What's wrong with my mill?
Monday, Mar 23 2020
I thought I’d use a chuck of steel left over from another project to make a V-block. It won’t be hardened, ground, or high precision, but should be good enough for my needs. The first step is squaring up the stock – making all faces perpindicular to the sides and opposite faces parallel.
Ooops. I got 4 of the 6 faces OK, but one pair of faces was off by more than 0.005" over about a two inch run. I should be able to do better than that.
The issue was one of my own making. I did not have the piece fully set in the vice or didn’t have the vice clean enough, etc. I tried again. But this time I used a fly cutter instead of an end mill because it would give a better finish and would be much faster. Except it didn’t solve the problem. A fly cutter is great at telling the machinist when the mill is not in tram; the z axis not 100% perpinduclar to the bed. I’d trammed the mill before starting in the X axis and it was OK. The issue seemed to be in my non-adjustable Y axis. Time to see what is going on.
Wednesday, Mar 25 2020
I 3D printed a part to hold a test indicator in the mill spindle. Twice. The first time I watched the print for a while then had to leave the room. As soon as I did the small part became unstuck from the print surface and the printer spent the next 30 minutes depositing melted plastic in the air. The second time I printed the part with a brim for a larger surface area stuck to the build plate.
Once the part was printed I used my test indicator and an square held in the vice to see of the column was square to the table. I’d already checked the X and Y directions. They were fine. The Z direction was off. It looks like the angle between the table and the column was something less than 90°.
Thursday, Mar 26 2020
I removed the table, the saddle, and the frame that holds the saddle to get down to the base. I also removed the motor and spindle. My first check was for nod. I verified that the dovetail that holds the spindle was parallel to the Z axis. It wasn’t. I removed a shim that I’d installed a while back and checked again. Now it’s OK. I think the shim was the wrong fix for a different issue – column lean.
I then checked for column tilt. Hmm, I’m not seing any. Could the saddle frame, saddle, or table have been misaligned? I cleaned the saddle frame and installed it without fully tightening the screws. The saddle and table were also cleaned and installed without their lead screws. I clamped a square to the table and checked Y movement to make sure the saddle frame was square. After some tweaking the screws that hold the frame to the base were tightened.
I installed the Y lead screws and clamped a 1-2-3 block to the table. Testing for column lean I got less than .0005" over 3 inches. Great.
Or is it? Everything being clean and freshly adjusted, but I’ve not yet mounted the heavy spindle motor. I’m wondering if the motor is causing the column to slightly sag. That will be the next thing to check after re-installing the lead screw for X table movement.
Friday, Mar 27 2020
I got the mill put back together. Mounting the motor did not cause the column to sag. Everything I checked says the X and Y axis are perpendicular to the Z axis. A cut with the fly cutter will verify my checks. It the alignment is off the cut will not be the same on the front and the back of the part. One will be lower.
I do have a slight issue with the table in the X direction. It takes more force to move the table at one end of its travel than it takes at the other. I’ll look at that later. Right now I want to check the alignment and the table is good enough for this test.
That’s a lot better. The 0 in the first picture is 0. It looks slightly off due to the angle of the camera to the dial. Total run out is less than 0.001". It’s closer to, but slightly more than 0.0005". I’m happy with that.
Tomorrow I’ll remove the table again and try to make it as smooth as it was without any play. I think the lead screw is binding. That is something I can adjust.