Page 2 of 3
Re: OAK linear encoder
Posted: Thu Aug 13, 2020 5:21 am
by tblough
Not going to get there. Glass scale accuracy is 3 microns.
Re: OAK linear encoder
Posted: Thu Aug 13, 2020 10:33 am
by martyscncgarage
Martin Zarnay wrote: ↑Thu Aug 13, 2020 1:40 am
Guys , maybe im over shooting with tolerances a bit. +-0,0025 it would be part tolerance. I just want to take part to hands every time and see 300,00 mm everytime, with no worried it can change. im shooting to 0,0025mm (0,005 mm max) not for small parts ( 20mm ) but in in all travel of machine 300x400mm. Bal screw C3 is rated at 13um ( 0,013 ) in 300mm it means that if I will machine 300mm part it could be 300,013 up or down in length and I would need to adjust for error, that's why linear scales .
In shop my friend works, he has DMG mori relatively new one. and in longer parts it can finish part with tolerance +-0,005 sometime part longer, sometime shorter. Of course , I know there are many many more factors to influence this I know . If is bad machinist - even linear scales wont help keep tolerances ...
I just wont to build best accurate machine I can, with very good ballscrews, if ball screws will be accurate enough for my needs good, and encoders will be safety measure.
Ive got email from zsinstruments.com so I will chat with them and will see. will share my results
Are YOU expecting to build this precision machine or do you have the machine already and you are retrofitting it. Mill? You still haven't said.
Re: OAK linear encoder
Posted: Sun Aug 16, 2020 3:11 pm
by BodeRacing
tblough wrote: ↑Wed Aug 12, 2020 7:14 pm
Mazak has linear scales as standard, Okuma offers them as an option.
Thanks. I am not very familiar with the high end machines, and was led to believe that they do not use linear encoders. But again, I don't have a lot of experience with those machines. I work with shops that use them and I see end results but do not know that much about the actual machines.
I do know a bit about metrology and don't believe you have any chance of repeatedly doing what you want. Way too many variables, and the scales are only one piece of the puzzle.
Re: OAK linear encoder
Posted: Mon Aug 17, 2020 4:21 am
by Martin Zarnay
No no im building new machine from scratch. In basement, with air condition and same temperature. If I would retrofit old machine, tolerances are gone, ( could not keep machine at same temp ).
Ive gotten reply email from zsinstruments.com and linear scales with tolerances of 1 microl will cost me for X and Y for about 1000 dollars.
not bad I think. now when I'm dealing with Heidenhein all in one solution system ( controller and encoders ) it is just fraction of price.
But again, Hendeinhein is industrial and Centroid kind of hobby...
My main question for Oak was, if is it full closed system so Oak is constantly adjusting servos based on encoder feedback? and if you think it could keep up with 1 micron scales? If anyone have experience with C3 ball screw?
Re: OAK linear encoder
Posted: Mon Aug 17, 2020 6:42 am
by tblough
I sincerely doubt those scales have a tolerance of 1 micron. They may have a RESOLUTION of 1 micron. But I bet they have an accuracy of either 3 or 5 microns. I have a C3 screw on my Z axis and a C1 screw on my X axis on my lathe. Both axis have 0.1 micron glass scales, which allows me to run parts in the +- 5 micron range. Oak has no problem "keeping up" at 200ipm speeds.
Re: OAK linear encoder
Posted: Mon Aug 17, 2020 3:04 pm
by AshG
most high end spindles have about 4-10um of play and that's before you even look at the rest of the machine, every cutter you use will need temp regulated through spindle coolant, cutter comp per tool, the list of things just goes on and on glass scales are the last thing you need to worry about, even batch to batch difference on the same grade of material will throw it off. the logical way of doing things is to machine to rough size then surface grind to finished size. that or stop being daft and go buy a wire edm machine.
Re: OAK linear encoder
Posted: Mon Aug 17, 2020 3:11 pm
by martyscncgarage
AshG wrote: ↑Mon Aug 17, 2020 3:04 pm
most high end spindles have about 4-10um of play and that's before you even look at the rest of the machine, every cutter you use will need temp regulated through spindle coolant, cutter comp per tool, the list of things just goes on and on glass scales are the last thing you need to worry about, even batch to batch difference on the same grade of material will throw it off. the logical way of doing things is to machine to rough size then surface grind to finished size. that or stop being daft and go buy a wire edm machine.
I'm not sure what the original poster is trying to build, nor if its within his capability to manufacture a machine to hold such tight tolerances or what he wants to make that will require them?
Marty
Re: OAK linear encoder
Posted: Mon Aug 17, 2020 3:55 pm
by AshG
the tolerance does seem a little nuts for a diy machine with a retrofit control lol, id say good on him if he manages it. At work we're machining pure copper particle accelerator wave guide cells to a <4um tolerance, all you need to do is walk into the production cell to piss off the operator. the vibrations of walking heavy footed across the floor in the room can throw off a part even with a reinforced concrete.
Re: OAK linear encoder
Posted: Tue Aug 18, 2020 2:54 pm
by Martin Zarnay
I´m very thankfull of all your responses guys. I dont think we will build such precised machine ( defenately not for production ) but why not to dream and pus for best result.I would be happy if i take calipers to hand, measure part - of course roughed and finished - and see on them for example 120.00 mm not 119.99 or 120.01.
Thank you AshG: C1 ball screw with doble nut is way to go, so i will do that.
And after talking with guy from Heidenhain, he explained me that scales are just safty part of machine, the most important is temp sensors on frame, screw, spindle... And intensive testing. so I won´t go for scales and rather go for controller what has temp comp.
But it cant be OAK
Re: OAK linear encoder
Posted: Wed Aug 19, 2020 6:15 am
by Martin Zarnay
hollow ball screw for liquid cooling , nice