BIG Dro Problem

Discussion on all milling machines vertical & horizontal, including but not limited to Bridgeports, Hardinge, South Bend, Clausing, Van Norman, including imports.

Moderators: GlennW, Harold_V

Mr Ron
Posts: 2126
Joined: Tue Dec 22, 2009 12:36 pm
Location: Vancleave, Mississippi

Re: BIG Dro Problem

Post by Mr Ron »

I adjusted the gibs on the mill and that has improved the readings a great amount. Instead of reading .010, it fluctuates only 2 or 4 tho's. I can tighten the table locks to just before where the reading jumps and it is fairly tight. I machined a part for my engine I'm building and was able to measure off from an edge on the X axis and after machining a step, I measured it with my Mitutoyo digital caliper and it was exactly in agreement with the DRO reading, so I guess everything is OK. I took some pictures, but I can't download to my PC,(XP) until i get the bluetooth adapter I ordered. Big problem not so big anymore. Thanks to all who have helped me along the way. Happy machining. :D :D :D :D :D
Mr.Ron from South Mississippi
User avatar
warmstrong1955
Posts: 3568
Joined: Thu Mar 18, 2010 2:05 pm
Location: Northern Nevada

Re: BIG Dro Problem

Post by warmstrong1955 »

I have a Jet JTM-4VS (Bridgeport Clone) I bought new.
Brand new, if I locked either X or Y, either would move a thou, up to two thou. It was under .001" after I snugged up the gibs.
I also found, that if I backed up the handwheel, just a touch, after I got where I wanted to be, before locking an axis, the movement is only a few tenths. The amount it changes, was always in the direction the screw is loaded up. Backin' it up....unloads it.
I think mine is pretty good for a knee mill.

Bill
Today's solutions are tomorrow's problems.
User avatar
WesHowe
Posts: 176
Joined: Sat Sep 19, 2015 9:52 am
Location: Huntsville, Texas

Re: BIG Dro Problem

Post by WesHowe »

I have an RF45 clone mill. It moves 2 tenths pretty much always when I lock either table axis. Locking the head moves as much as 2 thousandths, as well as a slight change in the tilt. I'm not sure that 2 tenths is worth worrying about. As for the tilt, not a problem, I just tram with the head locked. Perhaps I need to adjust the ways for the head to get better results here.

The quill (separate DRO) never seems to change when locked, I think it has a circular collar mechanism for a lock. Having a good (if Chinese) DRO is worth taking a little care when locking axes.

- Wes
Mr Ron
Posts: 2126
Joined: Tue Dec 22, 2009 12:36 pm
Location: Vancleave, Mississippi

Re: BIG Dro Problem

Post by Mr Ron »

Mr Ron wrote: Tue Aug 07, 2018 7:07 pm I adjusted the gibs on the mill and that has improved the readings a great amount. Instead of reading .010, it fluctuates only 2 or 4 tho's. I can tighten the table locks to just before where the reading jumps and it is fairly tight. I machined a part for my engine I'm building and was able to measure off from an edge on the X axis and after machining a step, I measured it with my Mitutoyo digital caliper and it was exactly in agreement with the DRO reading, so I guess everything is OK. I took some pictures, but I can't download to my PC,(XP) until i get the bluetooth adapter I ordered. Big problem not so big anymore. Thanks to all who have helped me along the way. Happy machining. :D :D :D :D :D
I must be losing my marbles. When I stated a DRO fluctuation of 2 or 4 tho's, I misread the display decimal places. When I lock down the table, the fluctuation is "tenth's or .0002 to .0004 not .002 or .004.
I would do a happy dance, but I don't think my heart would stand it.
Mr.Ron from South Mississippi
pete
Posts: 2518
Joined: Tue Feb 10, 2009 6:04 am

Re: BIG Dro Problem

Post by pete »

As already mentioned the table HAS to have some clearances for it to move at all. After your adjustments what you likely seeing is possibly the oil wedge being compressed and forced out as the locks get applied. If the work really did require the extra accuracy, then compensating for that slight table shift would need to be done. But at this point your still taking it on faith the dro is actualy giving accurate readings. I've learned the hard way to never blindly trust anything when real accuracy is needed. And the less you pay the more important it is to double check. Even the manual for my Newall dro points out when very high accuracy is needed then mapping the dro's scales and real world table movement using independant test methods might still be required. There dro's have a built in and programable error compensation setting. Laser measurements were mentioned but that's way outside what most home shops could afford to have done.

Good gauge blocks would be second best I guess, but even those are expensive. Any dro can still have fairly close linear movement checks done if you've got a few larger micrometers that can be trusted by stacking end to end various items on the table like 123 blocks, micrometer setting rods etc if there exact dimensions are known. Any mill is a very long ways from a jig borer or jig grinder for location accuracy of course. But if there's no checks done and there are errors there's no hope of ever compensating for those. How far your tests go depends of course on your expected accuracy. Most seem to think tramming the mills head to the table is all that's needed. It isn't. Buying, borrowing or even making 2 accurate cylindrical squares allows checking the spindle and knee axis plus checking that the Y axis is actualy square to the X axis. Depending on just how far those are out (and there will be some errors) they could change your part location over what the dro says it should be even if there's no major linear errors in the dro.

The more accuracy your trying for the more double checks you have to do to get it. When anything that can move does so there's 6 axis's of movement possible. Those can be seperate or much more likely have a slight combined affect on the accuracy. I found out the hard way after the warantee expired on my BP clone there's a .0015" twist in the ways from end to end, and after a couple of years of owning it a .0015" cup in the table surface showed up. :cry: However the Glacern vises I have use 3 large allen bolts on each side of each vise body to fasten them to the table. Using those the cup get's pulled back to zero. But I would have to remember and shim that cup out for any work held directly to the table. Without doing some in depth checking I wouldn't have known about either problem. .0015" table twist over it's length can mostly be ignored for what I do, but at least I do know it's there. By now I think I've learned enough I'd certainly double check even brand new Bridgeports, Deckel mills and Schaublin lathes if I could ever afford one of them. :mrgreen:
RSG
Posts: 1541
Joined: Fri Feb 04, 2011 9:59 am
Location: Ontario, Canada

Re: BIG Dro Problem

Post by RSG »

Good info Pete!

I just bought a new mill and one of the things I will be getting is a cylindrical square to aid in tramming it in once I place it in my shop.
Vision is not seeing things as they are, but as they will be.
pete
Posts: 2518
Joined: Tue Feb 10, 2009 6:04 am

Re: BIG Dro Problem

Post by pete »

What type of mill RSG? While there's hard copys available through Ebay etc there's also at least one free PDF version of Dr. Schlesingers book Testing Machine Tools if you run a Google search. He pretty much wrote the book (pun intended) about the required tests and most if not all machine tool certificates of accuracy today are still based on his work. If it's a knee or dovetailed column mill the cylindrical square is used and swept in X and y with a good dti while moving the spindle, knee, or head the full travels. Unless a guy has lots of hand scraping experience there's little that can be done about the knee being at less than a true 90 degrees to the spindle or even twisted. A dovetailed column mill can at least be shimmed true if there's no twist in the way's. But that's not always correct either. You may get the column dead on and then have the spindle out in the Y axis. If that happens hand scraping it into alignment is still required.

At-Man Unlimited on Youtube did a really great series on checking the basic machine alignments of a CNC bed mill and most of it directly applies to any mill CNC or not. He maintains that head tramming is about the last step you do not the first and I'd sure have to agree with that. You can check the Y axis with only one cylindrical square against the X axis if you set up a heavy STRAIGHT steel bar bolted to the table and aligned dead square to the X axis travel then butt the square against that bar and then run the Y axis in/out. Depending on what you find it might be better to run the tests when your already in a good or bad mood. :-)

Glenn showed a super clever idea here some years ago about how to check either the spindle, knee or head travel in the Z axis simply by single point boring a 1"-2" hole in 1/2" or thicker plate then sweeping that hole with a dti held in the spindle at different head, spindle or knee elevations. That will directly show any inaccuracy with no cylindrical square needed for that axis.
Mr Ron
Posts: 2126
Joined: Tue Dec 22, 2009 12:36 pm
Location: Vancleave, Mississippi

Re: BIG Dro Problem

Post by Mr Ron »

I did a check on table position with a known gage block and the readings were both on the mark, so I would say my mill doesn't have any "major" error. I do have a big error in my Z-axis though. If I set the Z for a certain depth and after machining it, a measurement with a depth mic will show an error of 3 or 4 thou's. It seems that my cut depth is always less than what I set it for. I do make sure the backlash is out of the screw before raising the table.
Mr.Ron from South Mississippi
pete
Posts: 2518
Joined: Tue Feb 10, 2009 6:04 am

Re: BIG Dro Problem

Post by pete »

Reading a couple of the books Moore Tools published was a real eye opener Ron. Before DRO's, CNC, laser measurement etc were invented they spent vast amounts trying to build as perfect as possible feed screws and nuts for there jig borers and grinders. They got pretty close, and while they could measure the lead and lag errors in the screws into the low millionths range even Moore tools couldn't manage perfection no matter how much was spent. Even if it were possible as soon as the machines were put into service then at those low millionths range wear would shortly throw all that expensive perfection out the window. The first jig borers and grinders only used the dials for table positioning much like we still do with our mills without dros today. They still accomplished some amazingly accurate work because of the high accuracy nuts and screws and the built in accuracy on the rest of the machine. Any wear in those parts was still a large problem. They then added a system of high accuracy dial indicators and a set of high precision length setting rods that were much like a micrometer setting rod today. That gave them an independant measurement system so inaccuracy's in those feed nuts and screws didn't affect the positional accuracy. Moore's jig borers and grinders and maybe a few others during that time period had seperate wells cast into the machines to accept the mounts for those indicators and to hold the length rods straight and square to the table travels. I've even read some high accuracy machines were set up to use standard gauge blocks and the same indicators where really high accuracy was needed. Today's dro's do exactly the same only a whole lot faster and easier plus having built in programs such as bolt circles etc. From my understanding the good CNC machines have built in abilitys to do error compensation because it's still impossible and not really needed now to make super accurate screws and nuts. They get the ball screws and nuts pretty close and then program the remaining errors out. As the machine slowly wears then measurement checks are run and adjustments to that error compensation is done again.

Our hobby level machines are in a much different world against anything Moore, P & W, SIP etc ever made and so they should be given the price difference. I don't recall you mentioning that you bought a 3 axis dro so am assuming your using the knee dial for the Z axis movement. If the Z axis feed screw nut is adjustable I'd double check it and the knee gib. But possibly your Z axis error is caused by a less than quality nut and screw? Off the shelf high quality feed screws and bronze nuts aren't hard to find at almost any accuracy level your wallet can stand. But the cheapest may not be any better than what you already have. At least your error isn't large and is on the plus side so that helps. I've read of more than a few offshore machines having metric screws with imperial dials but your error isn't large enough for that problem.
RSG
Posts: 1541
Joined: Fri Feb 04, 2011 9:59 am
Location: Ontario, Canada

Re: BIG Dro Problem

Post by RSG »

pete wrote: Thu Aug 09, 2018 10:44 pm What type of mill RSG?
I bought the Canadian version of the Precision Matthews 35" mill https://www.precisionmatthews.com/shop/pm-935ts-tv/

I was going to order the unit from PM but talked to my local dealer first and it turned out he brings the same model in so I stuck with local. I take delivery on Monday. So I'll be buying a four axis DRO from DRO Pros for it. I have a 3" x 6" Cylindrical square so I'll be testing it thoroughly as you mentioned in pervious post.
Vision is not seeing things as they are, but as they will be.
John Evans
Posts: 2366
Joined: Tue Jan 20, 2009 9:33 pm
Location: Phoenix ,AZ

Re: BIG Dro Problem

Post by John Evans »

Hope you bought the vari speed model ? For what I do that entails a lot of speed changes you could not give a step pulley mill .
www.chaski.com
User avatar
NP317
Posts: 4557
Joined: Tue Jun 24, 2014 2:57 pm
Location: Northern Oregon, USA

Re: BIG Dro Problem

Post by NP317 »

RSG:
Nice choice on the mill
I highly recommend you get the heavy duty power knee feed, especially if you are going to install a 4-axis DRO.
During installation of the knee DRO scale on my mill, I had to move the knee through its full range at least 10 times!
By hand? No way! Rotator cuff damage Here I Come! At least for me.
Share your set up process with us, please.
~RN
Post Reply