Howto Tram X4+ Head??

[john_t_h]

Does anyone know how to tram in the X4+ mill head?

[Martin Long]

John,
I assume you mean if the Z axis head is out of vertical alinement.
If either the head or the column are misalined it is a big job for someone who knows what they are doing. You first have to assertain if it is the column that is not square with the table, or if it is the slideways of the head that are out of line with the column.
To check the column, fix a dial gauge to the head and run it up and down along the edge of a very good square, do this both from the front and from the side. If this shows an error, the column will need to be removed and the base either shimmed (not prefered) or scraped to bring it into vertical.
Check the head by inserting a dial gauge in the spindle touching the table top surface, rotate the spindle and note the reading from front to back, and left to right. If an error shows up, first check the gib is making contact with the slideway vee over it's full length, if not then the gas strut could be pushing the head over to one side. This is most important because if the slideways are not square to the table the vee's will need to be scraped to bring the head into the same plane as the column. depending on how much material is scrapped from the vee's depends whether you will need to make a new gib strip.
You need to eliminate all other possilties before you start pulling the bedways apart.

Martin

[john_t_h]

Thanks for that Martin. I'll have to do the test with the square.

I have about 0.35mm difference left to right with a dial indicator in the spindle. I only learn yesterday on how to adjust the Z gib so I'll do that and do some more tests. I need to clean up the X4 first as it is covered in chips and cutting oil

[neilw20]

It is nice to have things perfect, but with my big CNC lathe I have been too lazy to test/adjust/test/adjust the tailstock.

Even if I got it right aligning the axis, it is serious work getting it correct to the correct height if it is wrong which it probably is.
On a job using the center I just do a test run and make a part, then add fudge factors, like a small taper into the program.
It makes perfect parallel parts all day and I still have clean hands.

Similarly I make parts on the SX3 and fudge the program for any errors, and just document them,
such as overshooting 0.032mm bringing the head down for a final cut, and backing off the 0.032mm to allow the head to settle at the correct Z height.

Even on large machines I have had had to compensate for Z final height in this manner.

Until you replace the Z strut with two 20Kg struts instead of one 60Kg strut you will have very significant Z repeatability issues, which gives an unexpected cr@ppy finish. With the strut at one side, as the head travels down the tip of the cutter has a small X+ error
and when you move the head up that tiny bit the Y+ error (due too sideways tilt) is removed.
Now when you go up you get a X- error. If the cutter is against a left face it will gouge slightly as you retract. You need a Z0.05 Y0.05 move for example to stop this gouging.

There are similar Y- Y+ errors. It will gouge on the face of a pocket nearest the front of the machine on a Z-UP move. Just document and fine tune the program.

As for rotating the head square, the spring loaded alignment pin works fine.
Just put the ahead back to vertical, grab hold of spindle with left hand and tighten the rotate lock nuts. That fudges it dead square.
You know when it is square. Surface milling has even criss-croos cutter marks. The error from head tilt and springing causes causes the criss cross error to change, and as the table machines across the X-Y range you can notice the table tilt as it over hangs either side.

The tilt in the X-Z plane means if you cut along the length of the table it will cut high in the middle. As the table goes off center the heavy bit hangs down and the other end pushes up slightly.

Of course, on my SX3 which has a quill and a Z micrometer it is really easy to fudge a small z compensation without a program change. I love my quill and readout. The (useless??) readout should not be pooh poohed. It is useful.

I have found the squareness of the machine (my SX3) to be fairly good and the deflection errors were much more significant in practice.

If you start measuring, flexing of the column and various parts become obvious, but for something that only weighs this much I just live with it.

I am doing jobs holding a Z tolerance of 0.01mm ALL DAY (on stainless steel) using a few fudges in the program to make a silk purse out of sows ear.
Ya gets what ya paid for, and its all not that bad when CNC can fix it.

[john_t_h]

Ok I did some more testing with a square mounted on the table.

Front and back were within 0.03mm, which isn't too bad

Left side: -0.29mm
Right Side: 0.29mm

Which is a real concern.

This fits in with what Neil thinks with the Z strut as it seems the head is being pushed to the side by the strut.

So I guess the solution here is to replace the single strut with two struts?

Perhaps Syil should be fixing this as it is a design fault and it does not deliver the 0.02mm accuracy advertised?

[neilw20]

Yep. 2 struts. MAX 20KG. The screw must be allowed to work with some pre-load.
With a bit of careful measuring you may get some more useful Z-travel too.
I haven't measured the X4 limits.
You can make the SX3 travel 385mm total with a few ugly attachments.

[ataxy]

do you have a post of the modification you made on your z axis

[neilw20]

http://www.cnczone.com/forums/showpo...54&postcount=3
http://www.cnczone.com/forums/showpo...11&postcount=4