[tmarks11]

That has got to be hard to even machine it like that. Much less unintended.

Not that hard... put it in a vise that isn't securely clamped to the table, than run an endmill through. The torque of the cut causes the vise to swivel, giving you a cut that is not straight.

Strangest thing to look at a curved line that you just milled.... by only cranking the x-axis....

[MBViklund]

Hi folks

Today I was out working with the mill again. The plan was to make a tool table with tapped holes. Before starting milling it I wanted to make sure my mill is square. Checked the tram with a indicator on an arm mounted in the spindle. It was of by 0.1mm in both x and Y. Adjusted it and got it close to zero by turning the head for the x axis and shimming the spindle for the Y. Then I moved the Y axis and got a different reading?! So I started measuring if the table surface actually was square against the column. I need to say I am very unhappy. As shown in the youtube video the x axis is of by 0.05mm over the entire travel (I can live with that) But the Y-axis is of by 0.2mm over just 340mm! This only matters if you are milling things attached directly to the table. I could always Shim the tool table to get it even. But I want the table to be flat and always at the same distance from the Spindle given that the spindle does not move.

Is there any way to adjust this? The only reasons I can think of getting readings like this Must be the table thickness is of by 0.05mm In the x axis direction. For the y it must be the table or the saddle. Is this really acceptable? To tram the spindle I need a square surface to measure from. Please advise.

https://www.youtube.com/watch?v=WOALp_5WsIM

[Hirudin]

Did you read this thread by chance? If not, it might have some things you could try.
http://www.cnczone.com/forums/tormac...cam-forum.html

Could you measure from the table surface down to the sliding surface on the base? You'll have to remove/move the bellows, I think. 0.016"/0.2mm should be easy enough to measure.
[edit]Oops, 0.2mm is close to 0.008", not 0.016".[/edit]

You could also measure between the bottom of the dovetails of the table to the top of the table. IIRC, you could probably do that without any disassembly.

Is the error consistent if you move the table in the X direction? In other words, if you move the table all the way to the right, take the measurement along the Y axis, and then move the table all the way to the left, would the Y axis measurement come out about the same?

[SCzEngrgGroup]

What Hirudin said. Most likely, your base is twisted, causing the problem. The thread he linked to shows the same problem, which was easily corrected by shimming the base mounting pads to remove the twist.

Regards,
Ray L.

[MBViklund]

Did you read this thread by chance? If not, it might have some things you could try.
http://www.cnczone.com/forums/tormac...cam-forum.html

Could you measure from the table surface down to the sliding surface on the base? You'll have to remove/move the bellows, I think. 0.016"/0.2mm should be easy enough to measure.

You could also measure between the bottom of the dovetails of the table to the top of the table. IIRC, you could probably do that without any disassembly.

Is the error consistent if you move the table in the X direction? In other words, if you move the table all the way to the right, take the measurement along the Y axis, and then move the table all the way to the left, would the Y axis measurement come out about the same?

I will go out and do some measuring now. I think I should be able to measure the distance to the table to the base pretty accurate. I have only done the measure shown in the video at the spots shown in the video.

Thank you for that link.I have not shimmed the machine on top of the stand, but I have thought about it. I leveled the whole thing without torquing the screws holding the mill. I could see some reaction on the level when tightening the screws. Will try that 3 point approach and shim the front feet When it is in level..

[MBViklund]

So I went out and undid the 4 screws holding the mill. Leveled the mill using a very accurate level. one or two degrees turn on one of the adjusting nuts on the stand feet moved the bubble 1 step. The reading was the same when rotating the level 180 degrees. I'd say is is dead on.

Remeasured the table as I was doing it in the video. It was still 0.15mm off in the Y-axis the table being completely to the left. 0.15 in the middle and 0.10mm at the right. A bit better but far from good. Then I tried lifting one corner at the time placing a 1mm shim under it. And then measure again. Tried all corners and it did nothing to the measuring results.

So I removed the bellows and measured the height from the table down to the dovetails. This is where the fault is!

Front right 202.595mm
Rear right 202.440mm
Left front 202.630mm
Left rear 202.450mm

Difference on the right side 0.155mm
difference left side 0.18mm

Will have to contact Novakon about it.

I have not measured if the thickness of the table is varying But I assume it does since the readings in all corners are different.

[SCzEngrgGroup]

First, having the machine truly "level" as indicated by a spirit level, is of zero value in the real world. The machine will neither know, nor care, if it's off by even several degrees.

I am very skeptical of the accuracy of your height measurements. There are many potential errors there that could account for the differences you saw.

Also, how are your gibs adjusted? When the table is at either extreme of travel, the opposite end WILL lift measurably, even if gib adjustment is perfect. If they're loose, it can lift more than a little. BOTH X and Y gibs will contribute to the lift, so both must be adjusted with great care (which is, in itself, a long, tedious, iterative process).

Given the very large size, and relatively low cost, of this machine, I would not expect perfection. Frankly, I don't think what you're seeing is all that bad. You're at about +/-0.0025" over 25" of X, 15" of Y. I'd be amazed if anyone managed to get this under +/-0.001" after a lot of "fettling". What you're measuring is "out of the box", after the machine has travelled, literally, half-way around the world. Doesn't seem that bad to me, and I'm sure it can be improved with enough time and effort put into really tuning it up properly.

Bottom line, if you haven't CAREFULLY adjusted the gibs, then spent considerable time playing with shimming the feet (not just placing shims under the feet, but using the bolts to pull the machine down onto the shims, to intentionally twist, or hopefully un-twist, the base), you really don't know where you stand.

Regards,
Ray L.

[UniqueMachining]

MB

May I recommend the following options assuming that purchasing a "super" accurate 50-100K machine is out of the question.

After performing the adjustments Ray L has suggested and tramming:

1. Install a surface plate (say 34 x 12) with holes drilled & tapped ( I use 3/8-16) every 1" on X, offset on Y by 1".

I had a local machine shop make one for me out of aluminum tool plate.
Shim this plate as required to obtain <.0005 variance across the entire surface.

2. Mount each vise on a 1/2" base plate, bolted to the bottom of the vise.
Mount the vise in a repeatable position on the surface plate and shim the vise to base plate for <.0005 variance.
I have 3 vises (Left, Center, Right) on plates which can be switched in and out with a minimum of fuss.

For a great deal less money I have a very accurate, repeatable setup which more than satisfies my needs.
I actually use the vices for 90% of my work but have a few fixtures which mount directly to the surface plate. Sure comes in handy when you need it.

Good luck, John

[vertcnc]

For parts machined in the vise, couldn't you just machine a step in soft jaws.

[UniqueMachining]

For parts machined in the vise, couldn't you just machine a step in soft jaws.

Absolutely, however each different set of soft jaws would require the step.
Although I have never tried this solution, I would be concerned that each time you switched soft jaws, you would introduce inaccuracy.
By virtue of the vices being "tuned in", they remains an accurate constant..

John

[kvom]

Over a 6" vise jaw the variance is 5 tenths or so.

I find it a challenge to tram the head that well; if I get it to .001" I call it done.

[MBViklund]

First, having the machine truly "level" as indicated by a spirit level, is of zero value in the real world. The machine will neither know, nor care, if it's off by even several degrees.

I only leveled it because I assume this is the way it was assembled and set up the first time and to have a starting point. It is also the way the manual tells you to do it.

I am very skeptical of the accuracy of your height measurements. There are many potential errors there that could account for the differences you saw.

The measurements are for sure not perfect but better than good enough to see if the table/sadle assembly is off. They are taken with the x axis being in the middle and the Y close to the middle. The measurements clearly shows that the table is not parallel to the base sliding surfaces. Redid the measurements today at another Y possition. I got the same resaults showing that the measurements are to trust. I also used feeler gauges to see if I had any gaps between the sadle and base (which I really should have if the base is twisted and the gibs are unadjusted) The smallest feeler gauges I have are 0.05mm and I could not fit them in anywhere! The only place I could get them in were between the table and saddle on the left side. Same place as where I earlier in the thread showed pictures of the gib being higher than the surface it it supposed to slide on. I could not get it in on the right side so maybe grinding the gib down 0,1mm would make those 0,05mm in the x-axis dissapear.

Also, how are your gibs adjusted? When the table is at either extreme of travel, the opposite end WILL lift measurably, even if gib adjustment is perfect. If they're loose, it can lift more than a little. BOTH X and Y gibs will contribute to the lift, so both must be adjusted with great care (which is, in itself, a long, tedious, iterative process).

The gibs are untouched so far. Yes the table will tilt when it moves from one side to the other caused by the weight of the table, this causes movement of the table and the saddle. But this is in my mind only true for the x-axis movement. As long as the table sits right over the saddle. I should not see any tilting/lifting moving the Y-axis since the entire saddle rests on the base the whole time, given that the base is not twisted.

Given the very large size, and relatively low cost, of this machine, I would not expect perfection. Frankly, I don't think what you're seeing is all that bad. You're at about +/-0.0025" over 25" of X, 15" of Y. I'd be amazed if anyone managed to get this under +/-0.001" after a lot of "fettling". What you're measuring is "out of the box", after the machine has travelled, literally, half-way around the world. Doesn't seem that bad to me, and I'm sure it can be improved with enough time and effort put into really tuning it up properly.

0,05mm over the full x travel is ok. That is also what I wrote. But 0.15mm over 343mm is not. I only get 343mm(13,5") of travel in Y being only 0,5mm from the physical stops, but that is another question. Today i measured Y with the table sitting in both extremes and in the middle. Did reset the guage for every measurement. the error was between 0.14 - 0.16mm every time. Also measured the thicknes of the table in front and back beside the saddle when the table was sitting in the middle. The back is 54,82 in that spot the front 54,87. So the saddle should be about 0.10mm high in the back.

Bottom line, if you haven't CAREFULLY adjusted the gibs, then spent considerable time playing with shimming the feet (not just placing shims under the feet, but using the bolts to pull the machine down onto the shims, to intentionally twist, or hopefully un-twist, the base), you really don't know where you stand.
For the shimming I did actually carefully tighten the bolts while having the shim placed under the feet in each corner. I did only measure flatness in the Y-axis doing this, but I saw almost no changes in the readings. To me it sounds like a bad idea starting adjusting the gibs on a machine that potentially have a twisted base. If this is the case and I manage to untwist it I would have to redo the gibadjustment anyway.



Regards,
Ray L.

Answered within the quote.

[MBViklund]

MB

May I recommend the following options assuming that purchasing a "super" accurate 50-100K machine is out of the question.

After performing the adjustments Ray L has suggested and tramming:

1. Install a surface plate (say 34 x 12) with holes drilled & tapped ( I use 3/8-16) every 1" on X, offset on Y by 1".

I had a local machine shop make one for me out of aluminum tool plate.
Shim this plate as required to obtain <.0005 variance across the entire surface.

2. Mount each vise on a 1/2" base plate, bolted to the bottom of the vise.
Mount the vise in a repeatable position on the surface plate and shim the vise to base plate for <.0005 variance.
I have 3 vises (Left, Center, Right) on plates which can be switched in and out with a minimum of fuss.

For a great deal less money I have a very accurate, repeatable setup which more than satisfies my needs.
I actually use the vices for 90% of my work but have a few fixtures which mount directly to the surface plate. Sure comes in handy when you need it.

Good luck, John

Thanks for the tip John. Actually i was going to make a fixture plate this weekend. I have a 500x400x100mm measuring table I am going to make a fixture plate of. I am only going to face the bottom parallel to the top and drill alot of holes in the top. That is why I started measuring the table.

To mount the vises on plates and shim them once is a great idea! I have been shimming them directly on the table every time...

I will continue working with the mill tonight.

[UniqueMachining]

MB

I forgot to include a link to some pictures of my surface plate & vises.

http://www.cnczone.com/forums/novako...30-bobcad.html

Enjoy

John

[engnerdan]

I only get 343mm(13,5") of travel in Y being only 0,5mm from the physical stops

Isn't that travel supposed to be 15" of travel int he Y axis?

-Dan

[Hirudin]

IIRC, that's about all I got as well (~350mm).

[MBViklund]

Isn't that travel supposed to be 15" of travel int he Y axis?

-Dan

Talked to John. He tells me it only to remove the screw in front to get the desired 15". i will try it tonight. It looks to me like i could move the rear screw a litte bit as well. But i am worried about the bellow if i do. If it is full of chips and the table go all the way in to the column it might brake?

John at novakon has been very helpful by the way.

[MBViklund]

Lately I have had a lot of work to do for others, but this weekend I have some time to do work for my self. Started making a fixture table from an old abused cast iron surface plate. See attached pictures to see starting point and what i aim to do with it. Used a Tormach flycutter to do it. Realize i need to get another face mill with smaller diameter and perhaps 4 inserts. The smallest diameter i can get with the flycutter is Ø70mm and it seems to me that i do not have enough torque at low RPMs to run it effectively in hard materials. With a smaller diameter i could run higher RPM and feed more with 4 inserts. It did the job but it took several hours. Ran 2,5mm cut depth, 800rpm @ 110mm/min feed. Removed 20mm of material in total. The finish came out beautiful!

https://www.youtube.com/watch?v=9IBDh6FPJlw


Attachment 273320Attachment 273322

[tmarks11]

That is hard service for the fly cutter insert, as the whole thing is one giant interrupted cut. As you are only cutting the webbing down, you could have taken it down with a 1/2" endmill, and then face milled it to get it smooth on the last pass. Probably would be quicker.

That is an interesting idea; I have a spare cast iron wing from a table saw that could probably be adapted in a similar fashion.

[MBViklund]

Almost finished with the table now. Faced the top, drilled alot of holes and tapped them with a chordless drill. Could not be happier with it.

https://www.youtube.com/watch?v=-6hJjLYDVWc
https://www.youtube.com/watch?v=z_lElauV518
Attachment 273814