#linuxcnc Logs

Jun 13 2019

12:05 AM skunkworks: _unreal_: https://youtu.be/xs-qVQtjOSs
12:09 AM skunkworks: hose clamps too :)
12:09 AM skunkworks: think that was running turbocnc
12:24 AM elmo40: CaptHindsight, really? All of it is py 2.7? Nothing ported to 3?
02:42 AM -!- #linuxcnc mode set to +v by ChanServ
03:14 AM XXCoder: finally home
03:14 AM XXCoder: all day drive south see my nephew graduate, then drive up home
03:14 AM XXCoder: nearly 6 hours sitting in my car lol
04:57 AM Tom_L: morning
05:01 AM XXCoder: https://www.youtube.com/watch?v=hHeki3AOy-g
05:01 AM XXCoder: fun lol
05:01 AM XXCoder: hey tom
05:02 AM XXCoder: jackman makes giant planer
05:18 AM jthornton: morning
05:18 AM XXCoder: hey
05:20 AM XXCoder: man it was nasty
05:20 AM XXCoder: i drove to vancouver for my nephew graduation
05:21 AM XXCoder: 5 miles took me a hour.
05:22 AM jthornton: traffic?
05:22 AM XXCoder: nope a truck rear cargo rolled over into other way road (way we were going)
05:23 AM XXCoder: so cars smashed into mail and stuff
05:23 AM XXCoder: so both sides of freeway was blocked
05:24 AM jthornton: opps
05:24 AM XXCoder: yeah
05:25 AM XXCoder: we arrived just in time to meet famuly at red robin
05:25 AM XXCoder: then went to graduation stuff
05:26 AM jthornton: graduation here was a couple of weeks ago
05:27 AM XXCoder: deaf schools last a little longer because of buses cut friday school time by half
05:27 AM XXCoder: school starts earlier and ends later
05:29 AM andypugh: XXCoder when you said “Giant planer” I was thinking of: https://www.youtube.com/watch?v=w4kPDhrhlgk
05:29 AM XXCoder: lol
05:29 AM andypugh: Though that’s actually quite a small one.
05:29 AM XXCoder: just flipping big machine for a small planer
05:30 AM XXCoder: but im pretty sure you meant machine itself also. hm lol
05:30 AM XXCoder: i cant imange how much stress machines takes when planing
05:30 AM jthornton: wow a giant planer
05:31 AM XXCoder: that actually works
05:32 AM XXCoder: i like jackman, he dont spam videos and all videos is good
05:33 AM andypugh: These are giant planers (I have shown these before) http://www.lathes.co.uk/waldrich-planers/
05:33 AM andypugh: Early photoshop.
05:33 AM XXCoder: wow
05:34 AM XXCoder: you mean last picture is a photoshop?
05:34 AM andypugh: The people look strangely fuzzy in all the pictures.
05:34 AM XXCoder: hmm yeah
05:34 AM XXCoder: double exposures
05:50 AM jthornton: Veronica (the sick chicken) is still alive this morning, I hope she recovers
05:53 AM XXCoder: yea
05:53 AM XXCoder: whats wrong?
05:56 AM jthornton: not sure, she has diarrhea and did not eat or drink anything all day yesterday until late in the evening then she started to drink some water
06:17 AM sync: andypugh: that is because most of the pictures are actually drawings
06:18 AM andypugh: Or heavily retouched photos for print use.
06:19 AM XXCoder: jthornton: sucks :( seperated to stop spread or?
06:19 AM jthornton: yea she is in a large dog kennel in the florida room
06:20 AM sync: they were really innovative machines back then
06:20 AM sync: too bad they have all been replaced with milling machines
06:20 AM XXCoder: it seems to me that those can still be useful
06:21 AM sync: no
06:21 AM sync: you cannot mill on them, which makes them unprofitable
06:21 AM sync: cnc shaping has its applications, but they are rare
06:21 AM XXCoder: not what I meant, more for doing stuff cnc mill cannot
06:22 AM sync: that virtually does not exist
07:09 AM Loetmichel: *GAAAH* sometimes $me is dumb like a brick... just CNC milled a bit of sheet alu... machine done... pulled the part from the vacuum table and it comes off in 5 parts... checked the G-code: i did make the "folding notches" 1,6mm deep instead of 0,75mm (for 1,5mm sheet alu) :-( *MAAAAAN* *head->desk*
07:26 AM andypugh: Most planers were converted to Plano-millers, with a rotating head where the too bit went.
07:27 AM andypugh: They were used for things like gearbox cases, but were really inefficient: cut wait wait wait wait wait cut return ……
07:28 AM Loetmichel: *hillarious* : a soldier just rang the bell here... "i want to retrive the repaired thin clients!" $me:"no idea what you are talking about." ... after some telephone calls it emerged that the german army had some defective tech that we made a few years ago, and decided to repair it at a cheaper company... didnt tell the soldier where to drive to though, so he assumed the orginal manufacturer...
07:28 AM Loetmichel: the actual repair company is 90km away... and its close to closing time here in germany already ;)
09:27 AM CaptHindsight: https://www.machinedesign.com/motion-control/linear-bearings-understanding-21-ratio-and-how-overcome-stick-slip-phenomenon
09:28 AM CaptHindsight: maximum distance from screw to linear bearing in gantry applications
09:29 AM CaptHindsight: and single vs 2 screw gantry
09:31 AM CaptHindsight: https://www.machinedesign.com/motion-control/five-heavy-duty-gantry-alternatives
09:38 AM CaptHindsight: https://www.youtube.com/watch?time_continue=10&v=D3-V6DKQnds
11:34 AM Deejay: hi
12:19 PM skunkworks: andypugh: how hard is it to figure out following error in your setup?
12:26 PM skunkworks: (hobbing)
12:30 PM andypugh: I haven’t tried.
12:37 PM CaptHindsight: gregcnc: did you find someone local for flat ground stock?
12:38 PM CaptHindsight: ?? http://www.nassteels.com/products.php
12:39 PM gregcnc: ups box address
12:41 PM CaptHindsight: was just at Amerimold yesterday, probably had a few distis these
12:41 PM CaptHindsight: these/there
12:41 PM CaptHindsight: http://www.toolsteel.com/ SW
12:44 PM gregcnc: the one I found a while ago was flatground.com but it looks like the site expired or was hacked
12:45 PM gregcnc: I wouldn't be surprised if that is the same place
12:45 PM CaptHindsight: Elgin P.O. box or similar IIRC
12:45 PM gregcnc: right
12:47 PM CaptHindsight: http://cintool.com/
12:47 PM CaptHindsight: http://cintool.com/page/Precision-Ground-Flat-Stock-PGFS
12:48 PM CaptHindsight: https://www.hudsontoolsteel.com/
12:48 PM CaptHindsight: wow Rockford is picking up some metal distis
12:49 PM CaptHindsight: even anew http://www.speedymetals.com/t-rockford.aspx
12:50 PM gregcnc: the rockford speedy has been there for a few years?
12:50 PM CaptHindsight: has it, just noticed them
12:50 PM CaptHindsight: http://www.speedymetals.com/pc-2796-8297-38-x-1-a-2-ground-flat.aspx not the best prices
12:51 PM gregcnc: a lot of places just change names or get purchsed. the hudson location shows some of steel supply if you search the address
12:51 PM gregcnc: there was a big auction in rockford of a former kennametal shop
12:58 PM CaptHindsight: gregcnc: https://www.cincinnatimetals.com/home.html not in Cincinnati
01:01 PM gregcnc: hah cintool was founded in rockford by a Mr. Cincinnati
02:20 PM miss0r2: 'evening
02:31 PM miss0r2: meh. a "professional" company couldn't make these splines... I just spend 20 minuts grinding on a 60degree shell mill... perfect fit. Makes you wonder how professional they are exactly
03:05 PM SpeedEvil: couldn't, or wouldn't ?
03:18 PM Deejay: gn8
03:25 PM miss0r2: SpeedEvil: Well, didn't have the cutter required. so.. I guess wouldn't
03:25 PM miss0r2: They could've just bought it and billed me :)
03:26 PM * SpeedEvil is planning on trying cutting GT3 soon.
03:27 PM SpeedEvil: (pulleys for)
03:29 PM miss0r2: GT3?
03:30 PM Tom_L: SpeedEvil single point cutter or hob?
03:30 PM ve7it: https://vancouver.craigslist.org/rds/tls/d/delta-monarch-engine-lathe/6899682848.html this looks like a nice old lathe.. close by
03:30 PM Tom_L: miss0r2, timing belt
03:33 PM miss0r2: okay. sounds like something that is atleast well documented
03:33 PM ve7it: a serious piece of american iron
03:33 PM SpeedEvil: Tom_L: possibly stupider - if i can get away with a small round cutter.
03:34 PM SpeedEvil: I am planning on trying composite pulleys of glued sheets.
03:36 PM Tom_L: be interesting to see how the composite wears against the belt
03:36 PM Tom_L: likely cheaper to buy the pulleys
03:36 PM miss0r2: I'm sure that is not the point, Tom_L
03:37 PM Tom_L: me too
03:37 PM miss0r2: Like most of what is going on in here :)
03:37 PM miss0r2: I just converted a iso 40 taper to be used with a hydraulic drawpar.
03:38 PM Tom_L: it would be nice to have a spindle with a drawbar and retention knob grippers
03:38 PM miss0r2: It has its moments
03:38 PM SpeedEvil: Err.
03:39 PM SpeedEvil: 'composite' as in 'made up from parts of aluminium sheet'.
03:39 PM SpeedEvil: Possibly a poor selection of words there.
03:39 PM miss0r2: but I am growing weary of the hydraulics on that old mill
03:39 PM Tom_L: ahh
03:39 PM miss0r2: is that what the chinese are doing? I always thought they were riveted or pinned or something like that
03:40 PM miss0r2: I have a few pulleys here that are made from sheets, it looks like
03:41 PM miss0r2: ouch its late.. I have to run - see you around
03:41 PM Tom_L: bye
03:41 PM SpeedEvil: by
03:43 PM andypugh: Tom_L: It is possible to convert a manual drawbar to penumatic. I did it with my milling machine.
03:43 PM Tom_L: mmm i've got a sprocket macro for my cad cam, i bet it wouldn't be that difficult to convert it for different belt pulley series
03:43 PM andypugh: And much to my astnishment it’s been reliable for nearly 10 years!
03:43 PM Tom_L: nice
03:43 PM Tom_L: i think i remember that
03:44 PM Tom_L: what air pressure do you run on it?
03:45 PM andypugh: I got the maths a bit wrong so need 140 psi.
03:45 PM andypugh: https://youtu.be/pxrzJ_KfcQ0
03:46 PM Tom_L: need a little bigger cylinder?
03:46 PM andypugh: Yes, ideally.
03:47 PM Tom_L: standard retention knob or one of your own geometry?
03:47 PM andypugh: Standard
05:23 PM tubes4branes: has anyone ever worked with OFDL-2 Kmollgren (Danaher Motion) servo amps? They hook up to Kmollgren 3phase servo motors with RESOLVERS. They were present on the 2007 Bridgeport CNC machines.
05:35 PM _unreal_: sweet... my machine is now tuned
05:35 PM _unreal_: step per mm
05:36 PM _unreal_: only thing I have left to get done at this is figure out how big my machine area is :)so I can enter that into the limits
05:51 PM andypugh: tubes4branes: No, I haven’t. But I am the resident Resolver lover here.
05:51 PM andypugh: In fact I have one taken an absolute serial encoder off of a servo motor and replaced it with a resolver.
06:08 PM tubes4branes: honestly I like Resolvers, and the servo amplifier takes the resolver inputs and converts them into quadrature outputs based on a divisor, etc, as the resolvers are resolved to 65535 counts per revolution.
06:41 PM andypugh: tubes4branes: What is it you want to do?
06:41 PM andypugh: LinuxCNC can work happily with the converted quadrature or with the actual reolvers.
06:42 PM andypugh: (my machines use the 7i49 card from Mesa)
06:44 PM andypugh: pcw_home: Hmm. talking of which, my 7i73 has disappeared from my HAL.
06:45 PM andypugh: That’s annoying as it is buried in the apron castings and basically have to take the saddle off to get to it. Poor planning, I suppose. A cover plate would not have been difficult to engineer in.
06:46 PM andypugh: Could easily be a broken wire in the cable chain, I suppose.
06:48 PM sync: andypugh: I wondered about that when you put it in
06:49 PM andypugh: So did I. But not enough.
06:49 PM andypugh: It’s more likely to be the cable than the board, I think.
06:50 PM andypugh: I will try moving the saddle by hand and see if it comes back.
06:50 PM andypugh: (I can’t jog, as the 7i73 does the jogging…)
06:50 PM sync: I thought you used chain grade cable
06:51 PM andypugh: I did except for the CAT5
06:51 PM andypugh: (As I recall, I might be wrong)
06:52 PM sync: well, that was a bad choice then
06:52 PM sync: (no shit)
06:52 PM andypugh: Maybe, time will tell
06:57 PM CaptHindsight: 7C80 RPI interfaced Step&dir plus I/O interface
06:57 PM CaptHindsight: just wait, the *akers will complain about how they could build it for $2 less :)
06:58 PM sync: I wonder how well a scraped cylinder head would work against a scraped block
07:19 PM CaptHindsight: sync: with our without glue?
07:20 PM CaptHindsight: with copper head gasket
07:34 PM andypugh: Why ot gold wires, or silk threads?
08:01 PM sync: well copper wire would work
08:01 PM sync: gold, probably also
08:11 PM andypugh: Gold wire is well known in vaccuum applications. Silk thread in aero-engines.
08:13 PM sync: yes
08:14 PM sync: probably aluminium wire would do ok as well
08:19 PM andypugh: mercury wire, for cryogenic applications
08:19 PM adamj12b: Good Evening! Anyone on that has successfully set up spindle orientation?
08:20 PM andypugh: I haven’t.
08:20 PM adamj12b: haha.
08:20 PM andypugh: I have written a bunch of docs on the subject, but never actually needed it :-)
08:20 PM adamj12b: I have it working but have a few bugs that are bugging.
08:20 PM adamj12b: it likes to unwind after the spindle has been running
08:21 PM andypugh: Are you using the orient comp?
08:21 PM adamj12b: yes
08:21 PM andypugh: That is intended to fix that.
08:21 PM andypugh: You may have it configured wrong, then.
08:21 PM adamj12b: i have theory's as to whats wrong but not positive.
08:22 PM adamj12b: Im using a 7i77 with encoder 5 set up for spindle encoder with a scale of 4096
08:23 PM adamj12b: I think the issue may be with the position of the encoder feeding into the orient component.
08:23 PM andypugh: The code is short and sweet inside orient: https://github.com/LinuxCNC/linuxcnc/blob/master/src/hal/components/orient.comp#L70
08:23 PM adamj12b: OO I will have to read through that later.
08:24 PM andypugh: Every time enable toggles it should latch the current position and deal only with partial turns
08:24 PM andypugh: So my guess is that you haven’t connected orient.enable correctly in HAL.
08:25 PM andypugh: 27 lines of code. You should be finished in later + a minute or two :-)
08:28 PM adamj12b: motion.spindle.orient connects to orient.spindle.enable
08:28 PM adamj12b: orient.spindle is the named orient comp.
08:29 PM Tom_L: ooo
08:29 PM Tom_L: just read.. yes i've done orient
08:29 PM Tom_L: but it wasn't real successful on my sherline due to the spindle motor
08:30 PM Tom_L: but i did follow andypugh's notes
08:30 PM adamj12b: here is the hal. It works, just not perfect :D
08:30 PM adamj12b: https://pastebin.com/sTgfNRd7
08:32 PM adamj12b: Here is what its doing. https://photos.app.goo.gl/tpVyz36Mnzh9oDAt5
08:32 PM adamj12b: I did start with andys svg diagram on the website.
08:32 PM Tom_L: i think i used the near component on mine
08:32 PM andypugh: Nice tidy HAL file at least
08:32 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/configs/sherline/sherline.hal
08:33 PM Tom_L: pretty sure that's the one with orient
08:33 PM adamj12b: I actually compiled the 2.8 component on 2.7 so I didnt have to use near as the 2.8 version has tolerance built in.
08:33 PM Tom_L: you probably understand it better than me then
08:33 PM adamj12b: Yes. I got the link from you a few months ago. I have your archive bookmarked. Lots of good stuff. :)
08:37 PM andypugh: Is your spindle encoder scaled for 1 = 1 revolution?
08:37 PM andypugh: If that was different (degrees?) then things would be off.
08:38 PM andypugh: I can’t see what is wrong, it looks OK. Halmeter on orient-out and orient.position might be illuminating.
08:40 PM Tom_L: andypugh you've not actually implemented it?
08:40 PM adamj12b: https://photos.app.goo.gl/gSHxGidXhBfe4sBw8
08:40 PM Tom_L: <andypugh> I have written a bunch of docs on the subject, but never actually needed it :-)
08:41 PM adamj12b: Encoder scale is -4096
08:42 PM Tom_L: mine seemed to wander around the orient point as well
08:43 PM adamj12b: the wandering is ok with me. I need to add a brake to hold it. The wandering is influenced by how stiff the belt is from sitting and the VFD
08:43 PM andypugh: Tom_L: Not on real hardware, no. I did do this, way before M19 and orient were added: https://www.youtube.com/watch?v=3oTJNEVpvYY
08:44 PM Tom_L: well it _does_ work. my sherline hardware just wasn't a good place to try it
08:46 PM adamj12b: Im using a sensorless vector drive with extremely high acceleration rates of 0.1 seconds per 1000hz with CW and CCW inputs to control the position of the drive.
08:47 PM Tom_L: once i did it, ich asked for the code and i don't know if he posted it anywhere or not
08:47 PM Tom_L: we may never know
08:48 PM adamj12b: andypugh: when you say orient-out, do you mean orient.command?
08:48 PM andypugh: Yes, probably.
08:48 PM andypugh: It’s 0245 here, I am not really up-to-speed
08:49 PM Tom_L: yeah that and position error are the only 2 outputs
08:52 PM adamj12b: Your up late. The pic I posted bit ago hows the command is 2557.319 with the encoder position being 2557.313, after orientating the spindle.
08:53 PM adamj12b: Im winding the spindle up now as it seems the more its out of place the initially, the more it wants to unwind. If I cancel the orient, and hit it again, it will position quickly.
08:54 PM CaptHindsight: no he's Andy :)
08:59 PM adamj12b: OK. so I slowed it down and its winding up +10 rpm, then ramps back down to the target position, even when its near the position from a previous orient.
09:07 PM andypugh: That sounds like it _might_ be a PID tuning issue
09:08 PM andypugh: And even I need to think about sleep eventually. And eventually is now.
09:08 PM andypugh: Goodnight.
09:09 PM adamj12b: Yes. It is PID error. PID is making the error worse before it corrects it. I will work on that.
09:10 PM elmo40: how do you calculate the PID?
09:11 PM adamj12b: Originally the PID was calculated with the spindle motor on the bench.
09:11 PM adamj12b: I just worked on increasing gains to end up with the least error with the stiffest output
09:13 PM tiwake: what is a good set of thread pitch micrometers?
09:14 PM tiwake: the one I used before, I don't remember the name to, but it had interchangeable heads on it for all the different thread pitches
09:24 PM adamj12b: Tom_L: I got it! It was PID
09:33 PM Tom_L: oh nice
09:34 PM Tom_L: tiwake sounds expensive
09:35 PM Tom_L: i've got an old one
09:39 PM Tom_L: https://ecatalog.mitutoyo.com/Screw-Thread-Micrometers-Series-326126-Interchangeable-Anvil-Spindle-Tip-Type-C1108.aspx
09:42 PM tiwake: Tom_L: how much is that?
09:42 PM Tom_L: alot
09:42 PM Tom_L: http://www.penntoolco.com/mitutoyo-screw-thread-micrometers-series-326-126-interchangeable-anvil-spindle-tip-type/
09:43 PM Tom_L: i like mitutoyo though
09:48 PM Tom_L: looks like about 950 with the anvil set
09:50 PM Tom_L: https://www.msi-viking.com/Mitutoyo-126-901-Screw-Thread-Micrometer-wInterchangeable-Anvils-0-1_p_768.html
09:50 PM Tom_L: 800 there
09:51 PM Tom_L: http://www.starrett.com/metrology/product-detail/575CP
09:51 PM Tom_L: mine looks like that