Worked through this in my head a couple times probably missing an easier solution. you could do the same thing with your indicator, choose a rail and shim behind the end of the rail of your choice until the indicator reads zero, then cut a 2X2 square and measure the diagonals on your square. If its good they will be equal and if you shimmed both the rails to cocked then the diagonals will not match. If you cut a larger square you will be able to catch more error. Limited to your caliper size. Can also cut a couple along the X axis to see if the error is the same in each portion
1 Like
To detect which rail has a bow I would put a 2-4-6 block on the fixture plate and square it to Y.
Go through gantry squaring process as best you can then sweep the 6in face of block wirh dial indicator mounted to the surface of top rail and then the bottom rail. Would be a pain to do the entire length of the rail with just 246 blocks but should be able to detect which rail is the worst.
I tried loosening up one edge of my top rail to see if I could pull the edge forward while tramming. I could deflect maybe .002 by hand but when I released pressure it would pull back to where it was. I do not have the clamps that would let me pull it into place while tightening the screws to keep it in place. Also fighting the edges machined into the X-gantry will be a chore.
2 Likes
Checking the upper and lower rails against a reference would prove to be difficult without a much longer indicator base as you canāt reach that far even with my NOGA. I did remove the upper linear rail again and sweep the machined surface of the gantry with the indicator. I put the NOGA base in the center of the two lower bearing blocks with the indicator tip centered above them that way I can get as long of a sweep as possible without resetting the indicator. So the readings below are a few inches short of each end of the gantry. Other checks I did at each end would indicate the number continue to fall as expected.
Left to Right on the flat between the screw holes the reference edge for the rail
0 -2 -3 -4 -5 -6 -6.5
Left to Right reading on the edge
-7 -5 -2 0 +2 +2 0
Again, Iām not sure the issue is really the upper rail, it could be the lower rail.
Not sure what method Daniel is suggesting. For the first set of numbers, I could shim under the rail to bring up to zero. Not sure how you would correct for the second set of numbers.
Your killed me Michael 
I just spent $250 on a 2.5" bit because this mirror you created looks so darn good. I am getting ready to try this fly cut. Any feeds and speeds you might share?
I took my first pass with a 1/2 and noticed a a small .002" ridge between the rows like the stepover is .502" Anyone else see this?
If its flat leave it alone. You may need that. 005 someday. Two or three projects and it will be dull and scratched up.
1 Like
Its not flat yet but soon to be. I want to be able to brush my 2 hairs off that base plate 
I already knew what you were going to do. 
I got the same and realized that these ridges are due to a slight inclination forward (nod) of the spindle. I will measure the amount soon. Itās funny because from the Langmuir calculator I got that I have to put 2 shims on the upper carriege of the z assembly, which I wonder if that was incorrect. I will eventually remove these shims and re-check.
1 Like
Good call. I decided to recheck after you posted your findings and you are 100% on. I have a low cut on the back end of the bit so the top of the spindle is setting to far back. mine suggested a shim in the lower carriage of the spindle which is showing if I remove that shim it would be right on. 
I will have to check this calculation again but I did the calculation 2-3 times before I added anything
Not sure what reference Langmuir take to come up with these calculation, but I think there could be many variables involved during the full assembly and these new errors cannot be predicted until fully assembled. I wonder if these shims should be placed at all at the stage of mounting the Z assembly, or just wait till the end when the plate is milled.
They said the calculator on the website was coded wrong. I assume itās fixed by now. I was meaning to re-enter my numbers to see the difference, but havenāt yet.
1 Like
Langmuir fly cutter setup at 2" diameter and 1.75" step over at 1100 RPM and 10 IPM. Wonderful finish. Just wish I could get the tram even to reduce that ridge. Iāve been side tracked so I havenāt been able to work on the machine for better than a week. I need to get it put back together.
1 Like
Great info. Thanks Michael. I am going to hold off until we figure out this tram deal so if anyone is seeing this keep posting your findings. Does anyone know if Langmuir has definitely fixed this issue and we are good to recalculate?? I dont want to do this again
The shim calculator has now been corrected on the website.
3 Likes
Langmuir fixed the issue. It only impacted some calculations (I think it was wrong in one direction). Mine for example was the same after the fix. I think it was Mike Downs that pointed out that the shim calc is only a starting point. After the base is surfaced then you need to fine tune it. Itās one the small things that should have more emphasis in the build doc along with checking spindle tram tilt side to side. I had heard they were doing a video but I havenāt seen it. My issue is I am seeing different tram numbers based on where the carriage is on the gantry.
1 Like
At the risk of speaking about something I havenāt done myself yetā¦getting the Z axis travel perpendicular to the XY plane and getting the spindle axis perpendicular to the XY plane will have similar effects on an operation like fly cutting the baseplate, but they are separate adjustments that need to be addressed one at a time. I feel like there is some confusion about this based on posts here and personally I find that the terms ānodā and ātiltā are being overused. my suggestion would be to use those for the adjustment of the spindle as that is familiar to many who may have used a knee mill where the spindle does not ride on the Z-axis. āsquarenessā is a more familiar term to me to describe the adjustment for the Z axis travel itself (i.e. XZ squareness and YZ squareness) and would help disambiguation between these two things. Also Iām not sure if the spindle adjustments have been detailed in assembly manual/youtube, but I believe they have been described on the forum, so some people may just be missing that information.
1 Like
It would also be worth checking tram at the 4 corners of the fixture plate as well as the middle.
As I got to within .001 across 4.5 inches on my nod, and faced with the fly cutter, you can see areas of travel across the fixture plate where there is a change in nod forward to nod backwards.
This is a discrepancy that is coming from the linear rails on the X-gantry likely being a few .001 out of co-planarity.
The swirl and ridge patterns in the fixture plate are plane as day while also going from detecting the scale/ridge with your finger nail in one direction but not the other on the left side and then it being the opposite on the right.
You can roughly estimate where in the gantry the transition is because there is a baby butt smooth area with imperceivable ridge between passes.
My final adjustment in tram is likely to occur with me using my SMW pallets as a base plate to test against on both X and Y directions but the final NOD is going to be a pain unless I start backing out linear rail screws and using clamps to pull the rail into alignment while shimming before tughtening bolts.
Y-axis coplanarity is within .0005.
@MichaelBarker Did you ever manage to get your machine squared up?
Iām having a similar issue while working everything in on my build.
Skimmed the plate; adjusted nod and tram within .001" over 6"; skimmed the plate again; adjusted the nod and tram down to .0002" over 6"; resurfaced the plate to nearly clean up and decided to check nod and tram on a whim before finishing.
The spindle was in a different location on the table and reading .004" tram over 6" and .002" nod over 6". Moving the spindle back closer to the earlier position brought those numbers down to .001" tram over 6" and .0005" nod over 6". So it looks like twist to me. Not looking forward to the idea of tearing the top end down again after all the shimming to get it this square, but it may need to be done.
Just looking for some more insight before I start unbolting everything to find the twist.
2 Likes
So the short answer is no I donāt have a solution yet. I did take it apart and fiddle with it and put it back together and it still wasnāt right. Its been long enough I donāt recall out far out it was. I got frustrated with it and just walked away. Since then Iāve had various other things that have kept me from getting back to it. I personally think the gantry has a twist in the weldment. Everyone talks about shimming it but thatās hard to do when the shim is required between the edge of the liner rail and edge machined in the gantry. This was one of the very first machines shipped and turns out I could have just waited. Sorry I donāt have a better answer. Iād love to know what you end up doing.
1 Like
Why fly cut the base? It will never look like it will again after your first project. Beside if your mill is not tramed dead nuts on you will have a mess. This is a garage style mill. Having it out in the hundred thousand range is maybe the best its going to get. I was lucky that mine came out almost perfect. There is just two many variations that could effect it. If its out beyond specs Langmuir should be eager to assist you. I think they are just a little over whelmed right now. Squeaky wheel gets the oil if you know what I mean.
Just my two centsā¦
3 Likes