On Thursday 30 July 2015 07:28:16 Roland Jollivet wrote: > If you're going to use micro-switches, rather use this > <http://uk.rs-online.com/web/p/microswitches/1031645/?searchTerm=103-1 >645&relevancy-data=636F3D3126696E3D4931384E525353746F636B4E756D6265724D >504E266C753D656E266D6D3D6D61746368616C6C26706D3D5E5C647B337D5B5C732D2F2 >55C2E2C5D5C647B332C347D2426706F3D313426736E3D592673743D52535F53544F434B >5F4E554D4245522677633D4E4F4E45267573743D3130332D3136343526> type; > > Notice how the button is almost directly under the roller, as opposed > to the 'normal' type of being near the lever pivot. > > The indicated type will have increased switching sensitivity and lower > hysteresis. > > Regards > Roland
Yes, thats a much better design. My x on the lathe was a button only since the crossfeed is a direct push, subminiature version of that. With the swarf brushed or blown away, repeatability is well under a thou. > On 30 July 2015 at 13:02, Gene Heskett <[email protected]> wrote: > > On Thursday 30 July 2015 03:02:45 Marcus Bowman wrote: > > > On 30 Jul 2015, at 03:51, Gene Heskett wrote: > > > > I bought a bag of small, roller tipped microswitches, and have a > > > > couple mounts laid out but not carved yet. > > > > > > > > Going thru searches on ebay just now, for both inductive, > > > > capacitative, and IR proximity switches, It sees like the > > > > majority of them are needing 6 volts or more, and around 300 ma > > > > each to function. Having seen some micropower capacitative > > > > switches in the textbooks of yore, it seems like the ball has > > > > been dropped in terms of being useful in a 5 volt logic circuit. > > > > > > > > Both the microswitch, and the inductive versions would seem to > > > > be sensitive to collecting swarf, particularly if its a rare > > > > earth magnet running a reed switch. > > > > > > > > My lathe seems to be relatively happy with a microswitch located > > > > under the back edge of the bed for Z homing where swarf tends to > > > > fall on past it, but x has to be kept swept clean because its > > > > actually on top of the carriage. > > > > > > > > I see a whole passel of stuff drilled onto GO704's in the hits I > > > > can get from google, but all of them preclude retaining the > > > > rubber swarf shields, which I'd say was not worth the tradeoff. > > > > There has to be a better way. > > > > > > > > So what sort of switches, and what swarf shielding to keep them > > > > relatively clean & accurate are others using? > > > > > > > > PM's with pix appreciated if you have the time. > > > > > > Actually, this is a really interesting question. Having spent ages > > > thinking about shielded locations for the mill, and about to do > > > the same for the lathe, it would be useful if folks could put the > > > pictures somewhere we could all see them. If anyone has any > > > recommendations about makes and models of switches/sensors which > > > provide better accuracy and repeatability than others, or useful > > > ways of providing adjustment on mounting arrangements, that would > > > also be very useful. > > > > > > Marcus > > > > My thoughts exactly Marcus. In my googling, something like 3/4ths of > > the sites that may have had helpful pictures, were behind a login > > requester. Since I am not a member of any of those Mach oriented > > groups, thats less than helpful. > > > > For X, I have considered removing the pointer which is also a > > mechanical limit stop, and likely demolishable to an out of control > > move as there are a couple OEM stoppers that would hit it. Once > > removed, then move one of the stopper knobs (they are in a 'T' slot > > in the front edge of the table) to someplace near the center of the > > table, and mount a switch that it would close on the way by which > > would function as the home switch. A similar setup is now on my > > lathe, with the switch under the back rail of what passes for a bed, > > located to be activated when no tool holder is mounted, and only > > 3/4" or so from the chuck face. X has already been homed and parked > > about 1mm from its outside limit. Having both switches well off > > center moderately well precludes its looking for the switch in the > > wrong direction as its z is normally parked well away from the work > > for gauge etc checking access to whats in the chuck. > > > > So this has worked well for a year for me. > > > > But I see no way to make certain that it looks in the right > > direction to find home on a mill without physically locating the > > switch well away from the center of the table. eg the switches trip > > point will need to be toward the end of the travel, and a > > HOME_OFFSET then used to put zero back at the center. From there, > > soft limits + & - would be used in place of the existing knobs > > hitting the heavy centrally located pointer. > > > > There are not currently available, any such Y stops, so all of that > > will have to be imagined and built. Again, off center far enough > > that from any usual powerup location that it will search in the > > correct direction to find the switch, then a HOME-OFFSET that > > establishes a zero on the dro's at the center of travel, and again > > setting software limits from this arbitrary home rather than actual > > limit switches. > > > > Z homing will either be completely arbitrary because of uncalibrated > > tool lengths in an R8 spindle, or some sort of a pcb pad laying on > > the table and a G38.2 move to find the top, sensitive side of the > > pcb. In that case, how do I guarantee that I will remember to place > > the contact pad? It needs to be automatic, and that gets complex. I > > often do that on my toy mill, jogging down slowly till I get the > > probe tripped message, and then home it after rising even slower to > > open the contact so I can home it there, typically about .064" above > > the table. This would be more accurately done IF I could home with > > the probe tripped, but LCNC won't allow that. And I'm too lazy to > > put a switch in the probe lead. ;-) > > > > Or just sense someplace on the column & call it home. That doesn't > > seem to be foolproof either, and I am absolutely the fool that will > > prove it. :) > > > > Thats why I ask for ideas & pix. If you PM me pix, include perms to > > post them on my web page, and they will then be available for others > > till I fall over. > > Cheers, Gene Heskett -- "There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order." -Ed Howdershelt (Author) Genes Web page <http://geneslinuxbox.net:6309/gene> ------------------------------------------------------------------------------ _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
