#robotics Logs

Oct 28 2019

12:32 AM flyback: https://www.liveleak.com/view?t=CH3RD_1571685536 <--- *CANUCKED**CANUCKED**CANUCKED**CANUCKED*
12:37 AM * flyback drags rue_mohr aginst the e-bike
07:25 AM veverak: so
07:26 AM veverak: again, I have the need to build new robot
07:26 AM veverak: veverak: but you have a lot of unfinished ones!
07:26 AM veverak: veverak: shut up
07:26 AM veverak: :)
07:26 AM veverak: the dilema!
07:29 AM Tom_L: just listen to the evil you
07:32 AM veverak: ooooh, I do regullary
07:32 AM veverak: :)
09:47 AM mrdata: veverak, did you finish any?
09:47 AM veverak: mrdata: not usually
09:47 AM veverak: but we did finish our car for NXPCUP
09:47 AM mrdata: ok
09:47 AM veverak: where finish = it participated the race
09:47 AM veverak: and ltely
09:47 AM mrdata: good
09:47 AM veverak: I finished a prototype for robotics course we are preparing
09:47 AM veverak: :)
09:48 AM mrdata: then you may proceed
09:48 AM veverak: however
09:48 AM veverak: ma legged quadruped is standing and blinking, not walking
09:48 AM mrdata: poor thing
09:48 AM veverak: and the tracked robot is collecting dust
09:48 AM mrdata: it will haunt you while you tinker with other bots
09:48 AM veverak: as I am not able to make robust raspberry pi setup
09:48 AM veverak: :/
09:49 AM mrdata: these are not failures; theyre just waiting for improvements to come along
09:49 AM veverak: yeah
09:49 AM veverak: hmm
09:49 AM veverak: I would like some small fast to finish project maybe
09:50 AM veverak: but half of the ideas I have screm "long term development"
09:50 AM veverak: touche
09:50 AM mrdata: you should get into the mind set of elon musk
09:50 AM mrdata: "if the schedule is long it's wrong; if the schedule is tight, it's right"
09:52 AM mrdata: he's doing a small fast to finish project himself these days
09:52 AM mrdata: it's called star ship
09:53 AM mrdata: costing only 5% of the business
09:53 AM mrdata: theyre getting ready for a 20 km hop
09:55 AM mrdata: my own project is a weird planetary gear thing, that could be doomed because i lack machining ability
09:56 AM veverak: yeah, I am aware o that
09:56 AM veverak: hence I want long project
09:56 AM veverak: *short
09:56 AM veverak: damn it
09:56 AM veverak: :)
04:41 PM bgilb: this robot im using expects 4 aaa batteries which are upregulated from 6v to 9.25v then down regulated to 5v. Would it be bad to use a 12v ACDC power supply?
04:42 PM Cracki: does anything in there need 9.25 volt
04:42 PM Cracki: 5v sounds like logic supply, what supplies the motors?
04:42 PM bgilb: the batteries
04:42 PM bgilb: its an m3pi, but the battery pack broke off and was lost
04:43 PM Cracki: then you wanna check if they can take 12v, otherwise you need to stepdown the 12v to 9 or so
04:43 PM bgilb: i noticed i can plug a 12v power supply into the charging port and it works, but didn't want to start a fire
04:43 PM bgilb: https://www.pololu.com/docs/0J21/5.b
04:43 PM Cracki: barrel plug
04:43 PM Cracki: they aren't coded
04:43 PM Cracki: you can plug gigavolts in there
04:43 PM bgilb: lmao
04:43 PM bgilb: and it will be okay?
04:43 PM Cracki: obviously not
04:44 PM Cracki: because something expects a tolerable voltage
04:44 PM bgilb: it says a switching regulator boosts to 9.25, then a linear regulator brings it back down to 5V
04:44 PM bgilb: "Instead, in the 3pi, a switching regulator first boosts the battery voltage up to 9.25 V (Vboost), and a linear regulator regulates Vboost back down to 5 V (VCC). Vboost powers the motors and the IR LEDs in the line sensors, while VCC is used for the microcontroller and all digital signals."
04:44 PM Cracki: sure, but does anyhting use those 9.25 volts
04:44 PM Cracki: motors don't like to be overvolted too much
04:44 PM bgilb: yeah the motors and IR leds
04:44 PM Cracki: aha
04:44 PM Cracki: check what the motors can take
04:45 PM Cracki: consider NOT giving them 12 volts when they're been designed for ~9 volts
04:45 PM bgilb: i didn't know if the "regulator" was smart
04:45 PM bgilb: as in it won't boost 12v to like 15v for example
04:45 PM bgilb: i guess i could test with a voltimeter?
04:46 PM Cracki: if it's a boost (step up) reg, it can't step down
04:46 PM Cracki: how about you give it 6 volts and pretend the wall wart is four batteries
04:46 PM Cracki: that's the smallest change
04:46 PM bgilb: i could find a 5v charger too right?
04:46 PM Cracki: sure
04:47 PM Cracki: it'll probably accept 5 volts into the boost
04:47 PM bgilb: okay
04:47 PM Cracki: that's ~1.2 volts per cell
04:47 PM bgilb: yeah because it said it handles from 3v to 6v
04:48 PM Cracki: work with the schematic if you want to interfere in its power https://www.pololu.com/file/0J119/3pi-schematic.pdf
04:49 PM Cracki: for some reason they send 9.25 volts into five optoisolators in series
04:49 PM Cracki: ah, optical sensors
04:50 PM bgilb: i plugged it into J7
04:51 PM bgilb: hmm only found a 7.5V ac/dc plug
04:52 PM Cracki: now you can find out what step-up regulator it uses, and read its data sheet
04:53 PM Cracki: boy that thing's "schematic" is being mysterious
04:53 PM bgilb: https://www.pololu.com/product/791
04:54 PM bgilb: it says input voltage 1.5V to 16V
04:54 PM Cracki: they always boost
04:54 PM Cracki: so if you need 9.25 out of it, you are limited to something upto around 8 volts
04:55 PM Cracki: there are buck+boost regulators but that's not one of them
04:55 PM bgilb: this isn't the one i have?
04:55 PM bgilb: i found a link on the forum
04:56 PM Cracki: wow, that stuff isn't open hardware at all
04:56 PM bgilb: yah he was really vague on the forum post
04:56 PM bgilb: but i see some of the same chips on the bottom of my board
04:57 PM bgilb: so does this mean 12V is okay?
04:57 PM Cracki: it means 12 volts is not ok
04:57 PM Cracki: NOT.
04:57 PM Cracki: if you give 12 volts into a boost reg, it'll output something that's 12 volts AT MINIMUM
04:58 PM Cracki: and then you need to figure out if EVERYTHING that used to get 9.25 volts is ok with 12+ volts
04:58 PM Cracki: motors may accept it, but look closely at the optical sensor circuit
04:59 PM Cracki: i.e. get data sheets for those optical sensors, specifically what supply they can take
05:02 PM bgilb: okay that makes sense
05:02 PM bgilb: i have a 7.5V power supply so i think that will just work anyways
05:02 PM bgilb: without any problems
07:40 PM logb0t_ is now known as logb0t
08:44 PM rue_shop1: I got a surprise that not even the current regulation can disable a boost under the input voltage,
08:44 PM rue_shop1: should have known if I'd thought about it
08:44 PM rue_shop1: another $1 lesson
08:44 PM Tom_L: woopsie
08:45 PM rue_shop1: cheap by a lot of measures
08:45 PM rue_shop1: even your least expensive bits are more than that eh?
08:45 PM Tom_L: this card has an intel chipset iirc
08:45 PM rue_shop1: :) I successfully remapped the serial port on the stm32
08:45 PM Tom_L: we won't talk about that :D
08:46 PM Tom_L: nice
08:46 PM rue_shop1: it turns out, even tho the clock to the uart is enabled, and the clock to the io port is enabled, you have to enable a clock to the remapped port system
08:47 PM rue_shop1: its like a house where there is a lightswitch every 3 feet and their all off
08:48 PM Tom_L: are any mounted upside down?
08:48 PM rue_shop1: not yet...
08:48 PM rue_shop1: I have all the components working
08:49 PM rue_shop1: I need to stitch them all togethor over the next ... weeks...
08:49 PM Tom_L: days
08:49 PM rue_shop1: many many distractions
08:49 PM Tom_L: you're on a time crunch.... remember?
08:49 PM rue_shop1: death? yea, always looming
08:49 PM rue_shop1: this should help me make awesome servos
08:50 PM rue_shop1: I can measure a system, and then make a complimentrary filter
08:51 PM rue_shop1: hmm
08:53 PM rue_shop1: so, next goal then is a serial controlled sinewave gen
08:53 PM rue_shop1: but for *right* now, a distraction
08:54 PM rue_shop1: wish my head were clear
09:03 PM * Tom_L puts a clearcoat of laquer on rue_shop1's head
09:27 PM flyback: sudo ./fluxengine read ibm -s :d=1 -o sony1.img | tee sony1.txt
09:27 PM flyback: Reading from: :d=1:s=0-1:t=0-79
09:27 PM flyback: 0.0: 191 ms in 182976 bytes
09:27 PM flyback: 36 records, 18 sectors; 0.96us clock (1037kHz);
09:27 PM flyback: logical track 0.0; 9216 bytes decoded.
09:27 PM flyback: 0.1: 191 ms in 182144 bytes
09:27 PM flyback: 36 records, 18 sectors; 0.98us clock (1018kHz);
09:27 PM flyback: logical track 0.1; 9216 bytes decoded.
09:28 PM flyback: \
10:40 PM marder is now known as katsumi