#avr Logs

Dec 31 2021

12:18 AM specing_ is now known as specing
12:18 PM specing_ is now known as specing
05:32 PM bpye1 is now known as bpye
05:47 PM qu1j0t3: .
05:48 PM qu1j0t3: happy new year, ahem
09:50 PM rue_mohr: happy new years, anyone still counting down?
09:50 PM qu1j0t3: yes.
09:50 PM qu1j0t3: 10.35pm here.
09:50 PM qu1j0t3: so i've pulled out the breadboard and scope.
09:51 PM Phantom: hally new year rue! and for qu1j0t3, let's celebrate later on together...
09:51 PM Phantom: happy*
09:51 PM qu1j0t3: \o/
09:51 PM qu1j0t3: indeed, happy new year rue_mohr
09:52 PM Phantom: hey rue_mohr ! question for ya, have you played alot with steppers and microstepping and mesured the precision vs current?
10:10 PM rue_mohr: precision?
10:10 PM rue_mohr: for microsteps??
10:13 PM Phantom: yeah like, let's say you take a big stepper and run it at 50% power, vs a smaller one at 100% power, both producing the same torque... do you have the same microstepping precision?
10:28 PM rue_mohr: there is no microstepping precision under load ever
10:28 PM rue_mohr: ever
10:28 PM rue_mohr: all it does is smooth out operation when running
10:28 PM rue_mohr: marketing would tell you otherwise
10:28 PM rue_mohr: but its all smoke
10:29 PM rue_mohr: I put a microstepped motor on a high resolution encoder
10:29 PM rue_mohr: I wanted to know how long it took to settle after a step
10:29 PM rue_mohr: as a benchmark for a dc servo
10:30 PM Phantom: that's what I tought...
10:31 PM Phantom: so for the 3d printer, the best move would be to use some big nema17 0.9°
10:31 PM Phantom: and basically ignore the microstepping...
10:31 PM rue_mohr: no
10:31 PM rue_mohr: close tho
10:31 PM rue_mohr: I use 8x stepping, thats reasonable for smooth running
10:32 PM Phantom: I need more torque to accelerate and deaccelerate the mass
10:32 PM rue_mohr: but add ateast a 2:1 gearbox
10:32 PM rue_mohr: yes
10:32 PM rue_mohr: the torque speed ratio of the motors IS wrong for direct drive
10:32 PM rue_mohr: your right
10:32 PM rue_mohr: I find 2:1 is ok
10:32 PM Phantom: gearbox have backlash... maybe a belted reduction?
10:32 PM rue_mohr: you dont want to tax the step rate of the microcontroller
10:32 PM rue_mohr: not all gearboxes
10:33 PM rue_mohr: look at gearboxes in cdrom drives
10:33 PM rue_mohr: what do they do
10:33 PM Phantom: the ones I saw was mostly direct drive
10:33 PM Phantom: worm drive for the head, direct for the spindle...
10:34 PM rue_mohr: you havn't see the rack drive ones?
10:34 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1220490.jpg
10:34 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1220491.jpg
10:35 PM rue_mohr: they split the gear, and use a torque spring that tries to sheer the teeth
10:35 PM Phantom: didn't seen those on cd
10:35 PM rue_mohr: fiar enough, make 2022 the year of cdrom stripdowns
10:36 PM Phantom: I'll probably just ends up using belted steppers
10:37 PM Phantom: that should work fine
11:30 PM Phantom: rue_mohr: so, I should instead find a good stepper which can do higher speed, gear it down via belt, which would increase the torque... yet not lose precision... sound right? like a 4:1?
11:43 PM rue_mohr: they can all do high speed
11:43 PM rue_mohr: have you ever used one of those drivers to wind up a stepper?
11:43 PM rue_mohr: its CRAZY
11:44 PM rue_mohr: I found 2:1 to be fine
11:44 PM rue_mohr: half steps area usually pretty reliable