#hazzy Logs
Mar 04 2020
#hazzy Calendar
02:06 AM CaptHindsight: I'm displaying SVG slices in a window now controlled by Z positions from a HAL component that is compatible with SVG files sliced using Slic3r
02:06 AM CaptHindsight: now to get it working with qtpyvcp
05:24 AM hazzy-m: CaptHindsight: very nice!
05:26 AM * hazzy-m has a 6:30 conference call :(
05:28 AM jthornton: I hate when that happens
07:15 AM dremeier[m]1: @pcw_home you are here?
07:16 AM dremeier[m]1: I´ve a problem with the 7i87, the analoge in from 4 to 7 does not working
07:17 AM * dremeier[m]1 uploaded an image: image.png (196KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/QJdIAmKKNwBdqKLMNdyVqvjZ >
07:17 AM dremeier[m]1: the sserial mode is 0
07:18 AM dremeier[m]1: on analogin7 is a poti, the value changed just once at startup of lcnc.
07:19 AM dremeier[m]1: It looks like the sserial is in Mode1.
07:20 AM dremeier[m]1: If I try to change the mode from 0 to 1 or 2, lcnc does not start
09:30 AM Lcvette: morning everyone!
09:30 AM Lcvette: \o/
09:30 AM Lcvette: its a beautiful day
09:31 AM pcw_home: dremeier: what LinuxCNC version?
09:31 AM pcw_home: (mode change behavior is expected, the 7I87 only supports mode 0)
09:33 AM Lcvette: o/ hi pcw_home
09:33 AM Lcvette: thanks for the help last night
09:38 AM pcw_home: Welcome
10:05 AM TurBoss: good morninng
10:07 AM Lcvette: Morning turboss \o/
10:08 AM Lcvette: I was reading back through the notes and saw JT mentioned making a servo component
10:08 AM Lcvette: Would that be easier?
10:11 AM TurBoss: a bit off
10:11 AM TurBoss: but I think I'm close ...
10:12 AM TurBoss: I'll paste here what i have when I go dev
10:13 AM Lcvette: Ok
10:49 AM dremeier[m]1: pcw_home: I´m using Lcnc 2.9
10:52 AM pcw_home: do you get the same numbers with halcmd show pin?
10:54 AM dremeier[m]1: will try, moment
10:56 AM dremeier[m]1: what will be the corret syntax, I forget it
10:59 AM pcw_home: halcmd show pin | grep *7i87* is one way
11:00 AM dremeier[m]1: Linuxcnc must be started?
11:01 AM pcw_home: yes
11:02 AM * dremeier[m]1 uploaded an image: image.png (305KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/PszhrxqmhZJvUkTHgIbsSHax >
11:02 AM dremeier[m]1: lookes better
11:05 AM Lcvette: \o/ hurray!
11:06 AM dremeier[m]1: but does not solve the problem
11:07 AM Lcvette: /o\
11:07 AM * dremeier[m]1 uploaded an image: image.png (736KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/JfFsAmAqubnpWpUVvTYSLSxh >
11:07 AM Lcvette: whats the problem
11:08 AM dremeier[m]1: no values from analogin4 to 7
11:08 AM Lcvette: oh
11:08 AM Lcvette: beyond my knowledge base
11:08 AM Lcvette: hal is voodoo
11:08 AM dremeier[m]1: in halcmd it shows the values but not in "hal Show" ^^
11:14 AM Lcvette: turboss did you see this?:
11:14 AM Lcvette: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?RC_Servo_Test
11:15 AM Lcvette: and this: https://forum.linuxcnc.org/24-hal-components/26993-controlling-a-futaba-rc-servo-with-pwmgen-type-0
11:18 AM dremeier[m]1: no, that is wunderfull, will have a look
11:18 AM * dremeier[m]1 uploaded an image: image.png (362KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/ywJYmkrEtOVJDAHPFEgrRqxr >
11:18 AM dremeier[m]1: I guess it is working anyway^^ even in hal show are wrong numbers /o\
11:19 AM TurBoss: Lcvette: good
11:24 AM Lcvette: \o/
11:24 AM Lcvette: is good?
11:27 AM pcw_home: dremeier: looks like a GUI bug
11:29 AM pcw_home: Ill try in axis later today
11:30 AM dremeier[m]1: oh, so turboss or hazzy have some extra work the GUI 🤣 or is this related on Linuxcnc
11:31 AM dremeier[m]1: * oh, so turboss or hazzy have some extra work at the GUI 🤣 or is this related on Linuxcnc
11:35 AM pcw_home: looks like the numbers are only wrong in halshow, does halmeter work?
11:37 AM dremeier[m]1: yes, Hal Meter shows correct values
11:48 AM TurBoss: dremeier: looks like a bug
11:53 AM dremeier[m]1: is the bug related to your work or is this Linuxcnc?
11:58 AM TurBoss: halshow, hal metter halscope they are external tools
01:18 PM * TurBoss uploaded a video: Peek 04-03-2020 20-10.mp4 (317KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/SDegzxnmyKtWFQEbYJfGPTTM >
01:19 PM TurBoss: Lcvette: this works with software pwmgen
01:19 PM TurBoss: now I need to implement on hardware pwmgen
01:23 PM Lcvette: Hurray!!!!
01:23 PM Lcvette: \o/
01:23 PM Lcvette: turboss that is great news!!
01:23 PM Lcvette: :D
01:34 PM JackDaviesDesign: sup guys! I've just finished my probe prototype but cant seam to get the probing routines to work in probe basic, got any tips?
01:37 PM Lcvette: what are you doing?
01:37 PM Lcvette: to try and probe currently?
01:37 PM Lcvette: have you set all the probing parameters?
01:37 PM Lcvette: set the correct probe tool number and entered it in the parameters field?
01:38 PM Lcvette: have you copied the subroutine folder into your config file and added the correct path for it in your ini file?
01:38 PM JackDaviesDesign: yes set all params, loaded tool 99 as probe
01:39 PM JackDaviesDesign: hmmm
01:39 PM JackDaviesDesign: that will be the reason haha
01:39 PM Lcvette: how did you install probe basic?
01:40 PM JackDaviesDesign: where can i find the sub folder?
01:40 PM Lcvette: depends on how you installed probe basic
01:40 PM Lcvette: did you use the dev install or quickstart?
01:40 PM JackDaviesDesign: using pip
01:40 PM JackDaviesDesign: quickstart
01:40 PM Lcvette: are you doing dev stuff?
01:40 PM Lcvette: from probe basic or qtpyvcp?
01:41 PM JackDaviesDesign: nope not atm
01:41 PM TurBoss: sadly pip doesnt copy the subroutines
01:41 PM Lcvette: https://kcjengr.github.io/probe_basic/quick_start.html
01:41 PM TurBoss: to linuxcnc/configs
01:41 PM Lcvette: if you are not doing dev then you should iuse this for installation
01:41 PM JackDaviesDesign: but would like to soon once i get some more time ahha
01:42 PM JackDaviesDesign: using the installer?
01:42 PM Lcvette: if you use the quickstart installer then the files are automatically installed in the linuxcnc/probe_basic config file
01:43 PM Lcvette: you would just copy over the subroutine folder to your machine config folder and then update the path to subroutines in your ini file
01:43 PM JackDaviesDesign: okay sweet!
01:43 PM JackDaviesDesign: epic i will try it out tomorrow!
01:43 PM Lcvette: but i would use the installer version of probe basic as it is the safer version not experimental
01:44 PM JackDaviesDesign: okay sweet
01:45 PM JackDaviesDesign: thanks a lot guys!
01:46 PM TurBoss: thank you and welcome
01:50 PM Lcvette: no worries!
01:50 PM Lcvette: TurBoss: that is great progress yes!?
01:50 PM Lcvette: on the servo thing?
01:51 PM TurBoss: today machine will be busy
01:51 PM Lcvette: feels like a break through
01:51 PM Lcvette: thats a good thing!
01:51 PM Lcvette: for bsiness!
01:51 PM TurBoss: its late to make more noise and need to keep zeroes
01:51 PM Lcvette: * for business!
01:52 PM Lcvette: i can test if you have something to test
01:52 PM TurBoss: than can work
01:52 PM Lcvette: got everything sorted out last night
01:52 PM Lcvette: new firmware is installed and everything is up and running
01:52 PM TurBoss: ok let me try to por that example to mesa pwmgens
01:52 PM TurBoss: *port
01:52 PM Lcvette: \o/
01:58 PM TurBoss: Lcvette:
01:58 PM TurBoss: > TurBoss: that is great progress yes!?
01:58 PM TurBoss: yes
01:58 PM TurBoss: * Lcvette:
01:58 PM TurBoss: > TurBoss: that is great progress yes!?
01:58 PM TurBoss: yes
01:59 PM Lcvette: \o/
01:59 PM TurBoss: xD
01:59 PM Lcvette: hurray!
01:59 PM Lcvette: :D
01:59 PM Lcvette: makes it easier to dow ith hardware now?
01:59 PM TurBoss: no
01:59 PM TurBoss: /o\
02:00 PM Lcvette: Oh /o\
02:00 PM TurBoss: x😷
02:00 PM TurBoss: * 😷
02:01 PM Lcvette: Lol
02:01 PM Lcvette: What's that?
02:01 PM TurBoss: for the corona virus
02:01 PM TurBoss: protection
02:01 PM Lcvette: Indeed
02:04 PM Lcvette: i was thinking, if the scale thing is an issue i could use a multiplier in the subroutine
02:04 PM Lcvette: to create the correct numeric output
02:05 PM TurBoss: in teory scale is a multiplier
02:05 PM TurBoss: in mutiplies the input with the gain and then applies an offset
02:05 PM TurBoss: if i did understand
02:05 PM pcw_mesa: Yep
02:05 PM TurBoss: hey pcw
02:06 PM Lcvette: so 0-180 would become the servo scale
02:06 PM Lcvette: pcw_mesa is here \o/
02:06 PM TurBoss: 🥳
02:06 PM Lcvette: 0-180 to 1-2ms
02:06 PM pcw_mesa: The halshow bug also appears when launched from axis
02:06 PM TurBoss: oh
02:07 PM pcw_mesa: you might want 0-->90 (0= straight down?)
02:08 PM Lcvette: i am using a 2:1 reduction gear
02:08 PM Lcvette: so servo 180 sweep moves the coolant blasters only 90
02:08 PM pcw_mesa: Yeah so at some point the range is 0 to 90
02:09 PM * Lcvette uploaded an image: image.png (509KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/rUJKaYcgAxdSUWamWqdRKubd >
02:10 PM Lcvette: 0 would be parallel
02:10 PM * Lcvette uploaded an image: aimed coolant nozzle macros calcs.png (16KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/EZewdZuiuTwjplKPnMjhGrvs >
02:11 PM * Lcvette uploaded an image: aimed coolant nozzle macros calcs.png (10KB) < https://matrix.org/_matrix/media/r0/download/matrix.org/GAZMBLhesFuprPUvwhzrtoRq >
02:28 PM TurBoss: Lcvette: I will try to test the pwm stuff tomorrow the machine will be free
02:28 PM Lcvette: \o/
03:02 PM pcw_mesa: same halshow bug here:
03:02 PM pcw_mesa: http://freeby.mesanet.com/hsbug.png
04:14 PM Lcvette: pcw_mesa: what's the bug from
04:16 PM pcw_mesa: No idea, the "show" pane of halshow is correct, the "watch" pane is broken but I have never seen this bug before it seem specific to the pin or pin name
04:17 PM pcw_mesa: I've used halshow for maybe 12 years and have never seen this kind of bug
04:18 PM pcw_mesa: very weird
04:22 PM Lcvette: Oh boy, that is a stumper then
04:22 PM Lcvette: Is it only on the latest master version 2.9? Or 2.8 and 2.7 also,?
04:35 PM pcw_mesa: Ill have to try2.7, my guess is that its an old bug
04:38 PM dremeier[m]1: good to know that is was not my fault :-)
04:38 PM dremeier[m]1: Could confuse other as well
04:40 PM pcw_mesa: its quite weird since the "show" pane of the same program works
04:45 PM Lcvette: i never was able to see anything change in show, only watch
04:46 PM Lcvette: but i thinkk someone said thats because show did not update and watch did or something like that
04:47 PM dremeier[m]1: lol ^^ the show must go on 🤣
04:47 PM pcw_mesa: show is not updated, it just samples the signal when you select it
05:10 PM Lcvette: ah
05:11 PM Lcvette: pcw_mesa did you see turboss got the pwmgen in software working?
05:13 PM pcw_mesa: you mean moving the servo?
05:13 PM Lcvette: no
05:13 PM Lcvette: :(
05:13 PM Lcvette: https://matrix.org/_matrix/media/r0/download/matrix.org/SDegzxnmyKtWFQEbYJfGPTTM
05:16 PM Lcvette: he got it working in software but says he needs to port it to hardware now
05:16 PM Lcvette: i don't know what that means though
05:16 PM Lcvette: lol
05:17 PM Lcvette: i need a hal for dummies book
05:17 PM Lcvette: all the existing material on it is too far away from my entry level position to get a grasp on it
05:18 PM Lcvette: everytime i think im beginning to see a connection it slips away
05:19 PM pcw_mesa: hal is mostly just wiring, connecting a signal source to a destination
05:20 PM pcw_mesa: (or multiple destinations)
05:21 PM Lcvette: i wish there were a diagram
05:21 PM Lcvette: showing a work flow for constructing a hal file with layman terms explaining what is doing what and why its important
05:22 PM Lcvette: but i have always been more of a visual learner
05:23 PM Lcvette: so if what is written on something doesn't allow me to formulated a visual mental example of it i tend to never grasp it
05:24 PM Lcvette: i have read the hal basics section probably more than a dozen times to try and force an understanding but it just never clicks
05:24 PM Lcvette: its like a crossword puzzle missing just enough letters to make sense of what the word would be
05:24 PM Lcvette: lol
05:25 PM pcw_mesa: I think its easy for electronics types since its mainly just adding components (loadrt somethng) and wiring the component pins to other component pins (net wirename somepin someotherpin)
05:25 PM Lcvette: yes m,uch easier for electronic types
05:25 PM Lcvette: its a small hop rom one understanding to another
05:25 PM Lcvette: * its a small hop from one understanding to another
05:25 PM TurBoss: e I'm a bit lost but I don't need the hardware pwmgen ?
05:25 PM TurBoss: http://linuxcnc.org/docs/devel/html/man/man9/hostmot2.9.html#pwmgen
05:26 PM pcw_mesa: if you think of "net" meaning "connect" does that help?
05:27 PM Lcvette: it does so9mewhat
05:28 PM Lcvette: from editing my hal file the net part is probably the one thing that seems to make some sense
05:28 PM Lcvette: sorry turboss
05:28 PM Lcvette: go ahead!
05:28 PM Lcvette: :D
05:28 PM pcw_mesa: TurBoss you can do it with a software pwmgen
05:28 PM TurBoss: ohhhh
05:29 PM TurBoss: ok
05:29 PM pcw_mesa: but the position resolution will be limited to the ratio of the servo thread time to the base thread time
05:31 PM pcw_mesa: say you have a 25 KHz base thread (so PWM width steps are 40 usec) you would have only 25 discrete positions
05:31 PM pcw_mesa: that may be OK for coolant
05:34 PM pcw_mesa: especially with some sweep/dither
05:35 PM Lcvette: how do we know the base thread?
05:38 PM pcw_mesa: its something you set on parallel port based systems
05:39 PM pcw_mesa: you typically run it as fast as you can based on the host computers latency
05:41 PM pcw_mesa: 25 KHz or so is typical for base threads, some PCs can run faster maybe up to 50 KHz at most
05:42 PM Lcvette: Ok
05:42 PM Lcvette: So the pc will determine the resolution of the servo
05:47 PM pcw_mesa: yes, with software PWM ist determined by the latency of the PC
05:47 PM pcw_mesa: With hardware PWM you have a basic 12 bit resolution (1 part in 4096) but you only use 5% of the PWM range at 50 Hz (1 ms to 2 ms = 5% to 10% duty cycle) so resolution is 1 part in 200
05:51 PM pcw_mesa: your servo can run up to 333 Hz (3 ms period) so could have 1 part in ~1365 resolution with hardware PWM (probably much better than the accuracy/repeatability of the servo)
05:54 PM Lcvette: Where are the 4096 snd 1365 numbers coming from?
05:55 PM Lcvette: I'm very familiar with duty cycle for fuel injection, and calculating flow rate vs pressure and injector pulse width and duty cycle
05:56 PM Lcvette: I'm sure runners are sons upstairs in there's that could help me understand this
05:56 PM Lcvette: * I'm sure there are some parallels in there's that could help me understand this
05:58 PM Lcvette: So 100% dc would be 20ms
05:59 PM Lcvette: Is that a standard chunk of time for electronics?
06:07 PM Lcvette: i see
06:07 PM Lcvette: 50hz = 20ms
06:07 PM Lcvette: 20ms = 100% duty cycle or no wave form
06:07 PM CaptHindsight: math winsn again \0/
06:08 PM Lcvette: \o/
06:08 PM Lcvette: math i get fine, love math, trig, geometry
06:08 PM Lcvette: electronics is a new spin for me
06:08 PM Lcvette: just need to aqcuaint myself to its temrinology
06:09 PM Lcvette: * just need to aqcuaint myself to its terminology
06:09 PM Lcvette: like bits
06:09 PM CaptHindsight: slic3r:z="5e-08" was generated for Z height in Slic3r
06:09 PM Lcvette: never dealt with bits before
06:09 PM CaptHindsight: should be 0 and why did they use notation?
06:10 PM CaptHindsight: well software people like to use different types of numbers all the time
06:11 PM CaptHindsight: often in the same app for more confusion
06:11 PM Lcvette: 8 bits in a byte
06:11 PM CaptHindsight: i prefer 1's and 0's myself
06:13 PM Lcvette: when i was in school we still used the dewy decimal system
06:13 PM Lcvette: * when i was in school we still used the dewey decimal system
06:13 PM Lcvette: there were not computers
06:13 PM jthornton: when I was in school we use our fingers and toes
06:13 PM Lcvette: in the whole of the schoole
06:13 PM Lcvette: * in the whole of the school
06:13 PM CaptHindsight: did they stop using it?
06:14 PM Lcvette: my first computer was a commodore 64
06:14 PM CaptHindsight: jthornton: did you get a chance to meet Gutenberg with his crazy ink on paper machine? :p
06:14 PM Lcvette: still have it
06:15 PM pcw_mesa: The hm2 hardware PWM generator has 12 bits so 4096 discrete PWM settings from 0 to 100%
06:15 PM jthornton: no, he was a bit before my time, however I was born in a Territory of the USA and not a state
06:15 PM Lcvette: ok
06:15 PM Lcvette: now things are making sense
06:16 PM pcw_mesa: so you only use 5% of that range (from 1 to 2 ms of 20 ms total) you only have 1/200 resolution
06:17 PM jthornton: Lcvette, 12 bits is 2 to the power of 12 or 4096
06:17 PM Lcvette: reading that now
06:17 PM Lcvette: or 3 nibbles
06:17 PM Lcvette: or a tribble
06:18 PM Lcvette: 10000 octal
06:18 PM Lcvette: 1000 hexidecimal
06:18 PM jthornton: or if your my age a dribble
06:18 PM Lcvette: lol
06:19 PM Lcvette: and with the reduction of 2:1 i bump that resolution up to 1/400
06:20 PM Lcvette: over a jthontons name is everywhere
06:20 PM Lcvette: * over a jthorntons name is everywhere
06:20 PM jthornton: I think my ladies have gone to roost time for a bedtime story and tuck them in
06:20 PM pcw_mesa: also with the hardware PWM if you raise the PWM frequency you get better resolution
06:22 PM Lcvette: so 50hz is a 20ms period
06:22 PM Lcvette: and 333hz is a 3.003ms period
06:23 PM Lcvette: so we are using 33% to 66%
06:23 PM Lcvette: am i getting this right?
06:24 PM pcw_mesa: Yes (and the 33% of the PWM range or 1/3 of 4096)
06:24 PM Lcvette: ok !!
06:24 PM Lcvette: \o/
06:24 PM Lcvette: hurray!!
06:24 PM Lcvette: im begging to understand
06:25 PM Lcvette: * im begining to understand
06:26 PM Lcvette: so 5% of 4096 is 2004.8
06:27 PM Lcvette: 33% of 4096 is 1351
06:27 PM Lcvette: so with a base thread of 333hz you get 1351 discrete positions?
06:27 PM pcw_mesa: 5% is 204.8
06:27 PM Lcvette: yes sorry extra zero in there
06:28 PM Lcvette: typo
06:28 PM pcw_mesa: not base thread, PWM frequency
06:28 PM Lcvette: pwm frequency
06:28 PM Lcvette: so wow 333hz makes a big difference in resolution
06:28 PM Lcvette: not that i think it would be needed
06:29 PM Lcvette: so question
06:29 PM pcw_mesa: not pointing a laser
06:29 PM Lcvette: when you say resolvable positions, what happens if the postioon comand is not an exact resolvable position?
06:29 PM Lcvette: it just goes to the nearest resolvable position?
06:31 PM pcw_mesa: yeah I think the driver rounds the floating point value to the closest PWM value
06:31 PM Lcvette: very cool
06:31 PM Lcvette: see i can now picture all of this and understand it
06:31 PM Lcvette: feels good!
06:31 PM Lcvette: :D
06:31 PM Lcvette: \o/
06:31 PM Lcvette: turboss are you getting all this
06:31 PM Lcvette: im learning electronics fu
06:31 PM Lcvette: waaaaaaa
06:32 PM TurBoss: yes
06:32 PM pcw_mesa: Yeah none of this should be too mysterious
06:32 PM Lcvette: its only mysterious until you understand it
06:32 PM pcw_mesa: same with hal
06:32 PM Lcvette: true
06:32 PM Lcvette: but im a numbers person
06:33 PM Lcvette: lol
06:33 PM Lcvette: i like math
06:33 PM Lcvette: so in hal loadrt is for loading a real time component that exists in the current linuxcnc database
06:33 PM Lcvette: or repository or whatever its called
06:34 PM Lcvette: thats what the loadrt's at the top do
06:34 PM Lcvette: and i read they sometimes must be inn a certain order
06:36 PM pcw_mesa: loadrts somtine depends on previous code being loaded first (like hostmot2 must be loaded before hm2_eth)
06:36 PM jthornton: which vette is the lc?
06:37 PM Lcvette: little chris vette
06:37 PM Lcvette: dad's is bcvette
06:37 PM Lcvette: big chris vette
06:37 PM jthornton: I thought it was a model or something
06:38 PM jthornton: I'm looking at a C5
06:42 PM Lcvette: nice those are great cars
06:43 PM Lcvette: we have C4's
06:43 PM Lcvette: the c6's are really really nice
06:43 PM jthornton: I like the C4 too
06:43 PM Lcvette: c6 z06
06:43 PM Lcvette: but they are still super high in price because of how sought after they are
06:44 PM Lcvette: i would stay away from the C4's they are maintenance nightmares
06:44 PM jthornton: really
06:44 PM Lcvette: the ls1 in the c5 vettes really changed all that
06:44 PM Lcvette: yeah
06:44 PM Lcvette: trust me we have 5 c4 vettes
06:44 PM jthornton: https://www.ebay.com/itm/2004-Chevrolet-Corvette-Commemorative-Edition/223926706394?hash=item34231260da:g:7g0AAOSwADReVCZe
06:45 PM Lcvette: and the geniuses at gm put an electronic distributor directly under the waterpump that leaks
06:45 PM jthornton: I believe youy
06:45 PM jthornton: oh crap
06:46 PM Lcvette: thats a nice low mileage edition
06:46 PM jthornton: yea and it's not that far from me
06:48 PM jthornton: chow time here
06:51 PM Lcvette: https://www.ebay.com/itm/2008-Chevrolet-Corvette-Z06/264650886279?hash=item3d9e6c0c87:g:SIIAAOSwtMVeWckV