#garfield Logs
Apr 02 2020
#garfield Calendar
12:07 AM Tom_L: so at what point can you start editing the code with the editor?
12:13 AM rue_mohr: :)
12:13 AM rue_mohr: I'm past it, but not doing it yet
12:13 AM rue_mohr: I think chapter 5 did save
12:13 AM Tom_L: part of the self replicating process...
12:13 AM Tom_L: learn
12:13 AM rue_mohr: indeed
12:14 AM rue_mohr: turned out the jib had worn right out
12:14 AM rue_mohr: and later a motor coupler let go
12:14 AM Tom_L: mine was as good as it was ever going to be
12:15 AM Tom_L: it's in a box now
12:15 AM rue_mohr: then, when I was making those little brackets, I found out the bit was worn from 3mm to 2.89-ish!
12:16 AM Tom_L: i haven't done anything that had to be very accurate
12:16 AM rue_mohr: 0.11mm makes for 1/8" holes that dont fit 1/8" rod
12:17 AM Tom_L: i was using a crappy dial indicator just to see tonight
12:17 AM Tom_L: i'll get the good one off my lathe and try again one day
12:17 AM rue_mohr: what did it say?
12:18 AM Tom_L: kinda hard to say. i was getting some flex in the holder setup i had
12:18 AM rue_mohr: hmm aluminum bed :/
12:18 AM rue_mohr: cant magnetize to it
12:18 AM Tom_L: i used a mag base on the vise
12:20 AM rue_mohr: ok, I need keycodes when in 'find' move for next and previous
12:20 AM Tom_L: i could give you scan codes
12:20 AM rue_mohr: no I get interpreted jibberish from the kernel
12:20 AM Tom_L: the fpga had a test program for the kbd scan codes
12:21 AM rue_mohr: yea I saw, thats nice
12:21 AM rue_mohr: I want to work on the fpga stuff
12:21 AM rue_mohr: but I'm caught in a horrid panic of have to run, but dont know which way
12:23 AM rue_mohr: so instead im working on this editor
12:28 AM Tom_L: https://www.w3.org/2002/09/tests/keys.html
12:28 AM rue_mohr: ShopMoon, I'd like you to play with this editor
12:28 AM rue_mohr: cool, I wonder how they managed that
12:34 AM rue_mohr: I need to change my supper plan
12:34 AM rue_mohr: I'm really realeaved when I'v finished eating it
12:36 AM rue_mohr: hmm
12:36 AM rue_mohr: there is something odd going on with his key interpreter
12:37 AM Tom_L: yeah
12:37 AM rue_mohr: int editorReadKey() {
12:37 AM rue_mohr: its not a pretty thing tho
12:39 AM Tom_L: ok i'm off to sleep
12:40 AM rue_mohr: gnight, dont let the virus bite
12:47 AM rue_mohr: damnit, this key reader is as airtight as a beer can at a shooting range
01:12 AM rue_mohr: huh, yea, I have a program that also shows codes, the mode I'm using cant do a shift modifier mixed with ctrl keys
02:06 AM polprog: weird. i was sure meta is alt
02:06 AM polprog: firefox must interpret that in a weird way. The deal is that meta and control clear (or set, i dont remember) certain bits
02:09 AM polprog: did you catch my imporved Makefile for stm32?
09:20 AM rue_mohr: how recent?
09:20 AM polprog: well, yesterday night's
09:20 AM rue_mohr: heh no
09:21 AM polprog: anyway the one i pasted up there is probably the old one now
09:21 AM polprog: https://gitlab.com/coniferous/stm32-template/-/blob/master/Makefile
09:21 AM polprog: this is the new one.. printf works !
09:21 AM rue_mohr: I dont like printf
09:22 AM rue_mohr: its really heavy
09:22 AM rue_mohr: here is a question tho
09:22 AM rue_mohr: can it print a floating point value right?
09:22 AM polprog: yeah, but it's handy sometimes. plus this makes all newlib funcs work fine
09:22 AM rue_mohr: yep
09:22 AM polprog: idk
09:22 AM polprog: it may be able to
09:22 AM rue_mohr: give it a try, I had challanges
09:23 AM polprog: since ive seen a newlib fork that claims to use libc-avr io code (incl printf) so this leads me to a conclusion that newlib uses heavier printf implementation
09:23 AM polprog: ill check it for you when i get to it
09:23 AM polprog: printing floats may be useful (despite floats being evil)
09:26 AM rue_mohr: https://www.worldometers.info/coronavirus/coronavirus-cases/ <- ad warning
09:28 AM rue_mohr: waiting for it to hit here
09:28 AM polprog: https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6
09:28 AM rue_mohr: by the numbers it looks like the local area will lose ~1k
09:29 AM polprog: 9k in canada confirmed (arcgis)
09:29 AM rue_mohr: much better sire
09:29 AM polprog: oh, they have per county data for US
09:29 AM rue_mohr: USA leading the way
09:31 AM polprog: always the best!
09:31 AM rue_mohr: I'm sure Trump will make them a world leader on this
09:31 AM rue_mohr: not sure how you save an economy when everyone in the country dies
09:33 AM polprog: the guy at #libopencm3 was surprised when i said that the example project from their github is too big for my liking :P
09:34 AM rue_mohr: heh
09:34 AM polprog: someone is selling "monochrome monitor IBM 3486" for $50 here. the listing has been up for a good year now
09:34 AM rue_mohr: its a bad bad example set
09:34 AM polprog: 3486 is a block terminal
09:34 AM rue_mohr: none of them properly isolate one thing
09:35 AM polprog: im thinking if theres any remote change of getting that junk to work at all without an ibm mainframe.. probably used twinax even
09:35 AM polprog: yeah. the examples are like 3 chained makefiles...
09:35 AM polprog: i mean the template project. the example codes are okay. but the makefiles are too complicated for someone who looks for an example
09:36 AM polprog: hehe http://www.twindata.com/IBM/ibm3486.htm
09:37 AM polprog: it's for an AS/400
09:37 AM rue_mohr: hmm
09:37 AM rue_mohr: something odd about this data
09:38 AM rue_mohr: I suppose its only the people who have been thru a medical system
09:38 AM polprog: yeah
09:39 AM rue_mohr: so if most people in the usa cant afford treatment
09:41 AM polprog: this epidemic will impact the view on healtcare in the US
01:33 PM zhanx: tube frame back packs, yes i have experience
02:19 PM rue: Log
02:19 PM rue: zlog
02:20 PM rue is now known as rue_work
02:24 PM ruetyu: Zlog
02:24 PM ruetyu: wtf
02:25 PM ruetyu: tom
02:26 PM ruetyu: this web vlient sucks now
03:03 PM Tom_L: rue_mohr, wtf was that?
03:04 PM Tom_L: ok i've got my x & y map data for the comp file
03:16 PM Tom_shop: rue_mohr, do you remember how we calculated the ballscrew pulses per rev on this mill?
03:17 PM Tom_shop: i'm wondering if i can fudge slightly on it and get it closer
03:17 PM Tom_shop: the Z axis is on a 3:1 ratio and is off approx 3x more than the other 2 axis which tells me i might be able to fix it with steps
03:20 PM Tom_shop: if i can gain .001 - .002" with the steps i think it would be good
03:24 PM Tom_L: i need to find the ballscrew mm per turn again...
03:26 PM Tom_L: i think it was 5mm
03:31 PM Tom_L: yes
03:58 PM rue_mohr: IU needed the link that polprog posted
03:58 PM rue_mohr: https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6
03:59 PM Tom_L: what client was that?
03:59 PM rue_mohr: webclient
03:59 PM rue_mohr: the new one really doens't work well on a phone
03:59 PM Tom_L: i'm wondering if we can calculate the steps per turn more accurately...
04:00 PM Tom_L: we came up with 10160 but it's off a tiny bit
04:00 PM rue_mohr: whast the screw pitch?
04:00 PM Tom_L: 5mm
04:00 PM Tom_L: and it shows up quite a bit on z since it's a 3:1 pulley
04:00 PM Tom_L: steppers are 200 steps/ turn?
04:01 PM Tom_L: 10 microstep on the drive
04:02 PM rue_mohr: so 5mm pitch -> 3:1 to 1.8 degree steppers, w/ 10 mircosteps
04:02 PM Tom_L: 1016002
04:02 PM Tom_L: might get it a little closer?
04:03 PM rue_mohr: the primer bulbs I got (last night?) from amazon arrived
04:03 PM rue_mohr: wrong size
04:03 PM Tom_L: :/
04:03 PM rue_mohr: cause, yea, cant just look at 'em on a shelf
04:03 PM rue_mohr: $20 *poof*
04:03 PM Tom_L: we figured 5.08 rev per inch before but i think it's actually 5.0800101
04:04 PM Tom_L: i wonder if that would make a .001- .002" difference over a .5" distance
04:05 PM Tom_L: the z is ~ .005" short on a .5" travel
04:05 PM Tom_L: given the reduction that amplifies the error
04:07 PM Tom_L: no way to adapt the bulb?
04:11 PM Tom_L: i thought you were talking about plant bulbs at first...
04:35 PM rue_mohr: no its just too big
04:35 PM rue_mohr: thats a lot of error
04:37 PM rue_mohr: I might tryi again
04:37 PM rue_mohr: I suppose if I were an average jo, I'd believing its shot and paying like $400 for a new one
04:39 PM rue_mohr: oh I'm hearin voices again
04:39 PM rue_mohr: must be falling asleep
04:49 PM Tom_L: mmm i can't figure out the .005" error on the Z
04:50 PM Tom_L: i'll attack it again another day
04:50 PM Tom_L: i did map the X and Y ballscrews every .5" and they're pretty close
04:51 PM Tom_L: close enough
04:52 PM Tom_L: i'm not exactly sure how the map file is supposed to work
04:53 PM Tom_L: whether you start from 0.00 each time thru the axis limit or progress thru the travel .5" at a time
04:55 PM Tom_L: i went in .5" increments recording the error from nominal each time both in the positive and negative directions
08:21 PM rue_mohr: maybe the ballscrew is stretched/compressed a bit
08:21 PM rue_mohr: you came at each measurement from the same dir, right?
09:22 PM Tom_L: yes all negative then all positive. 2 sets
09:23 PM Tom_L: i'm just not sure how they expect it to be measured
09:23 PM Tom_L: from 0 to each number or from 0 to the next to the next
09:23 PM Tom_L: i did the latter
09:51 PM rue_mohr: so, its out a few thou every half inch?
09:52 PM rue_mohr: can calipers pick up on it?
09:52 PM rue_mohr: (over a few inches)
10:03 PM Tom_L: but i don't know if it's collective error
10:04 PM Tom_L: i'll probably test some more