash compensation
Post by: Bob La Londe on August 10, 2012, 03:52:13 PM
Well, Jeff, I might owe you an apology.  I dialed it down to mostly the bearings and the pin couplers.  When I installed my ACBs (with my own spacer so they worked like paired ACBs) I got it down pretty close to .001.  I was on the phone with Taig this morning over an issue with another part and I discovered something.  There is supposed to be a similar spacer between the skate bearings that Taig uses that was never installed on my machine that makes the skate bearings perform similarly to paired ACBs.  It was never there on any of my 3 axis. 
Title: Re: Backlash compensation
Post by: Jeff_Birt on August 14, 2012, 12:56:44 PM
Glad to hear you are getting it sorted Bob. I figured something was amiss as what you were seeing was different that the norm on Taig machines. I would not have guessed as the spacer plates being missing though. On the coupling straws I have found that keeping the gap less than the recommended 1/16" really helps. Also make sure that the end of the leadscrew (or nylock nut) is not actually touching the stepper shaft. Sometimes it takes me a few tries to get the coupler half on the stepper motor in just the right place so as to ensure a small gap between them. If they touch you can actually put enough pressure on the stepper shaft to cause problems. I have also tried a few alternative couplers be have never been happy with the results.