g0704 / BF20L gravitational sag on X

[Nava]

Hi all,
I have a BF20L / G0704 (l for long table) that I bought in Germany and had pre converted to CNC. Upon tweaking the X axis I noticed quite some variations in the tables surface in the Z plane.

Im no expert but I deducted that for the most part its because gravity causes the long table to sag down at either end when extended causing the table to rise up a little in the center under the quill .

My tests so far went like this.

1. Placed a DTI(.01mm scale) directly under the quill in with X and Y in a center balanced position. Also tested the table for flatness as best i could
2. Checked with a level to see if the table was flat.
3. Locked the Gibs tight on my Y axis
4. Zeroed out my dial on the table
5. Moved the table along the X 24cm in each directions and took a reading

Here is what I got before adjusting gibs. By left and right I mean left and right facing the table.

X_right +.1mm (.0039")
X_middle 0
X_left +.06mm (.00236") (my stepper and belt are on the left side)

I then tightened the gibs a bit more and repeated steps 3 and 5 until I got the least sag in the table. I haven't checked my backlash yet but I'm pretty sure I have gone past my sweet spot and have to back off a bit

X_right +.05mm(.001968")
X_middle 0
X_left +.03mm(.00118")

Has anyone else tried to measure this in their equivalent machine? I'm thinking this is a bit crappy especially since its in the middle of the table and not the end of the gibs where I'm sure there is more of an offset.

Any other experiments that others can do to test their machines would be interesting. I'm considering to learn the art of scraping from a German guy and seeing as these machines are so popular for modding I'm sure that this is probably the ultimate mod for these machines to able produce some really precise work.t

[bjones]

So over a distance of about 20 inches you are worried about .001. that's already pretty dang precise. have you taken a micrometer and measured something .001 thick? a piece of paper is 4 times thicker. how long of a part do you normally work on, if it's 10 inches the error from side to side would be even less. too many people see numbers on a dial and if it isn't zero for everything they think it's no good. you need to learn about acceptable tolerances. you are better leaving well enough alone, you are likely to make it much worse scraping away yourself but it's your mill and you gotta have zeroes right.
bill

[Nava]

My longest parts are around 16" and in the worst case scenario a 2 sided part can be out by .2mm or 0.0078"
I probably wont start with my mill when I learn the scraping I have an old abused enco mill crosslide that is useless without some TLC ill start with that. But hopefully Ill be learning from nick in Germany who is a pro at scraping..this guy
http://www.youtube.com/watch?v=y2AUew5Evxc&feature=relmfu]Metal scraping: How to read the spotting - YouTube

He is a few hours from me so I can load up various old tools and go and learn from him if agrees Right now the mill is cutting some nice work so I'm not so worried but I keep hearing from others that these mills have such soft steel that at some point that gravitational sag is going to turn the table it into a banana shaped paper weight. So at that point Id consider re scraping ways or a linear rail option and as a requirement for that Id want some nice ground or scraped steel to put on the old ways to support the rails. Im not expert machinist so if you say the measurements are pretty good then its good news for me

[LUCKY13]

Try locking the X axis gib plus the Y and see if it gives better readings ( in each posstion)? If so then better fitting gibs may fix it.


Also you need to make sure the base is not twisting on you when you extend the table to one side. If its not bolted down this could happen real easy. The top of the table itself could even be the cause. A good surface plate will be needed to check the table, and to make new gibs.


The readings you got after ajustment are not bad for a benchtop machine, but that doesnt mean you cant make them better.



A good camelback master canbe bought here

Hand Scraping Tools

But you need to read about what your buying. They need machined, heattreated and hand scraped before using. A lot to go through, but one ready to go can cost much,much more.

Jess

[DonKes]

Do you have big vices at the ends of the table?

Do you leave the axes centered or at the ends when not using?

Don

[Nava]

I don't have any big vices on my machine and I usually leave the table in the middle when I'm done.

I wont get a chance to re measure with locked x gibs as Jess suggested until Thursday now(travel for work) Come Friday morning Ill be able to try that and any other suggestions that come up in between. The machine is bolted to wood at the moment but its not much larger than the base maybe 6" all around bigger and around 2" thick which sits on a large table. I can try and source a proper base plate for the table and get it there by Friday. When the gibs were in their loose position I could fit a piece of printing paper in the gap between table and the slide at the op of the extend end(being lifted). Also I got about .5mm deflection by lifting and then pressing down on the end of the table(with y locked down but not the x). Im eager to at least order a nice granite measuring base and blue the control surfaces and the camel back. The granite plate I can get from Google Translate but I cant seem to find a decent surface for the gib though apart from the one Jess gave a link for. But if im lucky Nick will have one to use if i manage to catch up with him

[michaelthomas]

A couple hours of lapping the ways on my ZX-45 were well worth the effort. It got rid of the most obnoxious high spots, and made the table travel much easier with tighter gibs. I just used Mothers aluminum polish. Apply, work back and forth, wipe off, repeat. Things got better quickly for me.