On Wednesday 09 September 2015 05:30:29 andy pugh wrote: > On 9 September 2015 at 04:13, Jon Elson <[email protected]> wrote: > > Ohhh, that sounds like trouble brewing! I can think of a > > high speed spindle running for 8 hours at 20K RPM with a > > 1000 count/rev encoder, that gives 9.6 billion counts. > > Then, a tool change that needs a spindle orient. Hmm, > > maybe a little electrical noise that causes it to pick up a > > few extra counts per minute, too. And then it is > > mispositioned for the change. > > The obvious answer is for you all to follow my lead and use resolvers > for your spindle encoders, then there is no danger of accumulated > miss-counts. > > The proposed M19.1 explicit home command could be used if there was a > real danger. > > Thinking further about accuracy, I believe that the position is > calculated from a 64-bit accumulator of encoder edges, I think that my > previous calculation might have been pessimistic.
And I think its much ado about nothing. The idea of having it running at 75 rpm as it is moved to the tool changer where it encounters a mechanical stop at the correct orientation, a stopped condition that one of the hal retriggerables could then power off the spindle and signal the tool changer to proceed with whatever it does to change the tool. That could get complex as my now R8 setup would need a rotation incorporated in the arm if the R8 itself was being changed for a different shank size. They are currently stored upside, tail down in a rack shelf I made. For those, there is the problem of getting the bolt activated too, as it spins with the spindle. The simplest way would be to drop the wrench, 8mm square and fixed to a pivot point, onto the bolt, selecting low range in the head and issue a 1 second long 1000 rev reverse, dropping the R8 into the changers gripper. The only orientation needed there is when dropping the wrench onto the bolt. And it has 4 legal positions. Properly adjusted, the spindle could be run fwd slow and let the wrench drop on. Tally switch that, and when the switch says the wrench is down, whang the spindle in reverse to unlock and unload. Raising the spindle at least 4.5" as the bolt is loosened and dropping the r8 and tool into the changer arm. Then have the arm turn it over and park it in the proper holder slot, move the strip of R8's to the correct position to pick up the next tool, do it, turn it over, and present it to the spindle and lower the head till the lock pin engages and the bolt can then be tightened to its normal torque again. Retract the wrench, wiggling it back and forth at low power to assure it does disengage. Move to the tool length setter and do a G38.2 to autoset the tool length off set as that is NOT fixed with an r8 spindle. And get back to work. Done correctly, I can't see it taking more than 30 seconds on my machine. Once done, and the motions calibrated, slow-mo gets edited to fast-mo of course. This machine can move at 1200mm/minute if the can is tipped high enough. :) The major problem might be dropping the tool bit out of the r8, my .375" drill bit fits that 3/8" r8 loose enough that it would drop out as soon as the bolt is loosened. And the shank mics at about .3745". The only place where exact spindle position is needed is when re-inserting the R8 and finding the drive key as the changers arm will have no knowledge of where it is. Then the quickest way would be have the changer arm spring loaded a bit, lowering the head till it is exerting some upward pressure, and running the spindle slowly while the arm is pressing up gently on the r8 until it snaps up, exactly as I do by hand right now. This without the wrench being engaged of course. Once it snaps up and engages the bolt, drop the wrench and tighten the bolt. So there would be two wrench engage/disengage cycles involved. Moving the carrier rack, and the arms motions are left as an exersize for the coder. Its all going to be custom made anyway. And a lot of that could be operated by a 12 volt battery power as I will eventually have that available to run the small compressor that runs the table lock release when I want to move a rotary table to a new angle. Stuff on my "bucket list" IOW. Or by relays running motors as I have been collecting small, intermittent duty but strong brushed universal electric motors for just such raw material. Most are bare franed unforch so would have to be swarf shielded. Like a windshield wiper arm racheting the parking strip (or wheel) for the tool holders. That would involve a custom written homing command that could be inserted in the homing sequence at powerup. At no point in this scenario, do I need precise spindle positioning, its all "found" by judicious, gentle turning of the spindle, both to find the wrench and to "find" the R8 drive key. Other spindle styles may, but an R8 is not married to the concept IMNSHO. It will probably get done if I don't fall over first. :) 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> ------------------------------------------------------------------------------ Monitor Your Dynamic Infrastructure at Any Scale With Datadog! Get real-time metrics from all of your servers, apps and tools in one place. SourceForge users - Click here to start your Free Trial of Datadog now! http://pubads.g.doubleclick.net/gampad/clk?id=241902991&iu=/4140 _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
