A Small Mill: My Sieg X3 Saga

Discussion on all milling machines vertical & horizontal, including but not limited to Bridgeports, Hardinge, South Bend, Clausing, Van Norman, including imports.

Moderators: GlennW, Harold_V

User avatar
Harold_V
Posts: 20227
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Post by Harold_V »

Bolsterman wrote:A question to the forum…I did a quickie check of my tram and I think I see that part of my “column leaning right” problem has returned to a lesser degree. I don’t have a cylindrical square to check, anymore; I just used an Enco-grade 4” angle, don’t know if I can trust it or not.
Compare it to a square you have, placing both on a surface plate. If you have no surface plate, use the mill table. If you find there is no gap between the square and the angle plate, it will likely serve well enough.
maybe I should buy a 2-4-6 block instead?)
There is no guarantee that such a block will be square, either. Unless you see a need to own one, it may not be a good investment.
I am suspicious my problem may be in “gunk from the factory” between the column and the table it’s bolted to
Rightly so. The two surfaces should have been examined before you attempted to correct error. I suggest you dismantle the assembly before you attempt anything in the way of correction. Otherwise you may be wasting your time.
So if I am able to buttress my skinny rafters to the point they can comfortably hold a couple hundred pounds,
I trust the ceiling is not covered with sheetrock, and you have access to the trusses? If so, there's no reason why you can't use a 4" x 4" that spans several trusses, to spread the load. If you are picking only 200 pounds, even if you span just two, each is carrying only 100 pounds. If you can hang from a single truss with your hands, the load for the machine components will be far less. Still, span as many as you can, for insurance.
do you think I can lift the column and head (firgure 200 lbs?) with a come-a-long I already own, and clean out the joint area?
I sure do, but it will be prudent to block such that it can't fall suddenly. You may be able to chain (or strap) the column/head assembly once it has been lifted.
Or would I have to invest in a block and tackle?

Only if you have need to own one.
And would it be OK to lift the column/head unit by wrapping a lifting strap choker-style around the head where it meets the column?
Assuming you can do that without trapping (and bending/breaking) any sheet metal or plastic covers, that's exactly where I'd recommend you attach your sling.
Make sure it can't slide off if the load shifts.

While 200 pounds is a considerable load for human hands, it's nothing in regards to mechanical means of lifting. A single ¼" bolt is very capable of handling that light load. Still, use caution. Should it drop and you are under the load, I expect it will ruin your day. Maybe your month.

Harold
User avatar
Bolsterman
Posts: 102
Joined: Thu Nov 12, 2009 12:36 pm
Location: Mexifornia

Post by Bolsterman »

Thanks for the good advise, Harold. I'll need to sister the joists in my garage before proceeding, but that's where I'm headed when I've accomplished that little chore...I'll pull the column and clean out the gunk, as I should have done earlier.

[Commentary and attached file removed; it was incorrect].
Last edited by Bolsterman on Tue Jan 05, 2010 6:57 pm, edited 2 times in total.
A Bona Fide Soggy Bottom Boy
User avatar
Bolsterman
Posts: 102
Joined: Thu Nov 12, 2009 12:36 pm
Location: Mexifornia

Post by Bolsterman »

A footnote. The GMT 4" vise I have came with T-slot keys on the base cut to a width of .499. That's too wide for the T-slots on the X3, published at .474 (although I measured them somewhat less). So I had to mill down the lower portion of the keys to about .472. This allows the vise to move just a little (a few thousandths) CW and CCW for precise tramming. Perhaps a more experienced machinist could tighten up those tolerances and not need any hammer-tapping at all.

The manufacturer (GMT) said the keys were hardened, but they in fact were not, Rockwell hardness was about 13, so it made for an evening's project.
A Bona Fide Soggy Bottom Boy
User avatar
Harold_V
Posts: 20227
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Post by Harold_V »

Bolsterman wrote:if so I'll correct and repost.
The only thing that is correct is the first set of instructions, which will orient a vise parallel with the travel of the table. None of those procedures will be an indicator of the position of the head. They will be an indication of how parallel the surfaces are with the table, however, but that does you no good in setting up the head.

In order to set the head, it's important that the DTI spin about a common, fixed axis (the spindle). Mount it such that it can trail the tip when it is rotated, and that it swings an arc that is the size of the width of the table, front to back. Spin the indicator (by hand) with the tip in slight contact with the table. That will describe the lean of the head and/or column.

It's very important that you correct any error in the column before making any adjustments on the head, so when the head is moved you don't lose orientation with the part.

None of the operations you will do when setting the column and head should include a second object. Do not indicate off anything but the table surface. By using any other type of device, even a flat plate, you risk introduction of error.

Does any of this help?

Harold
User avatar
Bolsterman
Posts: 102
Joined: Thu Nov 12, 2009 12:36 pm
Location: Mexifornia

Post by Bolsterman »

Yes, thank you...I need to lose the vise and tram referencing the table only. And it sounds like you're recommending I make myself a sweep bar (not sure of the term, that's what we called it in class). I don't own one, so I'll see if I can make one. But just to double-check: We're talking about chucking a horizontal arm that pivots from the spindle axis, that holds an indicator on the end. Correct?
Last edited by Bolsterman on Tue Jan 05, 2010 7:23 pm, edited 1 time in total.
A Bona Fide Soggy Bottom Boy
User avatar
Harold_V
Posts: 20227
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Post by Harold_V »

Bolsterman wrote:Yes, thank you...I need to lose the vise and tram referencing the table only. And it sounds like you're recommending I make myself a sweep bar (not sure of the term, that's what we called it in class). I don't own one, so I'll see if I can make one. But just to double-check: We're talking about chucking a horizontal arm that pivots from the spindle axis, perhaps 12" in length, that holds an indicator on the end. Correct?
That's pretty much it. The length should be determined by the width of the mill table. As long as you can sweep front to rear, near, but not on, the edges, that's all the larger it need be. A simple 3/8" rod bent @ a ± 90° angle that can be held in a drill chuck is more than adequate.

When you perform this operation, have the table set in the middle of its travel, with the spindle set @ the center of the table, front to back. It all makes sense when you apply the indicator, which should be in contact with the table only enough to give a reading through the full circle. That way there's very little tip movement when you go across the T slots. Also, insure that the indicator point is trailing off, which minimizes shock.

Harold
User avatar
Bolsterman
Posts: 102
Joined: Thu Nov 12, 2009 12:36 pm
Location: Mexifornia

Post by Bolsterman »

I’m not too proud to say it: I COULD USE SOME HELP. I think I am losing the battle with this mill, here. The help I need is at the end of this evening’s post. But let me first get you up to speed on how the project’s going.

Because tram changed when the mill was moved, I suspected “gunk” in the mating surface between the column and the table assembly. Needed to pull the column and clean it out. That required stronger garage rafters, so two days were spent “fortifying” the garage ceiling. You can see all the sistered 2x4s in the photo below. There’s a 1-1/2 square tube up in the top, which spans two beams. That’s holding the cable puller.

Image

Once I was satisfied I could hold the load, I put the location pins back in place in the head, (basically duplicating factory tram for the head), and pulled the column. It was a tense hour. I made one worrysome error, I didn’t adequately constrain the web wrapped around the head, to the column…so when I lifted, the column tilted more severely than expected. I should have bound the web around the head to the column tightly; I would have gotten a more vertical lift.

Image

For some reason this process spooked me. I was very nervous the whole time. I had to go in and stone the mating surfaces, and I could just imagine the cable puller (rated at 1 ton with a 2:1 safety margin, and I’m lifting what, 200 lbs? 1/10 the rated weight?) giving way while my hand was near (and occasionally under) the column. It didn’t, and I had no warning signs that the lift was dangerous, but still…I’m glad it’s over. All mating surfaces got a couple hundred short swipes with a fine stone. The surfaces were not pretty, but now there’s no obvious high spots.

Image

Image
A Bona Fide Soggy Bottom Boy
User avatar
Bolsterman
Posts: 102
Joined: Thu Nov 12, 2009 12:36 pm
Location: Mexifornia

Post by Bolsterman »

So when I got the column bolted back to the table assembly, I started tramming all over again. I put the factory locating pins in the base so it was how the factory had it.

Previously I had machined an extension to my “grasshopper leg” so I could use the sweep bar technique recommended by Harold_V. I put a dial test indicator on it, and swung it in a circle (turns out to be 4-3/4” in diameter). Factory tram was out .006, with the column leaning to the right. Then I pulled the locating pins, put my .007 shim back under the right side bolts at the base of the column (which had worked for me previously) and my left-to-right tram became a manageable .0005 to .001 (more or less, depending on where on the table you measure) over 4-3/4.” Still leaning slightly right. I could use a somewhat larger shim, it appears, I’ll try a .008 soon.

Image

Front to back (same diameter) is a little worse, hovering around .0015, and indicating that the head is dropping its “chin” a little in the nod direction.

What’s driving me crazy, is that my double-check method (using an angle plate) is not agreeing with the sweep bar method.

When I put my new ground 5” angle plate (supposedly with tolerance of .0005, parallel .005, square .001 over 6”) on the table, with the face of it facing the indicator (squared on the table using a machinist’s square) and run the indicator up and down the face of it using the Z axis crank, I am seeing .006 variance, telling me the column is still leaning way right! How is that possible, when the tram referencing the table is around .001? If I take the square and spin it 180 degrees, and place it on the other end of the table and try again, I get very similar “mirror image” readings, telling me that I’m not dealing with an out-of-square angle plate. For some reason moving the Z axis up and down 5” gets this .006 out-of-tram reading telling me the column is still leaning right and I need way more shim. (Running the indicator up and down the face of the angle plate with the quill downfeed is even uglier, up to .010 out.) Can anyone help me reconcile these two methods of checking tram? I'd think they should give similar results.

Also, I contacted Griz about the table, which is just around .003 variance overall and has a .001 dip in the right side (mentioned earlier in this thread), which makes it hard to check tram (you have to avoid that spot when measuring.) I wanted to know about their table grinding service they advertise. I learned they don't offer it anymore. Regarding the table, they said: “It’s out of spec, send it back. That table should be within .002 with no dip.”

So, given the dippy table that’s out of spec, and the fact that the base must be at least .007 out of register (seems like a lot of shim to me!), and my inability to get it trammed well, I am seriously considering packing it up and sending it back at this point. What do you think I should do, or try, before sending it back? Thanks, B.
A Bona Fide Soggy Bottom Boy
User avatar
Harold_V
Posts: 20227
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Post by Harold_V »

Bolsterman wrote:Front to back (same diameter) is a little worse, hovering around .0015, and indicating that the head is dropping its “chin” a little in the nod direction.
Get it right side to side, then that's easy. A thou shim placed along the front edge should bring it in. You can use a piece of aluminum foil.
What’s driving me crazy, is that my double-check method (using an angle plate) is not agreeing with the sweep bar method.

When I put my new ground 5” angle plate (supposedly with tolerance of .0005, parallel .005, square .001 over 6”) on the table, with the face of it facing the indicator (squared on the table using a machinist’s square) and run the indicator up and down the face of it using the Z axis crank, I am seeing .006 variance, telling me the column is still leaning way right! How is that possible, when the tram referencing the table is around .001?
That is the results of a column that leans, but has been corrected by rotating the head. You are experiencing the very condition I spoke of some time ago, one that has been chiefly ignored by many.

In order to determine the error, you must not use the quill. It's OK to mount your indicator in a collet, but make sure the quill is not moved. Move the head up and down, then shim accordingly. Once you have the head reading within an acceptable limit (I'd shoot for less than a thou over five inches), you can then correct the error you see when you sweep the table with an indicator. Once that's done, you should be able to get nearly identical readings moving the head, or by extending the quill.
Can anyone help me reconcile these two methods of checking tram? I'd think they should give similar results.
They will, once you have corrected the error.
Also, I contacted Griz about the table, which is just around .003 variance overall and has a .001 dip in the right side (mentioned earlier in this thread), which makes it hard to check tram (you have to avoid that spot when measuring.) I wanted to know about their table grinding service they advertise. I learned they don't offer it anymore. Regarding the table, they said: “It’s out of spec, send it back. That table should be within .002 with no dip.”

So, given the dippy table that’s out of spec, and the fact that the base must be at least .007 out of register (seems like a lot of shim to me!), and my inability to get it trammed well, I am seriously considering packing it up and sending it back at this point. What do you think I should do, or try, before sending it back? Thanks, B.
Getting one better is more or less the luck of the draw.
I have talked frequently about the concept of roughing, then finish machining. Castings are not exempt from the issues that come from unevenly removing metal from surfaces. In order for them to have produced a table that is free from distortion, it would have been required to be aged, rough machined, aged again, then finish machined. There's no doubt in my mind, they don't do any of that for inexpensive machines.

The dip you speak of can manifest itself by more than one way. The top surface of the table may or may not be straight, much the same way that the way surfaces may or may not be straight. Re-machining any one of the features may or may not improve its quality. I hesitate to suggest you send it back, although you may get one that is better.

I do believe it's important for you to come to terms with the fact that you did not purchase what might be considered a precision machine. It would be a hard call to expect it to perform to anything less than a thou, and even that may be taxing. I know it is on a BP, which is a slightly better machine. It's a far cry from the machine it's reputed to be by those that have never operated better quality machines, however.

We can walk you through this one if you're willing to tolerate the irregular table. If it brings you any comfort, my BP, which was purchased new, isn't flawless either, but you learn to work within it's capabilities. Unless you get involved in tenths work, I expect what you have will serve you adequately.

One thing that is quite important when doing these things is that the gib must be snug. If it's loose, you often experience shifting, which only adds to the problem, but may not be an indication of what the machine is about. Make sure your gibs are snug, but not so tight that the machine requires effort beyond reason to move.

Ball is in your court.

Harold
User avatar
DICKEYBIRD
Posts: 176
Joined: Sat Feb 03, 2007 10:45 am
Location: Collierville, TN

Post by DICKEYBIRD »

My X3 is a bit "iffy" in the head/column junction. The head doesn't lock into the same position consistently, depending on how much torque is applied to the lock handle. You might want to check yours in that area. Without consistency there, attempts to super-accurately tram the mill would be futile.

One of these days, I'll strip the head and gib apart to chase the problem down but right now I'm making decent parts and having too much fun with it.
Milton in Tennessee

"Accuracy is the sum total of your compensating mistakes."
gunboatbay
Posts: 134
Joined: Wed Nov 12, 2008 10:09 am
Location: Pender Harbour, B.C.

Post by gunboatbay »

Bolsterman,
I think before you go any further you should thoroughly investigate the gib situation that Harold mentioned in his last post. I have a Grizzly/SX3, which I think has exactly the same gib setup on the head/column interface. I had the same problems you're experiencing and finally tracked it down to the gib strip itself being rough and uneven. I lapped it smooth/even/level on a granite plate with wet/dry paper and lots of elbow grease. When I finally got it smooth/level/even, I'd removed so much metal that the taper allowed the lower end of the gib strip to project below the head when properly adjusted, but it had solved the problem. I have since obtained a new gib strip from grizzly. It came as a totally blank tapered gib strip, but was very straight and even. I had to machine the adjustment teeth on the top end, drill the oil port and mill-out the slot that serves as kind of a half-ass oil reservoir. After I did that, I lapped out the 'chinese-finish' on both front and back. I'm very happy with the results. I would also like to mention that adjusting this gib is nothing like adjusting the gibs on the x or y axis, or gibs on a lathe. The weight and lever-arm of the head puts a lot of pressure on the head/column interface (and gib), giving you a false feeling that the gib is properly adjusted, when in reality it's not snug enough.
User avatar
Bolsterman
Posts: 102
Joined: Thu Nov 12, 2009 12:36 pm
Location: Mexifornia

Post by Bolsterman »

First of all, thank you for your invaluable advice. You guys are my lifeline!
Harold_V wrote:I do believe it's important for you to come to terms with the fact that you did not purchase what might be considered a precision machine...Make sure your gibs are snug, but not so tight that the machine requires effort beyond reason to move.
Indeed. I would be pleased if I could keep the machine itself square within .001 over 5" and would settle for .002. As far as projects are concerned, I'm hoping I can hold .003 on small items (think 5" or less) and would grudgingly settle for .005. I have snugged the gibs until I feel binding and then let them off a scootch, which was my prof's recommendation, although I could tighten them up a bit more. While I did remove the X and Y gibs and stone them a little, I did not remove the Z gib for inspection.

I think I understand your advice about method sequence now (I misunderstood it the first time you gave it!) but just to make sure: First you get the column square left/right by using the indicator running vertically against an angle plate. You use the Z-axis crank only (not the quill feed) to move the head up and down. THEN you perform the sweep method referencing against the table, and rotate the head as needed to get that as good as possible. When it comes to correcting front/back, if the chin is dripping, I'd need to place a little shim under the bottom two bolts holding the head to the column, right?
DICKEYBIRD wrote: My X3 is a bit "iffy" in the head/column junction. The head doesn't lock into the same position consistently, depending on how much torque is applied to the lock handle. You might want to check yours in that area...
I've noticed that my head shifts position between .002 and .003 when tightened (with the "chin" coming up, obviously), and the amount you tighten it does make a difference! I've wondered how to deal with this when Z-axis movement is necessary to make the cut. With testing so far, it seems it'll cut up to .003 deeper if the head isn't locked. Well, you can't lock it when you need to plunge a cut, so...here's maybe an .003 I'll just have to live with.
gunboatbay wrote:Bolsterman,
...I had the same problems you're experiencing and finally tracked it down to the gib strip itself being rough and uneven....I would also like to mention that adjusting this gib is nothing like adjusting the gibs on the x or y axis, or gibs on a lathe. The weight and lever-arm of the head puts a lot of pressure on the head/column interface (and gib), giving you a false feeling that the gib is properly adjusted, when in reality it's not snug enough.
That sounds like an excellent lead, and what you say makes sense. I have not separated the head from the column yet, and it sounds like I should at least pull the gib and get it nice and flat. Do you have to pull the head to do that or does the gib just lift out?

I have been corresponding with Grizzly who on several occasions have said "Send it back" and am still leaning that direction. I'll keep you abreast in the next update. Thanks again for all the help, B.
A Bona Fide Soggy Bottom Boy
Post Reply