#linuxcnc Logs

Apr 04 2017

#linuxcnc Calendar

12:12 AM Cromaglious_: bbiaw... switching the hd from this hp dv4 into a hp dv2000
12:28 AM Jymmm: Damn... https://www.youtube.com/watch?v=FhgZ4hBGJkY
12:44 AM Wolfmetalfab is now known as Wolf_
12:50 AM Cromaglious_: sheeezh... need a sata connector to mate with the MB connector... grr... and the batteries are pretty much same form factor, but the move the connector on the battery so I can't swap them
12:52 AM Cromaglious_: and the dv2000 is a 1.6ghz the dv4 is a 2.1ghz... darn it... gonna try the other computer at the space that was donated
12:53 AM Cromaglious_: AND different power cords...
01:32 AM Crom: about ready to down grade this kernel to something not so bleeding edge..
01:32 AM Crom: Linux Linuxcnc 4.9.13-rt12 #2 SMP PREEMPT RT Fri Mar 10 03:54:44 PST 2017 x86_64 GNU/Linux
01:39 AM IchGucksLive: Morning from Germany
01:43 AM Cromaglious_: 2311 in California
01:44 AM IchGucksLive: so go to sleep europ awakes
01:45 AM IchGucksLive: with a sunny day
01:45 AM IchGucksLive: spring allover
01:45 AM IchGucksLive: cold night
01:45 AM IchGucksLive: Summer expected next week
01:52 AM IchGucksLive: till later
02:18 AM Cromaglious_: nite nite all
02:41 AM Deejay: moin
02:55 AM DavenportGuy: Good Morning DJ
02:55 AM DavenportGuy: Working on any projects?
03:04 AM Loetmichel: I am pleasantly surprised: After letting the 3dprinter work when i got home yesterday i came to a company, not a smoldering wreck this morning... and the print even finished decently! -> http://www.cyrom.org/palbum/main.php?g2_itemId=16528 http://www.cyrom.org/palbum/main.php?g2_itemId=16531&g2_imageViewsIndex=1
03:05 AM Wolf_: lol nice
03:28 AM JesusAlos: Hi
03:29 AM DavenportGuy: Loetmichel: nice run time?
03:30 AM DavenportGuy: Loetmichel: err.... Nice print. What is the run time?
04:54 AM Loetmichel: DavenportGuy: cura predicted 7 hours
04:55 AM Loetmichel: i dont acutally know because i left the company yesterdy after about 2 hours at 20% printed
04:56 AM XXCoder: Loetmichel: nice!
05:01 AM DavenportGuy: Im trying to write a software counter so I can know how many parts I have ran for tooling changes etc. I have most of my programs written in a fashion which are like o100 DO; gcode stuff ; o100 WHILE [#<_task> EQ 1]; M2 so most of my programs just loop infinitely. What I normally do is use a physical counter and its looped in here as m64p0; g4p[.1];m65p0. What is the best way to create a software counter which works within a DO, WHILE?
05:03 AM DavenportGuy: My current system is to fire off a M-code which runs a bash file which does $var + 1 > somefile; exit
05:04 AM DavenportGuy: I was wondering if there was a common practice for this kind of thing.
05:04 AM archivist: a user var saved probably
05:05 AM archivist: note possible queue busting effects of/on saving vars
05:05 AM DavenportGuy: ah yes, so I could actually call on this number if it reaches a certain point and redirect the gcode to finish cycle then pause and wait for tool change
05:07 AM DavenportGuy: most of the DO, WHILE cycle times will be under 10 seconds above 2. Would this be problematic for the queue?
05:07 AM archivist: I was saving stuff when I did a worm accuracy testing jig and hit an obscurity, the probing .ngc does something similar
05:07 AM DavenportGuy: im not really sure what you mean by queue busting* as it can hinder the gcode from continuing due to file writes?
05:08 AM archivist: nah its to do with when something happens and the scanning by the preprocessor
05:08 AM DavenportGuy: potential problems?
05:08 AM archivist: make sure it saves when you think it should, eg not on load time in axis
05:09 AM archivist: I wondered why I had a constant stored 4k times instead of real numbers
05:11 AM archivist: was storing encoder values on the input and output of a worm reduction to measure angular error
05:12 AM DavenportGuy: heh
05:13 AM DavenportGuy: I wish there was an irc room for people who run production machine shops
05:13 AM DavenportGuy: this place is close but.... not
05:14 AM archivist: the people running production rarely speak, too busy making
05:14 AM archivist: some are on the mailing list
05:15 AM DavenportGuy: well, the thing about production work is, you make the most money when you arent doing anything at all
05:15 AM DavenportGuy: If you got an operator with his legs kicked up and poking around on his cell phone, you are making money, if hes got his head in the machine you arent! =D
05:16 AM DavenportGuy: Im trying to start up a second shift and I need more pud jobs for some new guys to learn the machines with
05:17 AM XXCoder: production worker here heh
05:17 AM XXCoder: I run a61 often :P (but not setup)_
05:17 AM DavenportGuy: XXCoder: what kinds of machine?
05:18 AM DavenportGuy: oh horozontal cnc
05:18 AM XXCoder: cell actually. 2 horzional cncs with centeral computer and pallet system
05:18 AM archivist: some do one off production
05:18 AM DavenportGuy: you notice its a lot less work to run a perfect part than it is to run a borderline part? If you got everything dialed in you can just fire it up and pay it little attention
05:19 AM archivist: I mostly only make low qty
05:19 AM XXCoder: thats MUCH simplier with cell with standard tools
05:19 AM XXCoder: all diameters and lengths is known and long since tweaked close
05:20 AM XXCoder: +- .0005 parts tend to run just fine
05:20 AM DavenportGuy: low quantity is nice because things do not get boring, but the high volume is great because you can really focus on making yourself efficient and the profit per part almost always goes up after you ran 20,000 of something
05:20 AM XXCoder: heh low qualty but LOOONG run
05:21 AM XXCoder: nowdays its smaller and/or faster parts but sometimes run 6 hours per pallet, 2 parts on pallet
05:21 AM DavenportGuy: yesh, I got a job I been running for over a decade and it just gets easier to run every year
05:24 AM XXCoder: yeah easier each run on parts indeed
05:26 AM jthornton: morning
05:27 AM XXCoder: hey jt
05:28 AM jthornton: what's up today
05:28 AM XXCoder: it was a fairly boring shift
05:28 AM XXCoder: which is good
05:29 AM jthornton: yea if nothing blows up it is good
05:29 AM XXCoder: also so much less walking
05:29 AM XXCoder: lately I was running a61 and some other machine across shop lol
05:30 AM jthornton: I finally got the VMC back on line after the avalanche of steel last Thursday
05:30 AM XXCoder: avalanchge of steel? what happened?
05:31 AM jthornton: I stand my 10' and shorter stock up behind the CHCN and the VMC and was very careful not to bump it till Thursday
05:34 AM jthornton: after picking it all up and moving it out of the way I added a 2x4 near the ceiling with some 16p common nails to keep the steel from racking
05:35 AM XXCoder: ow heh
05:35 AM jthornton: the VMC had two air lines that were 50' long and just piled up on the floor so I fixed that mess while I was cleaning up
05:35 AM XXCoder: on my cnc front STILL waiting for parts :P
05:35 AM jthornton: what are you waiting on?
05:36 AM XXCoder: while 3d printer I upgraded extruder and found out I need stepper holder so ordered that and some bowen tubes and connectors
05:36 AM XXCoder: 90 degree 2020 connectors
05:37 AM jthornton: you got a dual hot end now?
05:37 AM XXCoder: nah plan to do Y type
05:40 AM jthornton: so you retract a bit and purge between colors?
05:40 AM XXCoder: retract a LOT more lol
05:40 AM XXCoder: like maybe 40mm retract for color change
05:40 AM XXCoder: depends on where I place Y
05:41 AM jthornton: what kind of extruder did you get?
05:42 AM XXCoder: hmm a second
05:42 AM XXCoder: https://www.aliexpress.com/item/42HD4027-01-3D-Printer-Aluminum-Extruder-kit-Set-w-NEMA-17-Stepper-Motor-1-75mm-1/32761224981.html
05:43 AM XXCoder: I didnt know till later than it needs bowden connectors with larger screw at end
05:43 AM XXCoder: ours come with tiny screw ones
05:44 AM jthornton: for $7 you can't beat that, prob better than the plastic one that came in the kit
05:45 AM XXCoder: I tested it. no shit
05:45 AM XXCoder: I really abused the filiment, pulling and pushing it across same spot constantly
05:45 AM XXCoder: it dont get destroyed though thinner
05:45 AM XXCoder: *though it does get thinner
05:46 AM XXCoder: old one would.
05:52 AM XXCoder: anyway hope that extruder fixes so printer can print retract hell parts like my bridge test chair lol
10:10 AM Roguish_ is now known as Roguish
10:56 AM Simonious: Where should a guy look for a rebuild kit for x/y on a bridgeport 2J?
11:14 AM CaptHindsight: spring cleaning http://www.ebay.com/itm/1983-matsuura-760-/201879814447 http://www.ebay.com/itm/matsuura-/201879823131
11:16 AM * djdelorie drools
11:17 AM CaptHindsight: http://www.ebay.com/itm/Hurco-BMC-4020-CNC-Machining-Center-for-Parts-Ulitmax-SSM-CNC-Control/252835503857
11:20 AM archivist: I saw a nice old manual hurco yesterday
11:26 AM CaptHindsight: Simonious: maybe http://www.machinerypartsdepot.com/
11:28 AM CaptHindsight: https://bptparts.com/ http://hqtinc.com/leadscrewassembly-4hp.aspx
11:29 AM Simonious: ok
12:29 PM IchGucksLive: hi
01:11 PM IchGucksLive: mot that many action today
01:29 PM IchGucksLive: Gn8
01:56 PM chopper79: afternoon all
01:58 PM JT-Shop: hey
02:11 PM Roguish: JT-Shop: hey JT
02:12 PM Roguish: check out this place: www.acornpd.com
02:12 PM chopper79: Putting a little time in on my spindle index again. Still having trouble getting it to read all the time.
02:17 PM jdh: I think someone here was using that sensor
02:19 PM chopper79: yes many are.... I just can not get it to pick up anything 100% of the time besides cast iron. Just strange
02:23 PM chopper79: I have tried many things. Bolts, nuts, sockets, 1018, 4140, weld rod, dowel pins, t name a few and they all give solid reading at 0-10rpm, then 11-40 it is hit or miss and then from 40 up it is never. Cast iron seems to allow it to work. I just do not have a piece of cast iron around to put on there to see what happens.
02:24 PM Cromaglious_: morning
02:24 PM chopper79: I can take the sensor an place it on the gear and it will read fine. So the only thing I can come up with is material issue.
02:26 PM chopper79: morning ish
02:26 PM Cromaglious_: still got 6 minutes here
02:26 PM chopper79: where is here?
02:27 PM Cromaglious_: PRK Peoples Republic of Kalifornistani, Court of Rio, Temecula
02:28 PM chopper79: ok
02:28 PM Cromaglious_: Riverside county, hence the Court of Rio
02:29 PM Cromaglious_: north east of San diego, South East of Los Angles
02:29 PM chopper79: Now you have 1 min
02:30 PM Cromaglious_: Tue Apr 4 11:58:17 PDT 2017
02:30 PM Cromaglious_: damn time to put batteries into the atomic clocks
02:34 PM Cromaglious_: or sensor need a big sample
03:07 PM Cromaglious_: chopper79, found this http://e2e.ti.com/support/sensor/inductive-sensing/f/938/p/385700/1362757
03:12 PM skunkworks: chopper79, how do you know it isn't reading?
03:12 PM Simonious: https://goo.gl/photos/ZRxWKnvuhYfkaR6L7 - so I tighted up the brass nut under the x-axis until it was pretty tough to turn the axis and the amount of slop (~50thou) didn't change, then another guy in the shop noticed this. Our current theory is that this may be occuring because the handle is missing on the left side. Running to the hardware store for a fine thread nut so we can tighten that down to
03:12 PM skunkworks: (how are you testing the index?)
03:12 PM Simonious: see, but I'm interested in other theories.
03:12 PM Simonious: (Bridgeport)
03:16 PM chopper79: back here now.
03:16 PM chopper79: skunkworks: Just using hal meter to see if the state switches
03:16 PM skunkworks: The correct way would be to set the index enable to true - then see if it goes false
03:16 PM chopper79: The state switches when using cast iron up to 6k (but then its to fast to read every index)
03:17 PM skunkworks: or a scope.. ;)
03:17 PM chopper79: It does when using cast and not when using any other material above 25rpm
03:17 PM skunkworks: hmm
03:18 PM skunkworks: how many teeth?
03:18 PM skunkworks: on the cast
03:18 PM skunkworks: you are using mesa hardware - right?
03:21 PM skunkworks: I don't know how consistantly you would be seeing it at 3000rpm... what is your servo period? 1khz? that would be with a 80tooth gear - 4000 transistions. (if I did the math right)
03:21 PM skunkworks: per second
03:22 PM Jymmm: If anyoen has a clue, I'm not understanding the what 3 /4 are doing/interacting with everything else? I suspect the cap/diode is voltage doubler.... http://imgur.com/nxDejsR
03:24 PM sync_: it is the heater
03:24 PM skunkworks: what are you using for the index?
03:26 PM skunkworks: biab
03:57 PM chopper79: skunkworks: 80 teeth, 7i92m+ 7i77, using ATS667LSG sensor for index, servo thread is 1,000,000 in the ini.
04:13 PM Frank_6: anyone knows how to diferentiate the step
04:14 PM Frank_6: step+- and dir +- when the servo manual says puls1 puls2 sign1 sign2
04:14 PM Frank_6: (sry pressed enter early)
04:14 PM Frank_6: im using a mesa 7i76
04:14 PM XXCoder: hey thats fine. its
04:15 PM XXCoder: not typing
04:15 PM XXCoder: like this like
04:15 PM XXCoder: some people
04:16 PM Frank_6: l
04:16 PM Frank_6: o
04:16 PM Frank_6: l
04:16 PM XXCoder: :P
04:17 PM XXCoder: anyway
04:17 PM XXCoder: hmm isnt step and dir for steppers only?
04:18 PM XXCoder: could be wrong, never seen servo controller and servos
04:18 PM Frank_6: this ac servo has the step/dir option
04:19 PM XXCoder: my guess?
04:19 PM XXCoder: step is pulse, and sign is direction
04:20 PM XXCoder: but its nothing more than guess
04:22 PM Frank_6: yes i belive so, the thing is i cant diferentiate between the pulse+ and pulse- as it refers to pulse1 and pulse2
04:23 PM XXCoder: yeah I wonder about that
04:28 PM chopper79: + = 1 and - = 2
04:28 PM XXCoder: but why label 1 and 2 not + -?
04:29 PM chopper79: Manufactures are all different.
04:29 PM Frank_6: chopper79: may i ask how you come to that conclusion
04:29 PM Frank_6: '
04:29 PM chopper79: This is there way of being a pain in the rear. Kinda the same thing when it comes to wiring codes.
04:30 PM chopper79: I have drivers that have the same lables ;)
04:30 PM Frank_6: ok thanks, btw, does something explode if i miss wire those?
04:31 PM chopper79: Should be protected. What manuafacture?
04:33 PM Frank_6: panasonic
04:37 PM chopper79: Which model are you working with?
04:37 PM chopper79: Do you have a manual?
04:38 PM Deejay: gn8
04:40 PM pcw_mesa: if you swap the DIR+/DIR- leads your motor direction will be reversed
04:40 PM pcw_mesa: if you swap the PULSE+/PULSE- leads, it will work but you may violate
04:40 PM pcw_mesa: the step timing
04:43 PM chopper79: ok i am back
04:43 PM pcw_mesa: On a 7I76, STEP+ is the active high output (high for the duration of the step pulse)
04:43 PM pcw_mesa: The drive manual should have some hint about the PULSE1,PULSE2 active polarity
04:43 PM chopper79: so where were we?...... nevermind pcw answered it
04:56 PM Frank_6: sry i was afk for a moment; here's what panasonic says, apparently pulsh1 and signh1 are the step+ and dir+. right? https://www.panasonic-electric-works.com/cps/rde/xbcr/pew_eu_en/dd_63789_0700_en_liqi_fp_npn_pnp.pdf
04:57 PM Frank_6: should i be looking at the pnp or npn page?
04:58 PM chopper79: Yes 1 should be +
04:59 PM Frank_6: page two makes it confusing
04:59 PM chopper79: From other manuals it appears that the inputs are high.
05:04 PM chopper79: Well.... I need to get back to fighting with the index on the spindle.
05:04 PM Frank_6: Good luck and thanks for your input
05:05 PM chopper79: Thanks and same to you
05:40 PM chopper79: Tried multiple target sizes and still no luck.
05:42 PM skunkworks: chopper79, again - how are you testing it? through halscope?
05:42 PM chopper79: I am going to pull my hair out on this one. Used a cheap china prox switch and the allegro sensor and still poor results. I really did not think that getting an index to work was going to be this difficult for me to do.
05:42 PM skunkworks: and what sensor are you using? one of the single ones?
05:42 PM chopper79: halscope does not work on my system. I can not upgrade due to no internet activity to this location.
05:43 PM chopper79: Tried a prox and a single allegro chip
05:44 PM chopper79: Take same sensor or chip and put it on the teeth and the index will change state. Move chip or prox to a steel target and will not pick up past 25 rpm on a consistant basis.
05:45 PM skunkworks: through halmeter?
05:45 PM chopper79: @ 3k rpm the index will change state very rapidly. Yes using halmeter as I do not have or have access to scope.
05:45 PM chopper79: @3k when on the teeth
05:46 PM chopper79: when on a target nothing above 25rpm.
05:46 PM skunkworks: I don't know of a good way of testing... With out a scope.
05:46 PM skunkworks: you are not going to see it with halmeter.
05:46 PM skunkworks: especially a single target.
05:46 PM skunkworks: one test would be to run the single target at 3000rpm and set index enable to true. then see if it goes false
05:47 PM skunkworks: or 6000
05:47 PM skunkworks: we are using the allergro single chip with a 3/16 hole for the index
05:48 PM skunkworks: if it changes to false every time you set it to true - then I would be optimistic.
05:48 PM chopper79: halmeter pick up changes in state @ 3k. Not 100% but a state change shows me it sees something.
05:48 PM skunkworks: yes - because there are enough transistions that halmeter see's it some times
05:48 PM chopper79: 8mm hole in some alum for the sensor mount
05:49 PM chopper79: How would I go about setting it to true for just one cycle?
05:49 PM skunkworks: you wouldn't
05:49 PM chopper79: Right now it is set to true but goes to false then true then false....etc.
05:49 PM archivist: real scope
05:49 PM skunkworks: no - the index enable pin
05:50 PM chopper79: enable pin is set true.
05:50 PM skunkworks: index-enable
05:50 PM chopper79: archivist: love to but do not have one or access to one.
05:50 PM chopper79: yes
05:50 PM chopper79: The enable is set true
05:51 PM skunkworks: hm2mumble.mumble_encodernubmermumble.index-enable
05:51 PM chopper79: so I will check it at 3k and see if it drops to false?
05:51 PM skunkworks: yes - set it to true - it should instantly go false
05:51 PM skunkworks: if that seems consistant
05:51 PM chopper79: ok I will check it.... brb
05:51 PM skunkworks: then try a threading cycle over and over
05:52 PM skunkworks: (put a shaft in the spindle - marker in the vise - see if it draws over itselfe
05:54 PM chopper79: ok
05:54 PM chopper79: setp for index-enable is 1 and in halmeter it shows false
05:55 PM chopper79: the index inout will change state but the false for index-enable never changes state.
05:55 PM chopper79: even though it is set to 1 in hal
05:56 PM skunkworks: so - you set it to true (1) - it stays true(1)?
05:58 PM chopper79: set it to true(1) and it stays false (0)
05:59 PM skunkworks: ok - without the spindle on - if you set it to true it should stay true until it passes the index - does it?
05:59 PM chopper79: let me check again
05:59 PM skunkworks: ok - without the spindle off - if you set it to true it should stay true until it passes the index - does it? - sorry
05:59 PM skunkworks: sorry - trying to feed the kids at the same time
06:02 PM Tom_L: it's a wonderful day when they can feed themselves
06:02 PM skunkworks: 1/2 way there
06:03 PM Tom_L: i certainly don't miss the diaper years
06:03 PM skunkworks: for some definition of feed themselves..
06:03 PM chopper79: ok
06:03 PM chopper79: restarted lcnc and yes spindle-index is true
06:04 PM chopper79: I will now try it with 6000 rpm
06:04 PM chopper79: to see if it goes false
06:04 PM skunkworks: wait - what pin? it should not be true when you first start linuxcnc. It is something that gets set manually - or during a spindle synced cycle
06:05 PM skunkworks: are you looking at the right pin?
06:05 PM skunkworks: or homing
06:05 PM skunkworks: (if it is an axis)
06:05 PM skunkworks: the pin is index-enable
06:06 PM chopper79: yes looking at encoder 05 index-enable pin
06:06 PM skunkworks: and it was true when you started linuxcnc?
06:06 PM chopper79: yes
06:06 PM skunkworks: I don't think that is right.
06:06 PM skunkworks: is it hooked to something?
06:07 PM skunkworks: index-enable - (Bit, I/O) When this pin is set to True, the count (and therefore also position) are reset to zero on the next Index (Phase-Z) pulse. At the same time, index-enable is reset to zero to indicate that the pulse has occurred.
06:07 PM chopper79: Yeah to the normal net spindle-index-enable
06:07 PM skunkworks: ok - you want to unhook that for testing
06:07 PM skunkworks: so you can set it manually with halcmd
06:08 PM chopper79: Can I just comment it out?
06:08 PM skunkworks: yes
06:08 PM chopper79: ok and in halcmd what is done to enter the command needed
06:08 PM pcw_mesa: its easier to just sets the signal that carries index-enable
06:09 PM chopper79: only used halcmd twice and I do not remeber much about it
06:09 PM pcw_mesa: (no commenting/uncommenting)
06:10 PM pcw_mesa: so if you have not changed anything, you can set index-enable with
06:10 PM pcw_mesa: halcmd sets spindle-index-enable
06:10 PM chopper79: so in halcmd I would enter setp hm2xxxxxxx.xxx.0. encoder 05 index-enable?
06:10 PM pcw_mesa: no that the hard way
06:10 PM chopper79: lol
06:11 PM pcw_mesa: you need to unhook it to do that
06:11 PM chopper79: setp spindle-index-enable?
06:11 PM pcw_mesa: no, sets
06:11 PM pcw_mesa: (set signal)
06:11 PM chopper79: sets spindle-index-enable
06:11 PM chopper79: do I use an =1
06:11 PM pcw_mesa: or true
06:11 PM chopper79: to set high
06:11 PM skunkworks: pcw_mesa, thank you
06:11 PM chopper79: ok
06:12 PM skunkworks: you can set a signal on a i/o pin? I didn't know that
06:12 PM chopper79: so now I am setup on that what is next step?
06:12 PM chopper79: run it at 3krpm and see if state changes?
06:12 PM pcw_mesa: on bidirectional pins, yes
06:13 PM skunkworks: cool
06:13 PM skunkworks: learn something new (or re-learn..)
06:13 PM pcw_mesa: a 3k RPM is will just stay low as far as you can tell
06:13 PM chopper79: ok so what rpm should I test at?
06:14 PM pcw_mesa: why do you think it does not work?
06:15 PM chopper79: The index input state changes rapidly at 3k rpm when reading the teeth of gear with index sensor. Then when I put it on the index target it will not read anymore.
06:15 PM chopper79: The index inut will only change states rapidly when on the index target at 25rpm or lower
06:16 PM skunkworks: he has no way of testing it other than halmeter
06:16 PM chopper79: no scope
06:16 PM chopper79: hal scope is broke in my version and no internet to use halmeter
06:16 PM skunkworks: the act of setting it to true - and it staying false - shows that atleast it is seeing it
06:16 PM pcw_mesa: It may be too short to capture reliable at higher speeds
06:16 PM skunkworks: if it stays true after setting it true - it isn't seeing the index
06:17 PM chopper79: Right now it is true and I am getting ready to run it at 3k
06:17 PM chopper79: any objections?
06:17 PM skunkworks: but - a threading cycle and seeing that it tracks probably is the best solution
06:17 PM skunkworks: sans scope
06:17 PM skunkworks: none
06:17 PM pcw_mesa: It should not be true unless you set it true
06:18 PM chopper79: ok...lets see if the state changes form true to false. Yes I set it to true
06:18 PM skunkworks: ok
06:18 PM chopper79: for testing. Then I will change it back to false as default when things work
06:18 PM skunkworks: while it is running - set it to true and see if it also falls to false
06:18 PM skunkworks: (should act like you never set it to true :) )
06:21 PM chopper79: ok it goes false at 3krpm
06:22 PM skunkworks: try 6k :)
06:22 PM CaptHindsight: http://imgur.com/a/sUbdH
06:22 PM chopper79: I have to go pick my kid up from the park. Be back in 15min tops
06:22 PM CaptHindsight: http://imgur.com/a/kbEpW
06:23 PM CaptHindsight: pics of his gear and tooth ^^^
06:23 PM chopper79: ignore the mess wiring. Its for testing
06:23 PM Tom_L: the sensor is pointing down?
06:24 PM chopper79: yes
06:24 PM CaptHindsight: I'd use a magnet vs that tooth, you're most likely not getting enough change in flux
06:24 PM Tom_L: that close to the other gear might interfere you think?
06:24 PM CaptHindsight: Tom_L: I agree
06:24 PM Tom_L: i would have mounted it sideways personally
06:24 PM skunkworks: what are you using for a circuit? the sample one in the datasheet?
06:24 PM skunkworks: you need a pull up
06:24 PM skunkworks: well - maybe the mesa card works
06:25 PM CaptHindsight: he can read the teeth just fine on the gear
06:25 PM CaptHindsight: it's that evtra tooth with the sensor in that location/orientation that it's not so goodly working
06:27 PM CaptHindsight: skunkworks: what circuit is he using? (link)
06:27 PM skunkworks: mainly because he doesn't have a way to see that it is working. It is all hardware side. nothing in the computer is fast enough to see it
06:27 PM CaptHindsight: add a one shot
06:28 PM CaptHindsight: how fast can that gear spin before he can't read it?
06:28 PM CaptHindsight: he said that it reads from the teeth on the gear just fine, but I'm not sure to what rpm
06:30 PM CaptHindsight: this would be much easier to debug if it was in front of me :)
06:30 PM skunkworks: loooking
06:30 PM skunkworks: the sensors we used would work up to 10 or 20k iirc
06:30 PM CaptHindsight: now I know how those Mars rover guys must have felt :)
06:31 PM Tom_L: no pressure
06:37 PM skunkworks: file:///home/skunkworks/Downloads/ATS667-Datasheet.pdf
06:37 PM skunkworks: wow - failed at internet
06:37 PM skunkworks: http://www.allegromicro.com/~/media/Files/Datasheets/ATS667-Datasheet.ashx
06:38 PM skunkworks: I think I used the test circuit example values
06:39 PM chopper79: ok I am back
06:40 PM CaptHindsight: ah the IC has AGC
06:40 PM CaptHindsight: so more fun
06:40 PM skunkworks: it works quite well
06:41 PM CaptHindsight: my rubels are on that it's not getting enough change in flux mounted in that orientations
06:41 PM chopper79: I placed a magnet on the gear and the results were the same
06:42 PM chopper79: Trying @ 6k rpm now
06:44 PM chopper79: ok state went form hight to low at 6k
06:44 PM chopper79: So maybe it is working then and I am just a pain in the ass
06:44 PM skunkworks: 90% sure it is working...
06:44 PM skunkworks: ;)
06:44 PM chopper79: So in hal set it to false now correct?
06:45 PM skunkworks: it should be false
06:45 PM chopper79: It was true for testing and in hal it is set to 1
06:45 PM chopper79: so I need to set it to 0 in hal
06:45 PM justanotheruser is now known as unmelanated_iris
06:45 PM skunkworks: no wait
06:46 PM skunkworks: when you set it to 1 while the spindle is running - it should have changed to false
06:46 PM chopper79: Lets start over....
06:46 PM skunkworks: how are you reading the pin?
06:46 PM skunkworks: :)
06:47 PM chopper79: In hal.... setp hm2xxxxxxxxxx index-enable 1
06:47 PM chopper79: This is currently the config.
06:47 PM skunkworks: yes - do a show sig hm2xxxxxx.index-enable
06:48 PM chopper79: My hal is currently set to 1
06:48 PM skunkworks: I bet it is false now (or it should be)
06:48 PM chopper79: which is true
06:48 PM skunkworks: what do you mean hal?
06:49 PM chopper79: actual hal file
06:49 PM skunkworks: oh
06:49 PM chopper79: in my hal file setp hm2xxxxxxxxxx index-enable 1 is current
06:49 PM chopper79: should this be 0 (false)
06:49 PM skunkworks: ok - can you open a terminal
06:50 PM chopper79: so index-enable goes true when index is passed?
06:50 PM chopper79: right now it is set true and when index is passed it goes false.
06:51 PM skunkworks: open a terminal - type halcmd -fk
06:52 PM skunkworks: now you can set and check the pin intractively.. (sets and show sig)
06:52 PM skunkworks: (while linuxnc is running)
06:52 PM skunkworks: because then you can test it at higher rpms - setting it true while it is running at 6krpm and such
06:53 PM skunkworks: (and seeing if it goes false
06:53 PM skunkworks: because if you where setting it in the hal file - it probably went false after the first rotation
06:55 PM chopper791: hey I reconnected
07:00 PM unmelanated_iris is now known as justanotheruser
07:00 PM chopper791: Lets see if I can stay connected this time
07:00 PM chopper791: not sure what I missed if anything.
07:12 PM chopper79: I am having a terriable time staing connected
07:12 PM CaptHindsight: chopper79: you haven't missed anything
07:12 PM malcom2073: Get yourself a bouncer
07:13 PM CaptHindsight: except for you disconnecting
07:13 PM chopper79: ok lets try again....
07:13 PM chopper79: Based on what I have read so far is that when the index is seen the index-enable pin should go high. If that is the case then I will need to change my hal file so index-enable is set to 0. This will enable the index-enable to go high (true) when index is seen.
07:14 PM CaptHindsight: chopper did you say that it reads the teeth on the gear just fine?
07:15 PM chopper79: Not sure what was happeing with my connection... Some cherryirc message kept popping up telling me that i need to verfy to use this server.
07:15 PM pcw_mesa: No index enable is normally low and stays low until set by LinuxCNC
07:15 PM chopper79: Ok so in my hal file I will set index-enable to 0
07:16 PM pcw_mesa: dont set it at all
07:16 PM chopper79: so it will be low and then get set high when a threading cycle is called correct?
07:16 PM chopper79: right now it is 1 in my hal file.
07:16 PM pcw_mesa: dont touch it in your hal file
07:17 PM chopper79: so leave it 1 even though it was set for 0 when I first did the hal file?
07:17 PM pcw_mesa: dont touch it at all
07:18 PM pcw_mesa: or set it low
07:18 PM pcw_mesa: so yes linuxcnc sets it when it starts a threading cycle and waits for it to go low
07:18 PM chopper79: ok so when I use a tapping cycle G84) lcnc knows to set the state so it can see the index and then count if I am understandning this correct
07:18 PM chopper79: got ya
07:18 PM chopper79: does this work with G84 ?
07:19 PM pcw_mesa: yes
07:19 PM pcw_mesa: to allows peck tap cycles
07:19 PM chopper79: ok as I have only seen g33 mentioned
07:21 PM pcw_mesa: So for testing assuming spindle-index-enable is wired normally in the hal file, you can just sets
07:21 PM pcw_mesa: spindle-index-enable and see that it gets cleared when you pass the index
07:21 PM chopper79: I have always tapped using g84 so I was not sure if G84 was setup in lcnc to allow for sync motion. As I said I have only seen some examples of people using g33.1
07:21 PM chopper79: understood
07:21 PM chopper79: about the sets
07:23 PM chopper79: No will tapping require a spindle at speed signal? The reason I ask is that all examples I have read about seems to use it and if I remeber correclty one post stated it was required for tapping.
07:24 PM pcw_mesa: I think so (threading does)
07:25 PM chopper79: ok then I will try to get that connected then
07:27 PM chopper79: It would make sense as you really would not want the tapping cycle to start when the spindle was not at speed.
07:31 PM chopper79: pcw_mesa: skunkwork: CaptHindsight: Thank you for your help.
07:32 PM CaptHindsight: it's all part of the service!
08:36 PM homeroot is now known as evilroot
09:18 PM chopper79: Got spindle at speed working for the most part. The at speed LED in panel seems to flicker a lot. So I set the difference at 3.5 not sure what this value means. Does anybody?
09:19 PM chopper79: near difference
09:19 PM chopper79: The man page I have pulled up does not really go into what it is
09:22 PM chopper79: I believe it is the difference of what is commanded to what is being read. Then is takes this difference into account to allow the at speed signal to work.
09:23 PM {HD}: My CNC computer just died and I have replaced it but my original setup no longer works due to the hardware and mach3 with 64bit and pcie parallel ports. I was going to install linuxCNC ontop of linuxmint but I see it is listed for 64bit as 'similuation only'. Does that mean linux cnc is also not well suited for 64bit?
09:31 PM {HD}: I guess Ill just use the default ISO to see how I get along...
10:25 PM {HD}: Well, thats not my favorite. I guess Ill need to run a network cable to my cnc computer to update linuxcnc. My display drivers were totally wack!
10:25 PM ApocalypseCow is now known as clownshoes
10:29 PM clownshoes is now known as Odo
10:31 PM Odo is now known as LtUhura
11:17 PM Wolf_: idk, I have lcnc on a 64bit mint install
11:18 PM Wolf_: but I haven’t run a machine with it yet
11:28 PM LtUhura is now known as clownshoes
11:34 PM CaptHindsight: 64bit RTAI has worked with LCNC for the past few years when we built it, not sure what is in the official tree or what works with what
11:44 PM CaptHindsight: as far as versions of Debain or Mint etc etc
11:45 PM Wolf_: seems to wrk on mint, just a matter of odd latency depending on what kernel you cook with