#garfield Logs
May 23 2021
#garfield Calendar
12:59 AM rue_shop1: polprog, from what I can tell I screwed up that simeple power converter schematic
01:06 AM rue_shop1: I have to find out what I did
02:19 AM rue_shop1: ok polprog I have a better feedback arrangement
02:19 AM polprog: what is it?
02:23 AM rue_shop1: bias the base of the transistor at about 0.5V and use a small capacitor (5nf - 47nF) instead of the resistor I used
02:24 AM rue_shop1: I used about 470 ohms to ground, and 10k to my 12V source
02:24 AM rue_shop1: for the bias
02:27 AM rue_mohr: its still a forward converter
02:30 AM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1190685.jpg
02:30 AM rue_mohr: add secondaries with ratios for whatever voltages you want
02:30 AM rue_mohr: just make sure their forward
06:00 AM polprog: that looks quite sensible
06:00 AM polprog: ill give it a try
06:00 AM polprog: maybe configure the 3kV transformer in forward mode :v
06:56 AM polprog: i got myself a nice bench DMM for bday
06:56 AM polprog: and a 2 channel power supply unit, so i can have a symmetric supply
06:56 AM polprog: they should be here next week (psu) and in two weeks (dmm, from belgium)
06:58 AM polprog: https://www.ebay.ca/itm/Solartron-Schlumberger-7150-6-1-2-DMM-tested-/143989122025?_trksid=p2047675.m43663.l10137&nordt=true&rt=nc&orig_cvip=true
07:01 AM polprog: damn, the measurement unit is fully floating
07:01 AM polprog: http://bee.mif.pg.gda.pl/ciasteczkowypotwor/Solartron/Solartron_7150_Maintenance_Manual.pdf
07:01 AM polprog: ciasteczkowypotwor, ofc
07:01 AM polprog: "cookie monster"
07:02 AM polprog: you can throw that directory URL in your bookmarks, it has loads of manuals for test equipment
07:15 AM spline: I used to do that w/manuals and books
07:16 AM spline: make a huge / and try to keep it sorted
07:31 AM polprog: rue and me have a small php script that searches for PDFs.. i have a local archive on the workshop PC
07:31 AM polprog: that i should back up somewhere, because they are all stuff downloaded from the net, but that doesnt mean they are forever there
07:38 AM spline: you always thought, thanks to the internet, that you'd never not be able to find a manual
07:38 AM spline: and the worst thing, now a days, is when you get a piece of equipment and the company who originally made it was bought out in, like, the 80s
07:38 AM spline: then they were bought out.. repeat 10 times
07:38 AM spline: I can recall, MAYBE ONCE, getting lucky there
07:39 AM spline: actually it was with the lathe we got at work. 3-4 transfers of ownership but the guy is like nope we thankfully have all this stuff on a specific server and even back up to a tape
07:39 AM spline: but, other times, one transfer of ownership and its like, "yeah we don't support X [click]"
07:40 AM spline: im surprised that archive.org or maybe the LOC hasn't taken up an effort to do some sort of archiving? there is just so much stuff and how would you decide what to archive and not
07:41 AM spline: you normally have to go to the specific group/area, like I have some older (80s?) Acoustic Research AR28s
07:41 AM spline: 9midsize speakers but just have such a nice sound and can take 120-150W)
07:41 AM spline: the audiophile folks archive _everything_
07:44 AM polprog: i think archive.org does scan and archive things
07:45 AM polprog: they have a whole warehouse of stuff to be scanned
07:45 AM polprog: But i never saw manuals there, they mostly do books/multimedia/software
07:45 AM polprog: elektrotanya is pretty good for hi-fi manuals, I got lucky with my tape deck and didnt do much repairs afterwards
09:01 AM spline: one thing I've never found is a manual for my Marantz SR-585 I think it is
09:01 AM spline: they have the schematic but not manual
09:01 AM spline: its close enough to another one of the models around then
10:31 AM Tom_L: one chore done
11:56 AM rue_bed2: spline, wrong, people are trying to monotize pdf downloads, lots of pdfs are slowly shuffling behind paywalls
11:57 AM rue_bed2: and archive.org is fighting dcma
11:57 AM rue_bed2: so
12:14 PM rue_mohr: who is virgin galactic and why does it look like they have a bunch of work done by spacex
12:24 PM Tom_L: Virgin Galactic is an American spaceflight company within the Virgin Group. It is developing commercial spacecraft and aims to provide suborbital spaceflights to space tourists. Virgin Galactic's suborbital spacecraft are air launched from beneath a carrier airplane known as White Knight Two.
12:25 PM rue_mohr: how long they been working on that?
12:25 PM Tom_L: he's just in it for the money
12:25 PM Tom_L: https://www.cnn.com/2021/05/22/tech/virgin-galactic-spaceflight/index.html
12:26 PM Tom_L: Founded: 2004
12:27 PM Tom_L: should be battery powered with a recharge station on the moon
01:13 PM rue_mohr: Sp P1.2 is 1 ?
01:14 PM rue_mohr: P1.7 is Kp
01:15 PM rue_mohr: P1.8 Ki
01:15 PM Tom_L: 100
01:15 PM Tom_L: ki
01:15 PM rue_mohr: I think you want to max out P1.9 and P1.10
01:15 PM Tom_L: 300 kp
01:16 PM Tom_L: no, 7-9
01:16 PM Tom_L: i don't care about decel
01:16 PM rue_mohr: P1.11 is your top speed (what is that?)
01:16 PM Tom_L: 12000
01:16 PM rue_mohr: P1.12 is the zero offset (whats your analog input range?)
01:17 PM Tom_L: 0-5v
01:17 PM rue_mohr: do you have 2.5V as stop?
01:17 PM Tom_L: tell me all this when i get there
01:17 PM Tom_L: no
01:17 PM Tom_L: 0-5v is 0-5000ish rpm
01:17 PM rue_mohr: how does it do reverse?
01:20 PM rue_mohr: they didn't describe P1.1 correctly
01:22 PM rue_mohr: does the motor itself actaully go to 12k?
01:22 PM Tom_shop: no
01:22 PM Tom_shop: around 5k iirc
01:23 PM rue_mohr: you got the settings you have now written down?
01:24 PM Tom_shop: i have the defaults, what they came set at and what i've changed all written down
01:24 PM rue_mohr: this has serial control too eh?
01:25 PM rue_mohr: ok, well me what you have for
01:25 PM rue_mohr: 1.2, 1.7, 1.8, 1.9, 1.10, 1.11, 1.12
01:26 PM rue_mohr: the translation on this document is bad
01:26 PM rue_mohr: like, P1.6 is serial bus address, WTF!
01:28 PM Tom_shop: 2-1,7-300, 8-100. 9-20000, 10-6000. 11-12000,12-0
01:28 PM Tom_shop: 6 well yeah
01:28 PM Tom_shop: it's got modbus but no docs for it
01:29 PM rue_mohr: so
01:29 PM rue_mohr: if your control is 0-5V for speed, how do you do reverse
01:30 PM Tom_shop: a separate rev signal
01:30 PM Tom_shop: it stays 0-5000 0-5v for speed
01:30 PM Tom_shop: and i run the analog out thru an absolute function
01:31 PM rue_mohr: but the gain looks like its out
01:31 PM Tom_shop: this is how it came to me set up
01:31 PM rue_mohr: if you ask it to do 5k, the analog line is probably at 2v
01:31 PM Tom_shop: they said you could use pwm, 0-10 or 0-5v with no changes
01:32 PM Tom_shop: uh huh
01:32 PM rue_mohr: instead of the whole 5V like it should be
01:32 PM Tom_shop: ^^^ that's why
01:32 PM rue_mohr: your using pwm or actual analog?
01:33 PM Tom_shop: analog ish
01:33 PM rue_mohr: ?
01:33 PM rue_mohr: pwm then?
01:33 PM Tom_shop: it's pwm to an analog chip on the board
01:33 PM Tom_shop: but analog to the vfd
01:33 PM rue_mohr: ok so its analog into the vfd then, ok
01:33 PM Tom_shop: and that also has a scale associated with it
01:34 PM rue_mohr: hey, can you hook a 0-5V pot to the CV?
01:34 PM Tom_shop: setp hm2_[HOSTMOT2](BOARD).0.pwmgen.00.scale 6000
01:34 PM Tom_shop: that's for the analog
01:34 PM rue_mohr: ok
01:34 PM rue_mohr: hey, can you hook a 0-5V pot to the CV?
01:34 PM Tom_shop: cv?
01:35 PM Tom_shop: i did when i first got it
01:35 PM rue_mohr: and P1.0 is set to 0, eh?
01:35 PM rue_mohr: cv is the control voltage
01:35 PM Tom_shop: yes
01:35 PM Tom_shop: i did when i first got it
01:35 PM rue_mohr: ok
01:35 PM Tom_shop: i'm not goin back there though
01:36 PM rue_mohr: and with the encoder hooked up, you can monitor the speed in linxucnc yes?
01:36 PM Tom_shop: yup
01:36 PM rue_mohr: will it chart it for ya?
01:36 PM Tom_shop: brb gotta fix something inside
01:36 PM Tom_shop: i doubt it
01:36 PM rue_mohr: darn
01:37 PM rue_mohr: how do you know if the speed is overshooting?
01:38 PM Tom_shop: i can hear the motor go past it then settle on the right speed no matter what speed as it is right now
01:39 PM rue_mohr: hmm
01:39 PM Tom_shop: under load it bogs down too much before it recovers as you can hear on the .250 drill
01:39 PM rue_mohr: so, I'd like to know this...
01:39 PM Tom_shop: it's useable as is, i'm just trying to make it better
01:39 PM rue_mohr: with the control votlage set to 2.5V, what speed does the motor run
01:40 PM Tom_shop: i've no idea without checking it
01:40 PM rue_mohr: this is to get an idea if the gain is right
01:40 PM rue_mohr: yea
01:40 PM Tom_shop: i need to make some edits before i can answer much else
01:40 PM rue_mohr: np I need a shower and breakfast
01:41 PM Tom_shop: 2 line edit
01:42 PM rue_mohr: at 2.5V see if the motor speed is 6000 or 2500
01:43 PM Tom_shop: actually 1 line :)
01:43 PM Tom_shop: i gotta get a meter
01:49 PM Tom_shop: ok wired up i think
01:51 PM Tom_shop: 2.52v commanded 2500rpm actual 2500rpm
01:57 PM Tom_shop: ok i'm pretty sure i got the analog PID out of the code now
01:58 PM Tom_shop: .811v commanded 500rpm actual 290rpm
01:59 PM Tom_shop: 4.03v 2500 commanded 4400 at the spindle
02:00 PM Tom_shop: ok
02:00 PM Tom_shop: 2.501v 1550rpm commanded 2450 actual rpm
02:01 PM Tom_shop: i'm pretty sure the parameters are off on the VFD for the input voltage and max rpm
02:04 PM Tom_shop: 6.4 actual V at max rpm and actual rpm is 5450 rpm
02:04 PM Tom_shop: so we'll say max rpm is 5000
02:04 PM Tom_shop: those are your numbers sir
02:05 PM Tom_shop: so if we get that closer i may not need the pid on the analog
02:06 PM * Tom_shop thinks rue_mohr is taking a very long shower
02:09 PM rue_mohr: huh
02:10 PM rue_mohr: so, 5000/(5/2.5) = 2500
02:11 PM rue_mohr: 5000/(5/.811) = 811
02:12 PM rue_mohr: oh
02:12 PM rue_mohr: wait
02:12 PM rue_mohr: ok I need this a bit clearer here
02:12 PM rue_mohr: when the control voltage was 2.5 the spindle ran at:
02:12 PM rue_mohr: 2450?
02:13 PM rue_mohr: so 2.5V causes 2450 rpm
02:13 PM rue_mohr: ?
02:14 PM rue_mohr: k/(5/2.5) = 2450
02:14 PM rue_mohr: k/2 = 2450
02:15 PM rue_mohr: k = ~5000
02:15 PM rue_mohr: strange
02:15 PM rue_mohr: so the desciption on param P1.11 is completely wrong
02:19 PM Tom_shop: 2.501v 1550rpm commanded 2450 actual rpm
02:19 PM Tom_shop: the 1550 was on the pc
02:19 PM Tom_shop: not the vfd
02:19 PM rue_mohr: dont want to know commanded, means nothing
02:19 PM Tom_shop: 2450 was the vfd rpm
02:19 PM rue_mohr: yup
02:19 PM Tom_shop: just saying
02:19 PM rue_mohr: 2.5V -> 2450rpm
02:19 PM Tom_shop: so the picture is xtal clear
02:20 PM Tom_shop: yes
02:20 PM rue_mohr: so there is a translation out
02:20 PM rue_mohr: which I expected
02:20 PM rue_mohr: thinking
02:21 PM rue_mohr: ok, lets say its really a 0-10V input
02:21 PM Tom_shop: what i meant was, i have to enter 1550rpm as the spindle speed to get 2.500v out
02:21 PM rue_mohr: actual hardware
02:21 PM Tom_shop: i think that's what it is
02:21 PM rue_mohr: then 5V will only be half the P1.11 setting
02:21 PM Tom_shop: and why maybe they set max rpm at 12000
02:21 PM rue_mohr: which would be 6000 rpm
02:21 PM Tom_shop: exactly
02:21 PM rue_mohr: and 2.5V should come out to 3000 rpm
02:22 PM rue_mohr: 3000 and 2450 are pretty far apart
02:22 PM Tom_shop: so i ran it at 4000 cause i knew that would give me max voltage and that was 6.xxv
02:22 PM rue_mohr: kp default was 1520
02:22 PM rue_mohr: which is pretty high
02:23 PM Tom_shop: they set that at 300 to me
02:23 PM rue_mohr: lets try this, set P1.8 to 0 (no integral) and Kp to 1520
02:23 PM rue_mohr: see what 2.5V actual speed is then
02:23 PM Tom_shop: you sure?
02:23 PM rue_mohr: should be 3k, if it oscillates kill it
02:24 PM Tom_shop: gimme a sec
02:24 PM rue_mohr: I think the Kp wasn't high enough for that test
02:27 PM Tom_shop: ok 1520 & 0
02:27 PM Tom_shop: 2.500v 2297 rpm
02:28 PM rue_mohr: ok ~2300
02:28 PM Tom_shop: and it didn't overshoot
02:28 PM rue_mohr: its... lower
02:28 PM Tom_shop: so the max rpm needs adjusted
02:29 PM rue_mohr: something seems wrong
02:29 PM Tom_shop: KP shouldn't be maxed out
02:29 PM rue_mohr: maxed is 99999
02:29 PM rue_mohr: 1520 is the default value
02:29 PM Tom_shop: oh ok
02:29 PM Tom_shop: that was factory default
02:29 PM rue_mohr: yea
02:29 PM Tom_shop: i like what it does there better
02:30 PM rue_mohr: shall we try dialing kp up?
02:30 PM Tom_shop: add in some I?
02:30 PM rue_mohr: no stay away from the I for now
02:30 PM Tom_shop: i'd like to try the factory defaults once just to see what they do
02:30 PM rue_mohr: its underdamped tho eh?
02:30 PM Tom_shop: i can undo it
02:30 PM rue_mohr: ? c
02:30 PM rue_mohr: cant undo?
02:30 PM rue_mohr: what?
02:31 PM Tom_shop: lemme try KI 320 once
02:31 PM rue_mohr: sure
02:31 PM Tom_shop: i gotta power cycle it each time
02:32 PM Tom_shop: it's got some damn big caps cause it just won't die right away
02:32 PM rue_mohr: ok
02:32 PM rue_mohr: dont get into a hurry and have an accident
02:33 PM Tom_shop: i'm double checking the settings after power cycle
02:34 PM Tom_shop: 2.500v 2450 rpm
02:34 PM rue_mohr: wtf
02:34 PM Tom_shop: i'm gonna leave KP at 1520
02:34 PM Tom_shop: i like that
02:35 PM rue_mohr: what is the base speed!?
02:35 PM rue_mohr: it seems to be about 10000
02:35 PM Tom_shop: what do you mean wtf? there was no P overshoot for I to output
02:35 PM rue_mohr: but the base speed isn't clear
02:35 PM rue_mohr: according to the docs its 12000
02:35 PM Tom_shop: 1550 commanded to get 2450 2.500v
02:35 PM rue_mohr: which is a long way from 9800 it seems to be
02:36 PM rue_mohr: yea, so your linuxcnc is also out
02:36 PM Tom_shop: you sure?
02:36 PM rue_mohr: there is a scale factor in there thats wrong
02:36 PM Tom_shop: i can change it
02:36 PM Tom_shop: i posted that line to you
02:36 PM rue_mohr: 2500 should result in 2.5V
02:36 PM rue_mohr: so that it results in 2450 at the fvd
02:36 PM rue_mohr: THEN you can close the loop
02:37 PM rue_mohr: but things are still out
02:37 PM Tom_shop: setp hm2_[HOSTMOT2](BOARD).0.pwmgen.00.scale 6000
02:37 PM rue_mohr: there are a few odd things going on
02:37 PM rue_mohr: ok, so just a sec
02:37 PM rue_mohr: linuxcnc -> control voltage
02:37 PM rue_mohr: what command speed results in 2.5V
02:37 PM Tom_shop: it doesn't help that there are 2 controls
02:38 PM Tom_shop: 1550
02:38 PM rue_mohr: right ok
02:38 PM rue_mohr: so, here is odd
02:38 PM Tom_shop: and that results in 2450
02:38 PM rue_mohr: its out by 1/4
02:38 PM rue_mohr: why
02:38 PM Tom_shop: so the 6000 scale is too high
02:38 PM rue_mohr: no
02:38 PM Tom_shop: in linuxcnc
02:39 PM rue_mohr: 6000 should be 100% duty
02:39 PM rue_mohr: aka, 5V
02:39 PM rue_mohr: but its not
02:39 PM Tom_shop: actually 5000 ish
02:39 PM rue_mohr: 1550 is 50%
02:39 PM rue_mohr: so 100% is 3100
02:39 PM Tom_shop: yeah
02:39 PM rue_mohr: not 6000
02:40 PM Tom_shop: so that scale is off
02:40 PM rue_mohr: there is another scale factor going on
02:40 PM rue_mohr: A -> B -> C -> D
02:40 PM rue_mohr: :)
02:40 PM Tom_shop: there are 2
02:40 PM rue_mohr: ok, so we want 5000 to result in 5V
02:40 PM rue_mohr: more
02:40 PM rue_mohr: there is another one hidden somewhere
02:41 PM rue_mohr: doesn't matter, lets assume its fixed and work around it
02:41 PM Tom_shop: here:
02:41 PM rue_mohr: so 1550 results in 2.5V
02:41 PM Tom_shop: ### Set pwmgen frequency and parameters for spindle pwm
02:41 PM Tom_shop: setp hm2_[HOSTMOT2](BOARD).0.pwmgen.00.output-type 1
02:41 PM Tom_shop: setp hm2_[HOSTMOT2](BOARD).0.pwmgen.00.scale 6000
02:41 PM Tom_shop: setp hm2_[HOSTMOT2](BOARD).0.pwmgen.pwm_frequency 5000
02:42 PM Tom_shop: those are the pwm parameters for the analog out
02:42 PM rue_mohr: the other factor is probably not linuxcnc or the vfd
02:42 PM Tom_shop: you know what?
02:42 PM rue_mohr: it doesn't matter
02:43 PM Tom_shop: if i leave the KP where it is and enable the pid for the analog i bet it would work ok
02:43 PM rue_mohr: hold on
02:43 PM Tom_shop: but it might overshoot
02:43 PM Tom_shop: i'm holding, just thinking
02:43 PM rue_mohr: we have 3 scale factors going on in a chain here
02:43 PM rue_mohr: lets get them all right
02:43 PM rue_mohr: maybe even 4 factors
02:43 PM Tom_shop: only 2 that i see
02:43 PM Tom_shop: maybe 3
02:43 PM Tom_shop: max rpm
02:43 PM Tom_shop: analog v
02:43 PM rue_mohr: but! 2.5V at the vfd should be 2500 rpm, and its not
02:43 PM Tom_shop: and what else?
02:44 PM Tom_shop: yes i agree
02:44 PM rue_mohr: 2500 cmd at linuxcnc should be 2.5V, but its not
02:44 PM Tom_shop: that's why i suggested an edit on the 6000 scale
02:44 PM Tom_shop: i've no idea what that number represents
02:44 PM rue_mohr: yes, but lets carefully work out the number
02:44 PM rue_mohr: I think the right nuumber is 3000
02:44 PM Tom_shop: kinda why i asked :)
02:45 PM Tom_shop: let's go in steps then
02:45 PM rue_mohr: but I'm on morning brain and i might be wrong
02:45 PM Tom_shop: i'll go to 5000 and see a result
02:45 PM rue_mohr: thinking thru
02:45 PM rue_mohr: right now, ~1500 causes 2.5V
02:45 PM rue_mohr: which is 1/4 of 6000
02:45 PM Tom_shop: if my 1550 rpm doesn't output 2.500v we made progress
02:46 PM rue_mohr: if we change scale to 3000, 1500 should be twice the value,?
02:46 PM rue_mohr: I need food
02:47 PM Tom_shop: i'm gonna test 5000
02:47 PM rue_mohr: scale should be the value that makes 100% output
02:47 PM rue_mohr: but its not, its the value that results in 50% output
02:48 PM Tom_shop: now 1550 gives me 3.000v and 3100 rpm
02:48 PM rue_mohr: it did give 2.5V
02:48 PM Tom_shop: so scale should go the other way
02:48 PM Tom_shop: no it gave me 3.00v
02:48 PM rue_mohr: ok, its inverse
02:49 PM Tom_shop: i'll try 7000
02:49 PM rue_mohr: easier numbers
02:49 PM rue_mohr: 2500 should result in 2.5V
02:49 PM rue_mohr: 2500 should result in 2.5V, which is 50%
02:49 PM Tom_shop: 1550 is 2.14v 2000 rpm
02:49 PM Tom_shop: i'll try 2500 now
02:50 PM rue_mohr: (5000/1550) = (5/k)
02:50 PM rue_mohr: yea do 2500
02:50 PM Tom_shop: 3.45v 3680 rpm
02:50 PM rue_mohr: ok, so, 2500 resulted in 3.45 when scale was ________
02:50 PM Tom_shop: 7000
02:50 PM rue_mohr: ok, so, 2500 resulted in 3.45 when scale was 7000
02:51 PM Tom_shop: yes
02:51 PM rue_mohr: so 7000/2500 = 5/3.45 ?
02:51 PM Tom_shop: umm
02:51 PM rue_mohr: there is still a factor of almost 2 going on
02:52 PM rue_mohr: ok, if scale is 5000, it'll go too high on the voltage, right?
02:52 PM Tom_shop: 7000 was too high
02:52 PM Tom_shop: i think 5000 was too low
02:53 PM rue_mohr: ok, idea
02:53 PM rue_mohr: you can turn the vfd off, what cmd results in 5V?
02:53 PM Tom_shop: want me to try 2500 with scale of 5000?
02:53 PM rue_mohr: I'm trying to work out what the heck results in a 5V signal!
02:53 PM rue_mohr: :)
02:54 PM rue_mohr: there is a factor of 2 still in here
02:55 PM rue_mohr: the pwm->analog board
02:55 PM rue_mohr: its rated to output 5V max or 10V max?
02:55 PM Tom_shop: was testing
02:55 PM rue_mohr: k
02:55 PM Tom_shop: 3675 is exactly 5.00v
02:55 PM rue_mohr: ok
02:56 PM rue_mohr: so, with a scale of 7000, 100% is 7350
02:56 PM rue_mohr: throw me a link for that pwm->analog board!
02:56 PM Tom_shop: interweb sucks out here
02:56 PM rue_mohr: I'll grab breakfast
02:57 PM Tom_shop: mesaus.com
02:57 PM Tom_shop: go to their store
02:57 PM Tom_shop: and search for 7i47S
02:57 PM Tom_shop: that's it
02:57 PM Tom_shop: the S means spindle out
02:59 PM rue_mohr: what do you ahve analog+ connected to?
02:59 PM rue_mohr: 5V or 10V?
02:59 PM rue_mohr: ;) its 10 isn't it?
03:01 PM Tom_shop: 5v
03:01 PM Tom_shop: wait
03:02 PM Tom_shop: no i bet it's 10 :)
03:02 PM Tom_shop: sorry
03:02 PM Tom_shop: can you tell from the pdf what pin? so i can check
03:02 PM Tom_shop: i tested both is why i'm not sure
03:04 PM Tom_shop: yeah it's 10v sorry
03:04 PM Tom_shop: i checked it on the control board
03:04 PM rue_mohr: no, its connected at design time
03:04 PM rue_mohr: ok!
03:04 PM rue_mohr: okok
03:04 PM rue_mohr: let me draw for a sec
03:04 PM Tom_shop: ok
03:05 PM * Tom_shop eats a chip
03:05 PM rue_mohr: atlk while I draw
03:05 PM rue_mohr: so linuxcnc takes your commanded speed, and divides scale by it to get a %
03:06 PM Tom_shop: wanna see that file?
03:06 PM rue_mohr: presumadly, that % is the pwm %
03:06 PM rue_mohr: no
03:06 PM rue_mohr: then, that % is converted to the same % of 10V
03:06 PM * Tom_shop hooks his meter back up to the vfd
03:09 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1190686.jpg
03:10 PM rue_mohr: does this seem right then
03:10 PM rue_mohr: scale is out
03:10 PM Tom_shop: browser takes forever to load.. hang on
03:10 PM rue_mohr: we want 5000 to result in 2.5V and it doesn't right now
03:10 PM rue_mohr: but
03:10 PM rue_mohr: 100% would result in 10V, not 5V
03:11 PM Tom_shop: no
03:11 PM Tom_shop: 5000 max rpm?
03:11 PM Tom_shop: should be 10v
03:11 PM rue_mohr: because the pwm% is converted to a % of 10V, not 5
03:11 PM Tom_shop: 5500 tops
03:11 PM rue_mohr: well
03:12 PM rue_mohr: shall we set it up for 5500 max then?
03:12 PM Tom_shop: i'm not sure it will go there but i can test that if you like
03:12 PM rue_mohr: that last box is wrong I think
03:12 PM rue_mohr: I think its 10V = 5000rpm
03:13 PM rue_mohr: oh, 10V = 16000RPM
03:13 PM Tom_shop: let me find the spindle speed that gives me 10v and then i'll turn on the vfd to see what max is
03:13 PM rue_mohr: see how all the numbers are out?
03:13 PM Tom_shop: ^^ ok?
03:13 PM Tom_shop: then we will know the max rpm it will do
03:13 PM rue_mohr: ok, lets start from the left
03:14 PM rue_mohr: we want 5500 rpm (I'm adjusting from 5k) to result in 100% duty
03:15 PM rue_mohr: at 100% duty, the analog output will be 10V
03:15 PM Tom_shop: ok that scale in lcnc won't let it go past 9.23v
03:16 PM rue_mohr: ok, alter scale till cmd 5500 results in 10V?
03:16 PM Tom_shop: i'll try
03:16 PM rue_mohr: tell me what you get...
03:16 PM Tom_shop: i'm not quite sure which way to go with it
03:17 PM rue_mohr: I kinda expect 2750 or 11000 (not sure which way it goes)
03:18 PM Tom_shop: 8000 scale is 6.52
03:18 PM Tom_shop: i bet it's 5500 :)
03:19 PM rue_mohr: possibly, something is throwing me for a loop here
03:19 PM Tom_shop: it won't go past 9.23v
03:19 PM rue_mohr: oh
03:19 PM rue_mohr: is analog+ 10 or 9.23?
03:19 PM Tom_shop: lemme try with the vfd on
03:19 PM rue_mohr: no, wont matter
03:20 PM Tom_shop: analog + maxes out at 9.23v
03:20 PM Tom_shop: not 10
03:20 PM rue_mohr: ok, thats important
03:20 PM Tom_shop: with it on it goes to 9.55v
03:20 PM rue_mohr: strange, ok
03:21 PM rue_mohr: let me think a sec
03:21 PM Tom_shop: that's probably tops
03:21 PM rue_mohr: lets say the converter is 8 bit
03:21 PM rue_mohr: and does not go to 100%
03:21 PM rue_mohr: it'd be 9.96...
03:22 PM rue_mohr: lets say its 7 bit
03:22 PM Tom_shop: i asked him but he may not be online
03:22 PM rue_mohr: 6 bit...
03:23 PM rue_mohr: it looks like its a 4 bit converter...
03:23 PM Tom_shop: 3000 puts out 9.55v too
03:23 PM Tom_shop: 3000 rpm
03:23 PM Tom_shop: scale at 5500
03:23 PM rue_mohr: ok let me catch up
03:24 PM rue_mohr: 5500 -> 9.55V
03:24 PM rue_mohr: oh its saturated
03:24 PM rue_mohr: ok
03:25 PM rue_mohr: so, lets work the numbers around 50% then
03:25 PM Tom_shop: i know
03:25 PM rue_mohr: 50% of our peak speed (5500 rpm) is
03:25 PM rue_mohr: 2750
03:25 PM rue_mohr: so, we want 7250 to result in 9.55/2 V
03:25 PM rue_mohr: 4.775
03:26 PM rue_mohr: cmd 7250rpm -> 4.775v
03:26 PM rue_mohr: via adjusting scale
03:26 PM rue_mohr: I still kinda expect 2750 or 11000
03:26 PM rue_mohr: I'm leaning towards 11000
03:27 PM Tom_shop: i'm getting close
03:27 PM rue_mohr: whats the number looking like?
03:29 PM Tom_shop: i'm getting confused
03:29 PM Tom_shop: 7350 scale 2500 rpm put out 3.xxv
03:29 PM rue_mohr: it might be a 4 bit converter
03:29 PM Tom_shop: i just asked him
03:30 PM rue_mohr: gosh I'd hope not
03:30 PM rue_mohr: it would be a good reason not to say in the docs
03:30 PM Tom_shop: he's very open about his designs
03:31 PM rue_mohr: ok
03:33 PM rue_mohr: Tom_shop, did you have a value on the other side of 4.775V?
03:33 PM Tom_shop: 8500 scale gets 2.8v
03:33 PM rue_mohr: perfect
03:33 PM Tom_shop: for2500rpm
03:33 PM rue_mohr: er, we want it just a bit off
03:34 PM rue_mohr: cmd 7250rpm -> 4.775v
03:34 PM Tom_shop: 9500 scale 2.54v 2509 on the vfd rpm
03:34 PM rue_mohr: cause it the half way mark scaled across the 10V
03:35 PM Tom_shop: i bet 10000 is the magic number
03:35 PM rue_mohr: we want cmd 7250rpm -> 4.775v, might be 11000
03:35 PM rue_mohr: we want cmd 7250rpm -> 4.775v, might be a scale of 11000
03:35 PM Tom_shop: no it saturates
03:35 PM rue_mohr: hu?
03:36 PM Tom_shop: at 9500 scale
03:36 PM Tom_shop: just a sec
03:36 PM rue_mohr: hu?....
03:36 PM Tom_shop: 4000 commanded gets 4.07v 4472 rpm on the vfd
03:36 PM Tom_shop: 9500 scale
03:37 PM rue_mohr: if you cmd 7250rpm, what voltage do you get?
03:37 PM Tom_shop: 2500 commanded gets 2.54v 2507 rpm on the vfd
03:37 PM Tom_shop: it will max out
03:37 PM rue_mohr: cause its over 5500?
03:37 PM rue_mohr: we are not worried about the vfd
03:38 PM Tom_shop: 7.4v but 54xx ish rpm
03:38 PM Tom_shop: at 7520 commanded rpm
03:38 PM rue_mohr: were just working on getting the linucnc commanded rpm to voltage right
03:38 PM rue_mohr: so at 7250 you get 7.4V?
03:38 PM Tom_shop: pcw_home> its normally 12 bit (and 5 KHz)
03:38 PM rue_mohr: oh good
03:38 PM rue_mohr: so at 7250 you get 7.4V?
03:39 PM Tom_shop: yeah
03:39 PM rue_mohr: we want that down
03:39 PM rue_mohr: to 4.78ish
03:39 PM Tom_shop: huh???
03:39 PM rue_mohr: cmd 7250rpm -> 4.775v
03:39 PM rue_mohr: this is the half way mark on saturation
03:39 PM Tom_shop: <pcw_home> PWM scale is usually just RPM at 10V
03:40 PM rue_mohr: yes, 100%
03:40 PM Tom_shop: but i'm already well above that
03:40 PM rue_mohr: I'm trying to get around a factor of 2 in your system
03:40 PM Tom_shop: yeah i know
03:41 PM Tom_shop: i'm pretty sure the KP and KI we changed are good numbers
03:41 PM rue_mohr: damnit
03:41 PM rue_mohr: I carried a type
03:41 PM rue_mohr: I carried a typo
03:41 PM rue_mohr: cmd 2750rpm -> 4.775v
03:42 PM rue_mohr: ugh
03:42 PM rue_mohr: <rue_mohr> 50% of our peak speed (5500 rpm) is
03:42 PM rue_mohr: <rue_mohr> 2750
03:42 PM rue_mohr: <rue_mohr> so, we want 7250 to result in 9.55/2 V
03:42 PM rue_mohr: ^^typo
03:42 PM rue_mohr: so, we want 2750 to result in 9.55/2 V
03:42 PM rue_mohr: cmd 2750rpm -> 4.775v
03:43 PM rue_mohr: how far off are we now?
03:43 PM Tom_shop: 4.83v
03:44 PM rue_mohr: cool!
03:44 PM rue_mohr: whats the scale factor?
03:44 PM Tom_shop: 5400 rpm
03:44 PM Tom_shop: 5500
03:44 PM rue_mohr: cute
03:44 PM Tom_shop: which is what he said was the normal
03:44 PM rue_mohr: ok, so if you wask for 2750 we get 4.83V
03:45 PM rue_mohr: I wonder if we can tweak that a bit closer to 4.775
03:45 PM Tom_shop: just a sec
03:45 PM rue_mohr: the difference is compentating for the 9.55 vs 10V
03:45 PM Tom_shop: i'm not sure it will ever reach the full 10v
03:46 PM Tom_shop: 5.32v @ 2750 scale 5000
03:46 PM Tom_shop: wrong way
03:47 PM rue_mohr: no it wont
03:47 PM rue_mohr: were scaling it around 9.55
03:47 PM rue_mohr: which is why 50% or the 5500 rpm we want is 50% of our available signal (9.55V) aka 2750 is 4.78V
03:48 PM Tom_shop: 4.74v
03:48 PM rue_mohr: great
03:48 PM rue_mohr: what did the scale come out to?
03:48 PM Tom_shop: 5300 rpm
03:48 PM Tom_shop: 5600
03:48 PM rue_mohr: ok
03:49 PM rue_mohr: ok, next the VFD, param P1.11
03:49 PM Tom_shop: but it's spinning way too fast
03:49 PM Tom_shop: yeah
03:49 PM rue_mohr: I THINK it should now be set to ...
03:49 PM Tom_shop: down to what?
03:49 PM rue_mohr: if this were 10V, it would be 5500, but its not 10V
03:49 PM rue_mohr: its 9.55
03:49 PM Tom_shop: i'll go a bit high and work my way down
03:49 PM Tom_shop: so gimme a number
03:49 PM Tom_shop: :)
03:50 PM rue_mohr: so, we want to be 10/9.55 of 5500
03:50 PM rue_mohr: 5759?
03:51 PM Tom_shop: 2750 commanded gets 4.74v just to double check with you
03:51 PM rue_mohr: yup
03:51 PM Tom_shop: i'll set it from 12000 to 7000 for a test
03:51 PM rue_mohr: I'm using up to 9.55 signalling in the middle
03:51 PM rue_mohr: between the analog output and the vfd
03:51 PM Tom_shop: or should i start higher on the test?
03:51 PM rue_mohr: cause they are both kosher with 10V, but we only have 9.55 avail
03:52 PM rue_mohr: at 7000, it should run slow
03:53 PM rue_mohr: should be 80% of commanded speed?
03:53 PM Tom_shop: ok i set it to 7000
03:54 PM Tom_shop: 4.74v commanded 2750 actual 31
03:54 PM Tom_shop: 4.74v commanded 2750 actual 3150
03:55 PM Tom_shop: now what?
03:55 PM Tom_shop: 500 rpm won't even turn it
03:56 PM rue_mohr: no worries
03:56 PM Tom_shop: i know i just wanted to see
03:56 PM rue_mohr: lets run numbers
03:56 PM rue_mohr: so P1.11 is 7000?
03:56 PM Tom_shop: i was getting 100 reliable
03:56 PM Tom_shop: yes
03:57 PM rue_mohr: ok, so 4.74V is 50% of top speed
03:57 PM rue_mohr: and were getting 3150
03:57 PM Tom_shop: yes
03:57 PM rue_mohr: so top is actaully 6300
03:57 PM rue_mohr: I think its really mushy
03:57 PM rue_mohr: oh but wait
03:58 PM rue_mohr: 80%
03:58 PM Tom_shop: i think i need the pid back on the linuxcnc side :)
03:58 PM rue_mohr: no hold on
03:58 PM rue_mohr: we need it to work right BEFORE the PID is on
03:58 PM rue_mohr: after all this, we turn on the PID
03:58 PM rue_mohr: AFTER
03:58 PM Tom_shop: i know
03:58 PM rue_mohr: so, hold on
03:58 PM rue_mohr: you got 3150
03:58 PM * Tom_shop grabs another chip
03:59 PM rue_mohr: at 50% of 9.55V
03:59 PM Tom_shop: at 4.74v
03:59 PM Tom_shop: 2750 commanded
03:59 PM rue_mohr: we kinda need to estimate 1.11 right cause its a pain to change eh?
03:59 PM * Tom_shop feels selfish and offers rue_mohr a chip
04:00 PM rue_mohr: ooo LM555 num num num num
04:00 PM Tom_shop: it's not too bad
04:00 PM rue_mohr: little crunchy
04:00 PM Tom_shop: just gotta power cycle it
04:00 PM Tom_shop: should i try 6300?
04:01 PM rue_mohr: ok, so, I expected 1.11 to need to be 5759
04:01 PM rue_mohr: ---
04:01 PM rue_mohr: if this were operating at 10V, then P1.11 should be 5500
04:01 PM rue_mohr: its 9.55V
04:01 PM rue_mohr: we want 9.55 to be 5500, and were testing at 50%
04:02 PM rue_mohr: 4.78 to be 2750
04:02 PM rue_mohr: if 4.78 is 2750
04:02 PM Tom_shop: but it's 31xx
04:03 PM rue_mohr: 5 would be 2877
04:03 PM rue_mohr: so 10 would be 5753
04:03 PM rue_mohr: so I keep getting 575x
04:03 PM Tom_shop: 4.96v
04:03 PM Tom_shop: :)
04:04 PM rue_mohr: if you use 5756? for P1.11 what does a cmd of 2750 result in?
04:04 PM rue_mohr: or am I being insane
04:04 PM Tom_shop: i did 5753 and got 9.55v
04:04 PM rue_mohr: no
04:04 PM rue_mohr: linuxcnc cmd 2750
04:04 PM Tom_shop: 5400ish rpm
04:04 PM Tom_shop: i know
04:04 PM rue_mohr: will result in 4.78V
04:05 PM rue_mohr: which we want to be close to 2750 on the vfd
04:05 PM Tom_shop: so change P1-11 to what?
04:05 PM rue_mohr: so when p1.11 5753
04:05 PM rue_mohr: you didn't change scale again did ya? :)
04:06 PM Tom_shop: no
04:06 PM rue_mohr: scale should be set so that a command for 2750 results in 4.78V ok
04:06 PM Tom_shop: scale is 5600
04:06 PM Tom_shop: on the pc
04:06 PM rue_mohr: ok
04:06 PM rue_mohr: set it in stone
04:06 PM Tom_shop: silicon
04:06 PM Tom_shop: done
04:06 PM rue_mohr: lets change P1.11 to 5756
04:07 PM rue_mohr: which should make everything match up
04:07 PM Tom_shop: i hope you remember this day :)
04:08 PM Tom_shop: 2750 commanded 4.74v 2602 rpm
04:08 PM rue_mohr: yay!
04:08 PM rue_mohr: ok, how can we tune that a bit more
04:09 PM rue_mohr: we want to lie..... up or down....
04:09 PM Tom_shop: 4000 commanded 6.92v 3949 rpm
04:09 PM rue_mohr: down, go down
04:09 PM rue_mohr: P1.11 to 5750?
04:09 PM rue_mohr: wait whats a percentage error
04:09 PM rue_mohr: 94%
04:10 PM rue_mohr: 98%
04:10 PM rue_mohr: so we want to reduce P1.11 by... 97%?
04:10 PM Tom_shop: huh??
04:10 PM rue_mohr: so maybe P1.11 to 5934?
04:11 PM rue_mohr: this will make the analog votlage seem higher
04:11 PM rue_mohr: its reverse
04:11 PM rue_mohr: P1.11 to 5934?
04:11 PM rue_mohr: no wait
04:11 PM rue_mohr: thats not 90%
04:11 PM rue_mohr: ----> 5583
04:12 PM Tom_shop: u sure this time?
04:12 PM rue_mohr: 97% of 5756 is 5583, right?
04:12 PM rue_mohr: yea
04:12 PM rue_mohr: just a hair
04:13 PM Tom_shop: 2750 is 2528
04:13 PM Tom_shop: 4000 is 3834
04:13 PM rue_mohr: ... did I got he wrong way?
04:13 PM rue_mohr: I did, wtf
04:14 PM rue_mohr: so we had to increase 5756 by 97%
04:14 PM rue_mohr: so then 5934 is the magic number
04:14 PM Tom_shop: pfft
04:14 PM rue_mohr: :D
04:14 PM rue_mohr: or the results are a bit random...?
04:15 PM rue_mohr: you ahve to admit, 5756 resulted pretty close tho
04:15 PM Tom_shop: 2750 is 2681
04:15 PM rue_mohr: close enough?
04:15 PM Tom_shop: 4000 is 4070
04:15 PM rue_mohr: average error = ~0?
04:15 PM Tom_shop: 5000 is 51xx
04:16 PM Tom_shop: the nice thing is it NEVER overshoots now
04:16 PM rue_mohr: 97% and 101.75% and 102%
04:16 PM rue_mohr: ok
04:17 PM Tom_shop: and gets to setpoint quicker
04:17 PM rue_mohr: ok!
04:17 PM rue_mohr: so, thats all worked out
04:17 PM rue_mohr: you can technically tune Kp and Ki if you want?
04:17 PM Tom_shop: naw i'm happy with that
04:17 PM Tom_shop: it's much better than it was
04:17 PM Tom_shop: we'll see what happens under load
04:17 PM Tom_shop: later
04:17 PM rue_mohr: ki = 0; dial up kp till it hunts a bit, dial it down from there, turn up ki
04:17 PM rue_mohr: k
04:17 PM Tom_shop: but those were factory settings
04:18 PM Tom_shop: i know
04:18 PM rue_mohr: dont forget to turn the linux cnc feedback back on
04:18 PM rue_mohr: ;)
04:18 PM Tom_shop: should i reenable the pid now?
04:18 PM rue_mohr: yea, it will need a re-tune
04:18 PM Tom_shop: and watch it runaway :)
04:18 PM rue_mohr: re-tune
04:18 PM rue_mohr: it has an autotune doesn't it?
04:19 PM Tom_shop: no
04:19 PM rue_mohr: ouch
04:19 PM Tom_shop: not that i know of
04:19 PM Tom_shop: i've never used it
04:19 PM Tom_shop: somebody said it didn't work all that good
04:19 PM rue_mohr: how did you tune the linuxcnc pid loop last time?
04:19 PM Tom_shop: just a sec
04:21 PM rue_mohr: oh!
04:21 PM rue_mohr: P1.9 and 1.10 to 20000 each
04:21 PM rue_mohr: if it can brake that hard
04:22 PM Tom_shop: i'm not that worried about that part
04:22 PM Tom_shop: accel is at 20000 already
04:22 PM Tom_shop: p1-9
04:23 PM Tom_shop: breaking is ok as it is
04:23 PM rue_mohr: (new values May 2021)
04:23 PM rue_mohr: .scale (5600)
04:23 PM rue_mohr: P1.0 = 0 (real speed)
04:23 PM rue_mohr: P1.2 is 1 ? [1]
04:23 PM rue_mohr: P1.7 is Kp (1520) [300]
04:23 PM rue_mohr: P1.8 is Ki (320) [100]
04:23 PM rue_mohr: P1.9 and P1.10 accel and decel, max out (60000) [20000] [6000]
04:23 PM rue_mohr: P1.11 is your top speed (5934) [16000]
04:23 PM rue_mohr: P1.12 is the zero offset (whats your analog input range?) [0]
04:23 PM rue_mohr: CV = 0-9.55V
04:23 PM rue_mohr: reverse = seperate ctrl
04:23 PM Tom_shop: ok i _think_ i got the pid bak in it
04:24 PM Tom_shop: #setp pid.0.Pgain .1
04:24 PM Tom_shop: setp pid.0.Igain 1.8
04:24 PM Tom_shop: setp pid.0.FF0 .75
04:24 PM Tom_shop: setp pid.0.FF1 .75
04:24 PM Tom_shop: setp pid.0.maxerror 200
04:24 PM Tom_shop: i wasn't getting anywhere using P the last time
04:26 PM rue_mohr: its hardware now saturates at 5500 rpm
04:26 PM Tom_shop: i think it's oscilating but it's hard to tell with 2 pid loops
04:27 PM Tom_shop: should i start over with just P?
04:27 PM rue_mohr: I suppose
04:27 PM rue_mohr: I will cause oscillation
04:28 PM Tom_shop: what should i start with?
04:28 PM rue_mohr: just P
04:28 PM Tom_shop: what value?
04:28 PM rue_mohr: dial it up till it overshoots, then back off a bit and apply P I think
04:28 PM rue_mohr: 1?
04:29 PM Tom_shop: 1500 rpm won't even start
04:29 PM rue_mohr: wow
04:29 PM rue_mohr: odd
04:29 PM rue_mohr: thats P?
04:30 PM rue_mohr: 10? 100?
04:30 PM Tom_shop: 10 i get about 1800 rpm
04:30 PM rue_mohr: it should settle out tho
04:31 PM Tom_shop: 10 is too high
04:31 PM rue_mohr: indeed, 1 is too low
04:31 PM rue_mohr: what are my objectives for today again
04:31 PM rue_mohr: oh, right, pass the butter
04:31 PM Tom_shop: 9 it goes nuts
04:32 PM rue_mohr: what?
04:32 PM rue_mohr: I'm confused
04:32 PM Tom_shop: oscillates
04:32 PM rue_mohr: how about 1.5
04:33 PM rue_mohr: 1 must be 100%
04:33 PM Tom_shop: 9.5 it oscillates then smooths out
04:33 PM rue_mohr: ok
04:33 PM rue_mohr: 5?
04:33 PM Tom_shop: i went from 10 to 9.5
04:33 PM Tom_shop: and you want 5?
04:34 PM Tom_shop: 9.75 was better
04:35 PM Tom_shop: but the rpm is high
04:35 PM Tom_shop: 1822 and commanded was 1500
04:35 PM Tom_shop: i don't get it
04:36 PM rue_mohr: P should do better than that given the sub-control
04:36 PM rue_mohr: well, go up and down till you get some overshoot
04:37 PM Tom_shop: overshoot is hard to tell
04:37 PM Tom_shop: too low and it oscillates
04:37 PM rue_mohr: ziiiingggg...wrrrrrrrr
04:37 PM rue_mohr: low and it oscillatoes?
04:37 PM rue_mohr: wtf
04:37 PM rue_mohr: try 0.8
04:37 PM rue_mohr: (80%)
04:37 PM rue_mohr: I think its wrapping and saturating
04:38 PM Tom_shop: .8 won't spin it
04:40 PM rue_mohr: ok doesn't sound like wrapping then
04:41 PM Tom_shop: no
04:41 PM rue_mohr: there must be critical damping in there somwhere
04:43 PM Tom_shop: how close were we without any pid?
04:43 PM Tom_shop: pretty close?
04:43 PM rue_mohr: +-5%
04:43 PM Tom_shop: i'm of the opinion they may fight each other
04:43 PM rue_mohr: its often hard for them to work togethor, but it shoudl be doiable
04:48 PM Tom_shop: ok it needs something because it won't start at 500 rpm
04:48 PM rue_mohr: yea
04:48 PM rue_mohr: why wont kp behave
04:49 PM rue_mohr: whats that chart of results look like?
04:49 PM Tom_shop: i couldn't get it to work last time either
04:49 PM Tom_shop: too big a mess to tell
04:49 PM rue_mohr: ok, from 1, table out results for command of 1000rpm
04:50 PM rue_mohr: say from 1 to 9?
04:50 PM Tom_shop: i took the pid out
04:50 PM Tom_shop: to try
04:50 PM Tom_shop: i'm gonna put it back... hang on
04:51 PM Tom_shop: you remember all the vfd parameters we edited?
04:51 PM rue_mohr: write it all down
04:51 PM Tom_shop: i'm going to
04:51 PM Tom_shop: but i need to stop for today
04:51 PM rue_mohr: later ask the guys in #linuxcnc how to tune the PID
04:51 PM Tom_shop: once i get the pid back in place
04:51 PM rue_mohr: we got the motor stuff all working good
04:52 PM Tom_shop: yeah
04:52 PM Tom_shop: i'll remember the settings
04:52 PM Tom_shop: i've got a book on this mill
04:56 PM rue_mohr: ok well, not a nice day for changing head gaskets
04:59 PM rue_mohr: that max output might need to be 100 tho
04:59 PM rue_mohr: aka 100%
04:59 PM rue_mohr: it cant go to 200% on the hardware as per how we reconfigured it
05:02 PM Tom_shop: until i can get back to this, i'm setting it like it was when i started but i wrote all the numbers down. if i don't do this i will forget where i was at
05:08 PM rue_shop1: yup
05:08 PM rue_shop1: prolly take the same timeframe to tube the Pid
05:09 PM Tom_shop: i know
05:29 PM rue_shop1: keep needing a cnc lathe :/
05:30 PM rue_shop1: I suppose a good start would be a hydraulic chuck
05:34 PM Tom_L: i did fiddle with the KP & KI settings after i set things back like when i started
05:37 PM rue_shop1: we had the settings right, we just need to know how to tune the linuxcnc side of it
05:38 PM Tom_L: i know
05:38 PM Tom_L: i just wanted to see what it would do as it was
05:38 PM rue_shop1: it really helps when there is a built in tuning thing
05:38 PM rue_shop1: the battery on the mowerbot is shot
05:39 PM rue_shop1: U1
05:39 PM rue_shop1: kinda annoying
05:39 PM rue_shop1: not sure it saves as much work per year as a U1 costs
05:40 PM rue_shop1: 6.5V, falls to 4V @ 4A load
05:40 PM rue_shop1: :S
05:42 PM rue_shop1: the U1 in the lawn tractor is ok, I'll top it up, but that good cause its only a year old
05:56 PM rue_shop1: so I have an idea for a discrete bushing
06:08 PM Tom_L: bushing for what?
06:13 PM rue_mohr: the 4'x8' cnc
06:16 PM Tom_L: oh
06:18 PM Tom_L: https://soylentnews.org/article.pl?sid=21/05/23/1632220&from=rss
06:18 PM Tom_L: somebody posted that
06:19 PM Tom_L: i
06:19 PM Tom_L: i'm not that concerned about any of it
06:22 PM rue_shop1: tho it would explain the religious context of the attacks over the last 20 years
06:25 PM rue_shop1: it would not surprise me if the new owner is the one behind the attacks all these years
06:25 PM rue_shop1: but I still dont understand why
06:25 PM rue_shop1: in either case they will be attacking liberchat
06:25 PM rue_shop1: this is the war started when he didn't get the channel after lilo passed
06:28 PM rue_shop1: it looks like hntop was a news channel of some kind?
06:31 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1190689.jpg
06:31 PM rue_mohr: http://ruemohr.org/~ircjunk/tempimage/p1190690.jpg
06:41 PM Tom_L: kbd stand?
06:42 PM Tom_L: i see now from the sketch
06:43 PM rue_mohr: I have some nylon blocks to hold the shaft in the middle
06:44 PM rue_mohr: it'll help via a bit of damping
06:44 PM Tom_L: ok so that guy advertises 5400 rpm
06:47 PM rue_mohr: polprog, did you register garfield on dalnet! :)
06:48 PM rue_mohr: hah
06:48 PM Tom_L: i think he said he did
07:20 PM polprog: yes
07:20 PM polprog: ill give you chanserv flags tomorrow
07:21 PM rue_mohr: np
07:21 PM polprog: i dont know if you registered the nick on nickserv
07:21 PM rue_mohr: yup
07:21 PM polprog: ok
07:21 PM rue_mohr: visit the people in #electronics too
07:21 PM rue_mohr: small fiendly group
07:21 PM polprog: oh i forgot about them
07:21 PM polprog: right
07:22 PM rue_mohr: dalnet tom, if you want to listen in for a while
07:22 PM rue_mohr: it doesn't really make sense to me to do #robotics and #avr when it was all just takeoffs from #electronics in the first place
07:22 PM Tom_shop: i'm just getting real tired of it all
07:24 PM rue_mohr: gotta have someone to stirr up the mine
07:24 PM rue_mohr: gotta have someone to stirr up the mind
07:27 PM Tom_shop: i don't need that kind of stirring
07:27 PM polprog: done
07:27 PM aandrew: oy
07:27 PM rue_mohr: Tom_shop, its huge to have someoen to bounce ideas off
07:28 PM aandrew: spent the entire day (well 11am - 8pm) doing a science experiment for the 8yo's school
07:28 PM rue_mohr: I need to go make huge ass holders for that bushing thing
07:28 PM aandrew: balloon rocket on a fishing line, maximizing force forward carrying a payload of two loonies
07:28 PM rue_mohr: hah
07:28 PM aandrew: it was supposed to be a 2h project or so but ended up just a total shitshow
07:28 PM rue_mohr: I remember baloons on fishing line
07:28 PM rue_mohr: not that well tho..
07:28 PM polprog: i have some BFX59 on the way, 1GHz NPNs
07:29 PM rue_mohr: go find a crt monitor and get the ones off the RGB drivers
07:29 PM aandrew: first is inflating them - have a little hand pump that worked well enough but how to inflate them consitently (number of pumps is easy)
07:29 PM aandrew: however trying to "hold" the balloons in a container so we can add fins/nosecone/nozzle/etc. was an utter and total failure
07:29 PM rue_mohr: heh, I hear rabbit holes
07:29 PM polprog: hard to get CRTs here.. and i dont have space to keep the rest of the CRT really
07:30 PM aandrew: must have spent an hour on that before giving up that route and also discovering the balloons I got were shit for this project
07:30 PM aandrew: (the long tube ones)
07:30 PM rue_mohr: just take out the RGB driver transistors and ditch the rest
07:30 PM aandrew: so went on a mad rush to get some other balloons
07:30 PM polprog: i keep that in mind whenever i see one in a ditch here
07:30 PM rue_mohr: MLE finds them on the side of the road all the time
07:30 PM aandrew: got them and similary they didn't work worth a damn in the containers
07:30 PM rue_mohr: polprog, keep a long shaft philips on you at all times
07:30 PM aandrew: finally got kind of desperate and taped the balloon directly to the straw on the fishing line... that worked pretty awesome so ok moving forward
07:31 PM polprog: :)
07:31 PM aandrew: then basically confirming that yes, trying to do anything CONSISTENLY involving a balloon taped to a fucking straw is impossible
07:31 PM rue_mohr: hmm, I think we taped the support straw to the side of the balloon
07:31 PM aandrew: and also the latex seems to adhere way to well to regular old cheapass scotch tape
07:31 PM aandrew: so we lose balloons through that
07:32 PM rue_mohr: server fan + lithium ion battery on straw?
07:32 PM aandrew: and trying to reinflate a wrinkly deflated balloon kind of half assed still taped to the straw is not possible
07:32 PM rue_mohr: hahaha
07:32 PM rue_mohr: compressor...
07:32 PM aandrew: and that trying to tape a tiny little plastic container for a nozzle to the ass end of a balloon is hard
07:32 PM aandrew: oh yeah
07:33 PM aandrew: about 3pm I said fuck this shit and dragged out the little camping inflator, which didn't do shit
07:33 PM aandrew: so I said ok fine, FUCK ALL OF THIS
07:33 PM aandrew: and grabbed my big tank compressor hose
07:33 PM aandrew: and that at least solved that problem nicely
07:34 PM aandrew: but it was about 4pm by this time, the kids are bored and a little disappointed that it's not as fun as they'd anticipated and are starting to wander off
07:34 PM aandrew: so it's really me doing the project
07:34 PM aandrew: anyway
07:34 PM rue_mohr: oh no
07:34 PM aandrew: it's done now, with 1 balloon to spare and not nearly as much fun as I wanted them to have
07:34 PM aandrew: but we did learn that the everyday regular 12" latex balloons are WAY better than the tube balloons for force
07:34 PM rue_mohr: need to do million monkeys
07:35 PM rue_mohr: set up a bunch of strings and let everyone take to it in different ways
07:35 PM rue_mohr: lots of it will be writeoff, but it could keep them busy
07:35 PM rue_mohr: and building things,
07:35 PM rue_mohr: any experience building anything is a + takeaway
07:36 PM rue_mohr: even its its a comlete fail
07:36 PM aandrew: and we also learned that if you poke a hole in a 4L water jug so the balloon inflating doesn't compress the air trapped in the jug, you can actually create enough suction to collapse the jug with a regular old balloon when you let the air out fast
07:36 PM aandrew: got a short slow-mo video of that which will be a nice anecdote for the presentation
07:36 PM rue_mohr: cool
07:36 PM aandrew: solution to that is just to make the hole for inflating air escaping much bigger so there's enough of an opening to replace the volume of air quickly rushing out of the balloon :-)
07:36 PM rue_mohr: bushing holders brb
07:36 PM aandrew: but anyway
07:37 PM aandrew: long tiresome frustrating but fun day
07:37 PM aandrew: and my 12yo is entirely becoming a teenager with the snarky comments, talking shit and hearing waht you say but doing whatever he wants anyway
07:37 PM aandrew: which is understandable, but kind of disappointing. he wasn't a normal little boy but looks like he's gonna be a typical teenager
07:38 PM aandrew: I love him just the same but told him that this behaviour will cause me to just send him away to his room or whatever. Just because you *are* a teenager doesn't mean you must act like one.
07:38 PM aandrew: I know it'll still happen and that's fine, but if he's aware and *trying* not to be a typical teen that is more than good enough for me
07:39 PM spline: online? tv?
07:39 PM spline: not even fb anymore.. everyone tells me "fb is for geezers"
07:39 PM aandrew: the 17yo likes to play the "I'm an adult, stop treating me like a kid" and "I'm only 17, you can't expect me to act like an adult" cards simultaneously
07:39 PM spline: but when I see many kids out there, its sometimes scary what they're already doing (or not listening to)
07:40 PM rue_shop1: yep, overlap or be stranded
11:13 PM aandrew: spline: hm?