On Thursday, 5 May 2022 17:32:54 EDT gene heskett wrote: > On Thursday, 5 May 2022 16:44:32 EDT gene heskett wrote: > > On Thursday, 5 May 2022 14:51:21 EDT gene heskett wrote: > > > > (M111) > > > > screw, but I'll try it again when thats done. > > > > finished, if it finishes the screw. > > > > > > And I noted that it was not returning to zero at the end of a > > > threading loop, so I changed the drive for another one that I think > > > is 50/1. > > > > If so, it will take a SCALE entry thats 5/4ths of the current one. > > > > > So > > > in prep for recalibration of it, I tried my M111. no error, and it > > > worked, so it that regard, I'm in business. Now for a recal run. > > > > And I am not getting sensible answers. > > I have two sample-holds made out of mux2's, both watching the > > stepgen3- counts which supposedly represents the number sent to the > > driver, like this: > > setp b_tally.gain0 -1.00000000 # b_tally is a sum2 > > setp b_tally.gain1 1.00000000 > > # count 100 trips of the b home switch > > setp comp_cal0.in0 4.00000000 # count 4 turns > > setp comp_cal1.in0 104.00000000 # count 100 more turns > > net bpos_counts <= hm2_5i25.0.stepgen.03.counts > > net bpos_counts => S32_float-cal.in > > net bpos_countf <= S32_float-cal.out > > net bpos_countf => mux2cal0.in0 mux2cal1.in0 # B steps made to both > > # sample-holds > > > > net bcalibrate0 <= mux2cal0.out => mux2cal0.in1 b_tally.in0 # make > > into sample-hold > > net bcalibrate1 <= mux2cal1.out => mux2cal1.in1 b_tally.in1 # make > > into sample-hold > > > > And watch b_tally,in0 for first count freeze > > watch b_tall.in1 for 2nd freeze > > At which point b_tally.out is showing the counts for 100 turns of the > > B drive. A rather large 8 digit value. > > > > Divide that by 100 to get steps per one full turn of the axis. > > > > Then since the axis is angular, and in degrees, divide that by 360 to > > get degrees. stop lcnc, edit the ini file and update that > > STEPGEN_SCALE with the number. Touch off b@0.00. run B either way to > > 360, but its short, I need to run it to around 390 to look like a > > full > > furn. > > > > Each time I run this, I get a bigger result. > > I should say a different result. This time it went from 146.xxxxx to > 144.xxxxxx. I'm going to plug in a dbounce on the b home switch.
I did that, using a dbounce module, and watching the switch and the dbounces output, 200ms per division on the halscope running at 2 kilohertz servo-thread, I could occasionally see a double pulse on the falling edge. And it walked right thru the dbounce at a delay of 2, 3 didn't help and I kept setp-ing its delay until I could see a complete delay in halscope that covered the bounceing on the falling edge, then doubled it for a delay of 87 in whatever units it uses to set delay. The man page could use some TLC. 87 seems unreal, but I think I have found that problem. It was counting both pulses about 10% of the time. Change subject: 3d printers use the cheapest way to home printers, often just running them against the stops, sometimes detecting the extra current to keep from sitting there banging the stop for seversal seconds. But they all, better ones anyway, use an inductive sensor to measure bed flatness. That sensor is uber cheap as a service part. What sort of drive does it use, and will it work against an alu frame as opposed to the steel bed sheet it normally senses. Anybody here know anything about how one might use them for home switches on our stuff? Thanks Andy and all. 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, 1940) If we desire respect for the law, we must first make the law respectable. - Louis D. Brandeis _______________________________________________ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
