Thank you guys again for your support! Well, I'm now taking a look at the ServoBelt datasheets and it looks more than promising.
I guess I can make a belt reduction from the servo to the little sprocket to gain torque. That was my original plan with the rack and pinion but this seems to be much cheaper and easier. So basicly what they do is use to pieces of belt one slightly longer than the other (to acommodate the sprocket and tensioners) and they are engaged with each other and well tensioned at the ends. Is that right? So the sprocket spins and makes the joint travel along the axis. From what I've seen on their datasheet, the heavy system can handle up to 500 Lbs of linear force so that's like 2200 N. I'm more than okey with that. I'm worried about the streching of the belts under acceleration, do you think I should use a special kind of timing belt? Like kevlar belts to avoid stretching under accelerations from 0 speed or under changes of direction? El dom., 21 abr. 2019 a las 0:40, Chris Albertson (< [email protected]>) escribió: > Wow, 3 meters! Another way I've seen is to use a cable. There are > pulleys at each end and the cable forms a loop. In the simplest design a > motor powers on of the pulleys. The cable takes the price of a long timing > belt. I've seen kevlar cord used for a cord and some steel cable also. > In the better design there is a drum where the cable takes a few turns > around and a spring to keep tension on the cable. The drum is powered. > > With a 3M long rack, I think keeping it clean will be an issue. Perhaps > mount is on the bottom of the track so chips fall down and off of it. or > use a continuous airblast > > With gears you will need to have a backlash budget tnad design for it. Or > just use a larger motor that still has torque at the low end > > > > On Sat, Apr 20, 2019 at 2:03 PM Leonardo Marsaglia <[email protected]> > wrote: > > > Thanks Andy and Chris for your quick response. > > > > I'll try to answer both messages into one because I lack the quoting > > function here in gmail. > > > > I plan to reduce the servos of the Y axis about 10 to 1 at least. I would > > love to use timing belts and pulleys but I preffer to avoid all the > trouble > > of making the steps for the reduction. The servos I plan to use are 750 W > > and 3000 RPM. > > > > About using a screw. The main problem is my Y axis is 3.5 meters long and > > I'm not sure I can get ballscrews that long so easy from China to > Argentina > > in a safe and cheap way (I don't even know if the Fedex or DHL would ship > > me something that long). In case of using a screw that long I guess is a > > must to have a rotating nut and that's not a simple task too (at least > not > > as simple as the planetary reducer). So that's why I'm not thinking about > > ballscrews although I would love to use them on all axis. > > > > > > > > > > > > > > > > El sáb., 20 abr. 2019 a las 17:50, Chris Albertson (< > > [email protected]>) escribió: > > > > > backlash is a big deal. Maybe not so much with wood routing as maybe > you > > > don't care about 0.1 mm error. But think about a how a backlash in > > > the 10:1 reduction translated to linear motion. Every time the servo > > > motor changes direction the cutter stops moving until the backlash in > the > > > system is taken up. I can not imagine a gear system that backlash. > > > > > > This is why most machine tools use ball screws. Chinese ball screws > are > > > now nearly cheap as dirt and have zero backlash and can be directly > > driven > > > with a motor with no reduction. > > > https://www.ebay.com/itm/Ball-Screw-SFU1605 > > > < > > > > > > https://www.ebay.com/itm/Ball-Screw-SFU1605-RM1605-End-Machined-BK-BF12-Ballscrew-Nut-Housing-Set/182984379186?hash=item2a9ab80b32:m:m0p8OTgikJ4HQd-zu4XytTg > > > > > > > > > > I have some of these kind and as close as my dial indicator can measure > > the > > > scre/nut is "perfect". The balls are spring loaded so they bear on > both > > > sides of the screw so there is zero "play" For mettal working these > > > screws are only "hobby grade" as they might have 0.03mm periodic error > in > > > the pitch but I can't measure 0.03 over a 1/2 meter distance. > > > > > > They also make some really nice Chinese linear ball bearing track as a > > kit > > > that makes building one of these things "plug and play" Hard to beat > a > > > full ball bearing linear motion setup for under $100. Here is one > > > example. They come os bigger and smaller sizes > > > > > > > > > https://www.ebay.com/itm/SFU1605-Ending-Machined-Ballscrew-Set-SBR16-Linear-Rail-Kit-For-CNC- > > > < > > > > > > https://www.ebay.com/itm/SFU1605-Ending-Machined-Ballscrew-Set-SBR16-Linear-Rail-Kit-For-CNC-US-Stock/401566069770?var=&hash=item5d7f33f40a&enc=AQADAAADAFjVrDbVsZ8oH%2F8PNHtt9VX4%2Fw7FZcmMuqsX8uaFEduVZdN2xO9SlmphozVEOH7g7jX%2Bxew6Qceghk4QQfAx5sBCl3mvk4eHWuVUGaj8C5BkivfcPYRvofBQwnqo2D6pzU%2BGUd3CvwVXAM1GCvgT5HKYTTJRaAeIZBdXqfWlMXTaon39bDBa3zC%2F1K9ZLnfOKqX05Hd3sFtOKZehDjsf0jFAfy7bTSEmDVD6AmIj6ajz1SzPHtc%2BnojN8s5utpzO3078Jp6elV34r23RIuEGPXeqA%2By3BsRGYYbRCMxUNK7Td2il%2BzXdMr%2BerE%2FeihOQd97OPlL94nECQyCWjqeVGtoxO232cboR1Wve0LY2A8zPL0z50lT0s9SeLuW3dajRLIHWtpgAzCKGxELJnOVNP9le4Jc%2FcbyRcoiJJv%2BGwcEgBY1vrr1dbjwhxAYnUJG43R4TRLGijViWapLTeb2Awsu1sDe2W5XmsXrIpqZ%2FwqZ%2BaPGBvQ72s1dSDJOAlRYgbqhBH%2F%2Fyr1dDEP47jY7CIzBtBCa%2BLIY3nrG49uIxuVA%2FdIYXzBneALLQgf3qNzOCTcTZJ%2B9mek7VnKpPd64lGwhjrGGem%2F28DybqcMFTeyb%2F9sue%2Bntf14YCADI5p%2FBJlGfEgY2ISNKVHriCM%2Bkm4XDW%2F6L%2FptM%2BOXTnhjf42Rhr%2B1Jkd051YRz59z6A0dps9OkNBtAYq4kVrOR6LYMxMWryqdbtzIrV0RnjPeAheOW10TvnQ0kWiXCV9JrZv1Mw4jKCTVIBt2lwm25r%2BgAFIAErGwpUxwOFdoZ4Dg1v6iDMdbmbqeST70Iy5AG1ccQhl69XQvcHN2GTJ8b3qhCzLkLsglFBXTY7%2Bin2x4io6%2FV4vWLAr6sw9%2BB050TxvCZMd0a67R1ZdV4dCLxgFFdc%2BxyiZwmNYtXWOoY04tjWSJJGG9ptIWM1rCArFJqn2kJc1Y1lKu9HjO6y1z4iz4N4HRY2%2B7rQwzoW%2BnmJoqmh%2Fa1c83xr0XP7XkA5jqL3Lwx3Gg%3D%3D&checksum=401566069770871222cfdded49168823fbf139903396 > > > > > > > > > > If you don't like ball screws then try long timing belts. They also > are > > > zero-backlash and come in any size you need at lower cost then gears > > > > > > > > > On Sat, Apr 20, 2019 at 1:27 PM Leonardo Marsaglia < > > [email protected]> > > > wrote: > > > > > > > Hello to all again! > > > > > > > > I'm defining my router design after some months of inactivity. I > > already > > > > defined my steel frame and purchased the iron to start building it. > The > > > > idea is to use the router mainly for wood and melamine boards. > > > Eventually I > > > > may use it for aluminum but that's not the main purpose of the > machine. > > > > > > > > Since you guys have lots of experience I would like to ask for advice > > to > > > > define two main concepts for my design: > > > > > > > > 1- Should I use helical or straight teeth racks and pinions? > > > > > > > > I plan to use some chinese rack and pinions (MOD 2 with 30 teeth > > pinions > > > > rotating at a max speed of 200 RPM) for the obvious reason of the > > budget. > > > > But I can get hardened and ground rack and pinions so I think that > > sounds > > > > pretty good. Anyway, I think I should stick with the straight rack > and > > > > pinion since it's the simpler solution and being hardened and ground > I > > > > think they will behave pretty well but I would like to hear your > > > opinions. > > > > > > > > 2- For the servo-to-pinion reduction: Planetary servo reducer or worm > > and > > > > gear reducer with timing belt? > > > > > > > > I'm thinking about the worm and gear because I can machine that in > the > > > shop > > > > and reduce at least the apparent cost (I say apparent because I still > > > will > > > > have to make more complex mounting plates for the servos and > > > reductions..) > > > > but also I'm tempted to buy something like this: > > > > > > > > > > > > > > > > > > https://www.aliexpress.com/store/product/high-Precision-Helical-planetary-gear-reducer-5-arcmin-2-stage-ratio-15-1-to-100-1/1155094_32867566227.html?spm=2114.12010612.8148356.28.1b5476fa9wYyAW > > > > > > > > > > > > Do you think they are worth it? I have no experience with this kind > of > > > > reducers but the seller has good reputation and sells Delta servo > > motors > > > > wich I'm familiar with and work great. What do you think? > > > > > > > > In any case, I plan to mount encoders coupled directly to each > pinion > > > so I > > > > can always track the position of the pinions no matter the backlash I > > > have > > > > in the reduction stage (but off course I want backlash to be the > > minimum > > > > possible). > > > > > > > > Those are my concerns for now. Please tell what you think so I can > > make a > > > > better choice. > > > > > > > > Thanks as always to all of you! > > > > > > > > Leonardo > > > > > > > > _______________________________________________ > > > > Emc-users mailing list > > > > [email protected] > > > > https://lists.sourceforge.net/lists/listinfo/emc-users > > > > > > > > > > > > > -- > > > > > > Chris Albertson > > > Redondo Beach, California > > > > > > _______________________________________________ > > > Emc-users mailing list > > > [email protected] > > > https://lists.sourceforge.net/lists/listinfo/emc-users > > > > > > > _______________________________________________ > > Emc-users mailing list > > [email protected] > > https://lists.sourceforge.net/lists/listinfo/emc-users > > > > > -- > > Chris Albertson > Redondo Beach, California > > _______________________________________________ > Emc-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/emc-users > _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
