Solving backlash problems on a Sherline mill is a common question and users have offered some very thoughtful methods of dealing with it. Simplistically, here we will define "backlash" as the element of "slop" in the screw mechanisms that move the milling machine table or head. When you crank in one direction, and then reverse direction, there will be a short interval while the handle turns (and the handle's scale reading changes) but the table or head does not yet move. If not corrected for, a cutting position error will result. Backlash is present on every make and model of metal lathe or milling machine; some have more, some have very little. The methods used for the Sherline may give you ideas that can be adapted to your own situation. If you got to this file directly from my HOME PAGE, return there by using your browser's back button. BUT if you came to this file as the result of a web search engine, see more than 70 additional files on my home page Machining and Metalworking at Home http://www.janellestudio.com/metal/ SAFETY WARNING BEWARE: DO NOT ASSUME that any subject matter or procedure or process is safe or correct or appropriate just because it was mentioned in a news/user group or was included in these files or on this site or on any other web site or was published in a magazine or book or video. Working with metals and machinery and chemicals and electrical equipment is inherently dangerous. Wear safety devices and clothing as appropriate. Remove watches, rings, and jewellery -- and secure or remove loose clothing -- before operating any machine. Read, understand and follow the latest operating procedures and safety instructions provided by the manufacturer of your machine or tool or product. If you do not have those most recent official instructions, acquire a copy through the manufacturer before operating or using their product. Where the company no longer exists, use the appropriate news or user group to locate an official copy. Be careful -- original instructions may not meet current safety standards. Updated safety information and operating instructions may also be available through a local club, a local professional in the trade, a local business, or an appropriate government agency. In every case, use your common sense before beginning or taking the next step; and do not proceed if you have any questions or doubts about any procedure, or the safety of any procedure. Follow all laws and codes, and employ certified or licenced professionals as required by those laws or codes. Hazardous tasks beyond your competence or expertise should also be contracted to professionals. Let's be really careful out there. (c) Copyright 2003 - 2008 Machining and Metalworking at Home The form of the collected work in this text file (including editing, additions, and notes) is copyrighted and this file is not to be reproduced by any means, including electronic, without written permission except for strictly personal use. ========================================================================== From: "Carol & Jerry Jankura" Subject: Y Slop [sherline] Folks: I've got a model 2000 mill that started out as a manual machine which was then converted to CNC using Sherline's coupler kit and motors. Everything's working well, but I've noticed that I get excessive backlash on Y. (You can move the carriage quite a bit, we're not talking thousandths here.) I can tighten the small nut to eliminate it, but after a while, it comes back. I'm thinking of using something like Locktite, but first thought that I'd ask here to see if there's anything else I should be doing. Carol & Jerry Jankura Strongsville, Ohio ------- Date: Fri, 21 Jun 2002 08:35:42 -0700 From: "Craig Libuse" Subject: Re: Saddle nut > Thanks for the comments, as always, Sherline has excellent support! I > did learn why I couldn't get my old backlash nut an pointer to work. I > had thought I'd removed the slack, but hadn't, so the leadscrew just > backed out of the nut. I'm not sure "what's in the table", but just > running the stepper to "take up the slack" wasn't doing it for me. I'd > almost think there was a sliding nut inside there! Oh well, all's > well now. Alan: Inside the mill saddle is a brass "nut" which is actually round with an internal thread. It is held in place inside the saddle with a set screw. This is the nut against which the antibacklash nut on the outside is tightened to pull backlash out of the leadscrew threads. You can see these nuts in the exploded view as P/N 50200 (X) and 40890 (Y). Makes sure these nuts are held securely with the set screws so they can't move inside the saddle. If you adjust the backlash too tight, the threads on both the nut and the antibacklash nut, which is also brass, wear faster. Eventually they will have to be replaced. CNC use also puts more wear on these components, because the process makes it easier to run the table back and forth more often and at higher speeds than would be possible by hand cranking. You can expect to replace these nuts eventually if you use your machine a lot, especially keep your backlash adjusted fairly tight (.003 or less). By design, the cheapest parts get the wear rather than the steel leadscrew. New X and Y nuts are only $2.00 each. Craig Libuse Sherline Products Inc. ------- Date: Mon, 24 Jun 2002 08:54:06 -0700 From: "Craig Libuse" Subject: Re: HELP!!!! [sherline] June 21, 2002 "santorelli333" wrote: >> I have a(2) problem, the feed on my 5410īs y axis is loose(there is play in the point where the feed-screw is attached to the black disk that holds it(behinde the feed handle)) the result is chatter and poor finish + movement where it apears thet the endmill tries to grab the work-piece and pull it along but the movement is only as big as the give in the feed screw attachement. my second prob is tht the brass bushing that the srew goes throug is worn in its treads(the outer ones that attches the bushing to the mill table, resulting in the backlash adjustment beeing effectively lost) as far as to which threads are worn out the ones in the aluminum (mill-table) seems to be the most likely since brass is stronger than alu... At this point I am thinking "why oh why didnīt I buy a taig-mill" any suggestions to what I can do to the mill to restore it and rid me of such thoughts is desperately sought.....San PS: to pre-empt suggestions about having mistreated the mill....I have seen many an example of sherlines that have the anodizing on the ways almost completely worn off(the anodizing on my ways are 100% and like new) due to religious cleaning and lubrication....and cuts are kept to max 1 mm in alu and 0.2 in iron (using a 10mm carbide end-mill). << Dear San: It sounds like you need a new y-axis nut and antibacklash lock. Fortunately, they are inexpensive and easy to replace. Order P/N 51200 (metric y-axis nut) for $2.00. If you have the older "pointer" style lock, order antibacklash nut P/N 51120 ($.2.00). If you have the newer "star gear" style lock, the part number is 51140. Follow the instructions on page 11 of the instruction manual to remove the backlash. Make sure you pull the saddle fully toward you to take out all space between the handwheel and the thrust before you tighten the handwheel set screw. Try this first before you replace the nuts. By the way, the X-axis nut is held inside the saddle with a set screw. You will have to remove the table to get to it. See the exploded view to understand how it is held in place. Craig Libuse Sherline Prouducts Inc. ------- Date: Tue, 30 Jul 2002 15:05:50 -0700 From: Randy Gordon-Gilmore Subject: Fun with Moglice [A FOLLOW UP TO A RECENT DISCUSSION IN THE SHERLINE GROUP AS TO ELIMINATING LEADSCREW BACKLASH] This weekend I cast Moglice nuts on my Sherline mill leadscrews and my Sherline lathe cross slide screw (due to the way I cast the nuts, I couldn't do the lathe saddle nut at the same time, unfortunately...) I'm extremely happy with the results. The Sherline mill column has a brass nut which kind of floats relative to the head. (The lathe saddle is the same.) The nut is held by a single 10-32 screw, and perched on the points of two setscrews that determine its alignment relative to the head. For the Moglice retrofit, I was not satisfied with this arrangement. Since the Moglice will have next to zero clearance, I wanted a solidly- and repeatably-mounted nut. I measured the distance between the nut mounting surfaces (lateral and fore-and-aft) and the leadscrew centerline. Then I silver-soldered brass shim on both mounting surfaces to bring the nut center in line with the leadscrew center, when it was bolted up tight to the head. I drilled out the nut to leave 1/16" wall, and roughened up the I.D. with a small slotting cutter in my Dremel tool. I turned up centering rings from some Teflon rod to hold the leadscrew tightly and accurately centered in the bore in the column (and thus parallel to the head travel) while I cast the Moglice. For the mill and lathe saddles, I milled out pockets for the Moglice. I used Ron Ginger's idea of using the original brass nuts and "anti-backlash" nuts to contain the Moglice from running out the ends of the bore. The leadscrews run in a clearance hole in the saddle, which is step-drilled to accept the nut, which is just a little cylinder of brass tapped for the leadscrew. I extended the enlarged bore to place the brass nut past the milled-out bore, and filed the flanges off the "anti-backlash" nuts to give the second containment nut. On the lathe, I used a 1/4-20 brass thread insert and turned it to the right O.D. Since the Moglice is described as "fluid", I was imagining how it would flow out any little gap, but was unsatisfied with the idea of using putty in the threads to seal. (By the way, I made the mill saddle recess a little too big, and broke through into the V-way area behind where the gib sits-I did plug that with plumber's putty.) After much thinking during my commute drive, I hit on the idea of using Teflon pipe tape to wind on the threads behind the containment nuts. That worked very well, and no Moglice leaked out at all. The Moglice release spray is wax dissolved in naptha and some other solvents. After degreasing the leadscrews with lacquer thinner and pre-placing the backup nuts, I sprayed two good coats (5-6 seconds per leadscrew) of the wax and let the screws sit overnight. The spray goes on translucent, but dries to a white waxy appearance. I assembled the leadscrews in the saddle bores, and placed the mill head and leadscrew in place on the column. I wound Teflon tape on the column leadscrew in place, and lowered the bored-out nut onto the mass of Teflon. This was enough to seal it. I did all this prep work on Saturday. On Sunday afternoon, I cast the Moglice. The filled component of the Moglice was the consistency of un-homgenized peanut butter-all the solids had settled to the bottom. I pre-mixed it with a Popsicle/craft stick, scraping the sides and bottom of the container well. I then discarded the stick, added the other component and mixed well with a second stick. This way, there would be less of a chance for unmixed resin. The mixed Moglice was about the consistency of honey. I filled the recesses in the mill and lathe saddles and then turned to the column nut. The clearance was so small that I could only apply a fraction of a cubic centimeter (using a small disposable plastic syringe) at a time, and then vibrated it down into the gap by pressing the platen of a palm sander (with the sandpaper removed!) against the nut. I repeated this 4 or 5 times, until the nut seemed full. I went back and vibrated the saddles too. Monday night when I returned from work, I freed the leadscrews. I had read of the difficulties of breaking the leadscrews free initially, and was prepared to use a lot of force. I wrapped layers of brass shim around the leadscrew, and gripped it with a Vise-Grip. To my surprise, the leadscrew broke free easily with relatively little torque (probably not enough to twist the head off a 4-40 screw.) and, after I removed the Teflon tape, the leadscrews all turned freely by hand. I probably was more generous with the release spray than others have been, but I can't feel any play at all by hand. I'm really anticipating reassembling the mill, tramming it in, and measuring the backlash directly. But I have a feeling that it will be well small enough in any case! :-) For the lathe, I could not cast the saddle nut at the same time, because it needs to be solidly bolted to the saddle, and the whole assembly tipped up on end. This conflicted with casting the cross slide nut, so I'll need to buy a second batch of Moglice to do the one nut (well less than a cubic centimeter.) In all, I used possibly a third of the Moglice. The rest, unfortunately, is a hardened mass in the container. If anyone in the area is planning to Moglice anything, let me know and we can share the next batch (as well as the spray release, of which I used a miniscule amount!) The Moglice I used is the 1000 Fluid, which contains Teflon. I'm really looking forward to using the "near-zero-backlash" Sherlines! Photos with captions are in a folder named "Fun with Moglice" in the photos area of the cad_cam_edm_dro Yahoo group. Best regards, Randy ------- Date: Tue, 30 Jul 2002 15:31:27 -0700 From: Alan Marconett KM6VV Subject: Re: Fun with Moglice Hi Randy, Nice work! Don't know if I'm ready to modify my old mill yet, 'tho. I see TWO threaded brass "nuts" per X and Y leadscrew, is that true? I only found one (it came out in the middle of cutting a part, ouch!) on my X axis (17 year old Sears/Sherline mill). The Y hasn't come out by it's self, ...yet. There is ONE setscrew for each (X and Y) axis. The brass "nut" being loose was probably the reason I couldn't keep the star washer "engaged" a few months ago (before U.K. and Squaw Valley trips). Best regards, Alan KM6VV ------- Date: Wed, 31 Jul 2002 13:42:29 -0400 From: "Ron Ginger" Subject: RE: Moglice I did my Sherline with a similar product from Philadelphia Resins. I believe it could be done with any good epoxy resin and either graphite powder or moly power mixed in. The directions say to mix the entire batch at once, but I have used lots of epoxy in boatbuilding, so I carefully measured out the two parts and mixed a small batch. That worked fine for me. My moglice nuts have been on a Sherline for a couple years and they still work fine. ron ginger ------- Date: Wed, 31 Jul 2002 15:43:13 -0400 From: "Van Lupo" Subject: RE: Moglice [WHERE TO BUY?] Well one place is: http://www.moglice.com/ Van ------- Date: Fri, 02 Aug 2002 12:49:27 -0700 From: Randy Gordon-Gilmore Subject: Re: Fun with Moglice At 03:31 PM 7/30/2002 -0700, Alan wrote: >see TWO threaded brass "nuts" per X and Y leadscrew, is that true? The second nut on each axis is the "anti-backlash" nut with the star-shaped flange removed (filed off by hand.) >it's self, ...yet. There is ONE setscrew for each (X and Y) axis. Yes, the recess I machined goes right through the Y setscrew, and the X setscrew is the one on the bottom that I plugged with the cut-down nylon capscrew. >The brass "nut" being loose was probably the reason I couldn't keep the >star washer "engaged" a few months ago (before U.K. and Squaw Valley My nuts kept coming loose too (actually, I'm more often described as having a screw loose!) but were the older style with just a little "pointer" to retain them, instead of the current retainer that is a lot like the safety retainer I've seen for use on externally-wrenching cap screws. Best regards, Randy ------- Date: Fri, 02 Aug 2002 15:02:17 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash, was Fun with Moglice Hi Randy: My "star washer" is on a very short threaded rod. I didn't even recognize yours, which looks like it's as long as the one secured in the middle of the leadscrew hole by the setscrew. The old mill star washer uses a "pointer" to engage it, and is less secure then the newer system that is similar to a pair of small gears (star gears?) that engage. The additional one is secured with a nice allen head cap screw (from descriptions I've studied). I've got a new Deluxe 5400 mill ordered, with a bigger y axis, spindle riser block and the improved anti-backlash hardware. I'm anxious to get it running CNC. Then I'll retire my old mill to manual work, or possibly sell it (list, or ebay). I'm of the opinion that the additional throat will enable me to get the rotary table in the proper position to allow me to cut timing belt pulleys (I've written a program to calculate them), and allow a little bigger work to be handled. Best regards, Alan KM6VV P.S. The mill paid off, I was able to cut some custom aluminum plates for modules we manufacture for our "hand held automotive test computer", which makes me more useful ($$) to the company! ------- Date: Sat, 03 Aug 2002 16:37:44 -0700 From: Randy Gordon-Gilmore Subject: Re: Sherline Mill backlash and Fun with Moglice At 03:02 PM 8/2/02 -0700, Alan Marconett KM6VV wrote: >uses a "pointer" to engage it, and is less secure then the newer system >that is similar to a pair of small gears (star gears?) that engage. The Yes, that's it. My memory was wrong about the new system (and after I mistakenly ordered a pair of the new locks a while back!) I have the mill mostly reassembled. An interesting thing happened--the Z-axis Moglice nut, which has never been off the leadscrew, is very free. The X and Y nuts, which have been sitting off the screws, were tighter when I put the screws back in--not more than hand tight, but a significant drag. The Moglice seems to have maybe a little long-term shrinkage as it continued to cure--I only left it to cure a day before disassembling from the leadscrews. I also had forgotten what a pain it is to reassemble and adjust the gibs, to get the axes shake-free but not binding. By tonight I'll have the stepper motors back on and do some trial back-and-forth jogging with CNC Pro. My workbench situation has been complicated :-) by the fact that I just inherited a Techno-Isel gantry table. X-Y-Z movement is about 20x20x4 inches, with about 4 inches clearance under the gantry. It is an older model (what they call Gantry I) with exposed linear shafting, but it does have ballscrews and is fitted with fabric bellows way covers. The leadscrews are metric, so there are 1270 steps per inch (I don't know yet whether the controller is in full or half step mode though.) It's all made from aluminum extrusions screwed together, and should be a great complement to the Sherline. I will buy a Sherline headstock/motor assembly when I have the extra cash. >P.S. The mill paid off, I was able to cut some custom aluminum plates >for modules we manufacture for our "hand held automotive test computer", >which makes me more useful ($$) to the company! That's great, Alan! So far my equipment has not started to "earn its keep" but I hope that day is not too far off... Best regards, Randy -------- Date: Sun, 04 Aug 2002 14:25:12 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash and Fun with Moglice Hi Randy, I'm once again having problems with the 'Z' axis. In working on a casting this weekend, a 1/4" endmill in an endmill holder dug in, continued to cut deeper and deeper until it ruined it. I had it happen more then once, before figuring out that saddle nut was shifting. I've always had a little problem with the 'Z' axis, it never really seems to allow close adjustment when at the bottom of the travel. Sure, I tightened it up again, but I'd really like to better understand it's proper adjustment. When my new mill gets here, I intend to study it, and hopefully a NEW machine in proper adjustment will suggest what's wrong with my 17+ year old Sears/Sherline. I do have a new saddle nut for it, although I'm not so sure that it's something like the leadscrew being out of alignment with the 'Z' ways. I'm hoping the "manual" for the 5400 will give some hints. I have no idea where my old manual might be, or if the Sears version even had the information. There is some on the Sherline website. I am envious! The Techno-Isel gantry table sounds like a great project! I still have the ways and hardware for a 12" x 12" x 3" PCB router table, but I never seen to get back to it. My youngest son was eyeing it, I think he has some more wood projects in mind. SO, I'd better get back to it! Best regards, Alan KM6VV -------- Date: Mon, 5 Aug 2002 06:14:35 -0500 From: "Nance, Tom" Subject: RE: Re: Sherline Mill backlash and Fun with Moglice Alan, I've had that same problem. I'd be cutting along just fine when all of a sudden the z-axis would drop about a 1/32 to 1/16th of an inch. It's a real treat when I'm using a fly cutter! You would think with the weight of the motor and head assembly there would be no way that the thing could fall; it would always be at the "bottom" of the leadscrew thread. I can't tell you how many things I have ruined. Whenever I position the z-axis, I place my hand behind the pulley and haul down on the z-axis to make sure it is "bottomed out". That can't be what the designers had in mind so there's something amiss. I'd sure like to know what is happening. Tom -------- Date: Mon, 05 Aug 2002 12:55:43 -0000 From: "kevin_sedota" Subject: Re: Sherline Mill backlash and Fun with Moglice I have the same basic problem. I've fiddled with it a bit but have never found anything loose or anything that I could explain it. I have about .011 play in my Z axis from where I set the reading to where the tool will pull down to. I just include this measurement in anything I cut where the exact depth is important. ------- Date: Mon, 5 Aug 2002 09:05:42 -0400 From: "Carol & Jerry Jankura" Subject: RE: Re: Sherline Mill backlash and Fun with Moglice Hi, Alan, Randy, Tom: When I bought my model 2000 mill, I found a similar problem, although not to the extent of 0.030 -> 0.060. One problem that I found was that the link between the saddle nut and the saddle of the Z axis actually moved left->right a little bit before the saddle moved up or down. It turns out that, I believe, the gib on the Z-Axis was too tight. I found this out when, after damaging the original gib by running the "free" end into a piece being milled, I had to replace and readjust it. After that adjustment, things worked a lot better. Also, if you're working with a manual mill, I'd suggest that you make sure that axis lock is engaged before making any final cuts. (I know it's tempting not to engage it for many of the roughing cuts, and that's ok as long as the "slop" does not move you past the "finish" surface) Another thing to be careful of is the "play" caused by using the Z handwheel as a surface that holds position. When you move over to CNC, Sherline provides a new leadscrew and leadscrew mount that provides a preloaded backlash free means of fastening the leadscrew to the axis. There's several photos and drawings in the files section of one person's way of building a preloaded Z axis for the manual mill. This might also be a worthwhile project. Jerry ------- Date: Mon, 05 Aug 2002 10:55:20 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash and Fun with Moglice Hi Jerry, It was making a cut along the long 3.5" X axis, and by the time it got to the right, it had dropped about 1/8". This is without any shift of the stepper. The cutting was in a critical area, a 1/8" thick flange. :-< It had been cutting fine before, but I think the tough raw edges of the casting might have aggravated it. It took me some time to figure out what was happening, by that time, the casting was "in the dust bin" (British casting). I have had that happen before, but not this severe. I had just nicely finished the other (bottom) side of the casting. I was through the "skin" of the casting, although I did notice a spot on one corner that "made more racket" when it cut, so I'm thinking it was a tough spot. I can get a replacement for the casting, but the real problem is curing the problem! I intend to carefully observe my new mill when it comes, and see if it could be that the 'Z' axis leadscrew is off axis, and is being pulled to one side at the bottom of the travel. Any thoughts appreciated. Alan KM6VV (wrecked my whole day) ------- Date: Mon, 5 Aug 2002 13:56:26 EDT From: joeb759x~xxaol.com Subject: Re: Re: Sherline Mill backlash and Fun with Moglice Hi Allen, I read your thread with interest as I also have had Z axis troubles. My 5400 mill is one year old with very little wear, if any. When I adjust the depth of cut on the Z axis and then lock the saddle nut while holding the handwheel to prevent it from turning, the depth of cut increases up to .010 at times. I don't know what to do to correct this problem as it varies from one setting to the next. I first noticed the problem after measuring my part. Then I placed a D.T.I. inthe spindle and you can see the increase in depth when I lock the nut. I would not even think of trying to make a cut with a mill cutter without the saddle being locked. I did it once but not twice! Wish I could have been of more help but as you can see I need a little of that also. Cheers, JoeB ------- Date: Mon, 05 Aug 2002 11:20:40 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash and Fun with Moglice Hi Tom, Yes, That's what I think I'm seeing. It sure is infuriating! This hasn't happened for a while, but then, I've been using 3/32" endmills, not this 2 flue 1/4" endmill. As the 'Z' depth was set by the stepper (which is much bigger now), I DON'T put my hand behind to test the 'Z' axis (like I did WAY before I got the CNC). Too bad Craig's not available to reply! I could use a little advice right now! Alan KM6VV -------- Date: Mon, 05 Aug 2002 11:35:50 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash and Fun with Moglice Hi Jerry, A "wiggle"! Yes, I thought I saw that on last week's project just before it got down to the part, but couldn't see anything to cause it. And I was "spot on" for the parts made (well, except the one that the X axis "nut" got loose on). So your observed left->right movement might be an indication. I don't see the gib as being too tight (more like too loose?), In my case, but a worn saddle nut, or miss aligned Z leadscrew as being the cause. The stepper on the Z axis takes care of the handwheel "play", and I put a new leadscrew on the old Sears/Sherline mill when I made the conversion a few years ago. I never did like (or understand) the Z axis adjustment(s). Alan KM6VV ------- Date: Mon, 5 Aug 2002 12:44:17 -0600 From: "Jim Hall" Subject: Re: Re: Sherline Mill backlash and Fun with Moglice Allen, I'm very interested in this thread since I also have Z axis problems although not nearly as severe as what you have experienced. My 5400 Mill is only about a month old. When using the fly cutter, I have found that sometimes a downward Z axis change of a few mils would result in no actual Z axis movement (fly cutter wouldn't remove any material). I might do this two or three times with no Z axis movement. Then on the third or fourth time, the fly cutter would take a big cut. I was locking the Z axis each time. Measurement with a DTI showed that sometimes, after adjusting the Z axis, pushing down on the spindle assembly would cause an additional downward movement of 4 or 5 mils. Since the above experience, I have been putting downward pressure (with my hand) on the spindle assembly as I adjust the Z axis downward. This has reduced (maybe eliminated) the problem. However, I don't understand why the weight of the motor isn't enough to eliminate the problem in the first place. Maybe it's because the weight of the motor isn't directly in line with the Z axis lead screw and therefore applies a torque to it. I'll be interested to hear if anyone has ideas for possible causes or solutions to the problem. Jim -------- Date: Mon, 5 Aug 2002 11:52:17 -0700 From: "Dave Hylands" Subject: RE: Re: Sherline Mill backlash and Fun with Moglice I expect that adding a counterbalance to the Z-Axis and/or moving the motor to the rear would help. I found some example mods to a Sherline Mill here: Shorter Version of link: Dave Hylands Vancouver, BC, Canada http://www.davehylands.com/ ------- Date: Mon, 05 Aug 2002 12:03:02 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash Hi Joe, Well after remembering what I saw on mine after Jerry mentioned his observations, and from what you're saying, I'd speculate that the saddle nut (the big square block riding on the Z axis) is tilting when you tighten the Z axis lock. I wish I had a Z axis lock! And I'm thinking that when my new mill gets here (it should have one), that possibly I can rig it so that I can run some CNC with it locked, and then do a "tool change" to allow me to loosen it for Z moves. Just an idea, I haven't worked out the details yet. The first thing to do is get zeroed in on Z axis adjustments! On my old mill, there are three points of contact with the saddle nut. One in the center is a 10-32 or so screw, and the other two (were) setscrews. These three points of contact are SUPPOSED to keep the nut running true to the leadscrew, without any unnecessary play. In my case, It SEEMS to be pulling the leadscrew over (at the bottom) if I set it up at the top, and so runs too loose. I don't know, I'm still trying to get it back into useable adjustment! Alan KM6VV P.S. Anybody know where a good description of this adjustment is? I've not been successful on several attempts to find it in the "adjustments" section. Cheers! -------- Date: Mon, 5 Aug 2002 12:07:36 -0700 From: gleex~xxica.com Subject: Re: Re: Sherline Mill backlash For what its worth, my Sherline has two long springs on it which apply a constant downard pressure to the spindle. This was done as part of the CNC retrofit that Bill Griffin did to my machine. Perhaps Bill can shed some light on why he did this. Eric ------- Date: Mon, 05 Aug 2002 12:58:37 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash and Fun with Moglice Hi Jim, Sounds like many of us have these problems! My motor (180 oz/in) on a Sherline CNC mount IS directly over the leadscrew, so I don't see why either. Jerry's "wiggle" observation suggests to me that possibly the saddle nut is getting "cocked" on the leadscrew, and thus takes a "nudge" (my rough casting?) to get it dislodged. I don't know! Ok Craig, we need you! Come back from vacation real soon! Alan KM6VV ------- Date: Mon, 5 Aug 2002 16:03:33 -0400 From: "Dan Statman" Subject: Re: Re: Sherline Mill backlash and Fun with Moglice Just as another point of reference. I have an approximately 1 year old Model 5000-CNC that has been used almost everyday for 5+ hours per day. I have NEVER had th problem that you are referring to. And I would be highly sensitive to this problem since I carve very small this in titanium with tiny cutters. An extra 0.005" on the cut will shatter my endmills. Am I just lucky and should I be worried about an upcoming event? Also, I hav never adjusted a single thing on the z-axis since I have been using it. Z-axis backlash is not a concern of mine since I am always downfeeding into the part, so I have never adjusted this axis. On the other hand, I adjust the x-axis backlash everyday before I start. Daniel J. Statman, Statman Designs www.statmandesigns.com ------- Date: Mon, 5 Aug 2002 17:15:32 -0400 From: "Carol & Jerry Jankura" Subject: RE: Re: Sherline Mill backlash Alan: I believe that Sherline's instructions are to move the saddle nut as close to the handwheel as possible, and then adjust the two set screws and hex cap screw so that there's "easy" motion when you turn the handwheel. Jerry |P.S. Anybody know where a good description of this adjustment is? I've |not been successful on several attempts to find it in the "adjustments" |section. ------- Date: Mon, 05 Aug 2002 15:36:56 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash and Fun with Moglice Hi Jerry: No, it's not excessive backlash, poor clamping, or the leadscrew turning. I'm using a 1/4" endmill in an endmill holder. The saddle nut "jumps" somehow! I'm still investigating. The 4017 part doc's describe a saddle nut adjustment. Might help, although I am of the opinion that the leadscrew's possible mis-alignment to the Z ways is part of the problem. Alan KM6VV ------- Date: Mon, 05 Aug 2002 16:12:03 -0700 From: Randy Gordon-Gilmore Subject: Z nut, was Re: Sherline Mill backlash The Z-axis nut is "held" to the carriage in a push-pull arrangement. The capscrew in the center is in tension, flanked by the two setscrews which are in compression. The Sherline instruction sheet is not very clear on the subject, but you can (and must) adjust both the lateral position of the nut, and its tilt or parallelism to the Z ways. When adjusted correctly, the Z-axis leadscrew should be positioned right down the center of the bore in the column, and running the saddle up and down its length won't push it off to one side or the other. By moving both setscrews together, you adjust the lateral position of the nut. By moving one in and one out (in minute but equal amounts) you adjust the "tilt" of the nut without disturbing its lateral position, and thus can make its axis parallel with the leadscrew's axis (that is what the instructions hint at by "minimum amount of drag") The center capscrew is mainly to lock down the adjustment provided by the two setscrews, but you can't tighten it too much or you just embed the setscrew points into the brass nut. The whole thing is pretty much a matter of balance and feel. The Sherline instructions say to move the saddle up close to the handwheel end of the leadscrew, but it is almost better to place it towards the free end, so it is easier to judge how closely the nut is holding the leadscrew centered in its bore (and then run it up towards the handwheel to confirm the adjustment.) On my mill and lathe (made years apart) there is a difference in distance from the nut mounting surface on the saddle to the column center of about .015" Thus the need for the adjustment for a "universal" replacement part. For my Moglice retrofit, I wanted a solid attachment of the nut to the saddle, and that is why I brazed shim material onto the nut so that the capscrew would hold it solidly to the saddle while having the bore centered on the leadscrew (which I held in place in the column with centering rings while I cast the Moglice) I had to put a shim on both the clamping surface, and the surface at right angeles to it, so the nut is solidly in the corner in the saddle. I don't think that this adjustment could account for your 1/8", Alan. There is not that much sideways play in the clearance hole for the capscrew, even if it was just left slack... The only time I have encountered a problem remotely similar to yours, is when I was milling with a 3/32" shank endmill and didn't tighten the collet enough. My cut was over 1/8" deeeper by the end of the cut than the beginning, purely from the endmill shifting in the collet! Best regards, Randy ------- Date: Mon, 05 Aug 2002 16:24:07 -0700 From: Randy Gordon-Gilmore Subject: Re: Sherline Mill backlash and Fun with Moglice 8/5/2002, Alan wrote: >Yeah, a counterbalance is a good thing, in fact, I have one of Randy's >old ones, except I can't mount it without limiting my spindle options. My counterbalance is only an extension of Alvaro Fogassa's motor relocation: http://groups.yahoo.com/group/sherline/files/sherline%20z%20axis.JPG He pointed out very correctly that having the motor weight centered on the ways would reduce friction (especially stick-slip "stiction") But surrouning the column with the spindle drive belt does keep you from being able to tilt the head too (not a problem for me, since I've been tweaking my mill as a single-minded CNC machine.) Best regards, Randy ------- Date: Mon, 05 Aug 2002 16:37:10 -0700 From: mark Subject: Z-axis backlash / unexpected movements 1. Take out ALL your gibs and take the rough spots off them with a nice, sharp file. Because they are a relatively soft composite material, they are compressible. If only 10 percent of the area of a gib surface is touching its corresponding way surface, it will be 10 times as compressible at a given pressure. Then, you will be able to adjust the gibs with far less "grab" for a given level of tightness. This isn't my idea, but it sure made an improvement in my mill. 2. Make sure the headstock and key are tight. Ditto for all the screws holding the ways to the mill column, mill column to base, etc. 3. IF, and only if, you have adjusted the z-axis leadscrew "tracking" with the two setscrews adjacent the lock screw on the z-nut, you can then adjust Z-backlash. In my opinion, it would pay to buy those radial needle roller bearings and washers for all handwheels, not just for the CNC version. This allows you to have lower friction between the handwheel and that silly 1-inch dia disk against which it bears, so that you can push the handwheel against it just a little bit more forcefully, and gain another thou or so in backlash reduction. I got mine down to a "few" thou. Shouldn't be 10 thou or above - but won't likely be 2 thou either. 4. The trick of putting a sawcut half way up the z-nut, then squeezing (or forcing apart) the two portions, is a great idea- not mine either. This has the effect of two opposing nuts, putting a portion of the leadscrew in either tension or compression, and one *should* be able to get down to a couple thou backlash without too great a friction penalty. 5. I discovered that high quality clock oil on the leadscrews makes quite a difference - a few inches/minute in a less-than-optimum-powered CNC situation. /mark ------- Date: Mon, 05 Aug 2002 16:47:38 -0700 From: Alan Marconett KM6VV Subject: Re: Z nut, was Re: Sherline Mill backlash Hi Randy, That's a pretty good explanation of the Z leadscrew and saddle nut! I AM of the opinion that there IS something going on that I'm not aware of! I'm not using a collet, I'm using the endmill holder (tight on the spindle's threads), but perhaps the endmill was "walked" down in the holder while I was cutting. It's a possibility. It could be a combination of the saddle nut getting loosened up by a little vibration (the rough casting), combined with the endmill walking (it didn't do it on the first side). I'll try to examine this whole thing some more as soon as I get time. I DO want to improve the adjustment of the saddle nut, 'tho. I can't afford to mess up another casting! Best regards, Alan KM6VV ------- Date: Mon, 5 Aug 2002 17:13:59 -0700 From: "Marcus & Eva" Subject: Re: Re: Sherline Mill backlash and Fun with Moglice Hi Alan: Are you SURE the Z axis is moving down??? One of the big problems that used to drive milling machine operators crazy (before carbide insert tooling was widely used to hog away material) was cutters slowly crawling out of the collets. Happens more with heavier cuts because the right hand spiral of the flutes exerts a large downward force on the cutter. I've seen guys milling pockets in injection mold bases and finding out that the pocket was 1/8" deeper at the end of the cut than it was at the beginning. Much cursing happened during those times, accompanied by scrapping thousands of dollars worth of blocks, or putting in underlay plates. An easy way to check is to mark the cutter with a felt pen when you start. Mark the column and the head too. You'll soon find out which bit moved when it wasn't supposed to. If it's the head that's dropping then, most likely, the Z axis slide is too tight. Check the gib and check the freedom of the z axis slide over its full travel by disconnecting the leadscrew from the nut and sliding the head up and down freehand. It should have the same drag top to bottom. If it doesn't, it's either burred somewhere or it's not parallel. Burs are easy to find and fix. Non parallelism requires replacing the ways. Cheers Marcus ------- Date: Mon, 05 Aug 2002 18:00:41 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash and Fun with Moglice Hi Marcus, Yeah, I'm starting to suspect the endmill "walking" down in the endmill holder, after reading all the great posts we've had today. And your example of the castings (molds) certainly fits in. This IS a new two flue endmill, and a new endmill holder, and a new material (cast iron) for me to cut. Oh well! I'm still *LEARNING* (painfully $$). I did see some irregular behavior in the Z axis, so I'm inclined to think the Z axis leadscrew IS out of parallel, :<( too. And you say the only thing to do is replace the Z ways? OUCH! As the Sherline stepper motor mounts mount at a limited plate on the top of the Z ways, I'm wondering if it's possible to "true" up this mounting surface? I believe the problem existed way back in the late 80's when I was building robot parts (manually). I haven't confirmed that the lead screw IS out of alignment, but I suspect it is. :<( Cheers, Alan KM6VV ------- Date: Tue, 06 Aug 2002 00:20:14 -0000 From: "tollgrade" Subject: Z AXIS endmill problem. When using a endmill in a collar make sure you use the top of the flat surface on the end mill when tighten up the 10/32 setscrew. If you use any other part of the flat surface and the setscrew loosen up the end mill will slip down cutting more than you want. Caused me problems till I figured it out. Larry M. ------- Date: Mon, 5 Aug 2002 21:23:07 -0400 From: "Carol & Jerry Jankura" Subject: RE: Re: Sherline Mill backlash and Fun with Moglice >I've seen guys milling pockets in injection mold bases and finding out >that the pocket was 1/8" deeper at the end of the cut than it was at >the beginning. Good Question. Alan - how do you mount your mills in the holder? With the lower end of the flat pushed against the set screw or the high end? If it's the lower end (closest to the workpiece), there's probably 1/8 inch of play to let the mill drop down. I think Randy's got the solution to your problem. Generally, I mount my mills so that they're pulled down as far as they'll go before I tighten the set screw. Jerry ------- Date: Mon, 05 Aug 2002 18:53:44 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash and Fun with Moglice Hi Jerry, At the low end. Is that wrong? I think it was Larry that was commenting on this, although I didn't quite get it (and perhaps still don't). WHERE are you clamping? I'd always thought you "backed up" the a bit or endmill against the shoulder, so to speak. It doesn't go IN any further that way (I'm cutting on the end). I hadn't thought of it PULLING OUT before. What is the 1/8" of play to let the endmill drop down? I'll check tonight to see if the endmill has shifted from where I clamped it. I've only had endmills push up (into) a collet before (only other time I've used my 3/8" endmill holder is for the Starret edge finder). This endmill should have been cutting on it's end, only a shallow cut (.005+), unless once it got started cutting on the side (or the tough edges of the casting), it continued deeper? And I didn't think there was anything to "learn" about an endmill holder. The saddle nut and leadscrew, and "all the usual suspects" will be examined thoroughly. Alan KM6VV ------- Date: Tue, 06 Aug 2002 02:01:59 -0000 From: "tauscnc" Subject: Re: Z-axis backlash / unexpected movements Hey Mark and Randy, great posts (still reading them all). From what I recall, Rich was the first one I read that sawed the Z axis saddle and then moved them apart to decrease the backlash. I tried this but was unable to and ended up breaking the saddle nut in half :-0. I later came up with my own way and have found that it's working well. It is also fully adjustable like Rich's. http://www.cuttingedgecnc.com/machinemods.htm (please go to the bottom). later tauseef ------- Date: Tue, 06 Aug 2002 02:29:30 -0000 From: "tauscnc" Subject: Re: Sherline Mill backlash Hey Alan, I just remembered, I am using a shim (broken piece of single edge razor blade, hey it works:) right at the base of the stepper motor mount to tilt it so its 90 degrees to the z axis. I orginally bought a manuel 5410 and converted it to CNC later. I would find no matter what I did, the saddle would stick at the top or bottom depending where I tightened it (it was smoothest in the middle of the leadscrew). I soon found (no saddle nut installed) when I tightend the stepper motor mount it would tilt causing the end of the leadscrew to be off by as much as 1/2 cm! After trying out some shims I got it to tighten down where I could not notice any leadscrew movement (you could file the end of the Z way to make it parrallel). Now its very nice. I also found this to be true on the Y axis and filed the end of the base just a bit (the X axis seems fine). tauseef ------- Date: Tue, 06 Aug 2002 02:35:35 -0000 From: "tauscnc" Subject: Mounting an endmill Hey guys, some very interesting information! I think I get it but want to be sure (and yes alan I'm with you, I didn't think there was much that I had to learn about endmill holders :-). I thought you wanted the endmill to be sticking out of the holder the least amount you could. When I use the 3/8" holder I put in the endmill, push it up as far as it will go, and then tighten the setscrew in the flat. Now I see, if I do this, when I am milling the endmill is *Pulled Out* in stead of being pushed in (makes sense) as it grabs on the material. Is this the right way? how about on Collets? Should i just put the endmill in as far as it will go and then tighten? thanks tauseef ------- Date: Tue, 6 Aug 2002 01:14:19 -0400 From: "Carol & Jerry Jankura" Subject: RE: Re: Sherline Mill backlash and Fun with Moglice |At the low end. Is that wrong? I think it was Larry that was |commenting on this, although I didn't quite get it (and perhaps still |don't). WHERE are you clamping? I'd always thought you "backed up" the |a bit or endmill against the shoulder, so to speak. Hi, Alan: Generally, I pull the end mill down, so it's riding on the top edge of the flat. BTW, I checked my Z leadscrew. It was difficult to measure using a caliper, so I used the sheets in a pad of paper. The right hand side (as you look at the leadscrew from the back of the mill) of the leadscrew measures 0.116 inches from the RH edge of the circular groove that contains the leadscrew. This is true whether the saddle nut is at the top or bottom of the screw, indicating that the leadscrew is parallel to the way. The paper is nominally 0.0035 thick, so I'd guess that the tolerance on the measurement is 0.007 or so. |And I didn't think there was anything to "learn" about an endmill |holder... Amazing what we all have to learn, isn't it. I think that's one of the best things about this list - everyone knows something more than everyone else, at least on one question, so we continue to learn from one another. |The saddle nut and leadscrew, and "all the usual suspects" will be |examined thoroughly. I'll be interested in hearing your results. Jerry ------- Date: Tue, 6 Aug 2002 09:18:08 -0700 From: "Marcus & Eva" Subject: Re: Re: Sherline Mill backlash and Fun with Moglice >And you say the only thing to do is replace the Z ways? OUCH! Hi Alan: What I was referring to is worn WAYS, not leadscrew misalignment. What you need to do is isolate the possible sources of problems one by one. The easiest place to start is to disconnect the Z axis leadscrew, either from the nut, or from the top of the ways. Then you can check the Z axis movement for smoothness. Drop the Z axis down as low as you can, and snug up the Gib until you can just smoothly slide the head. (I use the bottom of the Z axis for reference, because it's most likely to be worn from having chips slapped against it all its working life.) Now try to run the head all the way up just by pushing on it. If it won't go evenly, you have enough column wear that you can't adjust the gib to a position that's tight enough to prevent rattle at the loose places and sticking with unpredictable movement at the tight places. If all checks out at this step, then you can move on to the next step with confidence. While you're here, by the way, check that the cross hole in the gib is nice and snug on the locking wire, and that the setscrew holding the wire is nice and snug. (A loose gib will act as a wedge, and jam the slide occasionally.) The next step is to assess sources of unwanted movement in the leadscrew and nut assembly. The normal condition is that the weight of the head holds the nut down against the screw and is restrained from dropping by the screw. Of course, the nut is supposed to be a rigidly connected part of the head. If it's not, you will get an effect similar to that which you'd have with a worn nut...ie excessive backlash. Looseness in the mounting of the nut to the head is NOT likely to make the head stick and drop,(unless the gibs are too tight too) but it will make the whole head bounce around when a moderately heavy cut is attempted without locking the head to the Z axis ways. The next thing to check is the amount of looseness in the mounting of the leadscrew to the Z axis ways. Looseness here will also manifest as excessive backlash and will have the same consequences as a loose nut or a worn nut. The head will hop around under heavy cuts if it's not locked. The next thing to check is whether the leadscrew and nut are axially aligned. The easy and really only relevant way to do this is to run the head all the way up with the nut loosened slightly (so that it can find its own alignment. When you've reached the top of the stroke, snug down the nut. If the assembly runs hard, the nut is in the wrong place and is trying to bend the screw, or else the screw is canted relative to the ways and needs to be realigned by filing the top of the ways and deburring the mounting surfaces. I'd be very surprised if the screw is out of axial alignment, because the operation to make the seat is easy to do accurately, so it's difficult to imagine how the factory could screw it up. Misalignment of the screw is easy to see and the criterion for precision is not too high. Typically, the screw will wiggle a bit, side to side at the free end, and this is perfectly normal. The best way to fix nut misalignment, is to pot the nut in something like JB Weld. Open out the nut mounting holes a bit, so that there is no chance of the mounting screws hitting the sidewalls of the nut, and pulling it out of alignment. Clean and degrease the mating surfaces of nut and mounting seat on the head. Lube the mounting screws with grease. Slap on the JB Weld and assemble the nut to the head with the head at the TOP of its stroke, and the leadscrew securely in its proper position. DON"T reef down on the mounting screws...you want the nut to self align to its best position. Once the JB Weld is hard, you can tighten down the screws. Alan, these are really the only things that I can think of that could contribute to the problems you and others have experienced. One last thing to watch though...If your head runs very free on the ways and you are taking heavy interrupted cuts (such as using a flycutter) you can rattle the head down just from the harmonics shaking the screw and rotating it. So with a manual mill you really should lock the Z axis when you're taking a heavy bite. With a stepper driven rig, the motor will hold the screw immobile until you command a move. Hope all this helps Cheers Marcus ------- Date: Tue, 06 Aug 2002 10:20:30 -0700 From: Randy Gordon-Gilmore Subject: RE: Z-axis backlash / unexpected movements 8/6/2002, Kevin P. Martin wrote: >Someday I want to replace it with a nut mounting that attaches >to both sides of the saddle and adjusts using the usual >sloppy-clearance-hole method. It will take some care to arrange, since you need to adjust both side-to-side and front-to-back as well as the parallelism of the bore to the column axis. All are subject to tolerance stackup of the assembly. The cantilevered mounting is plenty stiff enough, if you can get solid metal-to-metal contact between the nut and saddle. An alternative to what I did might be to adjust the current nut, then smear in some JB weld or similar as bedding compound behind the nut. Best regards, Randy ------- Date: Tue, 06 Aug 2002 12:24:06 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash and Fun with Moglice Hi Jerry, Well, I did some detective work last night, and found a few things: With out touching things, I got .035 on the left side top of the lead- screw, and perhaps .055 at the bottom. It was visibly not in alignment. With the leadscrew pulled out, I could observe quite a bit of slop in the saddle nut on the threads, so I figure it's shot (and I had already purchased a new one when I got the backlash stuff for the X axis a few months ago). I couldn't tell if the endmill had pulled out, but it had no flat. It could have contributed to the problem. Not a good situation, I'll add a flat. The top end of the Z ways where the Sherline CNC motor mount bolts on is not really a finished surface. It has wide tool marks. The "original" Z axis handwheel drive round body mounts by one countersunk screw from the back, and there is no contact with the top end of the ways, just the round body contacts the ways. With new saddle nut installed on the Z axis leadscrew (no slop now!), I noticed that I could move the motor mount around a little, and thus I found a "sweet spot" where the Z leadscrew was parallel (by dial caliper). I also found that I could run the saddle nut up tight against the Z carriage, and keep the leadscrew parallel to the ways. I basically aligned it "backwards", that is, I matched the top position of the motor mount to agree with the bottom (parallel)! The saddle nut is not in the center (is this a problem?), but it runs up and down smoothly. I might experiment with it more. Do the new mills that are "CNC ready" (as far as the holes for the motor mounts) have this top end of the Z ways finished? I haven't cut any parts with it yet, but it's better then I've ever seen it before! (musta learned something.) I don't see how clamping at the top of the endmill's flat is any better, unless the flat is tapered (must be a reason)? I should think having the minimum length of the endmill extended from the holder would be best, same as for a collet. Thanks guys! Alan KM6VV ------- Date: Tue, 6 Aug 2002 16:25:11 -0400 From: "Kevin P. Martin" Subject: RE: Re: Sherline Mill backlash and Fun with Moglice Alan Marconett KM6VV [KM6VVx~xxarrl.net] wrote: >The top end of the Z ways where the Sherline CNC motor mount bolts on is >not really a finished surface. It has wide tool marks. I ran into this installing my CNC retrofit. No matter how I adjusted the saddle nut, there were tight spots in the motion. I did not notice the cause of this until a few weeks later when I was investigating why my Z leadscrew had snapped next to its bearing. The top end of the Z ways was rough and out of square--it looked like a bandsaw or cutoff-saw cut. The angled bearing meant the screw was flexing continually, effectively acting as its own universal joint! Metal fatigue quickly set in. I explained the problem to Craig, got a free replacement leadscrew, and milled the end of the ways square on my lathe. I seem to recall him saying that earlier production did not have the machined end on the Z ways since it was not originally a reference surface for mounting other parts. Kevin Martin ------- Date: Tue, 06 Aug 2002 13:35:28 -0700 From: Randy Gordon-Gilmore Subject: RE: Re: Sherline Mill backlash and Fun with Moglice 8/6/2002, Kevin P. Martin wrote: >I seem to recall him saying that earlier production did not have the >machined end on the Z ways since it was not originally a reference >surface for mounting other parts. Yes, on both the manual mill and lathe the leadscrew is mounted to a flanged sleeve that registers in the bore in the column/bed. (So does the drill jig for the stepper mounts for that matter, which is why I take the "default" position of the leadscrew to be centered in the bore...) I guess I was lucky--both my mill and brass-bed lathe had good square ends on them, and it never occured to me that it might have been otherwise. It would be good for Sherline to mention this possibility in their CNC conversion instructions. Best regards, Randy ------- Date: Tue, 06 Aug 2002 14:42:45 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash and Fun with Moglice Hi Marcus, Thanks for all your thoughtful and considered answers! Luckily, I don't think my ways are worn, although I DID find a worn saddle nut. And as I posted earlier, I also did determine that my Z axis leadscrew was out a little, and was able to bring it in. I didn't really start at the ways and gib's, so I might just go over it all again. They seem to be moving nicely with the stepper at this time, although with the weight of the spindle and stepper, I couldn't get a good feel for how "easy" it would glide. Much easier to feel it in the X or Y axis. I DO see considerable improvement after replacing the saddle nut and adjusting, hopefully this will work better! Cheers, Alan KM6VV ------- Date: Tue, 06 Aug 2002 15:26:24 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash and Fun with Moglice Hi Jerry, Couldn't the endmill still move down? Although it might get free of the setscrew. That could cause some problems! Is there some reason that the endmill no longer can move downward if clamped at the top? That's what I don't get. I'm all for keeping it from going DOWN! I'm almost wondering if the flat on the endmill shouldn't also be a bit of a V grove, such that the setscrew prevents movement along the axis as well as preventing any rotation. Alan KM6VV ------- Date: Tue, 06 Aug 2002 15:42:35 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash and Fun with Moglice Hi Kevin, That supports my suspicions! And yes, it DOES look like a cutoff-saw cut. I think I got around it this time. I did not notice it when I originally mounted the Sherline motor mounts. And trial fitting my old Z axis leadscrew while doing my "detective work" this time demonstrated that the manual handwheel drive didn't use that surface for contact (it didn't touch it). It didn't need to. I suppose it would be appropriate to go back and mill or file it off, but I believe I've got it working now. Thanks for the comments. Alan KM6VV ------- Date: Tue, 06 Aug 2002 17:53:34 -0700 From: Alan Marconett KM6VV Subject: Re: Z-axis backlash / unexpected movements Hi Randy, OK, 40910 (which I can't find on the exploded view) is called the Saddle, and I did find a "saddle gib", so it follows that what I might have called the "Z axis carriage" in my ignorance is properly called the saddle (and it probably corresponds on the lathe). I did remember to call the part I replaced a "saddle nut" (probably because I ordered one last month). I was just anticipating the possibility of more parts in the newer model! Best regards, Alan KM6VV ------- Date: Wed, 07 Aug 2002 03:00:16 -0000 From: "luisguillermo98" Subject: Re: Sherline Mill backlash and Fun with Moglice This is what I do: I loosen the gibs a little bit and keep the saddle lubricated so the motor can do its job with the weight. After that I have not had any problem at all. Also you can put some extra weight directly on the spindle to keep the spindle on the bottom of the backlash. ------- Date: Wed, 7 Aug 2002 15:44:07 EDT From: joeb759x~xxaol.com Subject: Re: Re: Sherline Mill backlash Alan, I think that the problem has to do with the backlash between threads in the Z axis. As you dial in a setting for depth the axis screw is pulling the head down. When you stop the downward travel you still have the backlash area between the threads giving play in the downward direction. Hence when you begin to cut without locking the axis nut the head will drop that amount. Most often it will also drop when you lock the saddle nut also. I have found that by dialing in a greater cut (.015 instead of what I want .010) then backing up on the setting to the .010 needed does eliminate this problem. I'm still fooling with it, seeing how it goes. Joe B ------- Date: Wed, 7 Aug 2002 16:45:32 -0400 From: "Dan Statman" Subject: Re: Re: Sherline Mill backlash absolutely correct! When dropping down to the depth you are moving the headstock with the bottom of the threads since the threads of the leadscrew are pushing downward. The backlash is the difference between the location of the bottom of the threads and the top of the threads relative to the saddle nut. As you cut the weight of the headstock will randomly shake itself as low as it can go, to the top of the threads. So what to do, measure the backlash and drop the cutter to the desired point and then back up the handwheel the amount of the backlash, then lock the saddle. Let me know how this works. Daniel J. Statman Statman Designs www.statmandesigns.com ------- Date: Wed, 07 Aug 2002 14:36:38 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash Hi Joe, As I mentioned in yesterday's post(s), basically I found a worn saddle nut, and misalignment of the Z axis leadscrew. And I'm sure the endmill in the endmill holder contributed to the "digging in" I experienced. I'm setting the Z depth by the CNC's control, and typically over the part. So as I plunge down, I can't very well go deeper, and then back it up. If I've started the Z axis above the part, and then moved down, then I HAVE "ran out" the backlash. The top surface of the part is "touched off" in Z, and the backlash is thus "ran out" in the downward direction. What I understand you to describe would seem to invite backlash or contain backlash. I've always heard that you remove the backlash by moving IN THE DIRECTION of the cut first (start far enough back so that upon getting where you want to be, the backlash is out). I cannot lock the saddle nut, as all axis are under CNC control. The Stepper motor holding torque and weight should hold it there. And what if one is doing a compound move? Then the Z axis may very well be following a contour, and obviously cannot be locked. Both my X and Y locks are REMOVED from my old Sears/Sherline mill (the old gal never had a Z lock). Hopefully my replacement of a worn out saddle nut, and the realignment of the Z axis leadscrew (now that I understand it a little better) will eliminate THIS problem from my milling. It should be noted that for most users unfamiliar with Sherline's endmill holder (as was I) that the endmills CAN pull themselves down INTO the work. I haven't heard, is this also a problem with collets? The 3/8" and larger endmills with Weldon "flats" apparently need to have the setscrew set against the UPPER taper on the end of the flat (I just learned this, thanks guys!), not just anywhere. I checked in my garage this morning, and sure enough, on a 3/8" endmill for the RF-31, there is a Weldon flat on it! I have no idea where I clamped it down in the R8 endmill holder for the RF-31. And I've been MISTAKEN, I've only used a 1/4" endmill holder (my brand new one) on the Sherline! Seems the only duty the 3/8" endmill holder has had on the Sherline mill is to hold the Starrett edge finder! So although I've been saying (from a failing memory) that I clamped on the flat of the endmill, actually I had really used a 1/4" endmill, which HAS NO FLATS, let alone a nice Weldon flat! Sorry about the false report. Perhaps I can grind a Weldon flat on the 1/4" endmill? OTHERWISE, I'm a bit puzzled, because the endmill holders are made very accurate, and as soon as I clamp down on a round shaft with a setscrew, I'll never get it out again! Yesterday I posted Sherline's list of new endmill holders for smaller size endmills, but I see a problem, HOW DO YOU CLAMP the endmills, Center drills, etc. into them? The edgefingder is fine, it doesn't really need to be "CLAMPED" in tightly. Thanks for the comments, do let us know what you work out Joe. Alan KM6VV ------- Date: Wed, 07 Aug 2002 15:04:27 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash Hi Jerry, Yes, that would probably have been better. I have used collets up 'till now, and thought I was going to have a superior setup by using the endmill holder. Might have been, if I'd used an endmill with the Weldon flat on it. I have had collets get loose during machining, because I didn't tighten them up enough (and severe vibration can loosen them too). Alan KM6VV ------- Date: Wed, 7 Aug 2002 15:36:32 -0700 From: "Dave Hylands" Subject: RE: Re: Sherline Mill backlash > What I understand you to describe would seem to invite backlash or > contain backlash. I've always heard that you remove the backlash by > moving IN THE DIRECTION of the cut first (start far enough > back so that upon getting where you want to be, the backlash is out). This would normally be true, and works well in a horizontal plane. Normally, on a lathe, there's nothing that will try to pull the tool bit into the part, rather the forces are pushing the toolbit away from the part. However, on the Z axis on the mill you have two forces working against you: 1 - Gravity. Since you're pushing the axis down, if everything is tightened up, it's possible that the saddle will be at the "top" of the backlash, rather than at the "bottom". Once you start milling and vibration sets in, gravity helps to let the saddle settle into the "bottom" of the backlash. 2 - End mills will try to pull themselves into the part. I was using a 1/4" endmill the other night, and had not tightened the set screw with the endmill all the way down. I happened to be looking at the endmill at just the right moment and actually watched the endmill suck itself deeper into the part. Sure enough, when I stopped and checked the endmill, the set- screw had loosened a bit and the endmill was down as far as it could go. Fortunatly, I still had lots of metal to eat away and it wasn't cause for a scrapped part. So, I'd speculate that if the gibs are too tight, then the head won't fall to it's lowest position naturally (due to gravity) and vibration may cause it to drop further once milling has started. If the gibs are loose enough, then the head should always ride at the "bottom" of the backlash. I don't know if this will cause other problems or not. You should be able to tell if this is the case by checking to see how much force is required when you change directions on the Z-axis handwheel. Get a feel for how much force is required on the handwheel when moving the head upwards. Then move the head downwards (should require less force). Then change to moving upwards again. Does it immediately require the same amount of force as when moving the head upwards? Or is there a couple thou where it's really easy to turn? If it's the latter, then gravity can pull the head down that extra couple of thou when the mill is operating. Dave Hylands Vancouver, BC, Canada http://www.davehylands.com/ ------- Date: Wed, 7 Aug 2002 20:27:33 -0400 From: "Dan Statman" Subject: Re: Re: Sherline Mill backlash Alan, I really think you are overlooking the obvious cause of the Z-axis dropping 0.005-0.010" during a milling cut. It is the backlash. You are not considering the pull of gravity and the potential pull of the endmill into the material. This type of DOWNWARD motion backlash is not "taken up" by your downward move. A downward move takes up the UPWARD backlash. If you think about the v-shaped space of a thread when you move downward you are rubbing the top of the "v" against the threads of the saddle nut. Therefore with gravity acting downward it is possible for the z-axis to shift to the bottom of the "v" by the amount of your backlash. All of your misalignment of the saddle nut and leadscrew may contribute to other problems, specifically accuracy of your moves if not moving perfectly parallel to the Z-axis ways. But why would any of it cause the headstock to drop. If you were the one see 1/8" drop then check your endmill before and after the move and see if the same distance is sticking out. If you drop 0.005-0.010" then it is backlash. Tightening of the gib may prevent the drop. Or backing the stepper out by the backlash amount right before you cut will also do it. It seems counter-intuitive, but this must be what everybody is seeing. Dan ------- Date: Wed, 07 Aug 2002 17:51:21 -0700 From: Dave Martindale Subject: Re: Digest Number 839 >What I understand you to describe would seem to invite backlash or >contain backlash. I've always heard that you remove the backlash by >moving IN THE DIRECTION of the cut first (start far enough back so that >upon getting where you want to be, the backlash is out). It's more complex than that. The advice you quote is the right way to avoid backlash on the cross slide of a lathe, where the toolbit will "push back" against the leadscrew once cutting begins. It's also correct for "conventional" milling, where the cutting forces push the table back against whatever leadscrew (X or Y) is providing the feed. But it doesn't work in a "climb milling" situation, where the cutting forces tend to *pull* the work into the cutter without the help of the leadscrew. If you have .010 backlash in the leadscrew providing the feed, then the cutter can move the work .010 all at once as the leadscrew goes from pushing the work to holding it back. With the Z axis, you're probably not actively feeding in Z while cutting. But the cutting forces from a normal right-hand spiral end mill want to pull the end mill down into the work, and gravity wants to pull the headstock and motor down too. So if you lower the Z axis and then stop, you've probably pushed the head down with the leadscrew and it can drop further by the amount of any backlash present. That's the problem. One fix that someone suggested is to overshoot in lowering the head and then bring it back up enough to ensure that the leadscrew is lifting the head, and so is prepared to resist downward forces. This doesn't work if there's no place where you can let the endmill drop too far before raising it. Another was to lower to the desired position, then crank backwards just enough to remove the play, but not enough to lift the head. This is probably not practical for CNC where the computer has no torque "feel" from the Z axis drive. Another was to use large springs (strong enough to overcome jib friction) to pull the head downwards at all times, so the leadscrew is always providing a "lifting" force on the head, even when it is being lowered. This method would work even if the Z axis is being driven during machining to follow a contour. The final option is ballscrews and nuts that have no backlash. Dave ------- Date: Wed, 7 Aug 2002 18:11:56 -0700 (PDT) From: Nhut Le Subject: Re: Digest Number 839 I agree that all the discussed factors contribute to the problem at hand, but no one (or I may have missed it) has brought up the role of the Z-axis handwheel. If this isn't kept in check, this could account for the .010 drop in the Z-axis. I've noticed over time that the little setscrew holding the handwheel to the leadscrew will slip and causes the leadscrew to drop lower, introducing even more backlash. Sherline acknowledges this and has instructions to fix the problem: http://www.sherline.com/Zaxisfix.pdf Nhut ------- Date: Wed, 07 Aug 2002 18:21:21 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash Hi Dan, Dave, Thanks again for the explanations. I hadn't been able to visualize gravity as being able to force a cutter further into the work then the stepper. So you both are saying that the 2 or 3 pounds of gravity will essentially add onto the cutting forces, which must be several hundred pounds, and move the cutter further into the work then the stepper does? Hard for me to visualize. But then, I've only thought of backlash while in the horizontal plane (as someone pointed out). Probably from my first lesson on an old South Bend lathe. "Always approach the work piece from a distance far enough away to run out the backlash". Or words to that effect. Hopefully I can improve over this last endmill example by using an endmill with a Weldon flat (3/8"), or going back to the collets. Should I also go back to 4 flue endmills for steel (cast iron)? Seems I just read that today in Sherline's "Vertical Milling Machine Operation". I THINK I've greatly improved the Z axis setup, and of course I've replaced the worn saddle nut. Dave's and Marcus' examples of checks to make for a manual mill may persuade me to remove the stepper, lay the mill on it's side, and experiment more with the gibs (I thought of putting it on it's side, as then gravity is out of the equation). The head dropped probably 1/8", and I don't belive there was that much backlash in Z at the time. Your .005" - .010" I could accept. Have I done enough? Will these steps prevent this from happening again? I really don't want to trash any more work like this! I'm tempted to "load up" the Z axis by increasing the gib tension, and somehow add another nut to limit the leadscrew backlash to the point of causing more friction on the saddle nut, to prevent gravity and vibration from doing as you describe. Best regards, Alan KM6VV ------- Date: Wed, 07 Aug 2002 18:33:06 -0700 From: Alan Marconett KM6VV Subject: Re: Z backlash Hi Dave, Yeah, I'm learning that it's a little more complex then my simplistic quote. I just responded to Dan and Dave H.' fine emails which are in agreement with yours. I believe it was Joe that was describing "coming from below" to get the leadscrew loaded up correctly, although I don't see how to do that with CNC. I am considering adding a second nut, and spring loading it to try to force out any backlash. The "climb milling" example is a good one. I HADN'T EVEN thought of it that way. You mention strong enough springs to override gib friction, why not just INCREASE the gib friction, to prevent it from falling? NO? Best regards, Alan KM6VV ------- Date: Wed, 07 Aug 2002 18:38:10 -0700 From: Alan Marconett KM6VV Subject: Re: Z backlash Hi Nhut, The Z handwheel is not a problem in my case, the handwheel is down in a drawer. ;>) I experienced the 1/8" drop while under CNC control. But your point is well taken. I think when I had a similar problem in manual operation (much before this CNC conversion, and I CERTAINLY didn't understand it then), I used my hand to tension the head, as someone has suggested. Alan KM6VV ------- Date: Wed, 7 Aug 2002 21:32:42 -0400 From: "Dan Statman" Subject: Re: Re: Sherline Mill backlash August 07, 2002 "Alan Marconett KM6VV" wrote: > So you both are saying that the 2 or 3 pounds of gravity will > essentially add onto the cutting forces, which must be several hundred > pounds, and move the cutter further into the work then the stepper > does? Hard for me to visualize. Alan, why would there be several hundred pounds of cutting force? Imagine if you have the milling headstock positioned over the work piece with a spinning endmill. If you could hold the headstock perfectly straight in your hands and lower it onto the work piece, and stabilize it from moving sideways at all (this is what the z-axis ways are doing. the force of gravity alone on a spinning cutter will tend to drop it into the workpiece and drill a hole with the endmill. once you have spun the cutter several revolutions all of the material is gone from under the cutter and it can drop again due to gravity. Only when it hits a physical stop will it stop dropping, this physical stop is the top surface of each leadscrew thread. Remember also that the side loads do not contribute to any of this since we are only concerned with the vertical vector component. The only things contributing to this vertical component are gravity, the leadscrew forces, and the potential for the cutter to grab the work in certain types of milling and certain materials. It is also a simple matter to program a backlash reducing upward z-axis move in a cnc program. With TurboCNC which has backlash compensation you just have to give a move 0.000250" above your bottom point. This will take up the backlash and not move your stepper (if you set the backlash compensation to be 0.000250" under the actual value), or you could probably live with the 0.000250" error. This is the amount of movement from a single step of a 1.8 degree motor with a 20 TPI leadscrew. Dan ------- Date: Wed, 07 Aug 2002 18:47:31 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash Hi Dan, You're right of course! If there is no downward cutting force (the Z axis is not moving), then it all starts to make sense! And of course the forces at right angles can't be added in either. The Z axis is free to float, and this explains why gravity is enough. I think I've finally got it! However I "fell" a lot further then .00025" (or is it .000125", as I halfstep? Microstep drivers?), that I don't have a problem with (I probably wouldn't even know it)! But 1/8"? That's a killer! Are you saying just to a move up by .00025 AFTER moving down to remove this backlash? Alan KM6VV (I've gotta add backlash comp. to my code...) ------- Date: Wed, 7 Aug 2002 21:45:26 -0400 From: "Dan Statman" Subject: Re: Re: Sherline Mill backlash If you tell the controller to move up the equivalent of 1 step, whatever that amount is (I think it is 1/4 of 0.001" since there are 400 steps per revolution when half-stepping) then the controller ACTUALLY moves the programmed backlash amount PLUS 0.00025". This will remove your backlash and only slightly alter your height, if at all. OTOH, if you set the backlash compensation to be 0.00025" less than the actual amount then it will be dead-on when it moves up the COMPENSATED 0.000250" Of course, you can't really measure anything this closely anyway. Dan ------- Date: Wed, 7 Aug 2002 18:50:58 -0700 From: "Dave Hylands" Subject: RE: Re: Sherline Mill backlash Hi Alan, I think that if you move up after moving down, you still run the risk of the cutter going too deep at the point that the Z-axis is being lowered. If you've measured the amount of backlash ahead of time, you should be able to stop that much short of the actual dimension, and let the backlash take out the rest of the way (I'm speaking from a purely theoretical standpoint, since I don't have my CNC working yet). I don't think that the amount of backlash is related to how for the leadscrew moves with a single step. It has more to do with wear and how well threads mesh together. Dave Hylands (who is also intensely interested in this particular topic) ------- Date: Wed, 7 Aug 2002 21:48:45 -0400 From: "Dan Statman" Subject: Re: Re: Sherline Mill backlash I forgot to mention, as I found out yesterday when I stripped my saddle nut. It is very possible for there to be 1/8" of backlash in a badly worn saddle nut and leadscrew. Especially on the z-axis where there is no anti-backlash nut. I use one of my manual Sherline lathes for rough cutting only and it has at least 0.075" backlash. I just do not care at all. When the leadscrew stops moving things I will replace it, and still not care. Leadscrews and saddle nuts are dirt cheap. Dan ------- Date: Wed, 7 Aug 2002 21:55:25 -0400 From: "Dan Statman" Subject: Re: Re: Sherline Mill backlash Dave, I understand that you are not running CNC, so I thought I would explain how it works a little more clearly. When you tell the machine to move down 0.050" into the part, it will move EXACTLY 0.050". If you have backlash compensation (and it is set to 0.010") then you tell the controller to move up 0.00025", the machine will move 0.0102500". This is the compensated backlash amount plus the amount of the move. If you measure the backlash and it is 0.00875" then the compensation will move that exact amount every time the direction of movement changes. So if you set the compensation on the z-axis slightly smaller than the amount of backlash you have when you command the upward move, nothing changes with the cutter and workpiece, but the backlash is removed. And it all happens in a fraction of a second with perfect accuracy and repeatability assuming your machine is set-up correctly. if however, you are doing some complex 3-D shape that required the upward movement of the cutter as it was cutting then you would need the exact backlash amount to be input. And, realistically if you need precision 3-D contours, then you aren't using a Sherline. Dan ------- Date: Wed, 7 Aug 2002 21:58:02 -0400 From: "Dan Statman" Subject: Re: Re: Sherline Mill backlash Download TurboCNC and give it a try. Or, if you are always milling down to the workpiece then you do not need automatic compensation, just program an upward z-move equal to your measured backlash right after the downward z-move, simple solution. You can even do a replace function in a word processor on your g-code files to take care of this automatically, or write a simple macro. Dan ------- Date: Wed, 07 Aug 2002 19:14:41 -0700 From: Alan Marconett KM6VV Subject: Re: Sherline Mill backlash Hi Dave, Dan, I wonder how I EVER got anything cut right in Z before! If the head doesn't drop because of gravity, then cutting down in Z shouldn't involve any backlash, as you are always cutting in the same direction. I know there IS gravity and endmills DO get pulled into the work. Perhaps if I take light enough cuts, so as not to wake up gravity, I'll have less "Z axis drop" to worry about. And I DID find a very worn saddle nut. I really think we should work on a backlash nut for Z, now that I realize we don't have one! Yeah, I like Dan's lathe example. Lotsa backlash, but who cares! You automatically remove it when you back off a tool, move it, and then travel back into the work. No worries! But then, putting CNC on a lathe still seems foreign to me, as it is difficult to predict where the cutting tip will be. I did learn that they use "standard" cutting tips, and they set them up with microscopes. I don't think I'm ready to CNC my 5" lathe yet. And I'm more inclined to agree with backlash having "more to do with wear and how well threads mesh together", otherwise we'd all have .000125 backlash! Alan KM6VV ------- Date: Wed, 7 Aug 2002 19:12:14 -0700 From: "Dave Hylands" Subject: RE: Re: Sherline Mill backlash Hi Dan, That makes sense - i.e. that every time the direction changes then the compensation would kick in. But that still doesn't factor in gravity. I would expect the backlash compensation algorithim to be slightly different for the Z axis than it is for the X or Y axis. For example, if you move the z-Axis down and stop moving (not change directions) you still need to compensate for backlash. So, a move in the downward direction followed by a stop should be considered a "change of direction". Dave Hylands ------- Date: Wed, 7 Aug 2002 19:20:35 -0700 From: "Marcus & Eva" Subject: Re: Re: Sherline Mill backlash Hi Dave: You've nailed all the important points perfectly. On a bedmill (Head moves up and down) the smoothness of the Z axis motion over the whole Z axis travel is important to allow the head to sit at all times on the bottom of the backlash. Worn ways will make the head stick in places and rattle in other places. With regard to endmill holders and plain collets for gripping endmills, cutters will pull out of either system. That's why the Brits developed the Autolock collet system that's used with threaded shank cutters. It's effective to prevent pullout but it's a royal pain in the ass to use, mainly because you can't clamp anything else and you can't control the depth to which you clamp a cutter. The only ways to get around the pullout phenomenon is to either use a positive clamp system like the Weldon flats on plain shank cutters as described already, or use the Clarkson or Dormer Autolock setup. For my own use, I do all my hogging with insert carbide tooling. The spiral angle on these cutters is so low that they generate little downforce and can be clamped with confidence in a plain collet even for hogging. The other thing I do is to avoid massively deep cuts. I find, (particularly with CNC) that I get a better job faster with shallower cuts at higher speeds. I rarely take more than 0.075" per pass in steel even with my Haas that has 7 1/2 ponies to make it go, and I find that my cutters stand up better, the machine doesn't hammer and howl, and the job is more accurate. I take full advantage of the spindle speed I have and crank the feedrates right up. I'm using small ER20 spring collets in CAT 40 holders, and I don't need to squeeze the piss out of them to clamp the cutters. I know lots of jockeys who beat the stuffing out of the machine spindles and collet chuck nuts with big lead hammers and then try to bury an endmill one diameter deep into the job. The tools get wrecked, the job looks like it was chewed out by a beaver, and the cutters still suck out of the collets and end up either snapped off or driven right through the stock and into the table. With just a little bit of finesse, none of this is necessary. Cheers Marcus ------- Date: Wed, 07 Aug 2002 19:24:15 -0700 From: Alan Marconett KM6VV Subject: Re: backlash > There is a gentleman here that has fixed this very concern. >He has a web site that gives us a "How to do it", only thing is, >I cannot remember who it was, Hi Smitty, That would probably be Tauseef, I think I was just looking at his Sherline tip a few days ago. I couldn't reach his website a few minutes ago with the URL I'd bookmarked. Seems like something to look into. My old Sears/Sherline mill didn't have one. I'm assuming my new one will. Tauseef had a plate such that the lock could be pinned in "halfway" position as you describe. I'd like to add spring tension of some sort. Too bad it's a LEFT HAND internal thread on the nut and lock! Alan KM6VV ------- Date: Thu, 08 Aug 2002 03:06:09 -0000 From: "tollgrade" Subject: Z Axis adjustment procedures for CNC machines The follow list is documents from SHERLINE for adjusting the Z axis. http://www.sherline.com/6700inst.htm also the ASSEMBLY AND INSTRUCTION GUIDE for the gib adjustment. For Tauseef modification goto http://www.cuttingedgecnc.com/machinemods.htm for his idea which remove the end play in the saddle nut by adding the Z-axis locking lever which is not equipped with CNC machines. Larry M ------- Date: Thu, 08 Aug 2002 03:31:04 -0000 From: "tauscnc" Subject: Re: Z Axis adjustment procedures for CNC machines Hey guys, man, lots of stuff on backlash. There are 2 ways that I know of to control Z backlash, Rich's and mine. Rich's can be found in the files section or: http://groups.yahoo.com/group/sherline/files/Rich%20D%27s% 20Views/SaddleNutSplit.jpg mine again: http://www.cuttingedgecnc.com/machinemods.htm The saddle locking level is not included with cnc but is only like $8 or so. The material is simple to drill through and even simpler to tap. Just make sure you put it on upside down, so that little hole where the ball would fit faces UP. This way you should be able to tighten down on it fully within the Z way. It's quick, easy and cheap. Mostly its adjustable and I have found it to work out quite well on my mill. As for the page not loading, wonder if the server I use was down or something? ummmmmm Either way, Z axis needs a backlash control and sherline should start thinking about this. later tauseef ------- Date: Wed, 07 Aug 2002 22:13:27 -0700 From: Randy Gordon-Gilmore Subject: Mogliced Sherline backlash report Tonight I had a chance to measure the backlash on my Mogliced Sherline mill. I have a Starrett Last Word 15-0-15 indicator, which I held in the spindle using the 3/32" collet. I'm using CNC Pro, which has fast, slow, and incremental jog. I set incremental jog to .010" First I turned the point horizontal and slow jogged the Z down until the contact point (sphere) touched the table by about .005" for preload, and zeroed the dial. I switched to incremental jog and jogged down a commanded .010" The dial then read .010". I jogged up one increment, and the dial read .001" So the Z backlash is .001" I jogged up and down one increment, and the dial read .000" (i.e. showing .001" backlash again in the other direction, canceling the original backlash accumulation.) I repeated this several times to verify that this was a repeatable reading. So the Z backlash is .001" I did the same for Y, getting .001" backlash. On X, the backlash is .0025" I'm very happy with the .001" Y and Z backlash, and pretty happy with the .0025" X. I'm assuming that I was more generous with the release spray than Ron Ginger was, who reported less than a thou backlash on his machine. CNC Pro has backlash "compensation" so I will enter these numbers in the program before I start using the mill again. Once I get some kind of ability to mount the indicator to the gantry's Z axis, I will do the measurements there. It is scary fast--300ipm rapids with CNC Pro on my 466MHz Celeron. I was unwilling to go faster, and will probably "keep" rapids to 100ipm. That machine will get a Sherline head assembly when I can afford it (i.e. the machines transition from "hobby" to "earning their keep" :-) For Dee and company at Devitt, photos of my Mogliceing are at: http://photos.groups.yahoo.com/group/cad_cam_edm_dro under "Fun with Moglice" Best regards, Randy ------- Date: Thu, 8 Aug 2002 07:40:37 -0400 From: "Ron Ginger" Subject: Backlash is BAD! All this long discussion of Z axis backlash and compensation just re-inforces my view that backlash is bad and must be eliminated, not compensated for in software. In the suggestion of doing a Z-axis 1 step up to get the backlash compensation neglects the small amount of time it takes for the comp move to occur. When you send the mill down to its desired position it may immediately grab and pull down to the limit of the backlash. Your next move up will then pull the tool back up, but you have already made the divot in your work, so its too late. In a manual case on a mill as small as a Sherline one can press down on the head while turning the crank and feel the backlash. Backlash must be eliminated. There is no other way to get good work. A spring loaded backlash device may work IF the spring load is great enough to cover all loads. Thats a big IF and it adds a lot of wear. ron ginger ------- Date: Thu, 08 Aug 2002 08:34:11 -0400 From: "Dan Statman" Subject: Re: Backlash is BAD! Ron, Sure you want to eliminate as much backlash as possible, but that is not always possible or economically feasible. Plenty of very high quality parts have been made with machines containing backlash. In your example it certainly is possible for the tool to grab or the headstock to settle in the 1/4 second between z-axis moves to take up the backlash. But in practice it most likely will not happen, and it is certainly a heck of a lot easier and cheaper to write some code for backlash compensation (or download TurboCNC and try it out)and see if this cures your problems. Sometimes perfect solutions for every posssible range of conditions are not required, and quality parts can be made immediately with the current state of a machine if you compensate for its limitations. It could take hundreds of dollars and weeks of time to remove the backlash, and it still won't be perfect. As always, this is just one man's opinion. Dan. ------- Date: Thu, 8 Aug 2002 07:10:15 -0700 From: "Yasmiin" Subject: RE: Backlash is BAD! Perhaps I am missing something but all machines have back lash, even the best ball screws have it. Compensation for back lash exists in all CNC machine tool controls. This is usually a configuration parameter and the inclusion of compensation requires nothing but the entry of a parameter. Of course most machine tools are using some sort of absolute positioning ( glass scales / DRO ) also. Therefore, I don't understand the drive for elimination of back lash. The back lash in the lead screws of my Wade 94 ( Super Slide ) is .005. Wade has told me that redoing the lead screws would have no real benefit and cost more than its worth. I mention Wade because it advertises tolerances and repeatability in millionths. Yes this is a manual machine but I still don't see the issue. It should be possible to move the lead screw to compensate for backlash without moving the slide. If you can't then perhaps the problem exists in the adjustment of the slides? Yasmiin ------- Date: Thu, 8 Aug 2002 08:34:27 -0700 From: "Marcus & Eva" Subject: Re: Backlash is BAD! Hi All: I think a fundamental point has been overlooked in our discussion of how to deal with backlash in Z on a bedmill. The critical "must haves" in a system that has backlash, is sufficient headstock weight to keep the head down as far as the screw will allow. This can be exacerbated or helped by the condition of the Z axis ways. We have all been assuming that the normal condition is that the screw actually DRIVES the head down, and that the head would sit passively in position wherever it was placed in Z, even if the screw was missing. I don't believe this to be the most desirable condition with respect to the backlash. As I see it, the way to improve the repeatability of the Z axis most reliably is to INCREASE the weight acting on the Z axis screw and REDUCE the friction between the ways and the head. This is how all large bedmills (mills with moving heads) operate. The ratio between the head mass and the coefficient of friction is very high on these machines so the mass of the head holds the cutter in the cut. (They are HEAVY and have relatively LOW friction). On these machines, the screw merely acts like an adjustable shelf that the head sits on. On a teeny machine like the Sherline, we have less of this advantage to help us. So, the way to improve performance with respect to Z axis repeatability is to reduce the coefficient of friction between head and ways. The best ways I can think of to do this are as follows: First, pay very close attention to the condition of the ways in Z and improve them with lapping or grinding if needed. Second, consider applying a low friction material like Moglice or Turcite to all the sliding surfaces. Third, keep the gib adjustments appropriate and the Z axis ways well lubed at all times. Fourth, as Alan has done, tune the leadscrew position so it doesn't contribute to stick-slip of the head. In addition, you can improve the Z axis performance by ADDING weight to the head. It should be properly balanced to reduce the cantilevered load on the ways that the wildly off-balance stock head produces. The extra weight can be easily compensated by beefing up the stepper. If the nut notices the difference, a longer nut more securely mounted is in order, perhaps accompanied by diamond lapping the screw to make it smooth enough not to chew up the nut so fast. Pressure lube in the screw and on the ways is good too and not difficult to rig up. All of the ingenious mechanisms to reduce the weight on the head with counterforce springs etc, are counterproductive unless the backlash can be reduced to negligible amounts and maintained that way indefinitely. Only a preloaded ballscrew has a hope of meeting these conditions. So, my advice to all who are wrestling with this problem: Pay close attention to the condition of your ways and make them the best they can be. Remove any weight reducing devices you may have on your Z axis and consider adding counterbalancing weight instead. Cheers Marcus ------- Date: Fri, 09 Aug 2002 08:38:25 +1000 From: "Charlie Lear" Subject: RE: Re: Backlash is BAD! ball screws Moglice, turcite, counterbalances, shifting the motor, ballscrews... I totally agree that we should do everything we can to make sure the sliding friction is reduced to a minimum. For the vast majority of us that will be oiling the ways and adjusting the gib (after taking off the gib's high spots by using very fine emery paper and a flat surface). I have never - touch wood - had the head drop on me during a cut. You take the backlash and stick/slip out by deliberately advancing the head further down than it needs to go, then bringing it back UP to the starting position for the cut. For me that normally means a full turn of the feedscrew with my other hand resting on the headstock, then bring it back on the dial to where it should be. (My mill doesn't have the lock nut, so I have to carefully watch the z axis doesn't start creeping during a cut. Any sign of it and I just hold the handwheel in place.) I don't see why it would be any different for CNC, except that it would be easier. Down 0.100, up 0.090 instead of down 0.010. Cheers Charlie Lear, Melbourne, Australia clearx~xxsteammachine.com Hutt Valley Model Engineer Soc. http://steammachine.com/hvmes Eastern Bays Little Blue Penguin Foundation: same site /penguins ------- Date: Thu, 8 Aug 2002 17:08:49 -0700 From: "Marcus" Subject: Re: Backlash is BAD! Hi Randy; You are correct when you suggest that balancing and lowering the loads on the leadscrew will preserve the screw and the nut. >From this perspective, unloading the head is beneficial. However, from the perspective of accuracy in Z, your method is disadvantageous unless you can substantially eliminate Z axis backlash. I spoke purely from the position of how larger bedmills achieve this, and pointed out that the weight of the head is helpful. (In fact, many of these mills use counterweights too, but they do so because some of these heads are outrageously heavy; far more than is needed to hold the head down. The counterweights preserve the screw and also keep the head from sagging.) When you are fully dependent on the lack of backlash to hold your position, you need to maintain it at a higher level of precision than when you make your system backlash independent. .001" is pretty darn good for a non preloaded system, and if you can maintain it over a long period of time, then your solution will be adequate for the vast majority of applications, and has the added benefit of being able to contour in Z for 3D milling. In an ideal world, we'd all have zero backlash systems...my solution is worth considering for those of us who want to improve the accuracy of their mill in Z without having to rely on the condition of the screw so much. I do have a comment regarding those who wrote in, advising backlash compensating upward moves. This is OK if you are hand writing a bit of code now and again, but it'll quickly drive you nuts if you have to do any reasonable amount of programming. None of the post processors I'm aware of, make provision for this, and of course, the move across the backlash needs to be commanded, so it needs an extra line of code for every Z negative move. Since the backlash varies in almost all instances along the length of the screw, it's not an intrinsically reliable method, and, as others have pointed out, it does not permit plunging onto a part very well. You will find that holding Z levels within 0.001 or even 0.002 will be hit or miss with this method. I don't recommend it. Cheers Marcus ------- Date: Thu, 8 Aug 2002 19:41:37 -0700 From: "Marcus & Eva" Subject: Re: Backlash is BAD! Hi Randy: I think a lube fitting right in the center of the nut would be your best bet. I haven't had a Sherline mill for a while now so I can't remember whether there's room for a Zerk type fitting. A plain hole will work too, but does not allow you to pressurize the lube when you pump it in. I'd go with a blind ended slot parallel to the long axis and stopping short a minimum of one turn before each end of the nut. Intersect the cross hole with this (of course) and if you've got the space, put on a Zerk. I like Vactra waylube (I've also got tons of it so it gets used by default, but it works very well too!) With regard to mosquitos and elephants...you are absolutely correct. When I still had my Sherline I recognized the backlash issue immediately (I was doing 3D contouring). My solution was to lap the ways till they were dead nuts parallel. I spent about half a day using a simple split copper lap and 1000 grit Clover compound. I lubed the screw and ways with nothing more than a squirtcan. I didn't add any weight to the head. I ran my programs so all roughing was done with waterline passes, and I constrained the toolpaths for finishing so they'd always start at Z0 and go down. I'd then rapid up, return to the start point of the last pass, stepover, and cut downward in Z again. I was (and still am) using Mastercam and Featuremill to do my programming...both these programs allow good toolpath control. No gouges, nice finishes (for a flimsy stepper rig). This was all in Delrin though...no steel!!! I have no idea of the longevity of my setup though...I got a "real" CNC machine about a month later in a repo sale, and the Sherline was sold to a gentleman who has never cut a part with it (but loves it anyway!!!) Cheers Marcus ------- Date: Sat, 10 Aug 2002 08:21:42 -0400 From: "Ron Ginger" Subject: RE: Moglice Question When I first heard of Moglice I arranged for the sales engnieer of the US supplier to speak at our model engineering club. He came with a slide show of some amazing and HUGE machines repaied with moglice. In fact its the huge machines they want to fix, where the repair cost by other means would be unreasonable. One example was a ball screw abuot 2" dia. The nut shell was bored out and cast with moglice onto the screw. He said in one case some mistake was made and they tried to press out the cured Moglice which took tons of pressure to do. So, I would not be afraid to try it on a big machine, but get the manual and look at some of the repairs to see about proper shape and forming. In the case you describe a small moglice kit (about US$40) would be a cheap experiment. If it failed you could always cut off the moglice and you would be back to where you started and not out much money. ron ginger ------- Date: Sat, 10 Aug 2002 09:20:51 -0700 From: "Yasmiin" Subject: RE: RE: Moglice Question Well I checked out the manual and it looks like it will do the trick. http://groups.yahoo.com/group/CAD_CAM_EDM_DRO/files/moglice.zip (Thanks Randy) The scraping seems doable as they are talking about 10-12 points per square inch rather than the 40+ that some of the old time books talk about. I do have a set of carbide tipped scrapers. They talk about 50-60% full contact so one is really just putting some dips in a flat surface and I know how to make a flat surface. The under cuts seem to solve my fear of the Moglice peeling and at the same time don't do anything that makes other solutions impossible at a later time. The use of a flat bar as an outside form seems to be their method for filling the hole. So now all I have to do is get the energy and nerve to do it. Yasmiin ------- Date: Sat, 10 Aug 2002 09:52:47 -0700 From: Randy Gordon-Gilmore Subject: RE: Backlash BAD! Balance GOOD! At 06:33 AM 8/10/02 -0400, Jerry wrote: >I'd think that the headstock is already balanced, so your weight ought to >be spindle motor + controller, right? You are right that the spindle-motor-controller weight is what you want to balance laterally, but fore-and-aft you want to balance the whole thing over the ways. The whole head assembly is cantilevered forward from the ways, and gives a wedging effect. The overhung weight produces a moment or torque, so that the bottom of the head is pressing back against the ways while the top of the head is pulling forward on the ways. You want it balanced so that if there was some clearance between the head and the column it would fall straight down rather than tipping forward. Best regards, Randy ------- Date: Sun, 11 Aug 2002 07:08:53 -0000 From: "tauscnc" Subject: Re: Backlash is BAD! ball screws [sherline] Tom Nance wrote: >> Holy smokes, guys! Come on, does anyone besides myself think we have completely gone over the edge on this thread? Look, the Sherline mill is a sweet little machine. But let's face it, it's a *little* machine. It has its limitations and its idiosynchracies but most folks learn what they are, accept them, work around them, and get over it. If you think you need .0001 accuracy, go buy a $10,000 mill. I actually think a great deal of the charm of the Sherline machines is the fact that they don't have all the bells and whistles and the precise accuracy of the giants. It makes one have to think a little harder and plan the cuts a little better. I think that's all part of the fun. Now, I admit, I don't make chips for a living as it sounds like some of you do. If I did I might feel differently about it. I make stuff for the enjoyment of making them. Sure, I've ruined pieces on the mill because of the problems with the z-axis and I wish I didn't have to go thru the pain, but I think that's all part of the game. If there was a simple fix (as I believe we'll learn from Craig when he returns), then I'm all for it. But I think it's crazy to attempt to change the mill into something it was never intended to be. Years ago, I bought the DRO Sherline offers. Before I bought it, I looked thru the instructions on their web site and read Joe Martin's feelings about the DRO. He basically said (paraphasing) "if you think you just HAVE to have this thing, I'll be happy to sell it to you, but you'll really be happier using the hand wheels." Well, I figured he was just an idiot, and bought the thing anyway. I mean, after all, what does HE know about the mill? I used the DRO about three months and then gave it away to a guy that felt he just "had to have it" to do anything good. I'm sure in a few months he'll learn and will pass it on to someone else. I think the same sort of thing applies with this thread. Yes, changing the saddle nuts, and redesigning the axis, and adding springs and weights may improve the performance of the mill, but it is also violating the very nature and spirit of a great little machine. Leave the sweet, little thing alone. Recognize its nature and celebrate it. Its not a Bridgeport, don't try to make it one. I need 3 yards of topsoil for my yard. Should I get my cutting torch out and modify my Toyota Camry so I can haul the dirt, or should I get a pickup? We're beginning to sound like Tim "the Toolman" Taylor. There are more posts on dicking with the mill than there are on actually *using* the mill for productive work. Okay, rant over. Now let's hear some really good flaming from you guys. Tom Nance Corpus Christi TX << Hey Tom: Well since you asked :):) I could not disagree more :):)! The type of people that generally buy these machines are the ones that have some internal buring desire since childhood to build, take apart, modify, experiment, and make what they have better in some custom way. What is the purpose of buying this stuff at much $$$ when we could more then likly buy almost the same part somewhere for a few bucks? Now come on, you enjoy taking things apart and seeing how you can make it just "a little better." Yeah, maybe we go overboard sometimes and should step back to see where we are coming from but "customizing" what we have is the name of the game. Its fun to try and figure out how to make something better or work more efficienty in a simplier way. If we just accept everything for the way it is, life would be boring. As for the topsoil, I say you get a BMW M5, you'll look cooler and get around faster :) j/k taus ------- Date: Sun, 11 Aug 2002 11:02:35 -0400 From: "Carol & Jerry Jankura" Subject: RE: Equipment mods, was Re: Backlash is BAD! ball screws Hi, taus: I agree with everything you've written, but would also add that those of us who add CNC to the mix are often involved more in writing and improving the program that runs the machine than we are in actually making parts. What's nice about this equipment is that it is so versatile that all of us can use the same tools (toys????) to achieve our needs, so there's a large enough market to make producing them worthwhile. Aside to Craig - Yes, I've noticed your hat at the N.A.M.E.S. show that reads "Tools, Not Toys" but still consider all of the metalcutting items I have (as well as my ShopSmith, tilting arbor table saw, surface planer, etc....) to be toys. As in "the only difference between the men and the boys is the cost of their toys." ------- Date: Sun, 11 Aug 2002 15:18:30 -0700 From: Randy Gordon-Gilmore Subject: Re: Re: Backlash is BAD! ball screws Scott wrote: >Maybe if Sherline see's that their is enough demand, they will >redesign the machine so ball screws can be added. Also beefing >up the frame so it is more ridged. Sherline (and Mr. Clisby, who originally designed the basic lathe) have really done a fantastic job of making a versatile machine that is well-balanced design wise--no part is much better or much worse than it needs to be, to work in harmony with all the other pieces. People have asked occasionally if this part or that part can be expanded or beefed up or "improved", but the whole thing is in proportion as it is. Folks like myself are just taking the basic machine and specializing it for our own individual purposes. I'd hate for Sherline to make a change that would improve one specific aspect of the machine, but detract from its general versatility. Best regards, Randy (now where are those superconducting magnetic levitating bearings I was going to try fitting to the spindle?!?) :-) -------- Date: Mon, 12 Aug 2002 13:14:25 -0400 From: "Van Lupo" Subject: RE: Re: Backlash is BAD! ball screws Marshall, It was I who suggested that the ball screw could be located along side the slide. I figured that the weight of the motor would balance out the thrust of the screw, besides if you think about it the current factory solution applies the thrust to the side of the saddle. It just happens that the screw is located in the middle of the slide. I am not sure how this would balance anything in the current design. So perhaps placing the ball screw along side the slide and mounting the ball nut in approximately the same location as is the current saddle nut attaches to the saddle you would end up with all the same loads. Only change is the leverage point of the lead screw bearing is now located offset to the end of the slide. Van ------- Date: Mon, 12 Aug 2002 16:37:43 -0400 From: Tom Bank Subject: Endmill Flats -- Catching up on my group digests... Fascinating discussion on the various means by which we can screw up a prized work piece, especially the business of setting the endmill in the holder so it won't (I have a hunch that "can't" is too strong a word) drop down and take more of a bite than desired. It has happened to me, and I am pleased to learn what happened and how to avoid it in the future. Actually, my experience with endmill drop goes back to my first attempt to mill -- a bronze casting that had a lump of excess material that had to be removed. I had a Sears drill press and (believe it or not) a Palmgren X-Y rotary table I picked up from Sears for $150.00 new on sale. I bought an endmill, stuck it in the drill press, locked the casting to the rotary table, and started milling. Sure enough, I soon had a hole through the casting. Fortunately I had put a block of wood under the casting, so the table wasn't ruined. I then went on to buy a Sherline lathe and the Z-column. Found it wasn't big enough in the table. Bought a Delta bench top drill base and column and welded up an elaborate means to put the Sherline Z-column on it (using the rotary table again). Found that it the column would rotate around the vertical pipe of the basic setup, which could be fixed by putting a screw through both the Z-column holder and the Delta column, but I didn't get that far. Instead, I bought a Sherline mill. Like everybody else, I love it. I will say that for those who need a slightly larger mill, that approach might be the answer. I still have the thing and could take pictures if there is interest in the idea. Now, the obvious first mistake I made was to try to mill with a drill press, although there was originally an endmill holder that Sears sold, but I could not find one when I needed it. As the many notes on this list pointed out, the bite of the endmill sucks the bit right out of the holder, and when it is chucked into a three jaw drill chuck there is even less to hold it in place. I guess it has the same effect on a drill, but there you want it to go down, not hold its vertical position and move sideways. It should follow that any sized endmill holder will experience the same problem, regardless of endmill diameter, and that one set screw will have less of a grip on the endmill than a three jaw chuck, making a flat in the side of the endmill a practical necessity. However, the Weldon flat on the 3/8" and up endmills is excessively large on our small equipment. Maybe we should be grinding a very short flat or even a V-groove in the side of our 1/4" and smaller shanked endmills. I have had the problem of endmill movement as have others. I didn't understand it at the time, but that had to be what happened, because I was holding the Z-axis handwheel (before I got the new Z-axis lock nut) and I know it didn't move. I had also dropped down and come back up to eliminate backlash. I think sometime this week I will go downstairs, set up a jig, and cut nice, neat V-grooves in all my little endmills. Regards, and sorry for the long post, Tom Bank ------- Date: Mon, 12 Aug 2002 19:16:22 -0400 From: "Ron Ginger" Subject: RE: Backlash is BAD! ball screws >While at this point in time I have no interest in putting ball screws on >my Sherline, someone previously mentioned hanging one off the side of >the Z axis. My thought is that this would create problems becaust of >the thrust not being centered in the middle of the axis. The most common add-on to a Bridgeport mill Z axis is a ball screw off to the side driving the quill through the stub that was meant for the depth stop. So, off center is not impossible. It does require a rigid mount and some careful work, but it will work fine. ron ginger ------- Date: Fri, 16 Aug 2002 11:55:09 -0700 From: Randy Gordon-Gilmore Subject: Re: Questions for Randy At 09:59 AM 8/14/2002 -0400, John wrote: >Now that you have had your counter balance modification and the spindle >motor relocated for several months, have these changes lived up to your >desires / expectations? Are you getting the performance improvements you >expected when you designed these changes? By the way, what length belt do >you use and where did you get it? I'm really happy with the way the Z axis performs. Admittedly, I only did the Moglice retrofit recently and haven't done any contouring since then, but the Z motion has been very smooth and responsive to small increments in Z since I did the prior changes. I can run the Z axis as fast as either X or Y in rapids (but am keeping all three to 20ipm) With the leadscrew nut disconnected I can raise and lower the head with two fingers. The belt is a 6mmx500mm Ametric brand (Made in Germany) that I bought through the local bearing supplier (Applied Engineering Technologies in Fairfield CA). I took in the Sherline belt and told them I wanted the same cross section and about 20" long. Ametric is the brand they carry, but I'm sure it is a standard cross-section. Best regards, Randy ------- Date: Tue, 20 Aug 2002 08:53:12 -0700 From: "Craig Libuse" Subject: Re: Re: Sherline Mill backlash and Fun with Moglice > The top end of the Z ways where the Sherline CNC motor mount bolts on is > not really a finished surface. It has wide tool marks. Alan, This is an old message from August 6th, but I just got back from vacation. All Z-axis columns, both CNC and manual, are now flattened on the top end with a pass of an end mill. Previously only the CNC-ready machines had this additional operation performed, but so many machines are now being converted we decided to do it to all the columns in case a manual one is converted in the future. Craig Libuse Sherline ------- Date: Wed, 21 Aug 2002 18:28:57 -0700 From: Alan Marconett KM6VV Subject: Old "Last Word" Universal Test Indicator (Starrett), CNC backlash Hi to the lists, I've been playing around with adding backlash to a controller program I've been writing, and I THINK it's working! What's not working is my antique "Last Word" indicator (possibly). I have it's 1/16" mount in a collet in the spindle, and the ball on the end of the mount allows the body of the indicator to still shift around a little. Nothing seems to tighten up on the little ball swivel, even if I apply considerable force. The "fingered cap nut" that holds against the ball doesn't seem to hold (threads look fine). Or maybe it's not supposed to? I could understand a little "safety" in allowing it to move if pushed too far. Perhaps it's just too old (yeah, I could use a new one...). I'm also curious how closely the .001" divisions on the indicator should agree with movement? I THINK I've got the pressure square on, that is, the indicator body is perpendicular to the body of the indicator. I could understand if I was at an angle, but I shouldn't be. I don't have any doc's on the indicator, as it was quite old when I bought it used. I do have another "plunger" indicator that is also old and second (third) hand, but it works sluggish, like the oil has dried up in it. I've had it open, but everything looks clean. I suppose it could be in the "spindle" part of it, but I don't think I'm going to take it apart... I'm not sure how "corrected" backlash should look. Oh yeah, I know what backlash looks like, you go one direction for a while, then when you reverse, it takes nn steps before the axis starts to move back. Then the axis moves for each step that is issued to it. Fine. But what happens (what is observable) with backlash correction? At the change in direction, the controller "provides" the steps need to get the axis moving again. Not sure what this looks like on the hardware. I suppose the axis "delays" while the backlash is taken up, then moves the required distance. If I turn backlash off, I "loose" distance compared to when it is enabled (when moving back and forth between two points, my test case). So I believe it is working. However I don't have a good way to "know", because I don't read the full .010" move I commanded on the indicator. Do I need a setup with two indicators, spaced a specified distance apart? would that help? This backlash now has me captivated, since I ran into the "casting problem" a week or two ago. Alan KM6VV P.S. Oh yes, the new boxbed casting and some "sundries" (taps and dies, and some steam pipe) came in from Stuart. Ģ86+, more then the cost of the 10V steam engine! ------- Date: Mon, 9 Sep 2002 21:24:53 -0400 From: "Carol & Jerry Jankura" Subject: Z-Axis Preload Nut Folks: Yesterday, I decided to do a little manual milling with my CNC Sherline Model 2000. While moving the Z-Axis by turning the handwheel mounted on the rear motor shaft, I noted what appeared to be excessive backlash. On investigating, I found that the preload nut was quite loose. This was not how I had set up the machine, so my questions: 1. Any ideas what caused the preload nut to unscrew? 2. Any ideas short of loctite or something similar to prevent this from happening? Carol & Jerry Jankura Strongsville, Ohio ------- Date: Mon, 09 Sep 2002 18:35:15 -0700 From: Alan Marconett KM6VV Subject: Re: Z-Axis Preload Nut Hi Jerry, I believe Sherline DOES Loctite the Z axis preload nut, and the coupler into the bearings. Alan KM6VV ------- Date: Sun, 17 Nov 2002 22:38:27 -0000 From: "smittys8002000" Subject: Positive "Z" axis stop I have read in the past that some have had trouble with the Z axis "Falling" or creeping down from a set point. I too had the same problem, and finally decided to fix it. It is a crude fix, but I no longer have any trouble with the Z axis. All I did was install a 3/4" 90 DEG bar with a 1/4 20 threaded rod through the bar. I made it long enough to be able to adjust it for different heights. I bolted it to the top of the 2 bolts that hold the mill column to the base block. I will post a picture of it later tonight. Smitty ------- [NOTE TO FILE: There is a series of pictures and instructions on the Yahoo Sherline group site in the photos section that shows a clever set of clamps/mods to deal with backlash on all 3 mill axes. Simple and relatively bulletproof. Could apply equally well to Taig lathes and mills as well as Sherline lathes and mills and similar small machines by other makers.] ------- Date: Wed, 20 Nov 2002 12:35:54 -0800 From: Dave Martindale Subject: Locking Z-axis [FROM THE TAIG GROUP] It seems to me that the Z axis is a different case than X and Y. All three axes have some mechanical backlash, but only Z has gravity acting on it. Suppose you are milling in the X direction. The X axis is feeding the cutter into the work, so you took up all the backlash in that direction before the cutter reached the work. If you are doing conventional milling, cutting forces are pushing the work against the direction of feed, so the backlash will remain taken up. (If you're climb milling, the cutter could pull the work into the cutter by suddenly moving the table to the other end of the X backlash, which is why climb milling is a potential problem). The Y axis isn't doing anything during this, but you had to initially feed the cutter into the work in Y, and this took up the backlash in Y. Nothing is going to change this during cutting. But in Z, things aren't so clear. If the last head motion you made was to raise the head, the backlash will automatically be taken up in the downwards direction, so the head can't lower any further, and gravity will hold it there. If the last head motion was downwards, the head will be at the top of the range of backlash if the sliding friction is greater than gravity, and at the bottom if gravity was stronger. If the head is still at the top end of backlash when the cutter contacts the work, cutting forces and gravity can suddenly overpower friction and pull the head down by the amount of the backlash. This can make a mess. To avoid this, you either need to always lower the head too far and then raise it back to where you want (can be a problem when cutting pockets), or supply enough downward force (extra weight, springs) that the head is always at the bottom of the backlash, or lock the head. The above assumes that cutting forces tend to pull the tool down. If you have a left-hand spiral tool, cutting forces would provide up force on the tool, which is the opposite problem. Dave ------- Date: Thu, 21 Nov 2002 03:35:57 -0000 From: "Robin S." Subject: Re: Locking Z axis [taigtools] "jumbo75007" wrote: > Thanks Nick for your info. > Thank you Dave for your explanation about the direction > of backlash/last movement vs unintended movement of the > axes during machining. I understand about backlash, but I did not > connect it to the walking of the axes. This may be old hat to some, > but this is very enlighting to me! This also explains why I > have not had a problem with the y axis walking on me. > BTW, until I have a much better handle on machining, > I will continue to lock the Z axis. Backlash is not the only cause of wandering slides. If you're running the tool reasonably hard, the vibration of the flutes impacting the work will creep the screws along. On a side note, one must also be aware of the cutter moving around. If something isn't right (or even if it is) a cutter can slide around in a collet and move (normally) up into the collet. This is especially true when plunging. It's very important to have the drawbar tight and there should never be any oil on the shank or collet when the cutter is installed (not to mention oil can and certainly does hold particles of grit and swarf which can mess up the clamping). I noticed Dave mentioned climb milling as being a potential problem. It can be a _serious_ problem. If you're machining a gummy material like aluminum, the cutter can grab the work and pull slide (due to backlash). This can result in the park coming out of the vice and possibly being flung at the operator. The only time it's alright to conventional mill on a manual machine is when you're reversing feed to get a better surface finish (usually necessary on aluminum) without changing the axial/radial depth of cut. Regards, Robin ------- Date: Tue, 24 Dec 2002 23:46:40 -0