#linuxcnc Logs
Jun 23 2019
#linuxcnc Calendar
01:24 AM Deejay: moin
05:29 AM jthornton: morning
05:52 AM Deejay: yo
06:26 AM Tom_L: morning
06:33 AM veek: what exactly is a core? You have a pattern and you fill around it, thus creating a dimensional negative in sand.. then? https://i.vgy.me/C2Q5jj.png
06:34 AM XXCoder: core is where you want a hollow
06:34 AM veek: if you want a hollow within the formed part.. you insert spacer shapes into the hollow?
06:34 AM XXCoder: suppose for some very weird reason you want to cast a pipe with sphere in middle
06:34 AM XXCoder: you would need mold as usual, but you want shift in middle and sphere hollow inside also
06:34 AM XXCoder: thats what core is good for.
06:35 AM XXCoder: so you make core using special sand that hardens by co2
06:35 AM veek: lemme think - sec
06:35 AM gloops: right XXCoder
06:35 AM XXCoder: you first set sand cast for external using mold, then you place core inside
06:35 AM gloops: hollow forms
06:36 AM XXCoder: typically you want mold to have extensions so you have space to hold core in
06:36 AM XXCoder: veek: https://www.youtube.com/watch?v=l3opkzAvVQk
06:36 AM XXCoder: very good video
06:36 AM veek: thanks
06:36 AM XXCoder: no problkem
06:37 AM gloops: with plastic this is often superceded with bubble injection moulding
06:37 AM gloops: like a plastic bottle
06:41 AM veek: man that guy is brilliant - i've already seen some of his videos - in fact my question was after watching one of his videos
06:41 AM XXCoder: hes quite old hand on machining and making parts
06:48 AM XXCoder: hopefully it helped :)
06:51 AM veek: yeah core's just a solid you insert into the mold, to make a hollow in your final solid-part
06:51 AM XXCoder: indeed
06:51 AM XXCoder: sometimes only way
06:51 AM gloops: green sand, or play sand mixed with kitty litter
06:52 AM XXCoder: that example isnt exactly complex, but he did others that cant be done in other ways
06:52 AM XXCoder: or can be done, but with long hours of cnc
06:52 AM veek: green sand = play sand + Sodium beanotite (yeah i saw his video on making greensand)
06:56 AM jthornton: I got a well pressure tank from a buddy to make an oil fired furnace with
06:57 AM XXCoder: fancy
07:02 AM gloops: veek cats litter stuff is the right clay, crush it into powder and mix the sand, which needs to be clean and soft - the stuff they sell for kids sand pits
07:02 AM gloops: then just moist enough - it will hold a form when pressed together, including detail
07:03 AM XXCoder: green sand seem to be good for details
07:04 AM XXCoder: though never cast stuff myself, though love myfordboy heh
07:04 AM gloops: just make sure it is fired and bone dry heh
07:15 AM veek: gloops, you got to have a lot of knowledge to even do simple things.. look at his cool furnace and all the cans and knicknacks he's got
07:20 AM veek: i've put up some of my notes online - in my very bad handwriting.. basically i'm going to condense all the crap i learn into a course on DIY :p https://www.reddit.com/r/indianripoff (it's just a collection of links with some notes on sed/bash/awk) (and I couldn't think of a better name :p)
07:45 AM veek: what does he mean by: 'Patterns also have core prints that create registers within the molds into which are placed sand cores'
07:45 AM veek: 'registers' ?
07:45 AM veek: 'core PRINTS'?
07:57 AM SpeedEvil: veek: It's foundry talk.
07:58 AM SpeedEvil: registers = features that let you align the sand core of a mold with the outer mold to align the two so you get the proper geometry
08:14 AM veek: thanks
08:16 AM gloops: yes, alignment, maybe holes a pin is inserted into, slots, grooves etc, the corresponding features 'register'
08:17 AM gloops: excellent work btw veek
08:20 AM * veek woofs
09:20 AM SpeedEvil: https://www.youtube.com/watch?v=nQaAKUAzK0w - also.
09:20 AM SpeedEvil: This colechester lathe video had footage of that.
09:20 AM SpeedEvil: core boxes with alignment features that the cores just slid neatly into.
09:21 AM SpeedEvil: @4:44
09:26 AM * SpeedEvil is moderately irritated to not be able to find contempory prices for colchester lathes in 1960.
11:14 AM jym: Morning folks
11:14 AM Deejay: evening jymmmmmm
12:28 PM Elmo40: i have a stepper motor question. if a NEMA17 and NEMA23 are both listed as 1.5A then they can both be driven by the same size stepper driver? Current is the limiting factor for drivers, right?
12:35 PM andypugh: Annoying, I don’t have a keyboard that works with my Pi.
12:35 PM Elmo40: really? that is odd.
12:35 PM Elmo40: the Pi should handle any USB kb.
12:36 PM jym: Unless he only has woreless keyboards
12:36 PM jym: wireless even =)
12:36 PM Elmo40: that is a limitation?? sad.
12:37 PM andypugh: The Pi does not work with the Appe wired aluminium keyboards. Of which I have two, specifically for plugging in to random SBs
12:37 PM andypugh: (SBCs)
12:37 PM andypugh: I do have a bluetooth keyboard, but I _think_ it only works with its own dongle.
12:37 PM jym: andypugh: THAT's totally weird... is it the full sized wired apply kybd?
12:37 PM andypugh: Yes, that one.
12:38 PM andypugh: I googled, and other folk have had the same problem.
12:38 PM jym: andypugh: totally weird, but good to know, I have two of them
12:39 PM andypugh: Yes, they are nice keyboards and availabe inexpensively.
12:39 PM jym: especially when you find one at a yard sale for $4, YAY BACKUP KYDB
12:40 PM jym: They replaced my IBM Model M's surprisngly
12:41 PM jym: andypugh: If it's BT kybd, it should work with any BT, they might have just included the dongle for conveinace
12:42 PM jym: If it's a "wireless" kybd (non-BT), then that's a different thing
12:43 PM jym: I like my 2" wireless kybd, then no BT drivers, etc need to be dealt with, works in BIOS/DOS even
12:45 PM jym: andypugh: It's funny, I pair my Apple BT kydb with my android phone, works great, especially since I've never used it with anything else =)
12:46 PM jym: Elmo40: At the same time? no, individually, yes.
12:46 PM jym: Elmo40: Steppers are usually driven at 20 to 25 times therm rated voltage though.
12:47 PM jym: their*
01:09 PM jym: Elmo40: This is a VERY GOOD READ, even if you might have to read it a few times to "get it"... https://www.geckodrive.com/support/step-motor-basics.html
01:16 PM andypugh: OK, after some guesswork about how to get in to pairing mode I have a BT keyboard connected to my Pi.
01:16 PM andypugh: Oh, and the Pi is now in the Mac-Pro style case
01:17 PM andypugh: https://photos.app.goo.gl/4bAHHSk3R24t2tev8
01:25 PM jthornton: are the holes on the bottom just offset from the top by 1/2 a hole?
01:26 PM andypugh: Yes and no
01:26 PM andypugh: The holes on top are the nodes of an equlateral triangle and those below are the centre of the triangle.
01:29 PM andypugh: Someone has quite a write-up here: http://saccade.com/blog/2019/06/how-to-make-apples-mac-pro-holes/
01:31 PM jthornton: ah I can see that now
01:34 PM jthornton: hmm I
01:34 PM Tom_L: andypugh that's aluminum?
01:34 PM phipli: I don't know Tom_L - I've never heard of it
01:34 PM jthornton: 've managed to find a 75mm cpu fan and harvested it without breaking it now to see if it fits the CHNC cpu heat sind
01:35 PM jthornton: hey phipli
01:35 PM phipli: evening jt
01:36 PM jthornton: stuck my finger into the CHNC cpu fan blade while trying to stop it to see if it was the noisy one turned out it's 75mm so not too common with the junk I have is mostly 80mm
01:37 PM Tom_L: hopefully it stopped
01:37 PM Tom_L: and you still have 10 digits
01:38 PM phipli: (or however many he had before this incident at least)
01:40 PM andypugh: Tom_L: No, it’s aluuminium.
01:41 PM andypugh: No it isn’t
01:41 PM andypugh: Bah!
01:41 PM andypugh: What a time for a typo!
01:41 PM andypugh: Aluminium
01:41 PM Tom_L: :)
01:42 PM Tom_L: did you use a ballnose and plunge cut?
01:42 PM gregcnc: extraneous letters be gone!
01:42 PM andypugh: Yes, just a 5mm ball-nose and spacing chosen to suit.
01:43 PM * Tom_L still isn't sure what material it was after that fiasco
01:43 PM Tom_L: kinda like making a turners cube
01:47 PM JT-Shop: phipli: a blade brokr off...
01:49 PM jym: andypugh: Oh you made the case after all?! NICE =)
01:50 PM andypugh: Yes. I couldn’t resist once I was fed the idea by perry_j1987
01:50 PM jym: andypugh: haha =)
01:50 PM jym: nice job!!!
01:50 PM JT-Shop: yippie the CHNC is back alive
01:57 PM jym: JT-Shop With Circa 1980's BIOS???
02:03 PM Tom_L: andypugh the cutter would somewhat be dependent on the material thickness to get the desired effect wouldn't it?
02:06 PM andypugh: Yes, you want a cutter radius such that a hemisphere is nearly, but not quite, all the way through.
02:06 PM andypugh: I used 3mm thickness and 2.5mm radius.
02:16 PM Tom_L: did you treat them like drilled holes? and did you use any dwell at the bottom?
02:26 PM andypugh: I didn’t dwell. Perhaps I should, as the ductility of the material meant that the holes where the spheres met didn’t clear. I had to poke them all out with a scriber.
02:27 PM andypugh: It does look really nice. Photos don’t really do it justice.
02:28 PM jym: andypugh: YOu'll just have to make 10 more and toss up on ebay now =)
02:29 PM andypugh: Apple might send the boys round….
02:29 PM Elmo40: jym, yes, individually. i was thinking of increasing the X-axis strength with a NEMA23 of similar amperage. Didn't know if I needed to increase the driver, too.
02:30 PM jym: andypugh: Just like a cars body design cna't be copyrighted,
02:31 PM jym: andypugh: maybe make the holes hexagon'ish
02:32 PM jym: andypugh: If you include a pi-hackintoch, then maybe you'll get a knock on yur door ;)
02:32 PM Elmo40: is there an acceleration loss with NEMA23 over 17 due to larger diameter (more inertia needed)?
02:34 PM Elmo40: andypugh, that is a lovely case! I'll take 5... ;-)
02:35 PM andypugh: Elmo40: I have seen an accel less going from 2 stacj to 3 stack NEMA23. So I think the answer is likely to be a strong “yes”
02:36 PM Elmo40: there isn't any formula or guideline on how much accel a motor can sustain? I know there are torque charts but how to use that to figure out acceleration? True, load is always a variable...
02:37 PM andypugh: The load inertia is the main variable.
02:37 PM DaViruz: angular acceleration = torque/moment of inertia
02:37 PM Elmo40: and inductance?
02:37 PM Elmo40: doesn't that also play a factor?
02:38 PM andypugh: But, apparently it is important to match the motor and load inertia. It’s like impedance-matching in electrical circuits,
02:38 PM DaViruz: figuring out the moment of inertia can be kind of tricky.
02:38 PM andypugh: I saw the maths once, but rather glossed over it so didn’t learn how to explain it.
02:38 PM Elmo40: yeah... impedance matching. i haven't done those calculations since university electronics course!
02:38 PM SpeedEvil: If you get resonance, that's a further issue.
02:39 PM Elmo40: resonance can occur at a specific speed. just how everything 'hums' together.
02:39 PM SpeedEvil: It only needs to have a rotary resonance of 1/4 step at any specific RPM (excitation frequency) before the torque falls to zero.
02:39 PM Elmo40: rubber mounts help. i added them and resonance really dropped.
02:40 PM Elmo40: don't want 0 torque ;-)
02:40 PM SpeedEvil: Depends what's resonating. Rotating mass can actually make your thing go faster.
02:43 PM Elmo40: so much can change when going form 17 to 23!!
02:44 PM Elmo40: non the least motor mounts and output shaft differences.
02:55 PM DaViruz: i think generally in machine tools, system inertia will be much greater than motor inertia, so higher motor torque always wins out even if the motor inertia increases more
02:59 PM Tom_L: http://image.sciencenet.cn/olddata/kexue.com.cn/bbs/upload/15714Calculating%20Motor%20Start%20Time.pdf
02:59 PM Tom_L: Elmo40
03:16 PM andypugh: Take it slowly, try a NEMA 18 first.
03:28 PM perry_j1987: haha andypugh told you people would be trying to put in orders
03:31 PM Elmo40: Tom_L, excellent read!
03:31 PM Elmo40: DaViruz, i figured as much... hence why a 23 is a win for me. same current load but more power.
03:33 PM Deejay: gn8
03:34 PM gregcnc: damper mounts for a stepper?
03:36 PM Elmo40: the ringing of the motor can translate into the equipment. it can be seen in a facemill's cut.
03:36 PM Elmo40: of course, depends on resonance...
03:37 PM gregcnc: and what about accuracy?
03:37 PM Elmo40: what about it? there is no twist motion in these couplers. there shouldn't be, anyways.
03:37 PM gregcnc: there has to be
03:38 PM Elmo40: find one and clamp it in a vise. try to flex it. they are very stable.
03:39 PM Elmo40: might not need them on larger machines with nema34 and up due to the equipment being moved has a large mass.
03:39 PM Elmo40: it is for lighter things. nema17, mainly. little mini-mills or 3D printers.
03:40 PM gregcnc: what kind of machine is this?
03:43 PM perry_j1987: hard to get myself going today
03:43 PM perry_j1987: guess i'll do some chip cleanup
03:44 PM Elmo40: clean-up is always a good thing.
03:45 PM andypugh: perry_j1987: If I was to make any more I would definitely take the time to put on the high-speed milling head
03:45 PM Elmo40: gregcnc, just a wood carver. want to make a large table system (4' x 8') and was thinking of a nema23 to move the gantry.
03:45 PM gregcnc: ask apple for permission before they come get you
03:46 PM Elmo40: is the cut sphere pattern copywritten?
03:46 PM Elmo40: i don't think they can copywrite a geometric shape.
03:48 PM perry_j1987: Elmo40 they patented square corners or something like that didnt they lol
03:49 PM perry_j1987: luckily im safe with my machines... they cant make a square corner :P
03:53 PM andypugh: perry_j1987: Ah, but, they don’t use square corners.
03:53 PM andypugh: Nor, even, corners with tangent circles.
03:54 PM andypugh: In mechanical terms they use finite-jerk circles, there is no abrupt change in curvature, so shadows and lighting look subtly different when they hit the parts.
03:54 PM gregcnc: I'ts possible they didn't get the design patent because cheese graters already exist
03:56 PM XXCoder: lol
05:56 PM skunkworks: https://www.youtube.com/watch?v=JsGqXhPGLiM
06:00 PM jthornton: afaik the main difference between a stepper and servo configuration for 2.8 mesa ethernet cards is the lack of step and direction for the servo...
06:20 PM * jym is gonna try this now... https://youtu.be/Kn3OMrxkfb8?t=1131
06:23 PM andypugh: This is a video I just made of modelling the grille pattern in Fusion360: https://www.youtube.com/watch?v=OfaQbrQO2AQ
06:23 PM perry_j1987: if usps delivers on sunday it'd be nice if they'd do it to the PO Boxes too...
06:37 PM mmachines: watching the grille video - thanks
06:39 PM mmachines: oops HAL: ERROR: signal name 'scopeinterface.microscope.rapid-scope-zero-link2' is too long(get_value): panelinterface.pyvcp.rapid-g5x-zero = False(from comp): panelinterface.pyvcp.rapid-g5x-zero = False(get_value): panelinterface.pyvcp.touchoff-xy = False(from comp): panelinterface.pyvcp.touchoff-xy = False(get_value): panelinterface.pyvcp.touchof
06:39 PM mmachines: f-xyz = False(from comp): panelinterface.pyvcp.touchoff-xyz = False
06:40 PM mmachines: bad copy/paste
06:40 PM mmachines: HAL: ERROR: signal name 'scopeinterface.microscope.rapid-scope-zero-link2' is too long
06:40 PM Tom_L: pastebin.ca etc
06:40 PM mmachines: guess I need to shorten these names
06:40 PM mmachines: ok
06:41 PM Tom_L: paste.debian.net or paste.ubuntu.com
06:42 PM mmachines: Tom_L: thanks - I intended to only copy/paste only the single line with the signal too long message
06:43 PM mmachines: I was building the net names automatically and they are getting too long
07:02 PM mmachines: max = 47 characters for a pin name apparently
07:42 PM jym: OMG so full
07:45 PM jym: Not bad, but didn't taste quite the same as the video, but I only used one stick of butter between two steaks, mybe better with two sticks of butter instead
08:58 PM Elmo40: anyone using the chinese TB6600 drivers?
09:00 PM Elmo40: Or the ST-4045
09:21 PM CaptHindsight: the linear profile bearing rail cut pretty easily with a bandsaw
09:22 PM CaptHindsight: i usually order it cut-to-length but it much faster when you buy full lengths and cut it yourself
09:22 PM Tom_L: i used my cutoff grinder on mine
09:23 PM CaptHindsight: yeah, i was trying to lose as little material as possible
09:24 PM Tom_L: .040" blade
09:25 PM CaptHindsight: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?WCH at some point somebody snuck the patch for these cards into Debian kernels
09:27 PM CaptHindsight: https://www.amazon.com/Parallel-Printer-Express-Adapter-Converter/dp/B00BAO55QY/ref=sr_1_1?keywords=WCH+CH382L&qid=1561342926&s=electronics&sr=1-1
09:27 PM CaptHindsight: https://www.ebay.com/itm/PCI-E-Express-DB25-Parallel-LPT-Printer-Port-Card-WCH-CH382L-Chipset-US-Stock/183132128667
09:28 PM CaptHindsight: US $11.49 shipped
09:30 PM CaptHindsight: https://www.ebay.com/itm/Parallel-Port-DB25-25Pin-LPT-Printer-to-PCI-E-Express-Card-Converter-AdaptG/293081576955 US $6.56 shipped from CHINA
09:44 PM CaptHindsight: running the Buster 4.19 preempt_rt kernel on Stretch, seems to behave on the celeron 784
10:55 PM * flyback bites CaptHindsight
10:55 PM flyback: pay attention to the chip used on it
10:55 PM flyback: that's the most important part
10:55 PM flyback: although never underestimate the power of chinease to fuck up a unbreakable chip
10:55 PM flyback: by skipping resistors, etc
10:56 PM flyback: I think oxford or somehing like that makes really good ones
10:56 PM flyback: and one that starts with a m
10:56 PM flyback: man I have this one boca serial card
10:56 PM flyback: it has it's own risc cpu
10:57 PM flyback: there is literally like 4 barrel diodes on it
10:57 PM flyback: 2 for every serial pin
10:57 PM flyback: surge eaters
10:57 PM flyback: shame the drivers don't go past nt4
10:57 PM flyback: it's rock solid it c ould probably take a small indirect ligbtning hit
10:58 PM flyback: solrry
10:58 PM flyback: 4/40
10:58 PM flyback: well 50 I geuss
10:58 PM flyback: something liek that
10:58 PM flyback: I can't remember how many pins the output was
10:58 PM flyback: if it's 25 it would be 50 diodes
10:58 PM flyback: two db9's would be 36
11:01 PM flyback: looks like linuxcnc has documented the chip on that ebay link CaptHindsight
11:01 PM flyback: so should be fine if they say it'
11:01 PM flyback: is fine
11:01 PM flyback: I realize you guys know 10,000 more than me on cnc stuff
11:02 PM flyback: i'm just telling you what I know from the pc hardware side
11:02 PM flyback: the good shit that can take some rough handling