Sherline 6 Hour Cnc Operation

j ferguson

Active User
Registered
Joined
Nov 28, 2013
Messages
196
I'm making Luiz Ally's slim vise. I'm cutting the pieces out of a 3.5 x 8 x 1/2 inch plate of aluminum.
I'm using a .250 HSS end mill, 4 IPM at 2500 rpm, .050 D.O.C., .120 step-over, and two operations, first a spiral roughing leaving .010 for a profiling finishing operation and then a two pass full depth profiling at .005 per pass. The time adds up to about 6 hours.

I'm using my approximation of Terry's excellent no-fog mister.

6 hours is a long time to sit there and watch it. I'm running with a pretty peppy PC and LinuxCNC.

The Sherline motor reads 140F in my non-a/c garage here in sunny Delray Beach. But it doesn't seem to get any hotter with time.

Questions:

1. Do I really need to give this my undivided attention?
2. Would any of you let this run with a look-see say every 1/2 hour?
3. What if I inserted pauses in the g-code say every 30 minutes?
4. What if I paused the LinuxCNC, turned off the mill, did somthing else, then turned the mill on and continued? I've tried this on other things and didn't seem to lose any steps, but ???

I encourage criticism here. I do want to cut the parts at once, mostly to improve chance of getting consistent dimensions on the two jaws.
 
Last edited:
I'm making Luiz Ally's slim vise. I'm cutting the pieces out of a 3.5 x 8 x 1/2 inch plate of aluminum.
I'm using a .250 HSS end mill, 4 IPM at 2500 rpm, .050 D.O.C., .120 step-over, and two operations, first a spiral roughing leaving .010 for a profiling finishing operation and then a two pass full depth profiling at .005 per pass. The time adds up to about 6 hours.

I'm using my approximation of Terry's excellent no-fog mister.

6 hours is a long time to sit there and watch it. I'm running with a pretty peppy PC and LinuxCNC.

The Sherline motor reads 140F in my non-a/c garage here in sunny Delray Beach. But it doesn't seem to get any hotter with time.

Questions:

1. Do I really need to give this my undivided attention?
2. Would any of you let this run with a look-see say every 1/2 hour?
3. What if I inserted pauses in the g-code say every 30 minutes?
4. What if I paused the LinuxCNC, turned off the mill, did somthing else, then turned the mill on and continued? I've tried this on other things and didn't seem to lose any steps, but ???

I encourage criticism here. I do want to cut the parts at once, mostly to improve chance of getting consistent dimensions on the two jaws.

I have a Tormach Running SprutCam generated G code with Mach 3. I stay with my machine when I am running long programs. If something goes wrong, at least I have half a chance to prevent a serious problem. I will generally set up some other tasks to occupy my time; lathe work, shop cleanup, etc.

Mach 3, at least the Tormach flavor, has the provision to single step through the program at any time so if I have to step away, I can hit the Single Block soft key. It will finish the G code loaded in and stop, waiting for me to hit the Cycle soft key for the next line of code (although the spindle is still running).

Tool changes are a convenient place to stop as everything stops running. A pause can be inserted as well. I would just use the Single Block feature in my case.

I will try to never shut down the mill once a program is set up. When the power shuts off, the steppers can lose steps. This is audibly observed by the loud "clunk when the power shut off. Similarly, when the power is restored. I have noticed position shifts of several thousandths in the axes. The Tormach Mach has a referencing routine but it uses the limit microswitches to home the machine. They can vary the referenced position by a thousandth or two. I have installed optical homing sensors on the Tormach which are repeatable to +/- .0001 but best practices, IMO, dictate performing all machining with a single setup if possible.

A few years ago, I had some complex machining which was rather lengthy so I had to re-reference the machine. I made a sub-plate for the work to accurately reposition it and a cylindrical top hat post (.600" dia. by .6" high that mounted in dowel pin holes that I added to my table. I then could find the center of the post with my edge finder and re-establish my reference. (I hadn't installed the improved homing at that time).

Your setup may be better at referencing than the Tormach. If you really think you don't lose steps, I would encourage you to do some critical measurements on all three axes. Establish three reference points and set your work zeroes for them. Shut the machine down and restart and re-reference the mill. Go back and relocate your work zeroes to see if they have shifted. Do this a number of times to be sure that you really can hold your reference.
 
It really depends on how comfortable you are with your machine and how stable you feel it is. If you have been running it on quite a few jobs and it is stable, then walking away is probably safe. At that point you're relying on the limit switches to stop the axis motors in case of a major failure. It's common in industry to run machines ''lights out'', but they have a number of error trapping routines built in that would not be common in a home shop machine.

Once I'm satisfied with the way a program is running, I walk away from my machine and check on it periodically. The most common problem would be a broken tool bit. But normally that just means you are cutting air. Loss of coolant would be the biggest concern when cutting aluminum, that will cause the tool to gum up almost instantly and probably break. Most of the time I'm in the shop while the machine is running, but on really long jobs, I may come into the house and relax while it's running, but I'm listening for something weird happening.

I have a tool path graphic on my machine so I am able to compare the actual tool path to the commanded tool path. When cutting a complex or high value part for the first time, I simply run the program with no tool bit or set the Z high to so I can make sure it doesn't rapid into a hold down or something stupid and to confirm the tool path, even on one-off parts. It takes longer that way, but I know the part will come out correctly when it is cut.
 
Thanks to both of you for your counsel. I think what I'm going to do is add a pause to every other Z move. I've left enough material after roughing to absorb any (I hope) minor anomalies. I won't interrupt the finish profiling.

john
 
I thought I was comfortable with running long jobs unattended on my Sherline mill until, as JimDawson mentioned, my coolant stopped and chips welded themselves to the cutter flute, resulting in a broken endmill and a ruined job. Now, I check on the progress every half-hour or so and I'm usually within ear-shot of the machine so I can tell if something is drastically wrong.
 
Back
Top