#linuxcnc Logs
Aug 30 2020
#linuxcnc Calendar
12:28 AM veegee: Something beautiful about a simple vernier scale caliper
12:34 AM veegee: digital needs batteries and LCD visibility/contrast is trash
12:34 AM veegee: dial type is cool, but reading vernier scales is fun
12:34 AM XXCoder: i prefer dial caliper
12:34 AM veegee: Sometimes I just want to visualize what 5.5mm is
12:34 AM veegee: vernier seems to be the easiest to set to that value
12:34 AM veegee: Or at least approximate value for the sake of visualization
02:07 AM Deejay: moin
05:16 AM -!- #linuxcnc mode set to +v by ChanServ
05:46 AM skec: Hi guys, could you suggest a 2 channel opto switch in one package that would directly work with linuxcnc for quadrature encoding? Thanks!
05:47 AM skec: would something like this work: https://uk.rs-online.com/web/p/slotted-optical-switches/1711830/
05:47 AM skec: ?
05:48 AM rs[m]1: skec: mouse experiments not successful?
05:48 AM skec: it is much harder than I anticipated to find wheel mouse :)
05:49 AM skec: I am still looking though..
05:49 AM rs[m]1: skec: i have not seen a mouse without scroll wheel for a very long time
05:49 AM skec: I know I had 1 at home... some 15 years ago
05:50 AM skec: you mean the middle button/wheel?
05:50 AM rs[m]1: skec: the product you linked to is discontinued, but something like that could be made to work with a slotted wheel
05:50 AM skec: I was looking for a mouse with a ball (not optic)
05:50 AM rs[m]1: skec: yes, wheel instead of button. wheel would also work.
05:51 AM skec: good point!
05:52 AM rs[m]1: time to find some cheese for your mouse traps!
05:53 AM skec: I was just wondering how often do I use middle mouse actually ... (hmm)
05:53 AM skec: button*
05:53 AM skec: but that thing would also give me just one signal (?)
05:54 AM rs[m]1: i'm talking about the scroll wheel
05:55 AM skec: yes, the scroll wheel.. I just opened it..and it looks like some closed-non optic thing is sitting at the wheel
05:55 AM skec: maybe some contact slider with many pins inside..
05:56 AM skec: something like those pot like looking pulse generators
05:57 AM rs[m]1: i guess it depends on the mouse model
05:57 AM skec: exactly
06:01 AM Tom_L: morning
06:02 AM skec: Morning
06:08 AM skec: please check this out:
06:08 AM skec: I found this: https://www.aliexpress.com/item/32926374797.html?ws_ab_test=searchweb0_0,searchweb201602_6_10065_10130_10068_10890_10547_319_10546_317_10548_10545_10696_453_10084_454_10083_10618_10307_537_536_10902_10059_10884_10887_321_322_10103,searchweb201603_6,ppcSwitch_0&algo_expid=d86d080c-c868-4cc0-88c8-5c3b3d5ef5a9-13&algo_pvid=d86d080c-c868-
06:08 AM skec: 4cc0-88c8-5c3b3d5ef5a9&transAbTest=ae803_4
06:08 AM skec: as well as this:
06:08 AM skec: https://www.aliexpress.com/item/32730823157.html?spm=a2g0o.detail.1000014.5.8f34756cpSNpFL&gps-id=pcDetailBottomMoreOtherSeller&scm=1007.14976.185485.0&scm_id=1007.14976.185485.0&scm-url=1007.14976.185485.0&pvid=bfdd8213-599c-4519-9e31-e3ea08e5a6da&_t=gps-id:pcDetailBottomMoreOtherSeller,scm-url:1007.14976.185485.0,pvid:bfdd8213-599c-4519-9e31-e3ea
06:08 AM skec: 08e5a6da,tpp_buckets:668%230%23131923%2385_668%23808%234093%2315_668%23888%233325%231_4976%230%23185485%230_4976%232711%237538%23934_4976%233104%239653%237_4976%233141%239887%231_668%232846%238111%23425_668%232717%237559%2373_668%231000022185%231000066055%230_668%233422%2315392%23389
06:08 AM Tom_L: don't post those long links here
06:09 AM skec: both look like opto switches, however, 1 has 4 wires, another one just 3
06:09 AM XXCoder: you could just copy url from http to before ?
06:09 AM Tom_L: use tinyurl
06:09 AM XXCoder: or that
06:09 AM skec: oh, sorry
06:09 AM Tom_L: by the time the link gets here it's split and does neither of us any good
06:10 AM skec: sorry, did not know that..
06:10 AM skec: here are tiny links: https://tinyurl.com/y36mttxn vs https://tinyurl.com/y4rh7rbj
06:10 AM Tom_L: you said you wanted 2 channel
06:11 AM Tom_L: https://www.digikey.com/catalog/en/partgroup/heds-9000-91xx-9200-series/40189
06:11 AM skec: Yes, this what linuxcnc expects for quadrature encoder
06:11 AM Tom_L: https://www.digikey.com/catalog/en/partgroup/hedm-55xx-56xx-and-heds-55xx-56xx-series/12510
06:12 AM skec: thanks!
06:13 AM skec: shall I count the slots in the metal disc that is on the motor?
06:15 AM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/maxbot/image006.jpg
06:15 AM Tom_L: count that one instead
06:15 AM skec: oh no..
06:15 AM Tom_L: that's from a heds 9000 encoder
06:16 AM skec: does it has 0.1 nm linear resolution?
06:16 AM Tom_L: 2048 quadrature
06:17 AM Tom_L: that's 512 slots in case you lost count
06:18 AM skec: meaning in 1 turn your motor can move 512 steps?
06:18 AM Tom_L: https://www.aliexpress.com/popular/nemicon-encoders.html
06:19 AM Tom_L: 1 turn is 2048 quadrature
06:20 AM Tom_L: i've got one of the nemicon encoders on my spindle
06:21 AM skec: in laymans terms, if there are 512 slots, then you have 512 ups and 512 downs, that is 1024 on/off per turn. and because you have 2 channels, its 1024x2 = 2048?
06:22 AM Tom_L: https://en.wikipedia.org/wiki/Incremental_encoder#Quadrature_outputs
06:26 AM skec: do you know if kuka or fanuc uses steppers or dc with encoders?
06:37 AM rs[m]1: skec: that sounds like there were only these two options ;)
06:38 AM skec: now I know what it actually is, so I can google if there is anything near local stores
06:39 AM JT-Cave: finally got internet back on
06:39 AM XXCoder: yay
06:40 AM XXCoder: i was stuck at home waiting for you to fix internet. ;)
06:40 AM JT-Cave: lol
06:44 AM XXCoder: been playing color zen for while. fn
06:44 AM XXCoder: fun
06:47 AM JT-Cave: time to brave the elements and open the chicken coops
06:48 AM XXCoder: watch out for lurking rain
06:48 AM * JT-Cave sticks his head out in the Florida room and decides the hens don't really want out of the run right now
06:49 AM JT-Cave: the pullets are still on the roost so no hurry for them either
06:50 AM JT-Cave: I'm still having fun coding the dawn to sunrise leds
06:55 AM JT-Cave: python datetime and timedelta are fun brain teasers trying to keep track of which one a variable is lol
06:55 AM -!- #linuxcnc mode set to +v by ChanServ
06:55 AM XXCoder: yay heh
07:13 AM _unreal_: yawn
07:15 AM _unreal_: so my duaghter and I have gotten most of the wiring done on her little machine
07:15 AM _unreal_: might even get it running today
07:30 AM SpeedEvil: :)
08:04 AM jymmmm: morning
08:09 AM _unreal_: going to take my daughter snorkling today
08:09 AM _unreal_: go see some coral
08:36 AM JT-Cave: yo
08:38 AM * JT-Cave wonders what will clean fly strip goo off of my keyboard???
08:40 AM jymmmm: I use naptha (lighter fluid) to remove adhesives
08:41 AM jymmmm: jt let me know if it works for flypaper glue
08:41 AM FinboySlick: Heard good things about Googone too.
08:42 AM jymmmm: googone is naptha among other things
08:43 AM jymmmm: and alcohol to remove pine tar/sap
08:44 AM jymmmm: I keep those lil alcohol wipes in the car to remvoe from windshield, works great..
08:45 AM jymmmm: coleman fuel is also napatha
08:47 AM jymmmm: ...with some rust inhibitors in it
08:52 AM jymmmm: Squirrel Ninja course https://www.youtube.com/watch?v=hFZFjoX2cGg
09:19 AM jymmmm: Ubuntu software updates were NOT working when set to US Servers this morning, switched to Main server and all is working again.
09:21 AM rs[m]1: Jymmm: a major ipv4 outage is currently affecting some transit providers (level3)
09:25 AM rs[m]1: https://www.thousandeyes.com/outages
09:26 AM jymmmm: rs[m]1: I JSUT saw that.... Level3 in LA and WA, then in TX, then two on the East cost. Fsk me, those are the major hubs
09:27 AM Tom_L: terrorists!
09:28 AM rs[m]1: never attribute to malice that which is adequately explained by stupidity
09:30 AM jymmmm: A construction work on a tractor digging up fiber lines on a sunday, eh possible, but doubtful
09:32 AM Tom_L: gotta get all those unpermitted jobs done on the weekend
09:37 AM rs[m]1: Jymmm: more something like botched BGP update
09:38 AM jymmmm: rs[m]1: Now, that is REALLY sad
09:38 AM jymmmm: rs[m]1: I'm guessing Level3 being the culprit
09:45 AM jymmmm: OpenDNS is down too
09:46 AM Tom_L: good thing we don't rely on the internet ehh
09:46 AM jymmmm: But you can switch to googleDNS 4.4.4.4 and 8.8.8.8 and they are working.
09:53 AM skunkworks: OMG - I wonder how much the 'I ran the screw out of the ball nut 2 times' was totaly mach4 doing something odd..
09:54 AM skunkworks: https://www.youtube.com/watch?v=WXGe48-MmOg
10:12 AM Tom_L: skunkworks, has he ever run linuxcnc?
11:02 AM Rab: What's good for a low-latency PC these days? I've been using an old HP DC7800p, but the power supply just blew up. Ordered a replacement, but this is the second part I've had to replace this summer...seems like any more money would be better spent on a newer system.
11:03 AM Tom_L: are you running buster?
11:03 AM Rab: Yep.
11:03 AM Tom_L: i've got a couple gigabyte boards that do ok but i just test with them
11:03 AM Tom_L: likely not "new" now
11:04 AM Rab: I use two parport BoBs with PCIe adapters, so I guess it wouldn't need a parport built in.
11:04 AM Tom_L: hmm
11:05 AM Tom_L: i'm still running wheezy on the mill on an old board
11:05 AM * JT-Cave still has two machines running ubuntu...
11:06 AM Rab: I guess I could look for any i3-i7 slimline business desktop on eBay and roll the dice.
11:06 AM Tom_L: one day the heat will get to it i'm sure and i'll be in your spot
11:06 AM Tom_L: these are i5
11:07 AM Tom_L: you can always update to mesa cards and it wouldn't be as big a deal
11:08 AM Tom_L: they're a little more forgiving latency wise
11:08 AM Rab: That is true.
11:08 AM * JT-Cave is still searching for the rgb color sequence to simulate dawn to sunrise lol
11:08 AM Tom_L: and in that case you could use a rpi4
11:10 AM Rab: Over ethernet?
11:10 AM Tom_L: i've tested ethernet and spi on mine
11:10 AM Rab: ahh
11:10 AM Tom_L: using a 7i90 and 7i47s daughter card
11:11 AM Tom_L: err 7i80 for ethernet
11:12 AM Tom_L: JT-Cave, dunno if this is even close: https://encycolorpedia.com/ffca7c
11:13 AM Tom_L: maybe they're just selling paint...
11:14 AM JT-Cave: looks that way
11:14 AM JT-Cave: I'm looking to simulate dawn to sunrise with RGB
11:14 AM Tom_L: not much out there..
11:14 AM JT-Cave: yeah I know
11:16 AM Tom_L: https://www.quora.com/What-are-the-exact-colors-of-the-sun-between-sunrise-and-sunset-in-RGB
11:16 AM Tom_L: dunno...
11:16 AM JT-Cave: pretty sure I've been to that page a few time but I'll look anyway
11:16 AM Tom_L: heh
11:17 AM Tom_L: would depend on the moisture content etc too
11:18 AM JT-Cave: all the answers seem to be hung up on some theoretical perfect color and none seem to really have an answer with a general progression from dawn to sunrise
11:18 AM rs[m]1: that question doesn't really make sense
11:18 AM Tom_L: nice way of saying they don't really know
11:19 AM JT-Cave: LOL yep
11:19 AM rs[m]1: what is different is the spectrum of sunlight
11:20 AM rs[m]1: human vision usually corrects for that and does it's own white-balance kind of thing
11:20 AM JT-Cave: I'm not concerned with the spectrum of the sun at all, just want a gradual progression from dawn to sunlight light level
11:20 AM JT-Cave: dawn being off and sunrise being 3000 or so kelvin
11:21 AM rs[m]1: just increase light level and color temperature gradually, i don't think your chickens will care very much for a proper dawn color
11:21 AM JT-Cave: correct, they could care less
11:23 AM JT-Cave: one program I found suggests that 1000k is 255, 67, 0 which is quite bright with the apa102
11:24 AM jymmmm: jt have you tried hue instead of rgb?
11:24 AM rs[m]1: blue light is scattered more than red light because of it's shorter wavelength, so light that travels farther through the atmosphere loses blue parts of light
11:25 AM rs[m]1: JT-Cave: just scale those values down proportionally to e.g. 128/33/0 to acceptable level if full on is to bright
11:25 AM JT-Cave: the apa102 you can set the level of each led BGR in each package from 0 - 255 and I think the first bit is intensity
11:26 AM JT-Cave: rs ok that makes sense
11:26 AM rs[m]1: what do you mean with intensity? vs. level?
11:28 AM JT-Cave: http://www.led-color.com/upload/201604/APA102-2020%20SMD%20LED.pdf
11:29 AM JT-Cave: the LED frame is 32 bits with the first 3 then 5 for something and the other three set the level of blue, green and red leds
11:30 AM JT-Cave: https://dpaste.com/2Z7NCM27J
11:30 AM jymmmm: jt maybe you can steal the sequence from here https://github.com/thehookup/RGBW-Sunrise-Animation-Neopixel-
11:30 AM jymmmm: jt http://www.thesmarthomehookup.com/diy-artificial-sunrise-rgbw-individual-addressable-leds/
11:32 AM rs[m]1: JT-Cave: if i interpret it correctly, the first byte is MSB 1 1 1 followed by 5 bits of current level
11:32 AM rs[m]1: JT-Cave: then followed by PWM values for Red/Green/Blue
11:32 AM JT-Cave: that's what I'm thinking as well
11:32 AM rs[m]1: JT-Cave: so it should be possible to scale brightness with the 5 bits
11:33 AM JT-Cave: I do know for a fact that if you pass [0,0,0,0] via spi nothing will happen to the current led settings
11:33 AM rs[m]1: changing current usually shifts frequency a bit, but that shouldn matter
11:34 AM rs[m]1: JT-Cave: [224, 255, 255, 255] should be all LEDs off
11:34 AM rs[m]1: because driving current is 0
11:35 AM Tom_L: after all JT-Cave needs to have a good sunrise in the cave :)
11:36 AM JT-Cave: rs[m]1, no that didn't turn them off
11:36 AM JT-Cave: >>> data = header + [255, 0, 0, 0] * numLeds + footer
11:36 AM JT-Cave: >>> spi.writebytes(data)
11:36 AM JT-Cave: that will turn them off
11:37 AM jymmmm: JT if you change 255 to 128, will they be half bright?
11:37 AM JT-Cave: they go off
11:38 AM rs[m]1: JT-Cave: you have to keep first 3 bits at 1
11:38 AM jymmmm: what turns them back on again?
11:38 AM rs[m]1: 244 255 255 255 doesnt turn them off?
11:38 AM JT-Cave: rs[m]1, ah ok I missed that
11:38 AM JT-Cave: no
11:38 AM rs[m]1: 224 255 255 255 doesnt turn them off?
11:38 AM JT-Cave: no
11:39 AM rs[m]1: what happens with 225 instead of 224?
11:41 AM rs[m]1: 255 255 255 255 should turn everything on to full brightness
11:41 AM JT-Cave: sorry was looking for a binary to decimal converter, yes 224 turns them off
11:41 AM rs[m]1: JT-Cave: so first value is 224 + brightness from 0 to 31
11:42 AM JT-Cave: yep, thanks for helping me sort that puzzle out
11:42 AM rs[m]1: JT-Cave: e.g. 50% should be 224 + 16 = 240
11:42 AM JT-Cave: >>> data = header + [225, 0, 3, 128] * numLeds + footer
11:42 AM JT-Cave: >>> spi.writebytes(data)
11:42 AM JT-Cave: that gives me a very low light level
11:42 AM JT-Cave: sweet
11:43 AM rs[m]1: 225 should be lowest light level
11:44 AM JT-Cave: this is making much more sense now :)
11:46 AM skunkworks: Rab: had really good luck with i5 old sff hp's 8200's
11:46 AM skunkworks: 8300
11:46 AM skunkworks: they have a printer port header on the motherboard
11:46 AM skunkworks: They run <10us latency on rt_preemt
11:47 AM jymmmm: jt https://www.thecrowned.org/base-conversion-ubuntu-decimal-binary
11:47 AM JT-Cave: jymmmm, https://www.rapidtables.com/convert/number/decimal-to-binary.html
11:48 AM jymmmm: JT heh, that works too :)
11:48 AM Rab: skunkworks, I saw someone posted good latency numbers for an 8300 to the latency test wiki--maybe it was you! I'm looking on eBay right now.
11:52 AM Rab: I already bought a replacement PSU, but it's pretty obvious I need a backup. Summer sometimes lasts through November here.
02:08 PM JT-Cave: rs[m]1, https://dpaste.com/5RAQUEBKY
02:31 PM rs[m]1: JT: seems your exact location is in this paste ;)
02:34 PM rs[m]1: JT-Cave: dtos seems complicated, why not self.intensity = int(31 * p / 100) or something like that?
02:35 PM JT-Cave: didn't think of it... good suggestion
02:35 PM JT-Cave: I was just stoked to get it working lol
02:41 PM rs[m]1: if it works leave it like it is
02:54 PM JT-Cave: not quite working as I get 0 up to p == 3.3 or so
03:09 PM Deejay: gn8
03:11 PM Tom_L: bye
03:41 PM rs[m]1: level3 seems to work again. seems to have been caused by some botched bgp flowspec.
03:42 PM andypugh: I made a thing. I haven’t made a thing for weeks.
03:42 PM Tom_L: andypugh, was hoping you hadn't given up making with all the work you've been putting in on 2.8
03:43 PM Tom_L: i for one appreciate it
03:44 PM Tom_L: so what's the 'thing'?
03:44 PM andypugh: A super-short ER collet holder for the mill.
03:45 PM Tom_L: must have a project for it..
03:45 PM andypugh: I used the shank part of a drill chuck that I destroyed, and a bought-in nut.
03:45 PM andypugh: Well, I am always running out of Z height, the curse of a small mill.
03:46 PM Tom_L: yeah that was one mistake i made on mine
03:46 PM andypugh: https://photos.app.goo.gl/mAyGoJHcypvf8HDb9
03:46 PM Tom_L: not enough z
03:46 PM andypugh: I bet you can touch the nose to the table, but that’s pointless?
03:47 PM Tom_L: did you print that wrench?
03:47 PM Tom_L: it looks plastic
03:47 PM andypugh: Yes, just to try out the model and to have something to fiddle with it with
03:47 PM Tom_L: but shouldn't be
03:48 PM Tom_L: i need to get better at broaching on the equipment i've got around here
03:49 PM Tom_L: i'll need a 5mm keyslot for the spindle motor soon
04:00 PM andypugh: Buy a keyway broach?
04:02 PM JT-Cave: Tom_L, you need a broach and a thingy I forget what they are called but for 5mm you could use a hammer to drive the broach
04:02 PM Tom_L: not for just one
04:02 PM andypugh: https://www.ebay.com/itm/5mm-B1-Type-Push-Type-Keyway-Broaches-HSS-Keyway-Tools-for-CNC-Machine-Tool/164137744291?hash=item26375f37a3:g:dl4AAOSwu2tcXCTt
04:02 PM Tom_L: yeah i've used them before
04:02 PM Tom_L: in a press
04:02 PM * JT-Cave wanders up to see what sizes he has
04:02 PM andypugh: I was pointing out that they are pretty cheap from that China
04:03 PM Tom_L: i should call my bud, he might actually have one
04:03 PM Tom_L: i got a reamer for it the other day
04:04 PM Tom_L: since i'm sure i'll have to size the pulley holes
04:04 PM Tom_L: andypugh, did you see my attempt from the other day?
04:05 PM andypugh: Though, if you have a CNC lathe: https://youtu.be/l_oC3p2lGMA
04:05 PM andypugh: No, I have been distracted
04:05 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/rotary/Broach/broaching2.jpg
04:06 PM JT-Shop: only metric duMONT I have is 8MM C
04:06 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/rotary/Broach/Test_Keyway3.jpg
04:06 PM Tom_L: still practiceing
04:07 PM Tom_L: i tried one on the lathe but it didn't work so well
04:07 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/rotary/Broach/Broach_holder3.jpg
04:07 PM andypugh: That first picture is wierd, it looks like you have a ghost drawbar with a square head.
04:08 PM Tom_L: heh couple things going on there
04:09 PM Tom_L: the square is the blank that holds the sherline spindle in place
04:09 PM Tom_L: behind the spindle
04:09 PM Tom_L: never noticed that
04:09 PM Tom_L: i just made a plate to mount it to
04:17 PM JT-Shop: cool broach mount
04:18 PM Tom_L: i need to change the angle on the cutter a little
04:18 PM Tom_L: that was just a stock blank
04:19 PM JT-Shop: need a tiny shaper head lol
04:20 PM andypugh: I went the other way with my sloter cutter: https://photos.app.goo.gl/5GaAES3BKAAcxjcX7
04:21 PM Tom_L: i thought about that but didn't have anything within arm's reach to make one
04:21 PM andypugh: I thought it was worth showing that picture as is it one way to not have to make a square hole.
04:21 PM Tom_L: i might still try it
04:28 PM andypugh: Pleased with this eBay buy: https://www.ebay.co.uk/itm/143698089381
04:29 PM Tom_L: kind of them to provide a magnifier :)
04:30 PM andypugh: Probably to identify the parts more than to read the scale.
04:30 PM andypugh: There is another on there, priced differently: https://www.ebay.co.uk/itm/333669412854
04:31 PM Tom_L: and no magnifier!
04:32 PM Tom_L: we would generally use go - nogo gauges
04:32 PM Tom_L: well, especially for threaded holes
04:33 PM Tom_L: i'm thinking of getting 2 pulley sets for the spindle, a 1:1 and something to step it up a bit
04:34 PM Tom_L: since i'm going from ~180w to 1.2kw motor i think it could handle it
04:34 PM andypugh: Yes, maybe 10x as fast?
04:34 PM Tom_L: i don't think the spindle would take that much
04:35 PM Tom_L: maybe up to 2:1 which would be around 10k rpm
04:36 PM Tom_L: the belts probably wouldn't like that very much
04:47 PM andypugh: What’s the encoder good to?
04:48 PM andypugh: (not that it really matters, you typically only care about spindle position at low speeds)
04:54 PM XXCoder: gem grinding. interesting https://youtu.be/BLnw4n-BX14
04:55 PM jymmmm: andypugh: Aww, I thought I was gonna see you made a square hole, cause that would be cool
05:04 PM andypugh: XXCoder: Mechanical, precise and difficult. Seems like a job for CNC. I know that the gem cutters insist that a machine can’t do it, but they would, wouldn’t they?
05:06 PM Tom_L: they'd be out of a job if they didn't
05:08 PM XXCoder: dunno
05:08 PM XXCoder: the part about meeting the point is interesting
05:08 PM XXCoder: youre probably right
05:08 PM XXCoder: its very visual thing though might leave pits and chips lol
05:13 PM JT-Cave: https://matrix.org/_matrix/media/r0/download/matrix.org/YfHYjbrBLvaMbBAnWvTjmDsc/20200830_160654.jpg
05:14 PM JT-Cave: https://dpaste.com/5KMYBBK9W sunrise with the apa102
05:14 PM XXCoder: funny how ancient and new is worked together
05:14 PM Tom_L: different part from the other day?
05:15 PM JT-Cave: no, just the finished part with the cover in place
05:15 PM skunkworks: finally back to radius on polygon.. one step at a time
05:15 PM Tom_L: heh
05:17 PM skunkworks: ./core_sim.hal:13: Unknown command 'Loadrt'
05:17 PM skunkworks: that threw me for a loop for a couple of seconds.
05:22 PM skunkworks: net
05:32 PM robotustra: https://imgur.com/a/Ix4A6dR
05:32 PM robotustra: preliminary assembled
05:37 PM robotustra: need to lower one plate to 0.05mm
05:43 PM JT-Cave: looks massive
05:43 PM JT-Cave: a gear box?
05:44 PM skunkworks: lathe head? Timkins?
05:44 PM robotustra: yes
05:44 PM robotustra: lathe head
05:45 PM skunkworks: nice!
05:45 PM robotustra: don't have a spindle though
05:45 PM robotustra: will use this stainless steel pipe
05:47 PM robotustra: now need to think how to merry this thing with the rest of the frame
05:48 PM robotustra: also need to know what grease will be good for these bearings
06:01 PM JT-Cave: getting close on the sunrise sunset apa102 led project
06:01 PM XXCoder: ya saw your program. getting nicely done!
06:02 PM JT-Cave: I can power 2 meters with the 10A power supply so I'm thinking of cutting the string in half... otherwise I have to add a second power supply at the 2m point
06:03 PM XXCoder: would it be bright enough?
06:03 PM JT-Cave: I think 2m is plenty bright at 4000k
06:04 PM Tom_L: next you need to simulate the moon phases for them
06:05 PM XXCoder: or just do vr lol
06:05 PM XXCoder: yes chicken vr was made
06:06 PM XXCoder: https://www.vox.com/2014/5/19/5729182/the-case-for-raising-chickens-in-virtual-reality
06:07 PM robotustra: why do you need a VR if you are alredy live in simulation?
06:09 PM JT-Cave: lol that's just nuts, you would never get a chicken to do that
06:09 PM XXCoder: seems that it did wotrk
06:10 PM JT-Cave: I doubt it, chickens don't like shit on their head or back lol
06:11 PM XXCoder: i did wonder about that too
06:11 PM XXCoder: and when do they pout it on?
06:11 PM XXCoder: probnably why its not widespready at all
06:12 PM robotustra: https://imgur.com/a/ycdbzfo
06:12 PM XXCoder: whats it?
06:14 PM robotustra: lathe
06:14 PM XXCoder: ahh cool
06:15 PM JT-Cave: nice, your getting there
06:15 PM robotustra: thinking if I have to go with 1/4-20 or M6 bolts
06:16 PM robotustra: slowly but yes, getting there
06:16 PM Tom_L: M6 are finer threads
06:16 PM Tom_L: might hold better
06:16 PM Tom_L: i did M5 M6 on alot of mine
06:19 PM Tom_L: how are you gonna cover your Z rails?
06:23 PM robotustra: I was thinking about rigid cover
06:24 PM robotustra: but it's not so much space in there, may be soft one
06:25 PM robotustra: but rigid cover should not be mothe than 3 cm
06:25 PM robotustra: not wider than 3cm
06:27 PM robotustra: 1/4-20 has bigger cross section
06:27 PM robotustra: than M6
06:55 PM skunkworks: omg - getting close
06:56 PM skunkworks: http://electronicsam.com/images/greenmachine/startrad.png
06:57 PM skunkworks: the logic is there.. Just have to add a bit more math to calculat the radius
07:10 PM robotustra: what the stange g-code?
07:18 PM skunkworks: no - spindle synced motion..
07:18 PM skunkworks: I think you have seen the videos..
07:22 PM robotustra: sync with what?
07:23 PM skunkworks: https://www.youtube.com/watch?v=C0Wz8vnFAQU
07:33 PM roycroft: i reached my goal for the weekend - my bench undercarriage is fully assembled, and the bench top pieces are dry assembled on the base unit
07:34 PM roycroft: https://www.roycroft.us/Bench/Bench-Front.jpeg
07:34 PM roycroft: https://www.roycroft.us/Bench/Bench-Back.jpeg
07:34 PM roycroft: https://www.roycroft.us/Bench/Bench-Shelf.jpeg
07:38 PM skunkworks: nice!
07:38 PM roycroft: thanks
07:38 PM roycroft: it's coming along ok
07:40 PM roycroft: i have 123+ hours of work into it this month so far, and 158 hours total
07:40 PM roycroft: and probably 50 hours to go
07:41 PM roycroft: maybe more, depending on how difficult the wagon vise and the right side end cap are to fit up
07:41 PM robotustra: what is the price of similar workbench?
07:41 PM roycroft: but today was the first time i had all the basic components assembled
07:41 PM roycroft: retail?
07:41 PM robotustra: yes
07:42 PM roycroft: at least $5000
07:42 PM roycroft: and it would not be as robustly built
07:42 PM robotustra: 30$/h
07:42 PM roycroft: less $1000 for the timber and $725 for the vise hardware
07:43 PM robotustra: it means that the work costs ~$20/h
07:44 PM roycroft: that might be about right
07:44 PM roycroft: the thing is, i would not be able to purchase a bench just like that
07:44 PM roycroft: unless i had someone make it for me as a bespoke piece
07:44 PM robotustra: and we didnt cound shipment
07:44 PM roycroft: in which case the price would probably be more like $8000
07:44 PM roycroft: or more
07:44 PM robotustra: yes
07:45 PM robotustra: it gives us already $40/h
07:45 PM robotustra: it means that to make it yourself will be cheaper than to buy it
07:45 PM roycroft: i haven't weighed all of it, but based on some parts that i did weigh i estimate the bench will weigh about 225kg when it's finished
07:46 PM roycroft: if i were to buy a commercial bench it would weigh more like 125kg
07:46 PM roycroft: and it would be nice
07:46 PM roycroft: and it would be pretty solid
07:46 PM roycroft: but that extra 100kg means that i can work as hard as i can on the bench and it won't move at all
07:47 PM unterhaus_: That's a nice looking bench. I built a scandanavian style bench. Next project is to put a cabinet underneath
07:47 PM robotustra: extra 100 kg means it will burn a bit longer
07:47 PM roycroft: which is exactly what i want - that was my first design creterion
07:48 PM roycroft: unterhaus_: about 20 years ago i started building a scandanavian bench, but after i started it i wasn't sure if it was the design i really wanted
07:49 PM roycroft: the roubo style, of which mine is a modification, is a lot more flexible than a scandanavian bench, imo
07:49 PM Tom_L: skunkworks, what are these odd shapes we strive to optain!!!
07:49 PM roycroft: i still have the bench screws i bought for the scandanavian ench
07:49 PM roycroft: i'm going to ebay them soon
07:49 PM Tom_L: roycroft, nice bench so far
07:50 PM roycroft: thanks
07:50 PM roycroft: i could go all pee wee herman and do the tequila dance on top of it and it would not move at all
07:51 PM skunkworks: Tom_L: I don't know - seems like there should be an option for a radius at the polygon corners.. On the lathe for sure it should help the instant change in direction..
07:51 PM Tom_L: if you decide to, just don't yt it
07:51 PM skunkworks: Not sure about the mill boring
07:51 PM roycroft: i don't do that stuff
07:51 PM roycroft: the tequila dance or the making videos stuff
07:51 PM Tom_L: skunkworks, are those arcs or line segments?
07:51 PM roycroft: i'm just saying i could
07:52 PM skunkworks: well - they are not tangent arcs yet - that is me getting the logic down so the program knows when to start the radius. Some trig...
07:52 PM Tom_L: i was out eyeballing my spindle and i think i can just barely fit 15mm belts on it and still have room for the encoder pulley
07:53 PM Tom_L: so i think i'll go with GT2 5mm pitch
07:53 PM skunkworks: so what you see in the screen shot is at the point the arc should start - it just keeps the radius of the current polygon..
07:54 PM skunkworks: *current radius of the polygon
07:54 PM roycroft: i think i need to design a caster system that i can attach to the bench when needed to move it around
07:55 PM Tom_L: roycroft, slide an L under each end and have a lever that lifts it up
07:55 PM roycroft: yeah, something like that would work
07:56 PM roycroft: i don't want anything permanently attached to it
07:56 PM roycroft: and i don't want to bore any holes for attaching the casters
07:56 PM roycroft: so something i can slip under and lift it is what i have in mind
07:57 PM Tom_L: https://geppertlumber.com/media/shop/product/gallery/0000/0697.normal.jpg
07:57 PM roycroft: i probably won't move it all that often anyway
07:57 PM Tom_L: only short
07:57 PM roycroft: that's a piano mover
07:57 PM Tom_L: it's adaptable
07:57 PM roycroft: yeah
07:58 PM Tom_L: raise the L up to fit under the shelf
07:58 PM roycroft: yes, i'd rather lift by the end stretchers than by the legs
07:58 PM Tom_L: then there is no lifting, not even to get it under the legs
07:59 PM roycroft: i need to focus on finishing the bench itself first though
07:59 PM roycroft: there's a *lot* of work yet to do
08:00 PM roycroft: but it sure is nice seeing the whole thing together for the first time
08:00 PM roycroft: well, the main parts together
08:00 PM roycroft: enough that it looks like a bench
08:03 PM XXCoder: LOL https://media.discordapp.net/attachments/749283134893457459/749796053737275422/image0.png?width=522&height=593
08:05 PM jymmmm: Well, it's orange, not red, so just a warning light.... low air maybe?
08:05 PM XXCoder: lol
08:21 PM -!- #linuxcnc mode set to +v by ChanServ
09:24 PM Rab: Today in noteworthy control surfaces: https://smile.amazon.com/dp/B01G3UTBHS
09:41 PM robotustra: does anybody know about taper bearings greasing?
09:45 PM Tom_L: robotustra, just auto wheel bearings
09:46 PM Tom_L: might be similar
09:47 PM robotustra: do you know particular one? because I have no idea which one could work up to 3000 rpm
09:48 PM Tom_L: https://www.skf.com/us/products/lubrication-management/lubricants/lubricant-selection
09:49 PM robotustra: oh, the bearings could be heating during the work
09:51 PM robotustra: I think the usual general purpose should work :)
09:52 PM Tom_L: so do i
09:52 PM Tom_L: if you're worried about it, get truck axle grease
09:52 PM Tom_L: regular is pretty 'stringy' anyway
09:52 PM Tom_L: sticks well
09:54 PM robotustra: is it flow out the bearings while heated?
09:54 PM Tom_L: not alot
09:54 PM roycroft: all i know is that you want to use a not moly grease with tapered roller bearings
09:55 PM Tom_L: nope
09:55 PM roycroft: if this is on a spindle, and you can stick an oil cup or ball fitting nearby, you can oil the bearings instead of greasing them
09:56 PM robotustra: https://www.amazon.ca/Allstar-ALL78241-Timken-Premium-Bearing/dp/B006K8VIAM/ref=sr_1_7
09:56 PM robotustra: looking into this
09:56 PM roycroft: most tapered roller bearings are rated at higher rpm when using oil instead of grease
09:57 PM robotustra: to make a permanent oil flow I need an oil pomp and plumbing, I'm not there yet
09:57 PM roycroft: if that stuff is clear red it's probably fine
09:57 PM roycroft: if it's "milky" red, i.e. opaque and whitish, it probably has moly
09:57 PM Tom_L: yeah i'd say that should work
09:57 PM robotustra: ok, I'll try that
09:57 PM Tom_L: timken makes roller bearings. i'm sure they know what they're doing
09:58 PM robotustra: yep, I have they bearings
09:58 PM roycroft: yeah, it has lithium, not molybdenum
09:59 PM robotustra: by the end I'll probably have a liquid oil coming to bearings, ball screws and rails
09:59 PM robotustra: but not for the beginning
10:01 PM robotustra: one more question - how heavy the bearing should be packed with the grease?
10:01 PM robotustra: does it change anything?
10:01 PM robotustra: or it will squeeze out excess of the grease?
10:03 PM Tom_L: i'd use plenty but not 'full'
10:03 PM Rab: robotustra, what's the application?
10:03 PM robotustra: bearings for spindle
10:03 PM robotustra: tapered bearings
10:04 PM robotustra: 29585
10:04 PM Tom_L: Rab, these: https://imgur.com/a/ycdbzfo
10:05 PM Rab: aha
10:05 PM Rab: That's a mighty bearing, what's the design spindle through diameter?
10:06 PM robotustra: about 40mm
10:06 PM robotustra: I'm planning to use 5 or 6" diameter chuck
10:07 PM robotustra: https://i.imgur.com/hre9pIE.jpg
10:08 PM robotustra: not sure yet how I'll attach backplate to this pipe
10:08 PM robotustra: but I will not run it faster than 2000 rpm anyway
10:09 PM robotustra: 6" chuck has 45 mm through hole in it, I'll use 4 jaw chuck
10:10 PM robotustra: and I still thinking how to place it as close as possible to the bearing
10:15 PM Rab: robotustra, weld a heavy flange onto the pipe and when the lathe is basically operational, turn it flat?
10:16 PM robotustra: I don't know any welder in Mtl
10:17 PM Tom_L: if you want it versatile, thread the pipe with a butt flange on the end then make a threaded plate to fit
10:17 PM robotustra: it should be welded in the lathe
10:18 PM robotustra: I have threaded back plate, the problem is 1) I'm not so experienced to cut threads
10:18 PM robotustra: 2) I will have only 1 chance to thread it
10:18 PM robotustra: 3) most probably I'll turn parts CCW and the backplate will unscrew all the time
10:19 PM Tom_L: not normally
10:19 PM Tom_L: both the ones i have are threaded on
10:20 PM Tom_L: the lathe chuck and the rotary head one
10:21 PM robotustra: how are they mounted?
10:21 PM Tom_L: threaded on
10:21 PM Tom_L: like i said above
10:21 PM robotustra: I wat thinking about taper + key
10:22 PM Tom_L: my lathe chuck has never come loose
10:24 PM robotustra: ok, gotta go, work tomorrow
10:25 PM robotustra: thanks guys
10:25 PM Tom_L: me too
10:25 PM Tom_L: bye
10:25 PM robotustra: go to sleep and think about grease
10:28 PM skunkworks: http://electronicsam.com/images/greenmachine/radcool.png
10:30 PM roycroft: grease is the word, is the word that you heard
10:30 PM roycroft: it's got groove, it's got meaning
10:31 PM Tom_L: skunkworks, nice optical illusion. if you look long enough, the outter lines look bent too
10:31 PM Tom_L: maybe they are...
10:31 PM Tom_L: at least the inner 2
10:34 PM skunkworks: it is adding cutter diameter - so yes - the inner lines start bending inward
10:38 PM skunkworks: with and without a .02 radius...
10:38 PM skunkworks: http://electronicsam.com/images/greenmachine/withwithout.png
10:39 PM skunkworks: (the overshoot was what I was hoping to help remove)
10:40 PM Tom_L: seems like it might have worked
10:40 PM skunkworks: seems..
10:40 PM skunkworks: needs a bit of work.. Now if I set the radius to 0 - it turns into an etch-o-sketch...
10:40 PM skunkworks: ;)
10:41 PM Tom_L: heh
10:41 PM skunkworks: so - the above screenshot was .0001 vs .02"
10:41 PM Tom_L: fine line there
10:41 PM skunkworks: I think the logic falls through at the points. I have an idea
10:42 PM Tom_L: what's the next best way to secure a pulley if you can't key it?
10:42 PM Tom_L: no room for a posi lock thingy either
10:43 PM skunkworks: shrink fit?
10:43 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/new_pulleys/spindle_enc8.jpg
10:44 PM Tom_L: i thought about turning a grub screw down and pinning it
10:44 PM Tom_L: just enough to remove the threads
10:46 PM Tom_L: it needs to be serviceable so i don't think shrinking it would do
10:49 PM skunkworks: the spindle pully on the emco is just 2 set screws.. I figured if I had issues - I would figure something else out
10:50 PM skunkworks: but that is only 400w or something likethat
10:50 PM Tom_L: there are set screws on those too
10:50 PM Tom_L: i'm afraid they won't hold with the new motor
10:50 PM Tom_L: if it was a solid shaft, i'd key it
10:51 PM Tom_L: even considered a half moon key