What's new
What's new

Spindle tram on a CNC

Adding to this: my VF-2 only has four leveling feet so this doesn't apply but, the VF-5XT I leveled after a move had six.

What the videos and procedures allude to is the middle feet are for taking sag out of the bed. That's only partly true. The true goal of those middle feet is to get the column vertical to the world. Once the table is level at the four corners, one would ideally put a precision granite block, cylindrical standard or something like that on the table and then sweep an indicator up and down in Z on both the side and forward positions. Dial up the torque on the two middle feet to get the Z traveling square to the table from both side and front.

Once that Z is dialed in to vertical, you'd finish with tramming the cartridge as we've been discussing here.

As a trick, I hooked up a GoPro linked to my phone so I could watch the bubble results (Starrett 199z) while laying under the machine and adjusting the feet. I didn't have a standard to check Z on that one so had to settle for a level bed. Took about four hours to get it within a few tenths front-to-back and side-to-side.
 
  • Like
Reactions: Ox
The remote vue is a swell trick if you are working alone!

You know how it goes: you want to take your time to get something right and finding someone with the patience to stand there for hours chasing a bubble around a level is tough. The remote camera meant I could lay on the ground, nudge the wrench and wait for the bubble to settle. Nudge it again, wait to settle.
 
  • Like
Reactions: Ox
Do I say something stupid if I would sugjest scraping ???
Peter
Maybe for someone braver than me. I don't know where the 'problem' is. It could be the spindle flange and changing this one in the future could tram differently. I'd hate to scrape the spindle face to match a non-square spindle flange.

Could be the mounting face of the head casting. Could be how the head casting was aligned to the Z axis guide-rails. Could be the face of the guide-rails on the column or even the column where it mates to the bed. I don't have the space or equipment to try to correct the many possible source(s). At the same time, I'm trying to maximize what I can without blueprinting the whole machine. Not saying that scraping isn't the answer but, it's beyond my talent and bandwidth.

Another related story: the other machine we trammed was a Haas GR-712 Gantry router with the extended Z axis. It's a 7x12 foot aluminum plate bed on a welded steel frame. It sits on four feet at the corners and that's it. No hope of getting it "flat."

What we were able to do was sweep in the tram on the spindle, then took a series of light skim cuts over the entire table. It was 2010 and my memory is fuzzy but, it may have taken as much as 0.020" material to come off before the entire table was surfaced to a constant Z. I didn't say it was flat but, it was at least as close as anything that could conform to the table bow it might have. I finished with a roloc pad and a criss-cross facing pattern over the whole thing.

With that done, we were able to hold stupid thickness tolerances on it with sheet materials. We went as far as ball milling tapered edges on sheet material that was held down to similarly matched vacuum plates. It also made large aircraft structural components it had no business doing. Not the best machine for the job but, a lot of value for what we paid.
 
Could be the mounting face of the head casting. Could be how the head casting was aligned to the Z axis guide-rails.


OK, here's a thought....

If the cartridge is able to be indexed, then maybe you clock it one hole, and check aggin?

Clock it aggin....

I'd rather try that as to try dicking with .0005 variable shims!


----------------------

Think Snow Eh!
Ox
 
OK, here's a thought....

If the cartridge is able to be indexed, then maybe you clock it one hole, and check aggin?

Clock it aggin....

I'd rather try that as to try dicking with .0005 variable shims!

If the spindle were out of the machine, the motor loose and the belt drive was off, this would be completely worth trying. I like this idea.

In the case of this spindle, there is an air/oil line (bearing lube) that snakes down below the drive belt. Rotating the spindle would change the location of where that plumbs in and changing that would require loosening the motor, removing the belt, etc.

Unless someone has another clever idea, this is what I had in mind:
  1. Mount a large block of aluminum and take a facing pass with an endmill no larger than an inch. Small stepovers to minimize any waviness. This will give a surface as close to flat and parallel to actual XY axes and eliminates table irregularities from the measurement.
  2. Loosen the screws and remove any factory shims from the spindle. Retighten.
  3. Mount a tenths indicator and sweep the newly machined surface. Find the high and low spot amount and a clocking angle of that axis.
  4. Make two half-circular shims out of maybe 0.005 shim stock.
  5. Taper-grind the shims from the high to low value.
  6. Install shims
  7. Something
  8. Profit!
Seems like a lot of work but, if that's the only answer...
 
If you are able to drop the spindle enough to R @ R shims, then you are able to spin it.

This is regarding the motor loose.

Now the air line thing, that may be a horsey of a 'nother color.

But I don't see replacing the shims w/o loosening the belt?
And I wouldn't expect that you would re-test shims w/o tightening the belt back up first - each time.

???


----------------

Think Snow Eh!
Ox
 
On the ones in the past, we didn't remove the screws completely. We loosened them enough to drop the spindle maybe 1/4". That made the shims reachable with tweezers or needle-nose pliers.

And yes, as I'm sure you're concerned too, if the spindle isn't located in a precision bore further up, is all that belt tension playing into how square this thing is? How about pulling the spindle harder under cutting load? Not a good design. Yes, all measurements with the belt installed and torqued to spec.

I even wonder how much fastener torque affects the tram with plastic shims instead of metal. Could another 5 ft/lbs on one side be enough to pull it that way? Would a proper, tapered metal shim result in better part finish because the spindle is better mated to the mass of the casting?
 
You don't need to "drop it" at all to spin it.
Pull the bolts and - spin it.

Doo you have a large chain wrench?


----------------------

Think Snow Eh!
Ox
 
Yeah, there is no way to not foul that air/oil line without removing the belt. I can't even see where it connects to the spindle. If I did change the clocking, the location of that would change. That's a whole new run of plastic tubing and new routing. To accomplish that, the belt has to come out.

Removing the belt means loosening a bunch of other stuff. This is a two-speed gearbox head so that comes loose, encoder belt gets disconnected, etc. This is already too much of a science project. I'd rather make a custom shim and be done. I was wondering what others have done. I can't be the only person to attempt this.
 
Most people aren't expecting perfection from their Haas.

If'n y'all git this all dialed in like you want, I would like to hear what it trams after running this big fat cutter for a while.

I take it that you are wanting to travers a plate in one finish pass - that you want it so perfect for a big basket cutter?
Shirley you're not roughing with such a big mill?


----------------

Think Snow Eh!
Ox
 
I have a particular 3", 45 degree, 6 insert facemill. Standard stuff. It would be nice to get a gasket-worthy, flat surface in as few passes as possible. Don't want to spend a ton of time doing 1/2" stepovers, rubbing the same surfaces over and over, trying to achieve acceptable flatness.

I did four composite molds about a week ago that needed a good face-to-face seal to limit material squeeze-out. It wasn't anything more than 11x12" aluminum plates with clamping holes drilled around the periphery. They needed a smooth, flat surface to close off the mold cavities cut into one side.

It wasn't a failure but, the cavities did leak. The squeeze-out is easily trimmed but, getting a tight seal eliminates the flash and nets a better part.
 
Last edited:
Donkey Hotey,

It might not help in your particular application, but I've found that the tolerances of shim stock can sometimes help me build stacks of shims with small differences in thickness, on the order of 1 or 2 tenths. I sort through my candidate shims with a mic and build a couple of stacks that "should be" the same height, but aren't. Obviously, shimming around your spindle flange is a harder puzzle than just building two stacks of shims.

The shims have to be nice, though. No wrinkles, no oxidation, and no burr from cutting it with scissors, etc. I pinch the shim stock between two pieces of fine wet-or-dry paper on a flat surface and pull it out, repeatedly, until it looks good and a mic cannot detect a burr where I cut the material.

Regarding the machine feet: Our 2019 VF-2 has 6 feet and I'm not convinced the middle feet are capable of changing the geometry of the machine in a helpful, predictable way. I think the machine is just too small for the middle feet to take sag out of the y-axis or to change the tilt of the z-axis. I have seen several different versions of Haas's official leveling advice, including one where they advised that the middle feet only lightly touch the floor. I find that even light pressure on the middle feet causes me to chase my tail trying get the thing level...while firm pressure on the middle feet does nothing to correct the apparent sag in the y-axis. So, I level with the 4 corner feet, then adjust the middle feet to gently kiss the floor.
 
Another proven way is to drill and tap some pushholes
Align the spindle with the pull and push bolts leaving a gap of about 2mm or less
Then fill the gap with a low shrinking hard epoxy using a sealant gun
The last couple of 0.001mm can be done by tightening the bolts more or less
But wow what a crappy design All american ?????

Peter
 
It might not help in your particular application, but I've found that the tolerances of shim stock can sometimes help me build stacks of shims with small differences in thickness, on the order of 1 or 2 tenths. I sort through my candidate shims with a mic and build a couple of stacks that "should be" the same height, but aren't. Obviously, shimming around your spindle flange is a harder puzzle than just building two stacks of shims.

Good tips. I haven't accumulated enough assorted shim stock to get there but, these all sound like possibilities. Thank you.


Regarding the machine feet: Our 2019 VF-2 has 6 feet and I'm not convinced the middle feet are capable of changing the geometry of the machine in a helpful, predictable way. I think the machine is just too small for the middle feet to take sag out of the y-axis or to change the tilt of the z-axis. I have seen several different versions of Haas's official leveling advice, including one where they advised that the middle feet only lightly touch the floor. I find that even light pressure on the middle feet causes me to chase my tail trying get the thing level...while firm pressure on the middle feet does nothing to correct the apparent sag in the y-axis. So, I level with the 4 corner feet, then adjust the middle feet to gently kiss the floor.

I have to wonder if Haas added the extra feet just in case? Or maybe they took material out to make them cheaper and lighter and the extra feet were a backup in case they sag? My VF-2 is 2004 vintage. It was leveled upon installation and it never changed after that. Impressively, it's still flat-flat all these years later.
 
I did not see it mentioned that the first step in tramming the spindle is to lock the spindle and travel to the sweep areas of the table to look for any tilt or slump in the table.
If your table is .001 low in the front don't try for zero when adjusting tram, the spindle tram should show the same table error to be square to axis movement.
 








 
Back
Top