On Friday 29 April 2016 01:25:08 Gene Heskett wrote: Continued. > On Thursday 28 April 2016 21:00:08 Peter C. Wallace wrote: > > On Thu, 28 Apr 2016, Gene Heskett wrote: > > > Date: Thu, 28 Apr 2016 20:46:36 -0400 > > > From: Gene Heskett <ghesk...@wdtv.com> > > > Reply-To: "Enhanced Machine Controller (EMC)" > > > <emc-users@lists.sourceforge.net> > > > To: emc-users@lists.sourceforge.net > > > Subject: Re: [Emc-users] New motor and driver in, runs w/o the > > > springs, at 10% of the big motors noise, but no faster. > > > > > > On Thursday 28 April 2016 19:38:19 Peter C. Wallace wrote: > > >> On Thu, 28 Apr 2016, Gene Heskett wrote: > > >>> Date: Thu, 28 Apr 2016 18:29:04 -0400 > > >>> From: Gene Heskett <ghesk...@wdtv.com> > > >>> Reply-To: "Enhanced Machine Controller (EMC)" > > >>> <emc-users@lists.sourceforge.net> > > >>> To: "Enhanced Machine Controller (EMC)" > > >>> <emc-users@lists.sourceforge.net> Subject: [Emc-users] New motor > > >>> and driver in, runs w/o the springs, at 10% of the big motors > > >>> noise, but no faster. > > >>> > > >>> Greetings; > > >>> > > >>> I can do 48.7 IPM all day, either direction, but one cliock > > >>> faster on the jog speed, to 51 IPM and I have a following error > > >>> in about 3/8" of movement, independant of the direction. Next > > >>> step up on the slider is about 53.x IPM, gives a following error > > >>> in about 3/16" of motion. > > >>> > > >>> The step timings are: > > >>> dir preset 6000 > > >>> Step_len 3000 > > >>> Step_space 3000 > > >>> dir hold, 6000 > > >>> > > >>> Worst case, 18 u-s to do a dir reverse and one step then revert > > >>> the dir. 12 u-s for dir_setup+dir_hold, and 6 more for the > > >>> step=18. > > >>> > > >>> Grabbing a copy of kcalc, enter 18 as .000018, do a 1/x to get > > >>> the max freq=55,555.5555555555556 or a bit over 55 kilohertz. > > >>> Drive claims it can do 200 kilohertz. driver is set for /8 > > >>> microsrepping. Screw is a 16x5mm. > > >>> > > >>> So how fast can it go when the pulses are at 200 KHz? > > >>> > > >>> FWIW, this tripping of the following error was taking place at > > >>> the exact same speed ( give or take a couple percent ) with the > > >>> previous driver & motor which was set for a /64, and the same > > >>> step timeings and this one is set /8, so no way can I convince > > >>> myself the problem is the maximum step rate the 5i25's stepgen > > >>> can do, or that the driver can eat and be happy. > > >>> > > >>> I have fiddled with the ini files maxvels, and maxaccels, > > >>> without a measureable effect. > > Found it, for some reason I had a pid.z.max-out set at 8, not 0. > Cleared that, tweeked the max_accels and maxvels up to 90 IPM, and > zero'd the stepgen_maxvels. W/O motor power, it just works while > jogging all 3 axis's at 90 IPM. That however, is faster than the nema > 23's on the x,y can move, so I know they'll stall under power.
This was the magic twanger. Even with the accels cranked up to double what the old driver/motor/psu could do, this motor/driver-psu does 90 IPM both ways with zero strain and 1/10th the noise that 1600 made. And while Y can do 90 IPM, x cannot when the table is way off-center, and while both will do 90 IPM, both have a heavy rumble sound when moving that fast, so Y is now slowed to around 75 IPM, and X is at 60 IPM & both now sound quite happy doing it. My torture test code, a G33.1 peck loop that taps air for a 4mm.7 screw but only advances about 1/4 turn per peck so I can let it run for quite a while, and has a 2 sec pause at the top for dusting off the tap and anointing it with a fresh drop of Buttercutt, can run almost un-eventfully while its tapping a 5" deep 4x.7mm hole in the air. Running at spindle revs is the 1100 range, in high gear, the spindle turnaround overshoot is still under 2 full turns of the spindle. I say almost because If it wasn't for the head gearing sounding like its eating its ball cages now and then, I'd be waking up the rest of this campsite whooping it up, problem solved. So I guess its time to see if I can put in the AC ball bearings Hoss made for his, and replace all the rest of the head bearings with better stuff. I may even drive up to his place and get some tool changer ideas, in addition to the AC conversion kit if he is making them for sale. That doesn't show on his web site though. He's up in the WV panhandle, nominally west and a hair north of Pittsburgh PA. About a 3 hour drive I'd guess, given that a 1/4 mile straight stretch of road is a rare occurrence in these here parts. :( > > >> Sounds like the stepgen maxvel or stepgen maxaccel settings are > > >> limiting you, so you are getting a following error. With 3/3 usec > > >> step timings you should get close to to 166 KHz. The stepgen > > >> maxaccel and stepgen maxvel parameters should be set to about > > >> 1.25X the machine acceleration and velocity limits > > > > > > Would this be more related to the TP asking for a higher velocity? > > > > No, the stepgen runs a feedback loop so must have headroom, your > > symptoms sound like you are running out of headroom > > > > >> If you have backlash compensation enabled, the stepgen maxaccel > > >> parameter must be set to about 2.5x the machine acceleration > > >> limit > > > > > > Playing last week, backlash is .0016", I found the stepgen_maxvel > > > was best at around 1.35, and currently the max_accel is 10 while > > > stepgen_maxaccel is about 35 > > > > That should be OK > > > > You will have to set stepgen maxvel higher if you want to go much > > faster and its not something you tune, in fact setting it to 0 (no > > limit) should work just fine. > > You are 200% correct. > > > If its picky you possibly have a hal connection or thread order > > error > > > > Also, what are the ferror settings? > > ferror_min = 0.0005 > ferror = .010 > I can now lower both I expect. Yup, NP at 0.00005 and 0.002, all 3 > axis's at 90 IPM. > > Now. I expect under power I'll have stalls, so I'll have to limit the > maxvels below the 1.5 (90 IPM) I left them at. If this DQ2722MA > driver wasn't so expensive, (and big, its taller than my box is deep > by about 1.5") I'd get 2 more as 40 volts isn't enough for the 472 > oz/inchers either. X is more limited than Y of course as it's home > switch is about 1/4" from the left mechanical stop, and that off > center weight binds the way's, a lot, so it takes pretty slow accel's > and vels to get it to come back toward center after homing at creep > velocity away from the switch. Quite lube sensitive too and I've been > contemplating ways to keep an oil soaked (vactra 68) way wiper on X. > Road to..... is paved with the best of intentions. > But we'll find our new stall speeds tomorrow, before I clean and lube > the X ways again. > > FWIW, the diff between these two drivers is amazing, that 920oz motor > is getting microstepped to higher speeds & runs amazingly quiet, where > that 1600oz, running on 60 volts thru that DM860 driver, was a rattle > and buzz you could hear across the street if the garage door was open. > At 1/2 an RPS! > > I'll also find out if there is a big diff in up vs down speeds as I've > not yet installed the counterweight springs. > And still haven't. And may not. I can't hear anything that makes it sound like its working harder when lifting the head. So Skullworks, Peter and Andy thank you all, you have been very helpful. FWIW, I checked the Docs, and the <= and => stuff is active if there and does allow out of order in the halfile, so I didn't putz with it. So I am sitting here with a VBWG. One more problem totally solved. 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> ------------------------------------------------------------------------------ Find and fix application performance issues faster with Applications Manager Applications Manager provides deep performance insights into multiple tiers of your business applications. It resolves application problems quickly and reduces your MTTR. Get your free trial! https://ad.doubleclick.net/ddm/clk/302982198;130105516;z _______________________________________________ Emc-users mailing list Emc-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-users
