Tramming the Sieg SX3

Thanks for all the replies. Too many points to reply to individually.

It has highlighted that I am making the beginner's error of just tramming the spindle to the table, rather than both the column to the table, and also the (rotatable) head's spindle to the table.

I keep forgetting that the head can rotate 90-degrees on this mill, as I have never dared to adjust it out of vertical!

It was correctly pointed out that if I only tram the spindle, but the column itself is crooked, then when I raise or lower the head then my calibration is invalidated.

I think I might try Dan's suggested method, as I don't have a big reference square, but I do have a mag-base holder for my DTI.

I'm also going to hunt around more for some thinner shim steel.

Feeler guage sets are cheap on eBay, so I might buy and sacrifice one of those.

Regards,

Nicholas Lee

Video on Youtube about tramming a milling machine column with metal impregnated epoxy as a shim material.

Paul

**LINK**

If you don't have a large reference square you may be able to make a large circular square by turning a piece of pipe in the lathe, turn the OD and face one end in one setting and you have a square (providing your lathe turns true!)

I don't have a big reference square

If you have a lathe, as Hopper said above make a cylinder square. A bit of round bar 2 1/2 - 3" diameter by 8 or 9" long - recess the end(s) leaving a ring at the o/d to sit on, then centre drill. Turn between centres to as parallel as you can get & face the ring at the end at the same time and you are good to go.

To use, sit the cylinder square on end on the table, set a dial gauge pointing in the direction you wish to check from the head & set zero by passing the square past the dial gauge and zero on the high point. (i.e. point the dial gauge along the X axis (table) & "wipe" the square past the gauge tip with the Y axis). Move the head up as far as practicable & pass the square past the dial gauge again, noting the high/low reading which is the error. Working this way rather than running the gauge up the side of the square only shows the error in the direction you are checking, rather than a compound error. Then repeat for the other direction.

I don't have access to precision granite squares any more, but did find a rusty cylinder square where I work now that had come in with a job lot of stuff from an auction. After de-rusting, a check between centres showed it wasn't parallel, so a couple of hours messing about with the tailstock position & taking light passes had it parallel with about 0.005mm over 350mm, near as I could measure. I know it isn't perfect, but a quick check on a couple on vertical maching centres showed no more than 0.01 mm / 300 mm squareness error. I am not confident enough in the setup to claim that the readings are gospel, but they did suggest that the machine alignments were in the right ballpark,

Nigel B

Edited By Nigel B on 16/02/2017 08:21:50

Quick question on the "solid" cylindrical square.

On a light hobby mill putting a 3" dia x 9" long bit of steel on one end of the table may well tilt the table, should a similar 8kg weight be placed on the other end to balance things out before taking readings?

J

Edited By JasonB on 16/02/2017 08:25:29

putting a 3" dia x 9" long bit of steel on one end of the table

No need to have it on the end of the table - centre the table on the underslide, then position the square with just enough offest from centre of the table to allow the head to pass it unobstructed. The dial gauge will be offset to one side naturally if you you use a magnetic base to attach it to the head casting.

It would not be difficult to put plugs in the end of a piece of thick wall tube, rather than use solid bar, if the weight bothered you - though a bit of heft means you don't have to worry too much about clamping the square down. The lump I cleaned up at work was solid cast iron.

Nigel B

I don't want to take this thread too far from its specific practical subject, but:

Some may find it useful to explore the use of the autocollimator for alignment ...

Commercial ones tend to be big, and expensive [when new]; but the underlying principles are simple enough, and there was a decent home-brew version described in MEW.

MichaelG.

Shim stock starts at 0.005mm, 5 um thick, then 0.01mm, 0.02mm, 0.0254(0.001 inch) , these will allow very fine adjustments. It's not very cheap, but is available.

This place claim to make shims from .01mm and up. They mention 0.005mm . The shim we use comes from Germany , is 0.005mm thick and is 11mm wide.

http://www.hollinbrow.co.uk/hollinbrow/final.asp?id=63 Their shimstock. 0.010mm
0.0127mm ,0.0150mm,0.020mm,0.025mm,0.030mm,0.035mm,0.040mm,0.045mm,0.050mm

Nigel: Nice idea, but sorry, I do not have a lathe. Just the mill.

Ok, here's an update.... Success!

I have now trammed in the mill, and have declared that a victory.

I had one of these Stainless handrail mounts lying around, which turned out to be very accurately machined, and it saved me the expense of buying a "circular square".

I mounted the "stainless handrail mount" on my rotary table, on the stage, and dialed that in with my DTI with respect to the stage. The column of the "stainless handrail mount" was then perpendicular to the stage.

Then I took the back cover off the mill, to allow a mag-clamp DTI holder to be attached to the moving part of the z-axis. This allowed me to assess whether the z-column was aligned to the column of the "stainless handrail mount", and tram it in using shims under three of the four M10 mounting bolts.

Then a DTI holder was mounted on the spindle, and used to check the spindle against the column of the "stainless handrail mount". The head was rotated until it was even, and a shim was fitted in the gap where the head rotates to raise the front-back angle of the head.

This completed the three-stages of tramming.

As a final check, the rotary stage was removed, and the DTI holder on the spindle was used to sweep a circle across the stage. This showed that the alignment was good to +/-10um (0.4 thou).

I have found that care must be used on this hobby-mill to keep this calibration valid. In particular I have found that depending which way the z-axis adjustment wheel was last turned will put enough bending force on the column to affect the front-back calibration. Also, doing up the the z-axis lock handle seems to affect it a bit too.

The machine lacks the basic rigidity to achieve any better, and it was as much as I could justify spending.

Still, I'm happy with the result, and it should be perfectly adequate for my purposes.

Thanks for all the helpful advice!

Regards,

Nicholas Lee

Hi Guys,

I've just joined your group and seen your difficulties in tramming the SX3.

My head was out by 8 thou over the table back to front. this was because some idiot had not relieved the ally measure on the from of the table, and had ground the front bed-way to correct this.

The first check to be made is with a clock in the chuck reading off the pillar front. Slide the quill up and down; it should be true. If not scrape the mounting face where it rotates round.

Next tram front to back with the clock on a horizontally bent bar in your chuck reading off the saddle bed-ways of the base casting, i.e. with the saddle removed.

Mine was spot on: obviously the bed-ways and the pillar mounting faces were machined at the same setting.

It was after this stage that I discovered my problem.

I would recommend tapping the pillar mounting right through and using M10 by 50 long capscrews

Hope this helps

Bob Swift