#robotics Logs

Dec 20 2020

03:08 AM rue_shop2: aluminum: Etch it in a 1:3 muriatic acid to hydrogen peroxide solution
03:46 AM Jak_o_Shadows: Just bought 3x v4 of my servo driver PCB
03:47 AM Jak_o_Shadows: plus some more parts, because I was running out of micro-controllers and 20 pf caps
12:10 PM rue_mohr: I wonder how many servos respond to better than 5 bits resolution
12:10 PM rue_mohr: (hobby servos)
01:20 PM mrdata_: 5 bits is 1/32 of 180 deg = 5.6 deg; you generate a PWM signal to control these, usually, and so a microcontroller can do better than the five bits; the servo is analog
01:20 PM mrdata_: i programmed one with a Ne555 timer
01:21 PM mrdata_: adding a potentiometer
01:21 PM mrdata_: the controls were quite smooth
01:22 PM mrdata_: so yes these exist
01:25 PM mrdata_: i believe you can count on 7 or 8 bits
01:30 PM veverak: not every servo is analog
01:30 PM veverak: and even analog has limits
01:30 PM veverak: and a lot of chinese cheap servos are 256 steps
01:32 PM rue_mohr: yea, I made the only library with stable 11 bits, but nobody wants to use it
01:33 PM rue_mohr: 24000 counts
01:33 PM rue_mohr: so actually a bit more than 14 bits
01:34 PM rue_mohr: but I also did it from 0.5ms to 2.5ms, cause lots of servos have range beyond the standard values
01:35 PM rue_mohr: but recently I made a switch-interfaced signal gen, and I noticed the servos barely responded past 5 bits
01:35 PM rue_mohr: as in, most of the time they didn't respond
01:36 PM rue_mohr: oh I didn't post that project?
01:37 PM rue_mohr: oh well I have to do that
01:37 PM rue_mohr: ok, time time