[keen]

I'm getting into more accurate work and worrying the Tormach isn't going to be up to it - so I decided to get some real facts to help with my decision.

I spend some time setting the X and Y trust bearings and gibs to what I thought was optimum - and then ran the following test - see attached Pic and G code.

The roundness was within 0.025 (0.001") total error - checked by micrometer and in lathe with dial indicator. (23.98 - 24.005 mm).

Good news!

[Art Pentz]

Hi Keen,

Your measurements are, indeed, good news! My Tormach continues to amaze me w/ its acccuracy.

Very nice job on the photo by the way!

Art Pentz (Tormach #1215)

[keen]

Not all Tormachs come out the same though - I have heard stories of machines coming out from the factory much less accurate than that. But if one can be that accurate it means at least the design is good and therefore the potential should be there. Oh by the way that test was in 4140 h/t steel.

[highspeedmazak]

I have had my tormach hold 4 tenths cutting slow and with tools made to hold these tolerences. I am sure that I could do even better with some tuning. I just ran a 4.250 c-bore with a 3.375 through hole and held a nice press fit of .0005 and it is 1.375 deep over all the c-bore is .375 deep. The machine is there you just have to use good tooling and take your time at machining. the material is 17-4 SS the cutter for finish is carbide 4 flute no coating. I believe that I paid around 120.00 for the cutter. The machine can do it. JUST TAKE YOUR TIME..........

[gep]

Have had very good results with my machine. Round cuts within .001 t0 .0005 as to the best I can measure. Must have good accurate tools. Some cheap milling cutters haven't been the best. I usually measure all tools before programming, or setting quick change tooling.

[RPS68]

Hi Guys
Introduce myself as new owner of tormach mill for about 2 months.
Everyone here seems to have good accuracy on their mill. The accuracy on mind is not as accurate, it is .07mm
of total runout when machine a 60mm pocket with brand new solid carbide endmill with two finishing cuts this was done after gibbs and thrust bearings adjustment has been made. The runout is all in the y axis backlash. The accuracy before the adjustment was 0.25mm of total runout in both X and Y.

I'm situated close to Keen , and he kindly spend over 2 hours helping me out last Sunday.
Sorry Keen you have to do pressed up to keep warm and thanks for the help.

[philbur]

Did you find where the backlash is. Is it in the ballscrew itself or in the thrust bearing. The adjustment on the thrust bearings is a bit fiddly and the preload can easily be lost as you tighten the lock nuts.

Phil

Hi Guys
Introduce myself as new owner of tormach mill for about 2 months.
Everyone here seems to have good accuracy on their mill. The accuracy on mind is not as accurate, it is .07mm
of total runout when machine a 60mm pocket with brand new solid carbide endmill with two finishing cuts this was done after gibbs and trusth bearings adjustment has been made. The runout is all in the y axis backlash. The accuracy before the adjustment was 0.25mm of total runout in both X and Y.

I'm situated close to Keen , and he kindly spend over 2 hours helping me out last Sunday.
Sorry Keen you have to do pressed up to keep warm and thanks for the help.

[keen]

I dont think it is that - RPS68 is experienced - and would have tried more than once. No there must be something else out on the Y. Could be something is not firm - perhaps the ball screw nut or thrust setup has an error or is rocking on a burr. hope its not the Y ball nut itself!
Any ideas welcome!

[philbur]

I've never tried on the Y axis but surely it is possible to use a DTI to find if the lost motion is in the ball/nut assembly or the thrust bearing assembly.

Phil

I dont think it is that - RPS68 is experienced - and would have tried more than once. No there must be something else out on the Y. Could be something is not firm - perhaps the ball screw nut or thrust setup has an error or is rocking on a burr. hope its not the Y ball nut itself!
Any ideas welcome!

[Zetopan]

Did you find where the backlash is. Is it in the ballscrew itself or in the thrust bearing.

There is another even simpler source of backlash on the Tormach. Check the
motor to ballscrew coupler machine screws, they may not be as tight as they
should be.

[RPS68]

If I assumed that the thrust bearings are in perfect condition and of high grade like P4 or p5. I can definitely eliminate this, but however the bearings when pulled out of the assembly and gave it a wash, there seem to be contaminated with metal shaving I picked metal shaving with a magnet from the bottom of washing container.
The metal shaving could be from contaminated grease or thrust bearing has been over tighten. I am expecting new bearings, I will give the new bearings a wash and see whether any metal shaving exist in new bearings. If it is good bearings I should be able to eliminate the thrust bearings.

The backlash is little better if the gibb is loosen fully.

I tried to detect backlash from thrust bearings by DTI the end of lead screw then give it a
good push and pull with gibb fully loosened , the movement causes the lead screw to rotate so this method is out.

(Zetopan) Yes I have checked the coupling. I broke the ball of 4mm allen key.


It is a challenge to detect where the backlash come from as flex is also some where in the equation.

[philbur]

It's not so much the quality of the bearings but more how well you set the preload.

If you jog the axis in a given direction then put the DTI on the end of the screw, set it and the DRO to zero. Then under power jog the axis say 5 mm on the DRO (in the same given direction) then jog it back to zero on the DRO and check to see if the DTI also reads zero (note the error if any). Then do the same with the DTI touching off the table. This should tell you if it is a problem with the ball-nut/screw or if it's in the thrust bearing preload, or less likely the motor coupling.

Flexing is easily distinguished from backlash as it will only show-up under load and will disappear after the load is removed. Backlash is a permanent error in position even with no load applied.

Just some thoughts and opinions.

Phil

If I assumed that the thrust bearings are in perfect condition and of high grade like P4 or p5. I can definitely eliminate this, but however the bearings when pulled out of the assembly and gave it a wash, there seem to be contaminated with metal shaving I picked metal shaving with a magnet from the bottom of washing container.
The metal shaving could be from contaminated grease or thrust bearing has been over tighten. I am expecting new bearings, I will give the new bearings a wash and see whether any metal shaving exist in new bearings. If it is good bearings I should be able to eliminate the thrust bearings.

The backlash is little better if the gibb is loosen fully.

I tried to detect backlash from thrust bearings by DTI the end of lead screw then give it a
good push and pull with gibb fully loosened , the movement causes the lead screw to rotate so this method is out.

(Zetopan) Yes I have checked the coupling. I broke the ball of 4mm allen key.


It is a challenge to detect where the backlash come from as flex is also some where in the equation.

[RPS68]

Phil
Yes, I did tried this method with DTI straight on the end of leadscrew then jog back and forth. The rough finish and shaft wobble influenced DTI reading.
I will make a threaded stud and screw this to the end of leadscrew and put DTI close to the center and hope this would only detect actual reading of the leadscrew axial movement.

[MichaelHenry]

I had about 0.003" backlash in my Y-axis and tried checking/adjusting bearing pre-load without any improvement. What did seem to help was adjusting the Y-axis gib, though I'm at a loss to explain why that should help. The amount of backlash gradually decresed with small adjustments in the gib "tightness" and by the end backlash was down to a shade under 0.001".

Mike

[fastlanecafe]

Loose gibs will allow the table to rock slightly when changing directions and show up as backlash. You should be able to adjust all the rocking motion out and still have free travel. However, if it begins to bind up when the rocking is adjusted out, then remove the gib and check it for straightness.