My 770m has always made tapered cuts on the y axis, specifically cutting deeper on the movable jaw face. When I try and square a block of material, I put a parallel on the back jaw, then I have a broken em shank that I use on the movable jaw, I’ll face one side, flip the faced side to the back jaw with em in movable jaw, face that side, and they’re already not square, if I try and make two side parallel, one side is normally .005 less than the other side.
I just re leveled the machine, to .0002, twist is .0006, trimmed head to within .0003, but I still cant understand why my cuts are tapered. I even went as far to think that the back jaw was moving when I clamped down on the part, its not by the way, or the vise is tilted somehow, also not by the way, to say the least, everything that I could think of to check is within a half thou, so if someone could elaborate on what’s going on that’d be awesome.
Could it be the vice? If it’s a cheap vice they aren’t always ground well. Also it could be the spindle needs a shim. I had to shim mine out of the box and always check the spindle bolts are tight any time I start getting any misalignment signs. If that’s not it you might see if you can borrow a better engineers level the super accurate ones are super expensive but make a difference. Also the floor it’s on can move during the year (mine sits on a garage slab and it moves depending on how much rain we get, so level this month is tilted in June. If you’re desperate you can also scribe a line in the table and hang a plumb bob from the head and from the roof. This can tell you if the machine is tilting while moving. Or if the table is tilting and tracking straight.
Just trying to understand your methodology here… what are the dimensions of the material you are machining… is that plate stock that is flatish on 2 sides and rough sawn on the other 4 sides?
Its a tormach 5in cnc vise, I had said previously I shimmed the spindle and made sure the mill was level, which my level is .0002 resolution over 10in, I find it hard to believe that that’s not good enough, but say when im trying to make a block parallel, I think my piece is around 4 x 2 x 1, and the one side is 5 thou lower than the other side, then that would mean its cutting .0025 deeper on one side always, which led me to think that something that I would have checked would be over 1 thou out of where it should be, and nothing is really. The block was a piece of cold rolled, saw cut on the ends, im not clamping onto any saw cut portions either. What would the plumbob method actually be showing other than the things I would have already checked?
There’s nearly always some ‘lift’ in a movable jaw or it wouldn’t move although .005" is a lot.
To be sure, try facing opposite sides of a piece of stock held down to the table with clamps (face one side in the vise or where-ever, then flip and clamp straight to the table. You don’t need to face the whole thing on the second op to make clamping easier), then indicate, remove and measure.
Also it’s usually easy enough to remove the movable jaw all the way and indicate the ways as well to be sure the vise is square if you hadn’t done that yet.
I’m inclined to think it’s the vice. You could bolt some stock straight to the table and cut it to see if the same error appears. If not you know it’s the vice. If it is the vice I would take it apart clean and reassemble. If it’s not just the vice then it could be a gib if all the bolts are tight. The plumb will show you if the whole thing is moving when the center of gravity shifts. I had some powdery concrete under one of the feet on mine and the mill would kind of roll in an imperceptible way.
You’ve levelled and removed twist so that’s the right place to start. Have you tried probing the table itself at extreme ends of travel? You can use a probe, a DTI, a haimer, etc. to do this. Just touch off on one end of the table, set that to zero, travel to the far end and touch again to see what, if any difference is there.
Assuming that checks out good, then I would try a piece of material bolted directly to the table instead of the vise as others have suggested. If that comes out good, then the issue has to be the vise.
From there I would take the material that you just made parallel, set it in the vise VERY loosely clamped, and touch off on one side to set your zero again, travel over and check the other side. That will tell you immediately if the vise is tilted. Assuming it’s not, the next step would be to tighten the vise and see how much the material moves on both sides.
With all of this it’s important to remember the tolerance stack. As you correctly said, 0.005" means one side is cutting 0.0025" deeper than the other. You already know you have 0.0006" worth of twist so you’re really only looking for 0.0019". If the table is off 0.0005, the vise is off 0.0005, and you’ve got 0.0009 of lift in the jaws when you clamp the part, that all adds up. Assuming it’s all on the same side of course.
This sounds familiar but correlation is not causation. The output you are getting is similar to the issue I had with my vise and it was the same size of stock being machined, probably clamped the same way.
I have checked many of the machine factory specifications and they all seem to come back as better than factory tolerances. Have not had a session to go back and set everything up again and run it, though…
Well, i did some more testing last night and realized some of it was my own stupidity after i releveled everything and such, i was still using the unsquare surface of the block that i had originally been cutting on, so i clamped it with both parallels, one was loose one was tight, then faced it and flipped it 180, and it seated on both parallels that time, so i did the same for the other two sides, and i wasent getting any tilt in the cut this way, so my block is parallel now, which it wasent before, but i got curious, and put the block back in with the em shank in the movable jaw and parallel on the back jaw, and it instantly cuts deeper on the movable jaw side again, so maybe it was actually making the part square to the back face then, but i didnt go the whole way around the block to see if it kept parallel, i wish i had a squareness comparator stand to be sure but they are quite expensive.