#garfield Logs
Apr 30 2020
#garfield Calendar
12:46 AM polprog: good morning
12:52 AM rue_shop1: ok, I got the change to the driver behaving
04:53 AM polprog: rue_mohr: cant say.. rise time till what? till 12V?
04:54 AM polprog: rise time till 12V is infinite :^)
05:23 AM polprog: hmm, trying to crack getopt usage
08:32 AM rue_mohr: heh
10:07 AM polprog: for some reason optarg is 0x00
10:07 AM polprog: 0x0 as in null ptr
10:12 AM polprog: ah
10:12 AM polprog: rtfm
10:33 AM polprog: looks like double doesnt have enough resolution for keeping nano and pico values
10:33 AM polprog: grr
10:47 AM zhanx_ is now known as zhanx
11:52 AM polprog: too bad nF and pF would be more useful than Gohm
11:53 AM polprog: i dont wanna reimplement floating point numbers..
11:53 AM zhanx: printing printing
11:57 AM Tom_L: bot parts?
11:58 AM zhanx: yep
11:59 AM zhanx: servo's are in and test fit was good so printing the rest
12:08 PM Tom_L: what happened to the first legged bot you had?
12:18 PM zhanx: the large spider one?
12:19 PM Tom_L: ya
12:20 PM zhanx: oh that was stolen a few years back, someone must have thought it was a toy or something
12:20 PM zhanx: but this little guy will be fun to play with
01:07 PM polprog: rue_mohr: https://polprog.net/rozne1/ircjunk/various/rclp.tgz
01:07 PM polprog: wrote it just now
01:07 PM polprog: what do you think? i feel like adding thermal noise across a single resistor could be useful
01:08 PM polprog: its just some formulas.. this is a good excuse to try numerical methods
01:08 PM polprog: so far its just formulas*
01:09 PM polprog: oh, i know
01:09 PM polprog: ill make it solve R or C for a given C or R and f3db
01:09 PM polprog: and pick the closest one from the std serie
01:24 PM polprog: it also has some stand alone functions for printing out stuff with SI prefixes
01:26 PM Tom_L: linux command to find parallel port address?
01:31 PM zhanx: cat /proc/ioports | grep parport
03:19 PM polprog: neat trick
03:19 PM polprog: looks like it needs sudo, ran as normal user there are zeros everywhere
03:32 PM polprog: i wonder if its easy to print food safe stuff on 3d printers...
03:32 PM polprog: https://en.wikipedia.org/wiki/Pythagorean_cup
04:12 PM rue_mohr: hey
04:12 PM rue_mohr: polprog, I was able to hit pico using double
04:12 PM polprog: hey
04:13 PM polprog: i was silly.. printf defaults to 6 decimal places ;)
04:13 PM polprog: its fixed now
04:13 PM polprog: check the code out
04:13 PM polprog: its ugly but i think its a good starting point.. trying to write a filter solver now
04:13 PM rue_mohr: I made a lib for standard values...
04:13 PM rue_mohr: just recently
04:13 PM polprog: hmm
04:13 PM rue_mohr: and an Si lib
04:14 PM rue_mohr: for input and output
04:14 PM rue_mohr: heh
04:14 PM rue_mohr: :)
04:14 PM rue_mohr: we must be doing the right thing if were both doing it
04:14 PM polprog: i could use one, im just writing a function that picks an e24 value N steps from a given one
04:15 PM rue_mohr: zippo:/morfiles/backup/blackie/files/My Logitech Pictures/Pictures and Videos$ ohm -v 400m -r 6G
04:15 PM rue_mohr: Power is: 26.667pW
04:15 PM rue_mohr: Current is: 66.667pA
04:15 PM rue_mohr: Voltage is: 400.000mV
04:15 PM rue_mohr: Resistance is: 6.000GOhms
04:15 PM rue_mohr: I did all the series'
04:15 PM rue_mohr: you pass it which one you want 6-(48?)
04:16 PM rue_mohr: 'ohm' is the only program I have that uses the SI stuff
04:16 PM rue_mohr: I have a test program for the standard values thing
04:16 PM polprog: i made this one so it takes R, C and E, or any one or two of these
04:16 PM rue_mohr: cool
04:16 PM polprog: and depending on what you give it it spits out different figures
04:16 PM rue_mohr: what about giving it t?
04:17 PM polprog: hmm, not yet.. good idea
04:17 PM rue_mohr: sv = stdResClosest(E24, v = 1200000.0 ); printf("sv: %f error: %f\n", sv, ABS(v-sv));
04:17 PM rue_mohr: thats how my standard value fn works
04:18 PM polprog: $ ./rclib -r 1k -c 1p
04:18 PM polprog: RC time constant (tau)= 1.000000 ns
04:18 PM polprog: 1st order knee freq = 159.154943 MHz
04:18 PM polprog: kT/C noise = 64.357960 uVRMS @ room temp
04:18 PM rue_mohr: interesting
04:18 PM polprog: $ ./rclib -r 1k -v 34m
04:18 PM polprog: I = 34.000000 uA P = 1.156000 uW
04:18 PM rue_mohr: your going to use it in a program for the op amp osc?
04:19 PM polprog: hmm, with the code i have i could
04:19 PM polprog: but its a pretty specific case
04:19 PM polprog: im trying to make it something like your programs
04:19 PM rue_mohr: heh, did you make it solve plain ohm stuff too?
04:19 PM polprog: no, it takes one R, C, and E only
04:19 PM rue_mohr: I can give you my source ya know...
04:19 PM polprog: doesnt do dividers
04:19 PM polprog: hmm
04:20 PM rue_mohr: zippo:/files/programming/c/StandardResistor$ ohm -v 120 -p 1.5k
04:20 PM rue_mohr: Power is: 1.500kW
04:20 PM rue_mohr: Current is: 12.500A
04:20 PM rue_mohr: Voltage is: 120.000V
04:20 PM rue_mohr: Resistance is: 9.600Ohms
04:20 PM rue_mohr: 'ohm' takes 4 arguments, -v (votlage) -i (current) -p (power) and -r (resistance)
04:21 PM rue_mohr: I have my own command line option parser I use
04:21 PM polprog: im using getopt.
04:21 PM rue_mohr: you set up call backs, and it triggers the functions as it goes
04:21 PM polprog: seems simpler for now..
04:21 PM rue_mohr: yea
04:21 PM rue_mohr: I think I learned about getopt after I wrote mine
04:21 PM rue_mohr: but, mine is nice in other ways
04:21 PM polprog: :)
04:22 PM rue_mohr: you can link multiple strings to an option
04:22 PM rue_mohr: --help -h -?
04:22 PM rue_mohr: dont really even need the dash
04:23 PM rue_mohr: I dont know where that saved to
04:25 PM rue_mohr: iiiinteresting
04:25 PM Tom_L: hoz the editor coming?
04:26 PM rue_mohr: zippo:/files/programming/src/polprog/rclp_git# make
04:26 PM rue_mohr: gcc -g -Wall -Wextra RClib.o main.o -lm -o rclib
04:26 PM rue_mohr: collect2: error: ld returned 1 exit status
04:26 PM rue_mohr: Tom_L, I still have to fix my kernel
04:26 PM polprog: hmm, on what
04:26 PM rue_mohr: ah
04:26 PM Tom_L: i learned some more about mine
04:26 PM rue_mohr: make clean, make
04:26 PM Tom_L: built a good one
04:26 PM rue_mohr: there was a .o that tripped it up
04:26 PM Tom_L: was name specific for that rtai patch
04:27 PM rue_mohr: :) you have unused variables!!!
04:27 PM rue_mohr: k
04:28 PM rue_mohr: haha scansci
04:30 PM rue_mohr: I wonder how I did mine
04:30 PM rue_mohr: heh printsci
04:30 PM rue_mohr: I think I'v been using fn's that return strings
04:32 PM rue_mohr: :) my number library is more complex
04:33 PM rue_mohr: for the exponenet, I wanted to cheat, and just pull it from the binary format of the double
04:33 PM rue_mohr: but it seems you cant
04:33 PM rue_mohr: there is all sorts of funkyness to the double format
04:33 PM rue_mohr: ok..
04:34 PM rue_mohr: polprog, nice
04:34 PM polprog: can i steal your closes e24 function..?
04:35 PM rue_mohr: yes, give me a min, I'm home for lunch and then need to run across town
04:35 PM rue_mohr: I'd like to show you my number collector too
04:35 PM polprog: sure
04:37 PM rue_mohr: odd
04:37 PM rue_mohr: zippo:/files/programming/src/polprog/rclp_git# ./rclib -r 1k -v 34m
04:37 PM rue_mohr: I = -nan A P = -nan W
04:37 PM polprog: weird
04:37 PM rue_mohr: did you break it and then send it to me? :)
04:37 PM polprog: no..
04:38 PM polprog: $ ./rclib -r 1k -v 34m
04:38 PM polprog: I = 34.000000 uA P = 1.156000 uW
04:39 PM polprog: maybe the tgz is an old version
04:39 PM rue_mohr: I suspect one of the libraries is not behaving right from my system
04:39 PM rue_mohr: we can check it out later
04:39 PM rue_mohr: (I'm kinda hurrying now)
04:39 PM rue_mohr: why does it take me an hour to cook noodles?
04:41 PM polprog: its a git repo, what is the latest commit?
04:41 PM polprog: fc1257bc
04:42 PM rue_mohr: there is an oddity for the standard resistor library
04:43 PM rue_mohr: there is a library and an application version of the header
04:43 PM rue_mohr: so that when you compile, the tables aren't included twice
04:43 PM rue_mohr: so p = private
04:44 PM rue_mohr: heh, I had an hour and a half to make noodles and eat them
04:44 PM rue_mohr: I'm rushing now
04:46 PM rue_mohr: main.c is a bit broken at the moment, you will need to adust it
04:46 PM rue_mohr: I'm working on a mod so that I can make a list of values from tables
04:46 PM rue_mohr: aka, I want all the E6 values from 10^2 thru 10^6
04:46 PM rue_mohr: so I can mod my programs that do searches
04:49 PM rue_mohr: http://ruemohr.org/~ircjunk/programming/c/stdres.tgz
04:49 PM rue_mohr: polprog, ^^
04:49 PM polprog: just when the closest picking function worked!
04:50 PM polprog: :P
04:50 PM rue_mohr: heh, so I wrote a file called siprint.c
04:51 PM rue_mohr: http://paste.debian.net/1144044/ <-- ;)
04:51 PM rue_mohr: heh
04:51 PM polprog: id argue mine is more elegant :V
04:51 PM rue_mohr: yes!
04:52 PM rue_mohr: for ref, here is my "number collector"
04:52 PM rue_mohr: http://paste.debian.net/1144045/
04:53 PM rue_mohr: full on evil state machine
04:53 PM polprog: wow
04:53 PM rue_mohr: but you can feed it stuff like 1n5
04:53 PM rue_mohr: or 5.6E+30
04:54 PM polprog: hmm, mine cant do that
04:54 PM rue_mohr: yea, I got on a thing trying to make the ultimate converter
04:54 PM rue_mohr: cause I keep needing to do it in my programs
04:55 PM rue_mohr: have to go soon, if you would like that code `proper` I can send it to ya
04:56 PM rue_mohr: I dont mind sharing my stuff, suppose I should put my name on it...
04:56 PM polprog: you can send stuff in
04:56 PM polprog: ill put you in the file header if i take something
04:56 PM polprog: i should already do so
04:56 PM polprog: i took the struct ;)
04:57 PM rue_mohr: :) np
05:01 PM rue_mohr: d /= 10000000000000000 <-- sometimes, you need to put .0 on the end of a number to make sure the compiler doesn't do stupid integer things with it
05:02 PM polprog: hmm, good point
05:02 PM polprog: i was thinking about replacin that with 10E(something)
05:02 PM rue_mohr: I kinda like it like that, you can read it easier
05:03 PM rue_mohr: m....u....n.....p......f.....
05:04 PM rue_mohr: nice ok, I... want to play
05:04 PM rue_mohr: gtg, have to play with a fire alarm panel!
05:04 PM polprog: have fun
07:40 PM polprog: http://paste.debian.net/1144054/
07:40 PM polprog: 02:39, jesus christ i have to be up tomorrow
07:48 PM zhanx: spent 6 hours printing a part only to realize i didn't export the new gcode and reprinted the bad part
08:26 PM rue_shop1: oops
08:26 PM rue_shop1: been there
08:27 PM rue_shop1: you know, before 3d printers, we used to fix mistakes
08:27 PM rue_shop1: polprog, it is tommorow
08:28 PM polprog: yes
08:28 PM polprog: gnight
08:28 PM rue_shop1: polprog, the start of the scan needed to start earlier
08:28 PM polprog: huh?
08:28 PM rue_shop1: for 6khz, you started at 6.001
08:28 PM rue_shop1: too high
08:28 PM rue_shop1: oh no wait
08:28 PM rue_shop1: read it wrong
08:28 PM polprog: its got some other bugs
08:29 PM rue_shop1: its needs more brute forcing
08:29 PM polprog: hmm
08:29 PM rue_shop1: were you going up and down from your ideal values?
08:30 PM rue_shop1: polprog, you need to wake up today eh?
08:30 PM polprog: i tried to make something that would reduce the error
08:30 PM rue_shop1: to wake up, you need to have slept
08:30 PM polprog: actually not.. its holiday today
08:30 PM rue_shop1: ok
08:31 PM polprog: so what that big table does is that it tries to pic the best C for each R in the e24 range
08:31 PM rue_shop1: how did you walk the tables? did you make a new fn?
08:31 PM polprog: yeah. a couple of
08:31 PM rue_shop1: ok
08:31 PM polprog: theres a fn that picks the best C for a given R and F
08:31 PM rue_shop1: ok
08:32 PM polprog: it has a while loop that attempts to minimize the error difference.. it took me too long to write that lol
08:32 PM rue_shop1: there is no work tommorow, so I'm gonna play
08:32 PM rue_shop1: well, at 3am, its bound to go slow
08:32 PM polprog: i was trying to do that not storing the error difference at all :v
08:32 PM polprog: yeah
08:33 PM polprog: maybe youll find why closest_E12 or closest_E24 always rounds up
08:33 PM polprog: aw crap
08:33 PM polprog: i didnt upload the new code to the server..
08:33 PM rue_shop1: oh?
08:33 PM rue_shop1: did my code have a bug?
08:33 PM rue_shop1: :)
08:33 PM polprog: no, i mean my functions
08:33 PM rue_shop1: oh
08:34 PM polprog: i havent looked at yours yet, have it saved
08:34 PM polprog: allright.. ill go sleep now
08:34 PM polprog: thanks :D
08:35 PM rue_shop1: np
08:35 PM rue_shop1: its cool to have two people on the same problem, compare notes when we get there
08:40 PM rue_mohr: you got me thinking about a more generic approach to the solver too
08:53 PM Tom_L: rue_mohr, how many times does it take to get a kernel right? :)
08:55 PM rue_mohr: well
08:55 PM rue_mohr: I'm on "g" of my new one
08:55 PM rue_mohr: how many is that?
08:55 PM rue_mohr: 7
08:55 PM rue_mohr: and I'm close
08:56 PM rue_mohr: between 8 and 15?
08:56 PM Tom_L: sounds about right
09:12 PM rue_mohr: https://www.thingiverse.com/thing:4063689
09:12 PM rue_mohr: oooooo
09:12 PM Tom_L: make one from angle iron
09:13 PM Tom_L: plastic will crack
09:13 PM Tom_L: alot of force on that thin V
09:14 PM rue_mohr: PLA is REALLY TOUGH
09:14 PM rue_mohr: I didn't tell you about the shredder fail?
09:18 PM Tom_L: no
09:22 PM rue_mohr: did up a big mean shredder, and when I put in a little peice of PLA and completely jammed up on it
10:41 PM rue_mohr: https://www.youtube.com/watch?v=WuY2-OrT9ig&feature=emb_rel_end
10:44 PM Tom_L: what i product had those vent holes that were offset?
10:44 PM Tom_L: aluminum case
10:44 PM Tom_L: like rows of offset holes and machined the same from the back but offset
10:45 PM Tom_L: was pretty cool looking but i can't find it
10:45 PM Tom_L: hackaday featured it once i think
10:45 PM rue_mohr: huh, those look like resin prints
10:45 PM rue_mohr: $$$$$
10:45 PM Tom_L: can't figure out what to search for
10:46 PM Tom_L: https://www.youtube.com/watch?v=518AWHLn3ZU
10:46 PM Tom_L: that pattern
10:48 PM rue_mohr: I dont know
10:59 PM rue_mohr: heh, most of his bending stuff doesn't need the 3d printer anyhow
10:59 PM rue_mohr: if I had the big cnc or a plama cutter, I could be making metal project boxes
10:59 PM zhanx: ok body is done
10:59 PM zhanx: legs are done and servo's in both
11:00 PM zhanx: dont like the hip joint so going to redo the design
11:00 PM rue_mohr: zhanx, ^^ you see the impressive 3d printed bendology?
11:00 PM zhanx: your link?
11:00 PM rue_mohr: yea
11:01 PM zhanx: nice
11:01 PM rue_mohr: not me, I posted
11:01 PM rue_mohr: https://www.youtube.com/watch?v=8cNeAOpR-Ws
11:01 PM rue_mohr: they require lots of skipping
11:01 PM zhanx: k
11:01 PM rue_mohr: but the posabilities are quite inspiring
11:02 PM zhanx: from the quick skim yes
11:03 PM rue_mohr: oh, his plama does a touch-off-per-start
11:04 PM rue_mohr: https://www.youtube.com/watch?v=M-fTY5L5uu0 <--- oooo cmon....
11:04 PM rue_mohr: srsly????
11:05 PM rue_mohr: I mean, a few cycles sure, but !!!!!
11:07 PM zhanx: I did forgot something in the design of this bot rue_mohr
11:08 PM zhanx: wires on the servo's and tolerance for them but at you can take them apart to move the wires
11:09 PM rue_mohr: so, if I have my head screwed on straight, i can form project boxes with printed dies
11:10 PM zhanx: you can
11:43 PM zhanx: ok printing