#linuxcnc Logs
Mar 19 2017
#linuxcnc Calendar
12:06 AM XXCoder: wo
12:08 AM XXCoder: http://507movements.com/mm_489.html
12:08 AM XXCoder: hm wonder why that was bolded
12:11 AM chopper79: So people could make ferris wheels
12:11 AM XXCoder: most ferris uses rotating seating setup
12:11 AM chopper79: Yeah but this is cooler
12:11 AM XXCoder: this one it dont need to have any, but still needs bearing
12:11 AM chopper79: This design is actuallt pretty efficient
12:11 AM XXCoder: for water pushing yeah
12:11 AM XXCoder: I wonder if I can design and build wheel boat based on that pripicle
12:12 AM chopper79: I do not see why not. It is more efficient paddle wheel
12:12 AM XXCoder: I bet i can bury wheels up to point below where it pushes wrong way
12:12 AM XXCoder: common ones have to move up becaue of pushing dowb and up drag
12:13 AM chopper79: Tis design does not carry the weight of the water since the paddles stay perpindicular to the horizontal (water)
12:13 AM chopper79: Smart people back then
12:13 AM XXCoder: hmm maybe little futher down because paddles is moving slowly at start and end
12:14 AM chopper79: Put a larger moment on there and you could.
12:14 AM XXCoder: maybe it failed becaus4 they didnt have rustproof stuff for bearing that constanually gets dipped in sea water
12:14 AM chopper79: lol..true
12:15 AM chopper79: Add more paddles also... increase dispalcement. Move some water
12:15 AM chopper79: Or increase surface are of the 4 paddles and increase displacment that way
12:16 AM chopper79: Then put a V8 on it for drive motor. Should be good to go ;)
12:16 AM XXCoder: lol
12:20 AM MacGalempsy: malcom2073: any insight on running a terminal command at startup?
12:22 AM XXCoder: https://www.youtube.com/watch?v=_BhmWRO2k1E nicde
12:34 AM XXCoder: lol myfordboy made a oopie https://www.youtube.com/watch?v=V5I5sLqv6u8
12:47 AM chopper79: Always wanted to do this with all my aluminum chips and chunks.
12:52 AM chopper79: Make castings that is
12:53 AM chopper79: Well I am off
12:53 AM chopper79: Have a good one all
12:53 AM XXCoder: later
12:53 AM XXCoder: yeah wanna cast em also but also dunno
12:54 AM chopper79: I may try it one day... majority of my jobs are all 6061 and I could use the chips and chucks to melt down and cast something.
12:55 AM chopper79: ok... I have to go or I am gonna get on a roll talking about stuff. Later
01:03 AM XXCoder: lol later
02:02 AM yasnak: anyone use autodesk featurecam?
03:32 AM Deejay: moin
04:08 AM MacGalempsy: hi dj
04:09 AM MacGalempsy: slow night, everyone must be recovering from St Patty's day
04:09 AM XXCoder: not me
04:09 AM XXCoder: https://youtu.be/Q4sSjd9UYpM lol apparently they remade classic games
04:09 AM MacGalempsy: me neither.
04:09 AM MacGalempsy: i quit the drink as a ny resolution
04:10 AM XXCoder: im not recovering, im dying
04:10 AM XXCoder: sorry youre same :(
04:10 AM MacGalempsy: i feel good, been lifting weights and eating right
04:11 AM mrpackethead: hello
04:12 AM XXCoder: heh guess joke failed
04:12 AM XXCoder: hey
04:15 AM MacGalempsy: booted.
04:36 AM XXCoder: hmm
04:36 AM XXCoder: Im wondering if its worth it to get dust collector
04:46 AM IchGucksLive: good morning from Germany
04:46 AM XXCoder: hey ich
04:46 AM IchGucksLive: https://youtu.be/5h5RUMfWd1E
04:46 AM XXCoder: I used estlcam to make a program and ran it
04:46 AM IchGucksLive: a quick on request gmocappy
04:47 AM XXCoder: it did decent, it was my own mistakes that made project fail lol
04:47 AM IchGucksLive: so what failed
04:47 AM XXCoder: https://www.youtube.com/watch?v=oQUSS9FO9kY
04:47 AM XXCoder: stock turned
04:48 AM XXCoder: you can see bunch of problems with fixture
04:49 AM IchGucksLive: fixure is real issue
04:49 AM XXCoder: it is.
04:49 AM IchGucksLive: even on chees mashine
04:49 AM IchGucksLive: this is the biggest problem in education
04:50 AM XXCoder: I still need bunch of stuff for my machine
04:50 AM XXCoder: like proper clamps
04:50 AM IchGucksLive: as they want to put 100lbs 5 sqft on a 10inch mashine
04:51 AM IchGucksLive: i realy did see a loose part staying and the mashine moved on table
04:51 AM XXCoder: ouch
04:52 AM IchGucksLive: the thinkiing of our giv is to put small mashies in the schools
04:52 AM IchGucksLive: but as they see wat 3 axis can do
04:52 AM IchGucksLive: they go grasy
04:52 AM XXCoder: grasy?
04:52 AM IchGucksLive: over head
04:52 AM XXCoder: ahh
04:53 AM IchGucksLive: and thats not the thing it is suposed to do
04:53 AM IchGucksLive: you may get your own on this
04:53 AM IchGucksLive: people seeing yust a bit of a mashine and thinking of there own big use
04:55 AM IchGucksLive: XXCoder: to your fixiing you did not folow rule 2
04:55 AM IchGucksLive: dont make the fixing one axis parallel
04:55 AM XXCoder: well its just learning stuff heh. at work we already has special fixtures and such. easy.
04:56 AM XXCoder: yah I thought so, didnt have anything long enough to reach Y axis part
04:56 AM IchGucksLive: do ist with at least 1/3rd od part diagonal
04:56 AM IchGucksLive: that woudt do it
04:56 AM XXCoder: like i said, machine really needs more stuff, like clamps
04:56 AM XXCoder: you know a good source of clamps?
04:56 AM IchGucksLive: i go mine
04:57 AM IchGucksLive: http://www.paulimot.de/spannmittel/
04:57 AM IchGucksLive: but there also on the net
04:57 AM XXCoder: hmm ok
04:58 AM XXCoder: maybe aliexpress heh
04:59 AM IchGucksLive: cand find a english translation on this
05:00 AM XXCoder: interesting https://www.aliexpress.com/item/Factory-Supply-65mm-Spindle-Diameter-Auto-Pressure-Foot-Fixture-Holder-for-CNC-Router-DIY-Accessories-CNC/32461888030.html
05:00 AM XXCoder: dont need it, but interesting
05:03 AM IchGucksLive: god vor PVC
05:03 AM IchGucksLive: but on plasma it will not work
05:03 AM IchGucksLive: as parts turn over and out on inlays
05:04 AM IchGucksLive: ok grand children arived i close down for now
05:04 AM IchGucksLive: have a nice Day
05:04 AM XXCoder: later
05:57 AM MacGalempsy: sooner
05:57 AM MacGalempsy: t minus 33 min
05:57 AM XXCoder: soonest
06:11 AM jthornton: morning
06:12 AM XXCoder: hey jt
06:13 AM jthornton: I had to relearn how to count in hex yesterday
06:15 AM XXCoder: why?
06:15 AM XXCoder: did the 6 sided overlords take over?
06:16 AM jthornton: the click plc does not have bit mask so in order to control 6 outputs with some clean code I used the copy instruction which uses hex to copy to binary and the binary turns the bits off and on
06:17 AM jthornton: hex is base 16 123456789ABCDEF
06:17 AM XXCoder: no, hex is 6 :) you mean hexadecimal which is 16 :)
06:17 AM jthornton: F = 16 10 = 17 etc
06:18 AM jthornton: well we tend to shorten things ya
06:18 AM XXCoder: :) but yeah cs here, I do know about that
06:24 AM XXCoder: jthornton: whats you working on
06:24 AM jthornton: the cover tester
06:24 AM jthornton: plc
06:24 AM XXCoder: ah yeah
06:30 AM XXCoder: you know of good source of clamps?
06:30 AM XXCoder: cnc rputer clamps so on
06:33 AM jthornton: make some out of hard wood
06:33 AM jthornton: if you run into one what the heck you just cut your way through
06:33 AM XXCoder: yea was thinking that, but wanted to buy lol
06:34 AM XXCoder: and making clamps mean I have to use shitty clampo setup lol
06:34 AM jthornton: only once lol
07:02 AM XXCoder: wow
07:02 AM XXCoder: I suppose can do that
07:02 AM XXCoder: anyway wow is for this https://hackaday.com/2017/03/17/are-you-down-with-mppt-yeah-you-know-me/
07:10 AM R2E4_bevins2: mornin
07:10 AM jthornton: yo
07:10 AM R2E4_bevins2: hey, M6 remap is coming along. hopeful;ly today I can auto change tools
07:11 AM jthornton: nice
07:11 AM R2E4_bevins2: Theres alot to learn...... WOW
07:11 AM jthornton: if you have any suggestions to the manual I'm all ears
07:11 AM * archivist has toolchanger on the end of his arms
07:11 AM R2E4_bevins2: I left it last night with the digitalins all returning 0's everywhere
07:12 AM R2E4_bevins2: then I saw last night have to poll it....lol
07:12 AM R2E4_bevins2: archivist, now thats a random toolchanger....lol
07:12 AM archivist: :)
07:13 AM archivist: it can even pick up a file
07:13 AM R2E4_bevins2: jthornton, , when I am done I will come up with some pointers
07:13 AM jthornton: ok great
07:15 AM R2E4_bevins2: python is very touchy with the indentation....... took me a while to figure it out...
07:16 AM jthornton: ah yea it is hmmm what do they call that hard indentation or something
07:17 AM R2E4_bevins2: I didnt think it was going to be this tough. I am at 600+ lines
07:17 AM jthornton: dang that is big
07:18 AM R2E4_bevins2: but I am not a programmer, so I am probably doing things not 100% proper
07:18 AM XXCoder: 600 lines for tool changer?
07:18 AM R2E4_bevins2: remapping M6
07:18 AM R2E4_bevins2: for toolchanger
07:18 AM archivist: erm, you could be doing something wrong at 600 lines
07:19 AM R2E4_bevins2: I probably am, but...... It is all in python, and there are alot of inits and all the error checking
07:20 AM R2E4_bevins2: theres also 49 digital outs and 18 digital ins
07:20 AM R2E4_bevins2: pluis dealing with three spindles
07:23 AM R2E4_bevins2: 725 lines in remap.py, but there is alot of comments
07:25 AM XXCoder: and novel or 2
07:25 AM R2E4_bevins2: http://pastebin.com/PffL7nXj
07:26 AM XXCoder: some vry specific code
07:27 AM XXCoder: like one for t11
07:27 AM XXCoder: and ranges 400-499 and 500+
07:27 AM R2E4_bevins2: http://pastebin.com/PerJsBvH this is the xml stuff.
07:28 AM R2E4_bevins2: I created xml file, parsed it, and using it with remap. XML is easy for me, plus it created my hal and it works....
07:30 AM R2E4_bevins2: http://pastebin.com/RrrFSFGi This is the xml file.... This may not be the best way to do it, but it works for me.....
07:31 AM XXCoder: interesting
07:32 AM R2E4_bevins2: http://pastebin.com/8TXNGnHZ This is my hal created by my xml to hal converter
07:33 AM XXCoder: wow
07:39 AM R2E4_bevins2: xml to hal converter http://pastebin.com/uvsfwcyf
07:39 AM R2E4_bevins2: so working on getting polling working this morning.
07:41 AM jthornton: looks pretty neat
07:42 AM R2E4_bevins2: once polling is done, have to get execute working...... then the axis move commands, those werent working yesterday for some reason
07:46 AM R2E4_bevins2: jthornton, thanks for the or2 suggestion yesterday. that worked great
07:47 AM R2E4_bevins2: once I got my head wrapped around using CL output for an input to the or2
07:49 AM jthornton: no problem, most of the time just knowing where to look is 90%
07:55 AM R2E4_bevins2: how often should the eroor channel be polled?
08:00 AM jthornton: as often as you like I guess
08:25 AM XXCoder: wow. https://translate.google.com/translate?sl=auto&tl=en&u=http%3A%2F%2Fwww.brikawood-ecologie.fr%2F
08:26 AM XXCoder: house bricks
08:32 AM SpeedEvil: XXCoder: yeah - saw that
08:33 AM SpeedEvil: XXCoder: there was an interesting thing on a related concept using plywood, which annoyingly was taken down by the creator.
08:33 AM SpeedEvil: Long story short, you take a full-sheet plywood router, and use it to make boxes with properly finished edges and stuff which clip together into a 3d house with insulated wallspace that is filled with insulation
08:36 AM SpeedEvil: https://vimeo.com/53932758 @16:20
08:36 AM SpeedEvil: (no audio needed)
08:37 AM SpeedEvil: 16:20-17:00 basically explains it all
08:49 AM R2E4_bevins1: i.net
09:06 AM malcom2073: That's an awful lot of plywood
09:07 AM malcom2073: I wonder how it compares pricewise
09:10 AM malcom2073: I vaugly recall this, I think I saw it when it was on TV, HGTV maybe?
09:20 AM Wolf_: thats pretty neat, but, lot of work to make a house that looks like it was made from shipping containers (and would have been cheaper and faster to stack/assemble lol)
09:37 AM R2E4_bevins1: self.command.mdi("G53 F%d X%f " % (self.feedSpeed, posX))
09:38 AM R2E4_bevins1: oops sorry
09:52 AM Jymmm: SpeedEvil: It's not a bad looking house at all. Not sure what the whole cubes were all about (alternative to stick built?); "simpler" construction? Less raw material waste? Plywood aint cheap these days, wonder what cost more ply vs stick?
09:52 AM SpeedEvil: yes, alternative to stick built. I can see it in principle has benefits with air-tightness and precision.
10:40 AM Jymmm: SpeedEvil: Yeah, okey. But mobile homes are already air tight, and they are toothpick built =) plywood boxes, ok, I guess, does seem like a bit of a waste, in a way, but there may be benefits that I'm unaware of.
10:47 AM jst_ is now known as jst
11:02 AM R2E4_bevins2: JT-Shop, one of the issues I had was figuring out how to deal with the DIO, sample code would be nice
11:03 AM JT-Shop: like the input example on this page ? http://gnipsel.com/linuxcnc/configs/7i77.html
11:05 AM R2E4_bevins2: in python code motion.digital IO
11:10 AM JT-Shop: not making sense python with motion.digital
11:11 AM R2E4_bevins2: I had a hard time trying to find out how to set outputs within python.
11:11 AM JT-Shop: are you talking about set_digital_output?
11:12 AM JT-Shop: the description is a bit terse...
11:13 AM R2E4_bevins2: There is alot of code examples for this, but not purely in python. I had to trial and error, and dig and finally got it. but it was tough finding it
11:56 AM R2E4_bevins2: I want to make sure all drills are up, there are 32 of them, so I send a command to raise all the drills. send a 0 to the motion.didital-outs, problem is it is taking about 5 secs for each one and I cannot find where the delay is.
11:58 AM archivist: air?
11:59 AM zeeshan: R2E4_bevins1: you could use my code as a ref!
11:59 AM R2E4_bevins2: where is your code?
11:59 AM zeeshan: http://wiki.linuxcnc.org/uploads/vfd.py
12:00 PM zeeshan: basically you setup pins like shown at the top
12:00 PM zeeshan: this is where you select the data type too
12:01 PM zeeshan: also to answer your question about the delay
12:02 PM zeeshan: i've found userspace components are slow
12:02 PM zeeshan: time.sleep(0.01)
12:02 PM zeeshan: i've tried changing that to be lower values
12:02 PM zeeshan: and i can still notice a delay
12:02 PM R2E4_bevins2: got all that stuff working, it just takes forever, I changed it to only raise the drills if they are down
12:03 PM R2E4_bevins2: it was taking up to 5 seconds to raise....
12:03 PM R2E4_bevins2: archivist, yes it is air valves
12:08 PM R2E4_bevins2: I'm not creating new pins, I am using signals
12:09 PM R2E4_bevins2: then accessing the digital ins/outs accociated with the signals
12:11 PM zeeshan: ah
12:14 PM archivist: is there enough air to shift them or is the software serial waiting for a previous one to get there
12:15 PM archivist: time for a scope
12:23 PM R2E4_bevins2: yes there is enough air
12:25 PM R2E4_bevins2: I was watching it in the terminal, I was printing it each one, and it was taking 5 seconds for Lcnc to send the command
12:26 PM R2E4_bevins2: I since changed it to only send the command if the drill is down, so it is not just sending all
12:27 PM archivist: you mean your code not lcnc :)
12:29 PM archivist: use halscope to see
12:31 PM R2E4_bevins2: yeah, your probably right
12:35 PM archivist: R2E4_bevins1, one thing I must say, you are using automation to create your hal, this can put stuff in the wrong order thus causing signals to be at a previous state not the state you expect
12:36 PM R2E4_bevins2: only creating the hal for the toolchange. nothing else
12:37 PM archivist: and your unexpected delay is in the toolchange
12:38 PM R2E4_bevins2: yes, but one thing funny, is when lcnc starts it stats the remap code for some reason
12:39 PM R2E4_bevins2: the file that is loaded has the first line as T1 M6 though, but why does it start when the machine is not even on?
12:40 PM archivist: I know nothing about the remap stuff at all
12:43 PM R2E4_bevins2: The hal for the tool changer is only mapping the pins and motion.digital-in/out, and a few or2's so should have nothing to do with delays.
12:43 PM R2E4_bevins2: no matter what order I put them in
12:45 PM archivist: it does, as they are seen once in order they are seen in the file during each period
12:46 PM Roguish_ is now known as Roguish
12:47 PM archivist: so if one depends on one further up the file the result wont be seen till next cycle
12:50 PM R2E4_bevins2: so looping through and sending the command would create that delay? i would have thought not.
01:02 PM zeeshan: when you choose "run from here"
01:02 PM zeeshan: is this a modal restart?
01:02 PM zeeshan: the only thing i need to do is mdi the speed?
01:08 PM R2E4_bevins2: zeeshan, you vfd code doesnt stop?
01:08 PM zeeshan: what do you mean doesnt stop
01:09 PM zeeshan: it loops infinitely
01:09 PM zeeshan: to allow communication
01:18 PM R2E4_bevins2: ok, thats what i thought
01:23 PM JT-Shop: my stinking windows pc keeps locking up and the speakers start to buzz...
01:24 PM R2E4_bevins2: zeeshan, what vfd do you have?
01:33 PM IchGucksLive: hi
01:35 PM IchGucksLive: R2E4_bevins
01:35 PM R2E4_bevins2: yeah
01:35 PM R2E4_bevins2: OK, I am stuck and need help.
01:36 PM IchGucksLive: there is a real cool way on remap toi change with 3 spindles on one mashine
01:36 PM R2E4_bevins2: when staring Linuxcnc, the program is running the remap and of course it wont work because it is in estop and the machine is not on.
01:37 PM R2E4_bevins2: IchGucksLive, Can you elaborate a bit more?
02:15 PM IchGucksLive: Gn8
02:33 PM {HD}: My thread gauge reads "10G 3/4". What kind of thread is that?
02:33 PM {HD}: I am trying to calculate TPI
02:37 PM CaptHindsight: http://electronicdesign.com/analog/what-s-all-brick-stuff-anyhow
02:44 PM {HD}: I don't know why I just read that.
02:44 PM CaptHindsight: one of the analog greats
02:47 PM CaptHindsight: {HD}: https://en.wikipedia.org/wiki/British_Standard_Whitworth
02:49 PM cncnoob: Any CamWorks users out there?
02:50 PM cncnoob: I've got a problem with cutter radius compensation
02:50 PM roycroft: i found another potential lathe candidate, up by seattle
02:50 PM roycroft: maybe this time i'll be the first in the queue, not the second :)
02:51 PM cncnoob: CamWorks is generating G41 D44, but the tool loaded is T24
02:51 PM cncnoob: I can change all the D44 to D24 and the program works
02:52 PM cncnoob: I'm unclear how to change CamWorks to make it use the right tool
02:52 PM cncnoob: I actually don't even have a tool #44
02:53 PM {HD}: CaptHindsight: guess I still don't quite understand. I did not measure thread diameter only thread pitch.
02:58 PM {HD}: Its 3/8" diameter and my thread gauge reads "10G 3/4" which I assume is 2 different standards that happen to line up.
03:22 PM R2E4_bevins2: stuck again...lol cnat get execute to work.
03:33 PM os1r1s: Why would a G53 Z0 move cause linuxcnc to tell me its going to exceed the positive limit for Z?
03:40 PM cradek: what are Z's limits?
03:41 PM os1r1s: Min is -8, max is 0
03:41 PM cradek: maybe you should make max +0.0001
03:42 PM os1r1s: cradek Is that normal to have to do?
03:43 PM cradek: I think you're not the first
03:43 PM gregcnc: yes i think so, there is no reason to try to use every last micron of travels
03:44 PM cradek: looks like lots of our sample configs have that
03:44 PM cradek: also numbers like 10.0001 etc
03:44 PM os1r1s: cradek I've adjusted it. I'll try to run the job again.
03:58 PM R2E4_bevins2: I am trying to use the execute within a python function during a remap, and anything I try fails. The axis just dowsnt move
03:59 PM {HD}: http://i.imgur.com/GLQSpp1.jpg where do I get another one of these?
04:03 PM cradek: dumpstercnc.com has a lot of similar delrin zerobacklash nuts
04:03 PM R2E4_bevins2: self.execute is not moving the axis's
04:04 PM {HD}: cradek: Thanks, I found this. http://dumpstercnc.com/images/3_8-12_ACME.jpg at least I know my search terms now.
04:04 PM cradek: if you get the right thread and are willing to machine
04:05 PM cradek: to machine/modify the flange you'll be set
04:06 PM cradek: I have a set of his on my desktop mill and they have lasted a shockingly long time - ten years maybe it's been
04:06 PM {HD}: cradek: Wow
04:07 PM {HD}: I think this is what I have https://www.mcmaster.com/#6350K53 but, I hate the pricing on mcmaster! they are so damned expensive!
04:08 PM JT-Shop: they are actually cheaper than some others
04:08 PM cradek: huh I didn't know they had those
04:09 PM {HD}: everything on ebay is metric...I need 3/8"-10
04:14 PM {HD}: I could buy a new metric lead screw w/anti backlash nut from ebay/china for half of just the nut from mcmaster...
04:14 PM cradek: I bet that'd be false economy
04:15 PM chopper79: Gear sensors came in and hooked a set up. I am reading spindle RPM but it is ot stable. The fluctuation is from 890 - 1164 when commanded RPM of 1000 is given. Anybody have nay pointers on what could be wrong or how to smooth it out?
04:17 PM R2E4_bevins2: Can I send a self.execute while in manual mode within a python remap?
04:17 PM R2E4_bevins2: self.execute("G53 F%d X%f Y%f Z0.0" % (10, 20.0, -9.0), lineno())
04:17 PM chopper79: If I bring the RPM up to 3000 then the my reading is between 43rpm - 3800rpm. Very unstable.
04:17 PM enleth: heh, I just had a ball nut accident, tried to slide the nut off the screw for a moment but the tube I used was a tiny bit too thin and it lost two balls in the process, they just fell out after a few revolutions back on the screw
04:18 PM R2E4_bevins2: do I need a G1?
04:18 PM enleth: I put them back in just fine, but now I'm wondering, as I had to open it anyway, I might just as well pull all the balls out to clean out old grease that's stuck inside the nut. I've been flushing it with a solvent a couple times and moving it back and forth over the screw before, but there's still a lot of old grease inside.
04:19 PM enleth: do you think it's worth the additional effort at this point? the old grease did almost solidify in a couple places on the screw, but it's much less exposed to the air inside the nut, so it might still be OK-ish
04:19 PM pcw_home: chopper79: take a look at your quadrature signals in halscope at low speed
04:19 PM pcw_home: they probably have significant quadrature error
04:20 PM pcw_home: so may require adjusting the sensors
04:20 PM chopper79: Ok.. I will do that. One second.
04:21 PM {HD}: enleth: Do it, clean it out. Now is as good a time as any.
04:21 PM {HD}: chopper79: I would try to help but I am notfimilare with your setup or what parts your running...
04:22 PM pcw_home: also if these are slow signals, make sure the quadrature filter is on
04:22 PM pcw_home: and perhaps lower the quadrature counter sample frequency
04:22 PM chopper79: halscope will not launch.
04:22 PM chopper79: Never had this issue before
04:22 PM enleth: {HD}: yeah, I figured as much. just wondering how likely it for the balls in those to not be all identical, for preload or whatnot
04:23 PM enleth: and if they aren't identical, how hard it is to figure the proper order they go in
04:24 PM enleth: the two that fell out were both 3.155mm according to a Mitutoyo mike I trust to be genuine
04:24 PM pcw_home: chopper79: what OS?
04:24 PM chopper79: {HD}: Using Allegro gear tooth sensors (magnetic) on a 1018 steel gear.
04:24 PM chopper79: Mint
04:25 PM chopper79: Tried in terminal and came back with a rtapi shem shared meemory error
04:25 PM enleth: (the mike might not have been zeroed accurately, but I couldn't be bothered to do that now, it's the difference that matters here anyway)
04:26 PM pcw_home: chopper79: This (mint 18/Ubuntu16.xx) permission issue is fixed in master (thanks Jepler!)
04:26 PM chopper79: shmem error due to shmget
04:26 PM chopper79: shoot... no internet connection to the machine.
04:27 PM pcw_home: its ok with Mint17/Ubuntu 14.xx/Jessie
04:27 PM {HD}: enleth: Mine are all identical size
04:29 PM gregcnc: enleth it really isn't a lot of work, I've done it many times.
04:29 PM pcw_home: also image-to-gcode is broken on mint 18/Ubuntu16.xx (not fixed yet)
04:29 PM chopper79: Well I guess I need to get internet to it somehow now.
04:31 PM chopper79: Yeah I am using Mint 18.1
04:35 PM {HD}: You guys have linuxcnc install on top mint?
04:35 PM {HD}: I like mint better than the default debian so, I might try that.
04:55 PM {HD}: Is there a way to tell good chinese bearings and ballscrews from bad ones?
04:56 PM R2E4_bevins2: I am trying to move an axis in a python remap but the interpreter is ignoring it it seems. It doesnt give me an error, it just ignores it.
05:02 PM SpeedEvil: {HD}: load and cycle them till they fail.
05:02 PM SpeedEvil: {HD}: price can help - but fundamentally if you're buying from aliexpress, you're hoping that they have a decent quality control system, and it's working that week.
05:08 PM chopper79: {HD}: it is a crap shoot. I have purchased large amounts of bearings form china and learned my lesson. Some were good, some ok, most just bad. I have tried 4 different bearing companies over there and results have been the same.
05:09 PM chopper79: So I now purchase from the USA so if and when there is an issue it is handled quickly and communication is good.
05:10 PM chopper79: cost more but do not have the issues I did before.
05:10 PM chopper79: Bearings are Hiwin, but serviced state side.
05:13 PM {HD}: Ill look into that. I have to go Chinese because they are a 10th of the cost of what I'm seeing available in the US.
05:14 PM chopper79: I understand... I just learned my lesson on using them for my product. Spend the money once and my customers are happy.
05:16 PM XXCoder: SpeedEvil: interesting video
05:28 PM {HD}: XXCoder: What video?
05:28 PM XXCoder: oh a second
05:28 PM XXCoder: https://vimeo.com/53932758
05:29 PM Deejay: gn8
05:34 PM R2E4_bevins2: Anyone know why the interpreter would ignore an execute and does not spit an error?
05:36 PM SpeedEvil: {HD}: on plywood houses, sane bits start at 16:30
05:37 PM XXCoder: WTF shaq thinks earth is flat
05:37 PM XXCoder: might be trolling but then dunno
05:37 PM XXCoder: Have that dude ever taken a plane ride?
05:39 PM R2E4_bevins2: ok for mdi
05:39 PM R2E4_bevins2: Waiting for completion next
05:39 PM R2E4_bevins2: Completed, send the MDI command G1 G53 F10 X15.375000 Y-41.545000 Z0.0
05:39 PM R2E4_bevins2: after the if ok for mdi
05:40 PM R2E4_bevins2: ITs ignoring the moves
05:40 PM R2E4_bevins2: and it didnt complain
05:40 PM R2E4_bevins2: smae thing if I use the self.execute
05:43 PM enleth: ah, shit. I almost finished reballing one of the nuts and only now realized that I'm out of the regular machine grease
05:45 PM enleth: as far as I can tell, the nuts were packed with that generic slightly translucent amber stuff you get when you ask for "machine grease"
05:45 PM SpeedEvil: Bacon grease.
05:45 PM enleth: smelled the same
05:45 PM SpeedEvil: (not really)
05:45 PM enleth: all I have at the moment is pure vaseline (not vaseline oil) and white lithium grease
05:46 PM os1r1s: cradek That fixed it
05:46 PM os1r1s: Makes no sense to me
05:46 PM os1r1s: But it fixed it
05:46 PM enleth: and I have absolutely no idea whether while lithium is OK for a ball screw
05:46 PM enleth: (I assume vaseline isn't)
05:47 PM enleth: technically the generic amber stuff is lithium-based
05:47 PM SpeedEvil: My more serious uneducated understanding is that vaseline is 'OK' - but will simply melt at about 40C or so, and run out.
05:48 PM SpeedEvil: It is not as good a grease, but it's a lot better than nothing.
05:50 PM enleth: OK, the internets seem to claim that white lithium with no weird additives is fine
05:55 PM enleth: I found that at the bottom of the greases box: http://lucasoil.com/products/grease/white-lithium-grease
05:55 PM enleth: that's the best stuff I have
05:57 PM enleth: at least one guy somewhere out there is using that on his lathe's ballscrews and claims it's fine, works for me
05:58 PM SpeedEvil: To a large degree, pretty much anything will work.
05:58 PM SpeedEvil: Certain lubricants might reduce wear or friction especially at high loads and rpm
06:14 PM mrpackethead: any one have a suggestion on edge-finders
06:14 PM mrpackethead: using a DTI takes so much time
06:15 PM mrpackethead: but the typical edge finders are not accurate enough
06:15 PM XXCoder: dunno my starret edge finder always works fine
06:15 PM XXCoder: and its all of 20 bucks
06:15 PM mrpackethead: i want to get down to abotu 0.005mm or so
06:16 PM XXCoder: sounds like probe land
06:16 PM mrpackethead: i cna do it with the DTI
06:16 PM mrpackethead: but it takes 20 minutes or so!
06:16 PM mrpackethead: i was contemplating getting a 3d taster
06:17 PM mrpackethead: but they are spendy $$$$$
06:18 PM chopper79: enleth: I know of somebody who is using that on their ballscrews.
06:18 PM chopper79: Not me
06:19 PM XXCoder: well pocket, it is true that if you want to move . left, price and time moves . to right ;)
06:19 PM os1r1s: mrpackethead I just picked up a haimer zero master that I've been using
06:20 PM os1r1s: Are you talking about that or a probe?
06:20 PM os1r1s: XXCoder cradek fixed my issue
06:20 PM XXCoder: yeah what was it
06:21 PM os1r1s: XXCoder The min limit on x/y needed to be -0.0001 and max iimit on z needed to be 0.0001
06:21 PM XXCoder: ah so it can go to exactly 9
06:21 PM XXCoder: err 0
06:21 PM os1r1s: Yeah
06:21 PM XXCoder: and max at whatever max it is
06:22 PM os1r1s: Seems silly, but I'm just happy it works
06:23 PM mrpackethead: whats a zero master?
06:23 PM XXCoder: os yea
06:24 PM mrpackethead: oh yeah.. "3d taster by anothe rname
06:24 PM os1r1s: mrpackethead https://www.haimer-usa.com/products/measuring-instruments/sensors/3d-sensor/zero-master/zero-master-with-integrated-short-adapter.html
06:24 PM mrpackethead: hows it working out
06:24 PM os1r1s: Actually very well
06:24 PM mrpackethead: i like that it only has a 10mm shank
06:24 PM os1r1s: Most repeatable X/Y I've ever head
06:24 PM os1r1s: had
06:24 PM XXCoder: 0.01mm
06:25 PM XXCoder: yeeah that probe way too big for my machine lol
06:25 PM mrpackethead: do they have a metric one
06:25 PM XXCoder: if I set z to 0 and remove probe, I MIGHT fit it in lol
06:25 PM os1r1s: mrpackethead yah
06:25 PM os1r1s: mrpackethead Its a German company
06:25 PM XXCoder: assuming it can be removed from cat40 toolholder of course
06:25 PM mrpackethead: :-)
06:26 PM mrpackethead: i have a converted Sieg SX3
06:26 PM XXCoder: *probe point
06:26 PM mrpackethead: i'm getting to a point where i'm thinking its time for an upgrade
06:26 PM os1r1s: XXCoder Squaring metal takes forever ...
06:26 PM XXCoder: lol
06:27 PM XXCoder: I made tap holder, squaring and sizing it took a while!
06:27 PM XXCoder: still have it, still unused :P
06:28 PM os1r1s: mrpackethead Mine is a PM25MV
06:30 PM mrpackethead: nice.
06:30 PM mrpackethead: i mostly hav ebeen using mine to make injection moulding cavities for low temp/pressure overmoulds
06:30 PM XXCoder: know any probes that is say a inch, from top to end of point?
06:31 PM sync_: the renishaw tp20 is quite small
06:32 PM mrpackethead: so, thats about 94mm long
06:32 PM mrpackethead: i can deal with that
06:32 PM XXCoder: stylus 10mm
06:32 PM enleth: OK, so there were 68 balls in each nut, I reloaded on of them, and it's suddenly got a lot of axial and radial play, I'm pretty sure there wasn't any I can feel before
06:32 PM enleth: I removed the balls again and counted them, all accounted for
06:32 PM enleth: any suggestions?
06:33 PM enleth: *one of them
06:33 PM XXCoder: maybe it orginially had too many
06:34 PM XXCoder: http://www.renishaw.com/cmmsupport/knowledgebase/media/img/gen/eb1bd02712244c249884f14a433aa2f0.jpg HMMM
06:34 PM XXCoder: guessing from chart its approx 20mm
06:35 PM XXCoder: plus attachment of another 20mm
06:35 PM XXCoder: thats almost 2 inch
06:40 PM mrpackethead: i'd love to have some of those quick change tool holders
06:45 PM XXCoder: "just for you, world's first smart ukelele
06:46 PM XXCoder: they dont know me. :P
06:47 PM chopper79: Down to a 20rpm difference from commanded at 3000rpm and 35rpm at 6000rpm.
06:48 PM chopper79: Is that ok for 80 tooth gear and gear sensors?
06:48 PM chopper79: Should reword that
06:49 PM chopper79: The RPM will fluctuate by 20rpm when at 3000rpm and will fluctuate by 35rpm when at 6000rpm.
06:50 PM XXCoder: dunno what accuracy expections are for rpm detection
06:50 PM XXCoder: I need to setup mine
06:51 PM chopper79: Always lower then commanded. So at 3000rpm the rpm will read from 2970 -3000
06:52 PM chopper79: The gain seams off also which is a simple fix as when at 6000rpm the rpm reading is about 5910rpm
06:56 PM chopper79: Basically I have a RPM fluctuation between 20 - 40 rpm at any given requested rpm. That is the best I can get the sensors dialed in.
06:56 PM chopper79: ignore the gain issue
06:57 PM XXCoder: you use some sort of pid to control speed>?
06:57 PM chopper79: How much fluctuation should one see using hall sensors and an 80 tooth gear? The ppr is only 320 so the resolution of the "encoder" is low anyway.
06:58 PM chopper79: All PID are set to zero in .ini
06:58 PM XXCoder: its not currently adjusting rpm by feedback?
06:59 PM chopper79: I will post that section of code give me a couple
07:03 PM chopper79: https://paste.ubuntu.com/24212125/
07:03 PM chopper79: Put hal stuff and then at bottom ini stuff
07:04 PM chopper79: right now I am just reading the rpm correct and not adjusting for what is commanded?
07:05 PM XXCoder: reading
07:06 PM mrpackethead: i need to get a camera for my mill
07:06 PM mrpackethead: so i can watch it from my desk
07:06 PM XXCoder: interesting
07:06 PM XXCoder: pid is all zeros so I guess yeah it is
07:07 PM chopper79: so it is adjusting for rpm difference?
07:07 PM XXCoder: "deadband" interesting
07:07 PM chopper79: just with pid set zero it has nothing to adjust
07:07 PM XXCoder: no, im saying its not doing anything atm
07:07 PM chopper79: ok
07:07 PM XXCoder: you ever seen pid video?
07:08 PM XXCoder: if not youre in for a treat
07:09 PM chopper79: Is the ini and hal at least setup to adjust for rpm based on feedback or am I missing something to do that?
07:10 PM XXCoder: im still hal newbie unfortunately
07:10 PM XXCoder: chopper79: https://www.youtube.com/watch?v=0vqWyramGy8 watch it :)
07:17 PM XXCoder: hey andypugh how goes your project
07:17 PM andypugh: Depends which one.
07:17 PM XXCoder: lol dunno
07:17 PM XXCoder: forgot what it was
07:18 PM andypugh: Odd weekend, I was popping in to work for an hour at 7am, 2pm and 9pm. I was expecting it to ruin the weekend, but actually it got me out of bed and got me thinking what to do with my time.
07:18 PM pcw_home: chopper79: the residual RPM jitter is probably due to quadrature error, if it bothers you, you can low pass filter it
07:20 PM chopper79: pcw_home: If it does not actually hurt anything then I will leave it. Also I am thinking that it is already lowpass filtered.
07:20 PM chopper79: net spindle-fb-rpm-abs abs.spindle.out => lowpass.spindle.in
07:20 PM chopper79: net spindle-fb-rpm-abs-filtered lowpass.spindle.out
07:20 PM andypugh: So, I did a bit of housework, made a ratchet ( https://goo.gl/photos/76JfbXWzoiv2az4e8 ) and did a fairly ambitious repair on my skis: https://photos.google.com/share/AF1QipMO_wq0s5HWGqeoqczqbCH0R3svTxVtlQTJRXMY_vtpPiseXqXO-8PEAHeIoPw0Lg?key=dkhvNDlPOHVKOVJ5SUU3ZU0zaW5jNXNNQWNHblJR
07:21 PM XXCoder: what happened? fif you ski over hot coals?
07:21 PM XXCoder: or is that repair gel?
07:21 PM andypugh: chopper79: pcw_home: deadband in the PWM of 4 encoder counts might be a neater way?
07:22 PM XXCoder: ah so it is. interesting
07:23 PM chopper79: Right now deadband is 0.0 in the ini. Would I just increase it form 0.0 to 4.0? andypugh:
07:23 PM XXCoder: I guess deadband is where it ignores changes that small?
07:24 PM XXCoder: or is it average of 4 measures so bad measures has less effect?
07:24 PM andypugh: I ripped the edge out on a rock in Chamonix. The ski shop there knocked it back in, but it was out of position and the base material was loose. I removed the loose base, fastened the edge back in the right position with epoxy and then melted new base material in with a soldering iron.
07:24 PM pcw_home: Slowing the filter would probably help if the RPM reading is just for looks
07:26 PM andypugh: chopper79: pcw_home I came in late. Is this servo-jitter or rpm display jitter?
07:26 PM pcw_home: RPM display jitter
07:26 PM andypugh: Ah, hust set the lowpass slow then
07:27 PM andypugh: I answered the wrong question. Sorry for the confusion.
07:28 PM chopper79: ok.. lowpass in active in hal. Not sure of the value though
07:29 PM andypugh: just change the setp number, if you like it less, change it the other way.
07:29 PM chopper79: I do not have one
07:29 PM chopper79: lowpass spindle gain?
07:30 PM chopper79: setp lowpass.spindle.gain 1.000000
07:30 PM chopper79: is that it
07:30 PM andypugh: Yes, but that’s a lowpass with no filtering
07:30 PM andypugh: 1.0 = no filtering. 0.5 = average of last two measurements. Try 0.001
07:31 PM XXCoder: how often does it measure anyway
07:31 PM pcw_home: 1 KHz
07:31 PM XXCoder: ah
07:31 PM chopper79: ok.. I was on the man page and did not see anyting about that. Thank you I will try it
07:31 PM XXCoder: so it becomes average over one second
07:31 PM andypugh: That’s a 1-second time constant. The mechanical tacho on my Holbrook has a 1 minute time-constant (that’s too slow, though)
07:32 PM pcw_home: manual page says gain of .0063 = 1Hz bandwidth
07:33 PM pcw_home: without filtering, the quadrature error will dominate
07:34 PM andypugh: I was being simplistic. The manual page is probably right.
07:35 PM pcw_home: Gene has the fancy pipeline filter that averages that last 4 readings
07:35 PM chopper79: http://linuxcnc.org/docs/html/man/man9/lowpass.9.html
07:35 PM pcw_home: anyway you almost certainly do not want 1
07:35 PM andypugh: I think that the calculation is (old_average * (1 - gain)) + (new_value * gain)
07:36 PM chopper79: So this filter will not mess up rigid tapping right?
07:36 PM pcw_home: the manual page in master is _much_ better
07:36 PM andypugh: So, gain of 1 is “no filtering at all”
07:37 PM chopper79: I set gain to .005 and works well. I will change it to .0063 as pcw mentioned.
07:37 PM andypugh: chopper79: You could configure it to mess up tapping if you wanted to. Do you want to?
07:38 PM chopper79: heck no
07:38 PM andypugh: But, tapping uses the position, not the velocity. You probably want to give LinuxCNC the unfiltered velocity, and keep the filtered one just for the GUI displays.
07:39 PM pcw_home: Yeah spindle synchronized motion uses encoder position, not velocity so it make no difference
07:39 PM chopper79: So right now I have a .5-.8 rpm fluctuation. I am happy with that
07:39 PM XXCoder: filtered is good for PID adjustments also I guess
07:40 PM XXCoder: you watched that video chopper?
07:40 PM chopper79: So rigid tapping is ok with the filter set like this correct?
07:40 PM chopper79: Not yet... buffering. Kids and wife doing homework online so with onlt 1.3mb connection videos take a bit to load.
07:40 PM andypugh: My point (perhaps phrased a little eccentrically) is that we can’t speculate on what your HAL tweaks do. You certainly could mess up tapping with lowpass, but you would have to try.
07:40 PM pcw_home: yes
07:41 PM XXCoder: cool
07:41 PM mrpackethead: just finished making this
07:41 PM mrpackethead: https://scontent-syd2-1.xx.fbcdn.net/v/t1.0-9/17362871_10155154011897661_8664972992658964722_n.jpg?oh=0ec27a6c1ccf98364da8fb5bc015b0ac&oe=5962825E
07:41 PM XXCoder: mrpackethead: fancy. whats it?
07:41 PM chopper79: ok..got it. I will be happy with the small fluctuation I have now and start setting the index of the spindle up then.
07:41 PM chopper79: Thanks
07:41 PM mrpackethead: bottom half of an injection mojudling die
07:41 PM XXCoder: assuming you actually know, I usually dont, for parts at work lol
07:41 PM mrpackethead: so pleased it all lined up first time
07:41 PM XXCoder: nice
07:42 PM mrpackethead: the injector pins where the challenging part
07:42 PM andypugh: New Project Binky came out this weekend. It’s a good one. Including the classic: “So, this is a small neat bush. Made in Brazil. Makes sense”
07:43 PM mrpackethead: designed in fusion360 and used Fusion360 to do the CAM as well
07:43 PM chopper79: So index on the spindle is what is seen when I do a tapping function. It is seen and then resets count and then the encoder is counted for synchronized motion. Or do I have this incorrect?
07:50 PM pcw_home: Yes, when waiting to start threading or tapping, the index is detected and the (software)
07:50 PM pcw_home: count is reset and the threading or tapping pass with spindle synchronized motion is started
07:51 PM chopper79: perfect... time to wire the index sensor. Thank you
07:53 PM andypugh: mrpackethead: I guess you have a faster spindle than me, that would have taken my machine weeks to get those radii.
08:04 PM andypugh: Anyway. Time to sleep.
08:08 PM mrpackethead: andy.. yeah, i have 8000rpm
08:08 PM mrpackethead: could do with more!
08:10 PM XXCoder: 12k rpm on mine but much weaker.
08:10 PM XXCoder: with spindle rpm and feedback it would be stronger.
08:10 PM mrpackethead: i have a 1200W motor
08:10 PM mrpackethead: it works pretty well
08:11 PM XXCoder: 400w lol
08:31 PM os1r1s: XXCoder You still around?
08:32 PM XXCoder: ye[
08:33 PM os1r1s: I'm still trying to figure out how to flip the piece and mill the other side well. It did better this time, but still a little bit of offset. Part of the problem is that the size of the material is a little different than cam is set to. I'm thinking about indicating off the left side when I mill the top, and indicating off the right side when I mill the bottom.
08:33 PM os1r1s: That sound plausible?
08:35 PM XXCoder: how ifferent from cam? your tools diameters may need adjustm,ent
08:46 PM os1r1s: .01-.02
08:47 PM XXCoder: mm?
09:23 PM chopper79: Hooked index sensor up and my index mark is only seen once and awhile. Not every rotation like I would think it would be. I have tried different offsets with no luck.
09:25 PM XXCoder: wonder if thats related to sampling speed
09:25 PM XXCoder: im now 3d printing mount for my tach sensor
09:26 PM pcw_home: Yess is a sampling artifact: would not normally expect to see the index reliably in halscope with just a 1 KHz servo thread
09:27 PM pcw_home: s/is/its a/
09:28 PM chopper79: So how would I monitor it to make sure it is picked up besides a scope. Maybe I should put an led on it to see if it picks it up.
09:29 PM pcw_home: you can set index-enable and make sure its always cleared at the index position
09:29 PM chopper79: net spindle-index-enable <=> motion.spindle-index-enable
09:29 PM chopper79: in my hal
09:30 PM chopper79: net spindle-index-enable <=> hm2_[HOSTMOT2](BOARD).0.encoder.05.index-enable
09:30 PM pcw_home: you can do this manually:
09:30 PM pcw_home: sets spindle-index-enable true
09:30 PM pcw_home: and check that its cleared when you pass index
09:31 PM pcw_home: halcmd sets spindle-index-enable true
09:31 PM pcw_home: I should say
09:32 PM chopper79: setp hm2_[HOSTMOT2](BOARD).0.encoder.05.index-enable 1
09:32 PM chopper79: look right?
09:36 PM chopper79: setp hm2_[HOSTMOT2](BOARD).0.encoder.05.spindle-index-enable 1 this looks correct.
09:38 PM chopper79: Well I will work on it again later tonight or tomorrow. Time to spend time with the wife. Thank all
09:43 PM pcw_home: no you cannot do that
09:43 PM pcw_home: thats why you sets spindle-index-enable true
09:48 PM XXCoder: 3d printing is so long and boring
09:55 PM malcom2073: Nonsense, I can sit there and watch it for hours
09:56 PM Crom: back home....
09:56 PM XXCoder: and hours and hours
09:57 PM Crom: wow found ciggibutts for $2.13 a pack...
10:02 PM enleth: OK, success, both ball nuts reballed and working properly
10:03 PM XXCoder: success! now your ball nuts can have kids
10:03 PM XXCoder: so what was issue?
10:05 PM enleth: there wasn't, those two nuts use external preload and do have some radial play when not preloaded, but not really any axial - I was inadverdently rotating them a little trying to tug them along the screw
10:06 PM enleth: after bolting them to the carriage mounting block, with the preload spring and all, they lost radial play too
10:06 PM XXCoder: ah nice
10:07 PM XXCoder: problem nonexistant then :)
10:07 PM enleth: well, there *was* an issue in that I lost a ball or two four times and forgot about this or that which was supposed to go onto the screw between the two nuts
10:08 PM enleth: so I actually reballed those six or seven times
10:08 PM enleth: in total, not both
10:08 PM enleth: on the plus side, I can do that really fast now
10:08 PM XXCoder: ow
10:08 PM XXCoder: well all good at end :)
10:09 PM enleth: the trick is to use a sligltly magnetic screwdriver and a thin brass bolt, held in the same hand a bit like chopsticks, while the other hand holds the screw
10:10 PM enleth: you use the screwdriver to pick up the balls and drop them into the nut's return port, then use the end of the bolt to cram it in there tight against the previous one
10:10 PM enleth: then rotate the screw with the other hand just far enough to make room for another ball, but not too far
10:11 PM XXCoder: wo
10:11 PM XXCoder: glad I never removed my ballnuts lol
10:12 PM enleth: if you rotate too far, you might create an empty space between the two newly inserted balls, that won't compress easily, and you won't have space to cram all the balls in
10:12 PM enleth: oh, and there were 68 balls in each nut
10:13 PM XXCoder: so thats why you dont go too far
10:13 PM enleth: just noticed that my second-previous statement sounds absolutely brilliant taken out of context
10:15 PM XXCoder: manly.
10:15 PM enleth: oh, and when you fill the nut itself, you need to put the rest in the return tube, then put a dab of grease on both ends of the tube to keep the balls from falling out as you rotate it upside down to insert the ends into the return ports on the nut
10:15 PM XXCoder: wow
10:16 PM XXCoder: I wonder how chinese ones they fill balls. because i dont see return tube
10:17 PM enleth: AFAIK you do exactly what happens when you unscrew a ball nut from the screw, but in reverse and in a controlled manner
10:17 PM XXCoder: but if balls fall out?
10:18 PM enleth: they can be hand-loaded one by one by positioning the nut right at the end of the screw, exposing just the last turn of the thread
10:19 PM XXCoder: crazy. im so not removing em :)
10:20 PM enleth: and I imagine it could be fast-loaded by using a tapered PTFE rod with the ball thread profile machined on it
10:20 PM XXCoder: https://www.flickr.com/photos/projectapolloarchive/
10:21 PM enleth: so you put the nut on that rod and kind of pour the balls into it
10:28 PM enleth: as for statements taken out of context, the topic on some machining forum named "Ball screw lube?" takes the cake
10:28 PM XXCoder: yep
11:14 PM enleth: http://imgur.com/a/5p2Ff - some photos from the ORAC overhaul
11:19 PM XXCoder: cool :)
11:45 PM mrpackethead: another 6 hours to finish this job
11:45 PM mrpackethead: :-)
11:46 PM xxcodercnc: heys
11:49 PM xxcodercnc: hmm
11:50 PM xxcodercnc: what volt do input on BOB use?
11:50 PM xxcodercnc: 12v hopefully? heh
11:55 PM xxcodercnc: hey Wolf_ ya in?
11:56 PM Wolf_: yo
11:56 PM Wolf_: I think it says on the board
11:56 PM Wolf_: :P
11:57 PM xxcodercnc: looking
11:58 PM xxcodercnc: inputs just has chinese words
11:58 PM Wolf_: lol
11:59 PM Wolf_: 12v or 5 or both I can’t remember
11:59 PM xxcodercnc: hmm will try find bob info