#garfield Logs

Jan 05 2023

12:25 AM rue_mohr: and the dc motor is 1880 counts
12:36 AM rue_mohr: skjdfksjdfh
12:36 AM rue_mohr: 2880
04:01 AM rue_mohr: I'll forever reffer to it as tom pwm
04:01 AM rue_mohr: whats funny
04:01 AM rue_mohr: is that I think I'll work better for the control method needed to make a dc motor act like a stepper than other things
04:03 AM rue_mohr: a number I didn't gather, and will need is to know, with a stepper, how many degrees the shaft will pull by under load
04:03 AM rue_mohr: the melatonin left me high and dry at the same point I would have been in without it, about 2am
04:04 AM rue_mohr: if It'd not put me out, I'd still be trying to get to sleep now
04:42 AM Tom_L: i didn't invent it
05:24 AM rue_mohr: no
05:24 AM rue_mohr: but moon didn't invent that oscillator he found online and I named it after him
07:52 AM aandrew: polprog: sure. There's also some good projects I found I'll share when I'm able to do it easier
07:54 AM rue_mohr: arg, positional data from encoders sucks!
07:55 AM Tom_L: moonscope
07:55 AM Tom_L: err moonillator
07:56 AM rue_mohr: ?
07:57 AM rue_mohr: ok I _THINK_ I see peak velocities from the stepper motor of 400 degrees/second
07:57 AM rue_mohr: which seems a LOT lower than I'd expect
07:58 AM Tom_L: 1.x rpm
07:59 AM rue_mohr: when the sensor jitters over a line, the stupid data spikes to 50000, which isn't true
07:59 AM Tom_L: 1.11
07:59 AM rue_mohr: this is the peak velocity during one step
07:59 AM rue_mohr: where the shaft stops between steps
08:01 AM rue_mohr: part of the problem is the concept of translating the encoder pulses to postiion
08:01 AM rue_mohr: instead of seeing a state change and saying "now I'm at this position" I need to to it like a racetrack, with a start and stop line
08:02 AM rue_mohr: and say" between these two lines, you took this long" so your velocity is:
08:03 AM Tom_L: find a resolver
08:03 AM rue_mohr: if you jitter over the start line, your not going anywhere because you haven't crossed a finish line
08:03 AM rue_mohr: same probelm with any sensor
08:03 AM rue_mohr: position/reading jitter
08:03 AM Tom_L: they're more analog
08:03 AM Tom_L: phase
08:03 AM rue_mohr: as soon as you take it to any integer position you have the same problem
08:05 AM rue_mohr: I kinda need a state machine that says "rolling back and forth over the start line doesn't count"
08:05 AM rue_mohr: but I need to work out how to define two lines
08:06 AM rue_mohr: in my code, I'v reduced the positions to floating point angles
08:06 AM rue_mohr: which doesn't help
08:07 AM rue_mohr: I need to filter the data so I can measure the time to cross over 2 unique transitions
08:08 AM rue_mohr: so I need a bit of a history
08:13 AM Tom_L: off to face the world..
09:25 AM polprog: aandrew: awesome!
11:05 AM aandrew: polprog: polprog: https://martein.home.xs4all.nl/pa3ake/hmode/index.html
11:05 AM aandrew: polhttps://www.eevblog.com/forum/oshw/open-source-hw-rf-signal-generator/?all
11:06 AM polprog: today i was experimenting with some IF generator
11:06 AM aandrew: polprog: some more here too: https://www.eevblog.com/forum/rf-microwave/how-to-control-ltdz-max2870-23-5-6000mhz-signal-source/
11:06 AM polprog: i made a 20MHz oscillator and a mixer
11:06 AM polprog: aw thanks
11:06 AM aandrew: that xs4all page is like pure RF geek porn
11:07 AM aandrew: every time I dip my toes into RF I realize I'm at the edge of the marianas trench
11:07 AM polprog: very nice
11:07 AM polprog: (ill post the schematics in a few moments, gotta do something in thehouse)
11:18 AM polprog: i basically have a butler oscillator that seems to work with 10-32 MHz crystals
11:18 AM polprog: and then its coupled into a single ended mixer
11:18 AM polprog: two transistors, one for the osc and one for the mixer
11:18 AM polprog: 2n2222
11:19 AM polprog: tho when i inserted a 4433kHz crystal into the oscillator it started oscillating at third overtone so there might be a need for a filter in the feedback pack
11:48 AM polprog: aandrew: https://polprog.net/rozne1/ircjunk/uhffun/oscillators.html
12:48 PM polprog: sroll down to the last one
07:03 PM rue_shop3: well I ahve an interetsing mystery
07:03 PM rue_shop3: I set up an encoder servo for bang bang with motor direction
07:03 PM rue_shop3: but
07:04 PM rue_shop3: even tho the setting of the direction is IN the encoder interrupt routine
07:04 PM rue_shop3: (which is fine for execution time)
07:04 PM rue_shop3: its not changing the direction for a few encoder counts
07:05 PM rue_shop3: it should be doing it right away
07:11 PM rue_shop3: oh wait I get it
07:11 PM rue_shop3: by the time the next encoder pulse comes in, its already picked up too much speed
07:12 PM rue_shop3: so it overthrows by a few lines before it turns around, then picks up too much speed in reverse
10:52 PM rue_mohr: I'm gradually adding this u
10:52 PM rue_mohr: p