[ngphil]

Hi all, I'm a newb to cnc milling and have several questions. I recently picked up a used sherline 5400 cnc mill on ebay for cheap. When I got the mill unpacked and setup I realized I didn't have any documentation on the steppers or the control board so it took quite some research and trial and error to get it configured in Mach3. Once I got the correct steps per figured out and the correct pins configured I was able to run some tests. Initially, on simple parts I designed in AutoCAD, then imported into MeshCAM for tool paths, everything seemed to be good.

Then I noticed the Z axis was "dropping" for lack of a better term. What was happening was the leadscrew was getting pulled up into the stepper housing slightly and then on the down movement the screw/spindle assembly would drop back down the distance it had been drawn up in. Obviously that was unacceptable. I machined a washer to go between the top threads of the leadscrew and the bottom of the sleeve where it mounts in the stepper housing which fixed that problem.

Next I tried a much more complex part which is somewhat indicative of what I intend to mill. This part had 30,000 lines of code generated from MeshCAM. I noticed once it got to the parallel finishing passes that the machine had obviously lost steps at some point. I discovered by reading that it was the driver board (old CNCgeeker/univelop TA8435) and that there was apparently a hardware problem with that board. So I ordered a geckodrive g540. I got it all wired up, configured etc. and continued to run tests.

Now I've found that I believe the table itself may be causing some step loss. If I grab the x table and pull on it I get movement back and forth in the x. The y seems solid. This mill seems to be an older model and either an older cnc kit on it or an after market kit. The stepper housings aren't like the modern sherline stepper mounts. They're rectangular boxes and the coupling between the steppers (vexta 2 phase 5v 1 amp steppers) and the leadscrew is done via a finned nut attached to the stepper shaft with a set screw, then a plastic collar of some sort which receives the finned nut and finally another finned nut on the leadscrew with a set screw. I initially discovered that the x axis leadscrew had threaded itself out of the stepper mount so I took it off and got it seated properly. The movement seems to be in the table and I guess the next step is to take the table apart and figure out why it's sloppy.

All this leads me to my real question. I had originally intended on buying a taig setup from deepgroove1 w/ the 4th axis. When this little sherline popped up on ebay I thought it was cheap enough that I could at least learn with it and get some things done. It's becoming apparent now that I'm probably going to end up spending as much or more on getting this sherline setup than it would've cost to just buy a complete 4 axis taig. I intend to cut fairly precise and complex shapes with the mill. If I was just cutting out rectangles with holes, the sherline would probably work fine with some final tweaking/tightening. I noticed yesterday though that my washer solution on the z seems to be slowly failing. I think the leadscrew is wearing the washer out so I'm getting movement in the z again. Probably the only way to fix it properly would be to buy new replacement leadscrews and housings and stronger steppers etc. Which would, again, be bringing the cost of the setup up to the cost of a brand new taig setup and that's without even buying the 4th axis rotary table yet.

So should I cut my losses and try to sell the sherline and just buy the taig or should I use the sherline and upgrade everything to try to get it tight? If I would've known how sloppy this mill is mechanically I wouldn't have bought it. It did say it was sold as is with no warranty or guarantee but it didn't say anything about the components being shot or flawed. I'm not going to try to get my money back or return it. I don't think I have any right to based on what the auction said.

If it helps, my intended use is milling brass parts for model master patterns to make molds. I need a fairly high resolution of performance. My goal is to be able to get very fine detail on very small parts (from 1/8" x 1" x 1" down). I obviously can't do that with slipping leadscrews and missed steps. So can the sherline be straightened up to handle high resolution/detail work or should I offload it? Would a brand new taig be able to handle high detail work? Sorry for such a wordy post, but if you've read this far you can see there's quite a few factors involved and I wanted to try to be thorough with my explanation.

Any advice is greatly appreciated.

[ngphil]

A few more details. I'm running the latest version of mach3 on a dedicated pc (3ghz w/ 4gb ram running winxp stripped down to nothing more than mach3 and all auto updates etc. all turned off). I'm missing steps even with the accel and vel set to 5 and .8 respectively. Using very low (1 ipm) feedrates etc. Again, I'm thinking the missed steps is a hardware issue, not software. If I cut simple parts it seems ok. It's when I run complex 2.5d parts that I'm getting missed steps. The fact that I can move the table by hand by grabbing it leads me to believe the hardware is crap. Also, I believe the motors are too weak. I can "twist" the table while it's driving and stall the motors pretty easily. I assume since they're 5v motors they don't have much holding power. So again, I'm looking at upgrading motors, leadscrews and couplers etc. which will put the cost of this mill over the cost of a ready to run taig before I even buy the 4th axis table. Just thought I'd add some more detail to the setup.

[NSR_CIVIC]

Shoot some pics of the machine. I have 2 sherlines and don't have any of the issues u listed.

Kevin.
Nsr customs.

[ngphil]

Thought I'd post some photos of the mill in case it might help explain what I'm talking about. This is really a horrible design for connecting the leadscrews to the steppers. The cast (zinc?) geared nuts have set screws that are mostly stripped from the previous owner. In fact, the one on the z lead screw won't budge and it's been completely stripped by the PO. So if I do go about changing these out I'll have to saw the sucker off.

So in the first photo you can see how the nut is on the leadscrew. What I just discovered is I can turn the leadscrew about 1/2 a turn of the nut before the screw actually turns. That's because there's slop between the nut and the retainer washer on the table. So when the nut starts to turn, for about 1/2 a turn it's just moving the nut until it hits the washer. Then the leadscrew turns. So no wonder I'm missing steps.



In the second shot you can see the "gap" between the back of the nut and the "washer" This is the play in travel.



Next image shows the intermediate "sleeve" that the toothed nuts fit into. That by itself would probably be fine so it's really just the play on the other nut and like I said, the set screws are so *****ed up I can't really do anything with them. I guess I could drill/tap them for new set screws.



Next image is the adapter block which like I said, looks different than all the sherline cnc adapters I've seen.



Finally just a shot of the motor. Like I said, they have no specs on them (the labels are missing) but from what I've found online they're vexta oriental motor 1 amp 5v 2 phase steppers.

[NSR_CIVIC]

I'm starting to think there was a multitude of companies that made conversion kits for these machines I have 2 different designs that are nothing like yours.. but I have looked at the actual sherline factory mounts and I believe that washer is supposed to be screwed to the table.. not totally sure but that may help with backlash on that. On the set screws have you tried the equivalent Allen wrench from sae. Or metric. Sometimes the slight difference is just enough. Or possibly a torks bit that fits tightly.

Kevin.
Nsr customs.

[ngphil]

I'm starting to think there was a multitude of companies that made conversion kits for these machines I have 2 different designs that are nothing like yours.. but I have looked at the actual sherline factory mounts and I believe that washer is supposed to be screwed to the table.. not totally sure but that may help with backlash on that. On the set screws have you tried the equivalent Allen wrench from sae. Or metric. Sometimes the slight difference is just enough. Or possibly a torks bit that fits tightly.

Kevin.
Nsr customs.

Thanks for replying! Yeah, the "washer' is screwed to the table. The problem is, there's a gap between the back of the toothed nut (for lack of a better term) and the face of that washer. So I can grab the table and pull it back and forth until the back of the nut hits the face of the washer. When the stepper turns, it simply pulls the nut toward the washer until it hits and then actually turns the leadscrew moving the table. The issue with the set screws is they're completely rounded off inside. No alan wrenches (metric or american) will do anything but spin in the socket. The PO (or someone who owned it before) actually cut a slot in one of them so you can use a flathead screwdriver to take it out. Looks like he struggled with the same issue and just tried tightening the nut onto the leadscrew further back. That x axis leadscrew is all kinds of chewed up from someone trying to machine it so that the nut could be positioned further back. At this point, looking at the damage done by trying to fix a bad design for a conversion kit, it looks like the only way I could get this to run reliably would be to replace the leadscrews and use a better coupler. That's sort of my problem. The mill plus the geckodrive board and now new leadscrews etc. is going to cost more than just buying a ready to use 4 axis taig. Of course now that I know this mill is a piece of garbage I can't sell it without disclosing the issues so I'll be taking a spanking on the sale price.

Here's a pic of the leadscrew end:


Here's a shot of the "modified" setscrew:

[NSR_CIVIC]

Ouch.. that's pretty tore up.. my sherline 5400 has similar motors but they are welded directly to the lead screw...

Kevin.
Nsr customs.

[ngphil]

Yeah, I'm pretty much "F'd" really. I basically have a manual mill at this point with bad leadscrews. So I guess I need to buy a cnc conversion kit and start over. The more I've talked through it, basically because I can't with good conscience sell the thing, I might as well use it. I guess I need to avoid buying things that are listed "as is". I'll take my lumps and buy a new cnc conversion kit for it. Any recommendations on a good/accurate set of leadscrews etc.?

[NSR_CIVIC]

I kno sherline sells their Cnc lead screws and saddle nuts fairly cheap. The conversion kit from them is gonna require more holes in the table ends but its a good design.

Kevin.
Nsr customs.

[DICKEYBIRD]

This part had 30,000 lines of code

:drowning: The poor lil' Sherline'll have trouble doing that kind of work accurately with it's tiny V-threaded leadscrews & bronze (brass?) nuts. You really need ballscrews for that kinda work and I don't think there's room for them in a Sherline.

You can help the situation some by modifying the X/Y slides per their website instructions showing how to drill some connecting oil holes and making a little oil reservoir. It helps a bit but there's only so much you can expect with plain screws & nuts.

Then there's also the problem of the anodizing wearing off the ways to worry about & the aluminum underneath galling.

I have one (Denford MicroMill w/CNC'd Sherline) with a lot of mods done to it and it's fun to learn with but I wouldn't dream of throwing 30,000 moves at it without stopping fairly often to adjust & lube copiously everything on it.

[ngphil]

Ok, Well I don't know if it will be a permanent solution yet, but I've got the play solved. I took it all apart and adjusted the nuts so their tight to the retainers that are screwed to the table. With everything squared up and tight I drilled the nuts and the leadscrews together. Then when I put them all back on I pinned the nuts to the leadscrews. Now the nuts are physically connected to the leadscrews so they can't pull off or loosen. Now I've got no measurable play in the table or column (within what's expected of a little mill) so it's time for another test run. I'll post back the results. If the pins hold and the nuts don't wear too much more it might work for a while.

Yeah, I'm probably asking a bit much out of a little sherline. Most of the master pattern stuff at 30,000+ moves will be in acrylic. I'm about to fire up another aluminum test with 18,000 lines to see how it does. Might as well push the thing. :devious:

[DICKEYBIRD]

Great! Glad you got most of the slop out of it. Here's the oiling mod I was talking about:
Milling machine saddle oil reservoir for - and Y axes

Here's a link to my oiler and a few mods I did on the Z-axis:
More Sherline mill fun & games.

There's probably some more if you're interested.

[NSR_CIVIC]

Very helpful links!.

Kevin.
Nsr customs.

[ngphil]

Great! Glad you got most of the slop out of it. Here's the oiling mod I was talking about:
Milling machine saddle oil reservoir for - and Y axes

Here's a link to my oiler and a few mods I did on the Z-axis:
More Sherline mill fun & games.

There's probably some more if you're interested.

Excellent! That counterweight is interesting. What is that clevis lookin thingamabobber just below the oiler on your z axis screw? Is that a lock? I do like the oiler lines etc. If I decide to keep this little bugger I might do that mod.

I always see people talking about hogging out steel and stuff with 3/8" end mills on these little guys. Talking about getting 99 million ipms out of them etc. It's tough for me because I want to go the other way. Like 1/64th endmills and 2ipm. I don't care if it takes the thing 3 days to make one part. I want tiny accuracy not speed and I'm not trying to mill engine blocks with the thing. Just curious if I can get the sherline tuned to do tiny precision work.

[wizard]

Hi all, I'm a newb to cnc milling and have several questions. I recently picked up a used sherline 5400 cnc mill on ebay for cheap. When I got the mill unpacked and setup I realized I didn't have any documentation on the steppers or the control board so it took quite some research and trial and error to get it configured in Mach3. Once I got the correct steps per figured out and the correct pins configured I was able to run some tests. Initially, on simple parts I designed in AutoCAD, then imported into MeshCAM for tool paths, everything seemed to be good.

Then I noticed the Z axis was "dropping" for lack of a better term. What was happening was the leadscrew was getting pulled up into the stepper housing slightly and then on the down movement the screw/spindle assembly would drop back down the distance it had been drawn up in. Obviously that was unacceptable. I machined a washer to go between the top threads of the leadscrew and the bottom of the sleeve where it mounts in the stepper housing which fixed that problem.

One of the issues with used is you don't know what you are getting and more so sometimes the owner doesn't know the condition of the tool in detail.

Next I tried a much more complex part which is somewhat indicative of what I intend to mill. This part had 30,000 lines of code generated from MeshCAM. I noticed once it got to the parallel finishing passes that the machine had obviously lost steps at some point. I discovered by reading that it was the driver board (old CNCgeeker/univelop TA8435) and that there was apparently a hardware problem with that board. So I ordered a geckodrive g540. I got it all wired up, configured etc. and continued to run tests.

Honestly even if you bought new I'm of the opinion you have the wrong mill.

Now I've found that I believe the table itself may be causing some step loss. If I grab the x table and pull on it I get movement back and forth in the x. The y seems solid. This mill seems to be an older model and either an older cnc kit on it or an after market kit. The stepper housings aren't like the modern sherline stepper mounts. They're rectangular boxes and the coupling between the steppers (vexta 2 phase 5v 1 amp steppers) and the leadscrew is done via a finned nut attached to the stepper shaft with a set screw, then a plastic collar of some sort which receives the finned nut and finally another finned nut on the leadscrew with a set screw. I initially discovered that the x axis leadscrew had threaded itself out of the stepper mount so I took it off and got it seated properly. The movement seems to be in the table and I guess the next step is to take the table apart and figure out why it's sloppy.

All this leads me to my real question. I had originally intended on buying a taig setup from deepgroove1 w/ the 4th axis. When this little sherline popped up on ebay I thought it was cheap enough that I could at least learn with it and get some things done. It's becoming apparent now that I'm probably going to end up spending as much or more on getting this sherline setup than it would've cost to just buy a complete 4 axis taig. I intend to cut fairly precise and complex shapes with the mill. If I was just cutting out rectangles with holes, the sherline would probably work fine with some final tweaking/tightening. I noticed yesterday though that my washer solution on the z seems to be slowly failing. I think the leadscrew is wearing the washer out so I'm getting movement in the z again. Probably the only way to fix it properly would be to buy new replacement leadscrews and housings and stronger steppers etc. Which would, again, be bringing the cost of the setup up to the cost of a brand new taig setup and that's without even buying the 4th axis rotary table yet.

So should I cut my losses and try to sell the sherline and just buy the taig or should I use the sherline and upgrade everything to try to get it tight? If I would've known how sloppy this mill is mechanically I wouldn't have bought it. It did say it was sold as is with no warranty or guarantee but it didn't say anything about the components being shot or flawed. I'm not going to try to get my money back or return it. I don't think I have any right to based on what the auction said.

If it helps, my intended use is milling brass parts for model master patterns to make molds. I need a fairly high resolution of performance. My goal is to be able to get very fine detail on very small parts (from 1/8" x 1" x 1" down). I obviously can't do that with slipping leadscrews and missed steps. So can the sherline be straightened up to handle high resolution/detail work or should I offload it? Would a brand new taig be able to handle high detail work? Sorry for such a wordy post, but if you've read this far you can see there's quite a few factors involved and I wanted to try to be thorough with my explanation.

I don't know precisely what you are doing but your interests in fine details on such parts, details requiring 30,000 steps of code, just seems out of the scope of this machine.

Any advice is greatly appreciated.

My suggestion would be to buy, rebuild or build a suitable mill. It really sounds like you need a machine with high quality linear bearings and ball screws. If you have the right tools you might be able to convert the machine you got to such. Obviously you don't need a huge work area so you can likely get by with shorter rails and lead screws to save some cash. It might be better to use this mill to build a suitable machine to which you can eventually can transfer the head and quill assembly. The Sherline should be able to manage such a build with little effort. I'm thinking a little mill with 4-6" travels in X & Y will do. Use high quality linear components and leadscrews to get reliable run time over those very large programs. In a nut shell you will make an CNC'ed XY table.

[wizard]

If the machine is already beat to hell you will have a tougher time keeping it running. That is true of any machine tool, especially if you are attempting to run long CNC programs on the machine. I'm certain with fresh components that the Sherline can be tuned to work well for general machining. However I'm not convinced that what you want to do is truly within the mills capability.

For example the 1 x 1 work area means a huge number of moves that never distribute oil properly on the ways. In the past I had issues with this on far more expensive diamond turning lathes working on very small parts with automatic lube systems. After a while the lathes where programmed to slew the axis full gravel every X number of cycles. This would force oil distribution across the ways, gibs and etc.

If you don't want to do a DIY machine your other option is to look at mainstream engraving mills or Mini mills with a suitable spindle. Obviously these aren't cheap relative to a Sherline.

Excellent! That counterweight is interesting. What is that clevis lookin thingamabobber just below the oiler on your z axis screw? Is that a lock? I do like the oiler lines etc. If I decide to keep this little bugger I might do that mod.

I always see people talking about hogging out steel and stuff with 3/8" end mills on these little guys. Talking about getting 99 million ipms out of them etc. It's tough for me because I want to go the other way. Like 1/64th endmills and 2ipm. I don't care if it takes the thing 3 days to make one part. I want tiny accuracy not speed and I'm not trying to mill engine blocks with the thing. Just curious if I can get the sherline tuned to do tiny precision work.

[ngphil]

Thanks wizard, this is sort of what I'm looking for. My thought was because I'm working with such small parts it didn't make sense to buy a big honkin mill. One thing I considered was buying a G0463 or G0704 etc. and converting it to cnc. I hear what you're saying on the linear bearings and ballscrews etc. Question is whether it makes sense to use the sherline as a starting point or to build something from scratch. I'd really like to get to the tasks I want to do rather than embark on a long mill build. On the other hand, I'm obviously cash conscious so maybe the only way to get a mill that I need is to build one/modify one.

Out of curiousity, with what I want to do, is there an off the shelf package in the $2,500-$4,000 range that would handle high precision work off the shelf? I have a feeling I'd be into the build/modify mill for around that same amount. I'm not opposed to building something or modifying one but I'd rather spend my time doing what I'm good at rather than building a mill. At least now I know you can't just take a sherline and get high detail work out of it. Or at least I shouldn't be expecting to throw 30,000 move nc files at it! lol

[ngphil]

So can someone recommend a machine that would be able to do what I want to do and how much I would need to spend? I need something that can provide very fine repeatable accuracy on large files (30,000 range). I can't imagine a big haas machine or even their mini mill is my only route. Am I to believe that the only way to get the kind of accuracy I'm looking for is with a $30,000 center? Like I said, I"m not too keen on the idea of building one. I'm sure I could design it and build it, but I would hate to spend a bunch of money and be right back where I am now. I'd rather buy something that's proven. My issue is I need to be able to run the thing pretty often. That's the only reason I'm looking for my own mill. I've already spent a fortune on outsourcing the work and it's getting harder to find people to do the work (seems like a lot of shops have been getting very busy). I've been looking at mills all day and I'm more lost now than I was before. I was looking at the Tormach packages, roland machines etc. Is the little MaxNC mill any good? I like that it's a closed loop/servo system. Even the tormach isn't a servo based system. Any advice is appreciated. I'm willing to spend, I just need to know that what I'm buying can do what I need. Need a few examples of machines/prices that would do what I need.

[awerby]

I don't think you need to give up on the Sherline quite yet. It was designed to do highly detailed, precise work. It's perfectly capable of executing 30,000 lines of code, which isn't that much for modern CAM programs to generate, since they tend to write a new line for every few thousandths of an inch of movement. If it's really going out of adjustment before the program's done, then use some (reversible) Loctite to seal the gib screws in place.

Since you seem to have solved the issue of your wonky leadscrew attachments, what's the problem, exactly? Have you tried running a program and seeing what happens? What tolerances do you really need to achieve? The Sherline should be accurate to within a few thousandths of an inch, with any error mostly due to backlash, which can be compensated for to some degree. (Is yours fitted with the newer style of backlash adjusters? Backlash Adjustment on Sherline Tools ) To get much better than that, you'd need to get a much more expensive machine, in the $6,000+ range. Forget about the MaxNC; it sounds good, but its mechanical shortcomings far overshadow any increase in accuracy you'd get from its "servo" control system. Properly adjusted, steppers work just fine, unless they're overloaded by trying to run too fast or too hard.

To use really small tooling effectively, you need a high-speed spindle. If you want to spend more money now, that's a good place to start. The stock Sherline spindle is pretty fast, and can be boosted to 10k RPM with a change of pulleys. That's fast enough to cut with a 1/64" tool if you go slowly. To go faster, you can swap in an air spindle, or a (more expensive) high-speed electric spindle. Basically, you need to calculate the amount of material removed per revolution of the tool, and make sure it's out of the way before moving forward, since those little tools will break instantly if that's not the case.

Andrew Werby
www.computersculpture.com

[691175002]

Provide an example of what you are trying to cut, or endmill diameters and desired accuracy.

Backlash and stick slip becomes a huge problem for very small tools because any jerking of the table will snap them.

If you are doing true micromachining it becomes harder to get by with a sherline/taig/maxnc level machine. There is a certain point where you need zero backlash ballscrews instead of threaded rod.

If I ever needed to do detailed work with small endmills I'd probably pick up a GX but this kind of stuff gets expensive fast.
Minitech Machinery