What's new
What's new

Help with alignment between base ways and reference surface

addman20

Plastic
Joined
Nov 24, 2020
Hi guys, this is my first post here in this forum and great to be part of this community. Right now I'm in the information gathering phase, I am trying to scrape my ways and dovetails on my mill before I convert it to cnc. I've been practicing how to scrape for flatness and I've been doing some reading, mainly the Machine Tool Reconditioning book. I've tried to search for videos or articles on alignment and setting up the tools and dial indicators properly to get the proper readings. However I am having trouble figuring out how somebody would check and test for and scrape the ways on a milling base so that they are perpendicular to the back face of that base where the column attaches. I've attached a rendering image showing what the base of my mill looks like and the black line is flat up against the surface that the column attaches too. The red lines are the hypothetical errors of having the ways not perpendicular to that black line.

You would want the ways to be perpendicular to the back of the base, I am using the back as my datum or reference point.

The problem I am having is figuring out how exactly I would even go about measuring this. I have considered using a granite square lying next to the column base. But without any moving parts how would I articulate a dial indicator to make sure the granite square and the back face are perpendicular to each other. I was thinking maybe I could put some wheels under a dial indicator and run it on the back side of the base with the granite plate sitting next to it.

Once I know that the granite square and the base are parallel and square to each other then I can use something like a kingway tool to traverse the ways and have the dial indicator up against the granite square to tell me how much out of square they are.

The second challenge would be figuring out how to scrape for squareness and not just flatness and bearing surface.

Here is the link to my image

Base-1 — ImgBB
 
It would help if you showed us a picture and the brand of the machine. If it is a Chinese made mill it might have been screwed up from the factory, but if it was made in Japan, USA or Europe one can assume it was built right. In that case you look for original surfaces that are not worn and as a rebuilder you follow the new machine builders unworn areas as your guide. On a mill column the extreme top and bottom in most cases are not worn and you scrape it straight down hitting both ends the same to bring in the middle.

Another rule of thumb is if the head stock can be removed like a Bridgeport you scrape the ways square to each other and don't worry about it being square to the mounting surface where the head/ram swivel bolts on. If your machine has a fixed spindle like you see in horizontal mills you start at the spindle and scrape out from there so the ways are co-planer to the fixed spindle bearings.(Connelly book page 323, fig.27.7a. How about taking some actual machine photo's showing us what your working on, so we don't have to guess.
 
Hi Richard thanks for taking the time to replying to my post.

I'm embarrassed but the mill is a Chinese G0704. The mill actually has had very little use. I'm just getting around to being able to use it in my shop. So I'm not sure if there are going to be any wear patterns from use itself.

The head stock can be removed because it sits on a "saddle" that rides the column. The mill basically can come apart totally.

My question is that if the ways on the column base aren't perpendicular to the back side of the column base. When I move the table in the y direction the table will not move and be perpendicular to the spindle. Just trying to figure out a metrology setup to get the column base and granite square aligned.

I don't trust the grinding and ways on this mill in some places it looks like it was hacked. So now I don't trust anything on the mill and basically assume I'll have to scrape and align all the ways and dovetails

The column base, the column, the table, the table saddle and the headstock saddle and the headstock itself.

Please see the pictures below.

Base-1.jpg
IMG_1271.jpg
IMG_1272.jpg
IMG_1273.jpg
IMG_1274.jpg
 
Last edited:
I don't click on shady-looking third-party image links. Can you attach the pics to your post?

You need a reference square, something good like a granite master square or at least a Starrett no. 20. Put it on the y-axis ways a few inches from the z-axis ways. Attach an indicator to the column and indicate against the square as you ride the knee up and down.
 
I don't click on shady-looking third-party image links. Can you attach the pics to your post?

You need a reference square, something good like a granite master square or at least a Starrett no. 20. Put it on the y-axis ways a few inches from the z-axis ways. Attach an indicator to the column and indicate against the square as you ride the knee up and down.

Thanks still getting used to this forum, didnt think I had permission to upload pictures. I attached pictures to my previous post. Please took a look there
 
Here is a link to a PDF of the parts diagram of the mill. https://cdn0.grizzly.com/partslists/g0704_pl.pdf


I would approach this using a square against dowel or gage pins in the dovetails of the base and parallels against the ways of the column. this requires that they are bolted together. I would also check it with the head in place so the bend from gravity is already there. If it is out, then you know how far out it is as it would be used.

Do you have a surface plate big enough to put the base on, and have room to use a square?
Your mill breaks down into small enough parts to be handled. That will help.

It would be helpful to know what measuring tools you have.
 
I have micrometers, a surface plate but I don't think it's big enough to hold the base. I have gage pins, I have a dial test indicator that measures down to .0001 and a height gage. I have steel parallels.

I'm going to probably buy a larger surface plate either a 24x18 or 24x36. I'm looking at getting a 15" granite square as well.

So if I understand you correctly I'll be checking to see if the dovetails on the base are perpendicular to the dovetails on the column. I shouldn't get the base dovetails perpendicular to the back of the base and then proceed on the column.

When you say bolted together you mean bolting the parallells together into one unit? Or bolting the column to the base
 
It seems that you are concerned about the y axis being perpendicular to the column on a horizontal plane? That is to say, when the y axis moves in or out, there is not also a shift to the left or right. Am I understanding your concern correctly? If this is your concern, then stop worrying about it, it simply doesn't matter.
On the horizontal plane, the y axis need only be perpendicular to the x axis, not the column. Obviously, it also needs to be perpendicular to the z axis, but that's a different issue if I understand your questions right. You don't seem to be talking about up and down movement, only left and right.
NO AXIS needs to have any rotational orientation to the column. The alignment you are trying to check can be off by any conceivable amount and the accuracy of the machine not be affected.
 
I agree with Derek. I have seen a few of these Green beauties and they are a mess. You will never be able to hold tenths with it I am afraid. It's a shame people waste money on them, it should be a crime in my opion to sell them as precision. Your putting the cart ahead of the horse. Buy a bigger plate so you can set the base on the plate and be able to move a height gage all around it as the 4 leg supports under the base will not all hit at the same time. It looks ground, and I bet they ground a twist into the base. Once you get all 4 legs touching then indicate the top corners of the top of the base and they should be all the same height Then check to see if the ways are not twisted to each other or co-planer. The saddle scraping looks typical to those green beauties(embarrassment to scraping) and if you expect the other areas to be better your sadly mistaken. If I were you I would take the machine apart and fic it the right way or put it back together and use it. Rebuilding a machine is like building a house and if the foundations are not level then the whole house will be cockeyed. I will do my best to help you but with that where the column bolts on the back it will always move. The better models are the ones where the column sets or bolts to the base.
 
It seems that you are concerned about the y axis being perpendicular to the column on a horizontal plane? That is to say, when the y axis moves in or out, there is not also a shift to the left or right. Am I understanding your concern correctly? If this is your concern, then stop worrying about it, it simply doesn't matter.
On the horizontal plane, the y axis need only be perpendicular to the x axis, not the column. Obviously, it also needs to be perpendicular to the z axis, but that's a different issue if I understand your questions right. You don't seem to be talking about up and down movement, only left and right.
NO AXIS needs to have any rotational orientation to the column. The alignment you are trying to check can be off by any conceivable amount and the accuracy of the machine not be affected.

Thanks Derek for explaining that. But I guess I'm having a hard time wrapping my head around why that would be the case. Say the y axis ways on the base are not perpendicular to the column. And when I turn the wheel it moves the table little left for example instead of straight backwards square to the column so the work piece on the table will approach the cutter at an angle less than 90 degrees.

Thank you again Richard, I'll take your advice and buy a bigger surface plate and proceed with measuring the heights and making sure the foundation of the "house" as you mentioned is solid first.

Through all this I seem to be having fun.
 
Addman20

Think of it this way. For your mill he rotating cutter is fixed in space. It turns about its axis, but IT does not move. When it cuts, it is because the work moved into it. The work is moved by the table. If the x and y of the table are not square TO EACH OTHER, the parts will not be either. The head and column have nothing to do with that. They do not have to be attached to the table at all.

Visualize 3 Bridgeports sitting in a small circle facing away from each other. Turn one around to face the center of the circle. Now take the other two and rotate the heads 180 or so degrees, so they stick out the back. Move those two "backwards" mills so the heads are over the table of the other mill, and tram them square to the table and put vices under them. You can now put work in the 3 vices and make 3 parts by moving one table.

There are also rotary cross slide tables. I have not watched all of it yet, but this video shows one and what it does. No CNC needed!!! setting up a cross slide rotary table part 1 - YouTube
 
Read thru this one....it's the book all modern standards are based on. For your machine save some money and buy a precision blade square. You will be wasting money buying a granite square. There are several B&S and Starrett used blade squares in Ebay. I am also sure once you get the column aligned and square to the top if the table. It will get knocked out taking a heavy cut. http://totallyscrewedmachineshop.com/documents/Testing Machine Tools (Dr.Schlesinger).pdf
 
Addman20

Think of it this way. For your mill he rotating cutter is fixed in space. It turns about its axis, but IT does not move. When it cuts, it is because the work moved into it. The work is moved by the table. If the x and y of the table are not square TO EACH OTHER, the parts will not be either. The head and column have nothing to do with that. They do not have to be attached to the table at all.

Visualize 3 Bridgeports sitting in a small circle facing away from each other. Turn one around to face the center of the circle. Now take the other two and rotate the heads 180 or so degrees, so they stick out the back. Move those two "backwards" mills so the heads are over the table of the other mill, and tram them square to the table and put vices under them. You can now put work in the 3 vices and make 3 parts by moving one table.

There are also rotary cross slide tables. I have not watched all of it yet, but this video shows one and what it does. No CNC needed!!! setting up a cross slide rotary table part 1 - YouTube


Thanks! that helped. Makes sense now, like a giant lightbulb that went off.
 
Read thru this one....it's the book all modern standards are based on. For your machine save some money and buy a precision blade square. You will be wasting money buying a granite square. There are several B&S and Starrett used blade squares in Ebay. I am also sure once you get the column aligned and square to the top if the table. It will get knocked out taking a heavy cut. http://totallyscrewedmachineshop.com/documents/Testing Machine Tools (Dr.Schlesinger).pdf

Thank you Richard, will do
 
I bought a used Brown and Sharpe 540 square, with an 11 inch base and 20.5 in blade. The blade is .097 in thick and has a slight curve in it. The curve does not affect the accuracy, but in using it the tip of the indicator can run off the edge. Connelly mentions this on page 92 of his book, so it is not just me... It is usable, but you have to be careful. I later bought a similar sized mitutoyo. Its base is 11.75 in and the blade is 19 3/4 " long- and .315 inches thick. It is clearly heavier, but easier to measure from as the blade is over 3 times as thick. 3 months ago I got another Mitutoyo off craigslist It its blade is 29 3/4 long, and .392 " thick. At a bit over $100, the price was right.

Read up on how to test a square for accuracy using an indicator. With some creativity you can make accurate temporary squares.

Getting back to your mill, and Richards "heavy cut" comment
The column holding the head is not very thick, and from the diagram is only held in position by bolts providing friction (no pins). The spinning cutter will act like a wheel, pulling the head sideways. The higher up the column you are, the more leverage it has.

You want the spindle axis of rotation to stay the same as the head is moved up and down the column, and be square to the table in all axis. As the head is raised up higher on the column, the column will bend, and the axis will shift away from the column and out of true. You may want to take measurements and decide if it is worth scraping or shimming in a slight backward lean to account for this.
 
A few years ago when those machines were popular on another forum I was a moderator for in the reconditioning forum, many hobbyists complained about not being able to keep them square. A few years ago those green machines were banned on this forum. Many poured concrete in the bases, mounted the machine on thicker steel plates, made braces from plate to column. All that weight on the back of the base was a poor Chinese design IMHO.
 








 
Back
Top