Milling machine head stabilizer - video

[Jon]

Milling machine head stabilizer. By Lets Roger That. 25:24 video:

[Paul Alciatore]

I have a dovetail column mill in my shop now, but in the past I have worked with at least two round column ones. So I am very familiar with the vertical alignment problem. I had given it a lot of thought and reviewed a number of methods for preserving the position when moving the head up or down. I will comment on this one and then about what I would do.

Overall I like this approach. It allows the head to be raised or lowered with no attention to this alignment feature: that is a big advantage. There would be a temptation to not lock the head down after moving it but that should be ignored as this alignment method will not resist the cutting forces. The head should always be locked in place to prevent HORIZONTAL movement while milling. Gentlemen, ALWAYS LOCK YOUR HEAD! Well, lock any movement that is not needed for the current cut.

I always think about the accuracy of these methods. The designers of every one of them always says that they have good accuracy, but few, if any, of them ever give any numbers to quantify that. Here are my thoughts on this and many of the other methods that rely on a guide rail that is mounted at a small offset from the round column. The columns on these mills are approximately 4" in diameter which means about a 2" radius. The radius is what I use because it is the center of the horizontal circle that the head can rotate about. And the distance from the column to the center of the quill is about 8". Adding the radius to that number, I find that the center of the quill's axis is about 10" from the center of the column. I must now resort to an approximation, but on the generous side, I estimate that the guide rail in this and the guide rails in many other such methods is about 2" to 3" from the column. Again, adding the radius of the column the guide rail is about 4" to 5" from the center of rotation. The 5" number is the better end of this range so using that and some simple trigonometry, I find that an inaccuracy of 0.001" at the guide rail translates to twice that at the spindle's axis or 0.002". This is the fundamental flaw that I see in this and all the other methods that use a guide rail that is offset from the column by a short distance. The error at the guide rail is MULTIPLIED by at least 2 times and in some of the designs by 3 times as much as 5 times when the guide is mounted on the surface of the column. This version of a guide rail is at the better end of this range of errors and that is in it's favor. But neither this nor any of the others that use a guide rail at a short distance from the column are going to offer anything better then +/-0.001" accuracy and, with other factors taken into account, even that number is going to be difficult to achieve.

In the video a common, carpenter's level is used to set the guide rail vertical. But no mention is made about the orientation of the mill itself nor of it's column. My observation of two of these round column mills is that the column is not set perpendicular to the movements of the table with any great accuracy. They just are not made that well in China or the other Asian countries where they are manufactured. But this and the guide rails in other designs must be parallel to the column. Instead, in the video the guide rail is set vertical in only ONE plane (left-right/up-down plane) and the other, more important plane (front-back/up-down plane) is completely ignored. The angle bracket on the head has slotted holes for this adjustment, but they are covered by the guide rail and would be difficult to access for adjustment. I say that this second plane (front-back/up-down plane) is the more important one because it is the one which actually determines the rotational position of the head while the one which is actually adjusted only keeps the guide rail in approximately the same position in the lower bracket (the one with plastic slides). This second plane adjust was only made when the angle bracket was bolted on to the head and that seems to have been just an eye-ball operation. Finally, even if you do make the leveling operations in both planes, using a carpenter's level which does not even come with accuracy specs., leaves the set up open to large errors. And NOT checking the column first allows even more errors to creep, no JUMP in.

I fear that this and all the other guide rail methods can easily have errors of 0.003" or more.

Finally, what method do I think would be best? From geometry, it is OBVIOUS that the longer the baseline (radius) of the method used to maintain this angular movement (rotation) of the head, the BETTER the accuracy at the quill's center line will be. I would obtain an optically flat, first surface mirror that is about 2" wide and about an inch or two greater than the vertical distance that the mill's head can travel. This mirror would be carefully mounted on a shop wall as distant from the mill as the shop permits. I would use a three point mount to avoid any twisting of the mirror. An LED light would be mounted on the mill's head and pointed horizontally at the center of that mirror. The reflection from the mirror would be directed to a right angle prism just above or below the LED light. That prism would have a scale on the face that points to the machinist with vertical lines. The LED light and wall mounted mirror would have adjustments that would allow the reflected spot to be centered on the center or reference vertical line in the prism when a part is first started in the mill. When the head is moved up or down, all that would be needed to keep the vertical alignment would be for the for the head to be rotated to bring the dot of light back to that reference line. Then the head would be tightened in that position and machining could begin again.

This optical method has several advantages. First, the long base line (twice the distance from the mill to the mirror) would increase the accuracy, not decrease it as a closely mounted guide rail does. Second, alignment would be easier as the adjustments would NOT rely on the orientation of the column. A mask with a small vertical slit would be placed over the mirror and the mirror would be adjusted to allow the dot to be seen at both the top and bottom of the head's movement range. This would be needed only when first installed and at times when use of the mill dictates. And this mirror adjustment would only involve rotating the base the mirror is mounted on: this would compensate for any error in the vertical alignment of the mill's column. These mirror adjustment times would only be needed at widely spaced times. The second alignment would be of the light source and that would be easily done with a single adjustment when each job commences.

The biggest problems with this optical method would be the size of the dot of light the LED source can project and, of course, the size of your shop. But modern instruments, like laser levels, are routinely used for leveling construction where the dot is around 1/8" in diameter over many feet. That 1/8" dot can be centered on a line with, perhaps 1/64" or even 1/100" accuracy. Lets say your shop allows an 8 foot distance between mill and mirror. 2 x 8 x 12" (the light travels that 8' distance twice) = 192" or about 200". With the same 10" distance from the center of the column to the quill axis, that means that the accuracy is multiplied by a factor of 200/20 = 20. And 0.010" / 20 = 0.0005". If you can get a better LED light source or have a bigger shop, that only gets better. And the simple adjustment of the mirror completely eliminates any need for adjusting things parallel to the column's axis.