Serious alignment issues, twisted Y?

Community Laser Talk
1

I need a validation or rejection of my theory that this machine has a twist in Y, non-parallel rails.

Trying to sort out a 900x600 80w rdc6445 machine. Spent 2 full days trying to align and shim it with no success. I’ve done quite a few of lasers, but this one throws me for a loop.

Location index looking from the front of the machine, beam comes in from top left:
1 - top left
2 - bottom left
3 - bottom right
4 - top right

The person who got this machine explained to me that doing usual alignment he’s getting beam travel between location 3 and 4, while 1-2 and 2-3 are perfectly centered and spot on. For me its a clear sign of twist in Y, non-parallel rails.

So i went to see myself. I, foolishly, assumed tube set from factory would be more or less ok (on that later).
Upon trying to align it I, as well, would get perfect alignment in 1-2-3 locations, but 4 would always strike high, by quite a lot (~5-6mm). And no matter mirror twisting i couldn’t get it out.
So i resorted to shimming. Tried all 4 rail mount locations, but the only location that made any difference was top of the left rail (position 1). So i started shimming it up by sticking ~1mm thick shims under the mount foot. By sticking 4 shims i reduced travel between pos3-4 by half (to about 3mm). At this point the work day was at an end, so i made machine workable, tightened down and left it as is.

Second try involved me completely rechecking all beam components (bar mirror1). What i discovered was that tube mounts were wayyyyy off parallel to the machine in both vertical and horizontal positions, by significant margin. So i re-spaced tube to strike center of mirror 1 and tried realigning again (after removing all the shimming, to put machine the way it came from factory). Exactly the same thing happened - pos1-2-3 perfect, 3-4 walk away by ~5-6mm. So i shimmed it again to make the machine somewhat workable again (walk of ~2mm), as it was close time…again… The only thing i never moved was mirror1, its still as it was set in the factory.

By my estimation i would need to shim 8mm (total, ±), but i cant do that for tube mounts don’t go high enough and i can’t adjust mirror2 and 3-head vertically. So now i’m just scratching my head and trying to figure out what next steps i should take.

Machine is sitting on concrete floor, feet had been leveled.

All of this is pointing at machine being twisted, or rails not parallel. I need someone else to either confirm or deny my thinking. With constructive criticism, preferably…

HELP :smiley:

Some pics:
As the machine sits now, with 4 shims in…

image
image1199×899 56.9 KB

The gradual shimming test result, gap gets smaller more shims i add, all the shot columns are of pos3-4, then tape moved over, shimmed up and next test. (Tests were done with realigned tube)…

image
image674×899 45.5 KB

Next pictures illustrate how badly tube mounts were set. Note screws in the foot and gap between base and vertical adjustment difference. Effectively tube was sitting looking up and towards mirror1 (looking from back of the machine). I have now leveled and spaced the tube properly.

image
image1199×899 50.9 KB

image
image1199×899 64.5 KB

2

No one ? This thing doing my head in…

3

Do you have a dial indicator you can use to check runout on your rails? Just a thought
Charlie

4

I do have a dial indicator but no mag base for it :smiley: Will be visiting machine next week, reset the tube precisely (made a ton of spacers/gauge blocks in 1mm increments out of mdf). Will be doing a ton of measurement.

What i’m surprised is how little information there is with regards to 4th corner problem solving.

My “gauge blocks” below, 5-60mm in 1mm increments, 2 sets :smiley:

image
image1199×899 157 KB

image
image1199×899 124 KB

1 Like
5

Hi
The 4th corner mystery.
To start with watch this video by Russ Sadler:

and

1 Like
6

Thats not the situation. My beam doesnt drift horizontally at all. All 4 corners are vertically perfectly on center, but one corner gets high burn compared to opposite (in Y).

7

OK, sorry. Not sure what you mean by high burn.

8

If position 3 (Bottom-R) is spot on, position 4 (Top-R) will be above what pos3 shot was. Always centered but above “good alignment” position. It does not travel horizontally at all, only vertically. No diagonal adjustment helps any of this.

9

Is the frame twisted somehow? Either built that way or the box got twisted in transit?

Does it have leveling feet? Is it possible the frame isn’t rigid enough and raising or lowering a corner with the adjustable feet twists it a little?

10

Is there a possibility that your table is a bit higher/lower in Position 4?

11

It does have feet, and verified exactly the same distance case to floor. I’m thinking i might try to misalign top left foot on purpose, as thats the position i’m shimming the rail at to reduce vertical walk of the beam.

Table doesnt even come into this at all. I’m not leveling table until there is perfect(ish) alignment of the beam. No point in trying to get focus distance when beam is all over the place…

12

Well, many more of ideas were tried now.

I specifically unleveled the feet of the machine until it rocked on 2 points only, no deformation in chasis whatsoever, no effect on beam alignment. Tried to “lift” the left back corner of the machine (where we shim to lessen the vertical walk) - no effect.

I’ve rechecked all motion components for loose parts, nothing rattles, nothing’s loose… Gantry, idlers, rails, gantry-to-rail blocks, etc…

Then, as per many, many, MANY suggestions i straightened the tube to within a millimeter of the tube compartment (51mm from bottom, 19mm from the back wall), then strike mirror1 and go from there… No effect to the problem.

Then we received a video-call from an American Photonics gentleman, who, after receiving all the information confirmed that the machine is twisted. I also was assured that, really, the positioning of the tube doesn’t really matter (as i suspected all along, unpopular opinion :smiley: ) as long as 1st mirror is struck in the center properly. With this in mind he gave me an idea on how to raise the beam without raising the tube higher than the mounts will go. It made a bit more difficult to steer the beam in diagonals, but it is fairly simple, if tricky. None of this helped to alleviate the problem.

In the end i put the machine with ~2mm vertical misalignment back as good as i could with 4mm shimming on TL rail, aligned perfectly at middle of right rail (instead of bottom right as i usually do) to spread the error a bit better throughout Y and left it there.

Now, ALL of the attempts to align the machine ended with the same and CONSISTENT result - TL perfect, BL perfect, BR perfect, TR high. Tube nice, tube crooked, unshimmed (error ~6+mm), all the same result - TR high.

This is proper tube aligned and 1st mirror strike:

image
image1199×899 52.3 KB

image
image1199×899 71.7 KB

image
image1199×899 68.6 KB

image
image1199×899 93.6 KB

image
image1199×899 62.9 KB

Then there’s this - mirror1 holder plate (not the holder itself) mounted off 45deg at the factory. QC my butt, someone just eyed the installation of that. I corrected it before starting the tube position experiments:

image
image1199×899 52 KB

I was thinking of getting the reverse alignment tool from American photonics, but upon thinking on it all it would do in this situation is prove decisively that there is twist in Y, but show up on the left side (instead of right).

At the moment i’m at a complete loss here…

13

Well if the dude from American Photonics says something is tweaked, why continue? Replace or straighten the said part. I also find that their reverse mirror alignment is by far easier and faster if you don’t have a beam splitter (that has been properly adjusted). Even with a beam splitter I still do it in reverse.

14

I have the exact same issue, did you ever resolve? After aligning M1 and M2, the scorch mark at m3 drifts several mm as I test bed top right vs bed bottom rt. Similar to yours it moves down, but unlike yours it moves to the right as well. I’ve attached a photo of scorch marks at bed top right (top left scorch) vs bed bottom right (bottom right scorch), and with reverse beam alignment tool shining up the center of the lens tube. It seems like I need to shim my right side Y rail to match the left (which doesn’t look easy on my machine). Also I don’t see that fixing the horizontal drift, only the vertical.

Further, this whole solution doesn’t actually make sense because since the M2 is mounted to the gantry, even if the Y rails weren’t parallel, the mirror (and thus the beam) should still follow the gantry as it dips or rises? Bewildering problem but one that reduces the usable area of my bed (1400 x 900) by at least half.

IMG_5517
IMG_55171920×1440 239 KB

15

I have this issue too, and have come to the same conclusion - as the drift is on the X axis, with the mirrors effectively not moving in relation to eachother how on earth does the scorch/alignment change? Mine moves only in the vertical, by about 6mm. I’ve even raised and lowered the tube in the hope that it might make a difference, but once the first mirror is lined up tube position has no impact on anything.

16

Just looking through Russ’ excellent videos I found this one, and will be trying it shortly. I’ll poste my results…

17

I thought Russ’s idea of using Mirror 1 to fine tune alignment in the 4th corner was a stroke of genius, sadly it did not eliminate my problem, only improved it slightly. His videos do clarify that shimming rails is pointless as I suspected. Russ works on small format machines which are more forgiving of minute alignment errors. My 1400 x 900 machine provides almost 2.5 meters over which to propagate error.

Something I’ve not seen anyone mention is doing the initial alignment of M2 with the gantry hard to the back of the bed. This minimizes any error in the M1 alignment. If you do it at the front as is typically recommended, you are compounding whatever small error exists in your M1 alignment, and it should be no surprise if you find your 4th corner in disagreement with your 3rd.

Anyways let me know if you find a way to agreement. I’m considering buying a whole new set of rails, if not an entirely new machine.

18

I have the same issue.
Did you managed to resolve it?