#linuxcnc Logs
Apr 13 2019
#linuxcnc Calendar
02:14 AM Deejay: moin
04:34 AM _abc_: Hi. Does anyone know if the PRU's in the Beaglebone are used in the cnc-interpreter versions which run on beaglebone?
04:35 AM _abc_: Also if those can be "slaved" to linuxcnc? I.e. is there some way to use a cnc controller fw running on a bb with a suitable "shield" for cnc drive, to linxucnc on a pc?
04:35 AM _abc_: And if that fw on the bb uses PRU based realtime code?
04:37 AM _abc_: Wow, Satudays kill freenode :)
04:39 AM _abc_: And I meant "cape" not "shield" heh
04:42 AM _abc_: http://www.righto.com/2016/08/pru-tips-understanding-beaglebones.html reading this it makes total sense to use a bb as cnc driver, the PRUs were made for this!
04:58 AM _abc_: Ha there's even a FORTH compiler for PRU https://github.com/biocode3D/prufh
04:58 AM _abc_: </me will lurk waiting for intelligent comments>
04:59 AM * XXCoder keeps hiding
04:59 AM XXCoder: heh
04:59 AM * _abc_ does not notice XXCoder at all
04:59 AM XXCoder: whats up? unfortunately i dont know about your questions so dunno
04:59 AM _abc_: Thank you for adding bytes to the lack of information :)
05:00 AM _abc_: No, thanks for being nice.
05:00 AM XXCoder: lol
05:02 AM _abc_: See what I did with that judiciously placed comma there? :)
05:04 AM XXCoder: indeed
05:04 AM SpeedEvil: PRUs are moderately featured microcontrollers.
05:05 AM _abc_: Yes I just read about the PRU programming
05:05 AM SpeedEvil: Or just use one core of an ESP32.
05:05 AM SpeedEvil: (45% kidding)
05:08 AM _abc_: SpeedEvil: that would be 50% kidding, no?
05:08 AM _abc_: One core of 2?
05:08 AM SpeedEvil: - ESP32 has ethernet if you don't want to trust wifi machining, quite a lot of RAM, and one core entirely free at 240MHz.
05:08 AM _abc_: I just discovered that qemu has qemu-system-xtensa
05:09 AM SpeedEvil: The determinism of timing of that core is not as good if you use off-chip RAM and ROM, and it has about ten pins, but still, it could be very useful.
05:09 AM _abc_: wifi machining just like ether machining, unless on dedicated segment, will be used to pump data buffers on the realtime embedded side, not real realtime
05:09 AM SpeedEvil: Internal RAM is about the same as the PRU core IIRC
05:10 AM SpeedEvil: It's been a while since I read up on the PRU
05:10 AM _abc_: dedicated segments means 1:1 ether cable, interface
05:10 AM _abc_: SpeedEvil: PRU has 8k program ram 8k data ram and is Harvard
05:10 AM _abc_: brb
05:19 AM _abc_: SpeedEvil: make that: Sitara as used in bb has 8k+8k, other devices differ, lowest is 4k program + 0.5k data
05:20 AM _abc_: Also apparently PRUs are optimized for write, read has latency. http://processors.wiki.ti.com/index.php/AM335x_PRU_Read_Latencies#
05:20 AM _abc_: Which is sort of a bummer for bit banging in the usual sense of the word, as known from avr and pic and so on.
05:27 AM _abc_: SpeedEvil: also re: ethernet realtime on beaglebone: it has Industrial Ethernet extensions made for this http://www.ti.com/processors/sitara-arm/amic-industrial-ethernet/getting-started.html
05:27 AM _abc_: Of course the other side (linux pc) must also have them to work right.
05:27 AM SpeedEvil: I meant ethernet on the ESP
05:29 AM _abc_: ESP32 is definitely not aimed at industrial. I am pleasantly surprized the cpu in the bone was not selected by "consumer" features only, like hdmi and other hipster factors.
05:29 AM _abc_: Did not know it has advanced industrial features.
05:29 AM _abc_: I just read one needs to disable HDMI to get more PRU pins :) :)
05:29 AM SpeedEvil: In a large way it's a damn shame the Pi came out when it did.
05:29 AM SpeedEvil: It killed a lot of interesting boards and stifled development.
05:30 AM _abc_: Pi is a toy. A dog slow toy with the default linux too.
05:30 AM _abc_: I assume someone did put QNX on it at some time, am curious if that is also slow.
05:30 AM SpeedEvil: yeah. But it's $35. Wait, it's $5. (but you can't buy one)
05:30 AM XXCoder: QNX is $35/$5?
05:31 AM SpeedEvil: The loss leader thing they did for ages with the $5 one could almost have been (and I'm not sure it wasn't) calculated to kill other boards.
05:31 AM SpeedEvil: XXCoder: Pi
05:31 AM _abc_: SpeedEvil: indeed.
05:31 AM XXCoder: CHIP by nextthing tried to ne... next thing
05:31 AM XXCoder: but failed
05:31 AM XXCoder: its stats were underwealming
05:31 AM * SpeedEvil looks at his CHIP
05:32 AM XXCoder: i have a few and 2 pocketchips - one is dead
05:36 AM _abc_: Is the CHIP the BBC one?
05:37 AM _abc_: Is it good for anything? Obviously, nobody thought about it being google-able. The name. As usual. Brilliant for 1850, Lady Ada would have been proud. We're in 2019.
05:37 AM XXCoder: BBC? nah its from NEXTTHING like I said
05:37 AM XXCoder: proper name is c.h.i.p.
05:38 AM SpeedEvil: ChIP is like Pi done on a tiny budget by several guys.
05:38 AM SpeedEvil: And without someone high up in the company making SoCs leading it.
05:38 AM XXCoder: they had massive budgeting problems and crashed
05:38 AM _abc_: http://www.chip-community.org/index.php/Main_Page was
05:39 AM _abc_: Well, R8 is a good cpu afaik
05:49 AM XXCoder: maybe i didnt really get into playing around with it
06:15 AM Tom_L: morning
06:34 AM jthornton: morning
06:34 AM XXCoder: hy jt
06:40 AM * jthornton slept in today
06:40 AM XXCoder: i tried to do that. failed.
06:41 AM jthornton: last time I looked at the clock it was 3am lol
06:42 AM XXCoder: last time i looked at mine its 4:38 am
06:42 AM XXCoder: which is now few seconds ago lol
06:44 AM jthornton: I mean before I woke up lol next thing I knew it was 6:30
06:44 AM XXCoder: pretty deep sleep
06:47 AM XXCoder: talking about sleep, later
06:51 AM jthornton: night
09:29 AM gregcnc: Tom_L jthorton I'm sure you both know just the line that minor diameter controls final pitch diameter is not entirely correct
09:30 AM gregcnc: It works for full profile multi thread mills, but single pitch will be more difficult to setup correctly
09:32 AM gregcnc: I've only caught some of the messages about this, but radius comp is a given to allow adjustments
09:47 AM JT-Shop: it seems to me that pitch diameter should control the tool path but still struggling with that
09:48 AM JT-Shop: also all the examples that I've found turn on cutter comp after moving to the hole location and that won't work with lcnc as it requires a move after G41/42 to be longer than tool diameter
10:02 AM gregcnc: Pitch diamter does control the tool path in the end, but the hard part is having the details of the too to define it correctly
10:02 AM gregcnc: Tool
10:04 AM gregcnc: I use an excel file to create threadmill code and use rad comp in linuxcnc, but i use a diameter offset, not too diameter, which has screwed me more than once when i forgot
10:08 AM JT-Shop: if you don't mind sharing that excel file, I'd like to look at it
10:09 AM JT-Shop: I have a spreadsheet that does a good job of finding the sharp V of the tool based on the diameter and crest
10:09 AM JT-Shop: I'm popping in and out splitting firewood
10:10 AM gregcnc: it's metric for more fun
10:13 AM jthornton: I need metric as well
10:42 AM boddax: i need help about ramp 1.4 to connect
10:44 AM boddax: A-C B-D the stepper coils , ramp 4 pin connector has 1B-2A-1B-1A somebody can tell me how to wire correctly
10:45 AM Tom_L: that sounds like a #reprap question
10:51 AM Tom_L: thread calculations: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=3&cad=rja&uact=8&ved=2ahUKEwjuztqyss3hAhXina0KHRlmDzQQFjACegQIAxAC&url=https%3A%2F%2Fwww.actx.edu%2Fmachining%2Ffilecabinet%2F15&usg=AOvVaw0TWuifiyAgAkLDeFph4YaV
10:53 AM boddax: ah ok Tom_L thanks
10:55 AM Tom_L: PD = (D - 0.64952 p ) where PD=pitch circle D=maj diam p=1/tpi
10:56 AM Tom_L: ISO 898-1:1999
10:57 AM Tom_L: https://www.engineersedge.com/thread_strength/thread_bolt_pitch_circle_diameter.htm
11:01 AM pink_vampire: Hi
11:22 AM _unreal_: low
11:42 AM lcvette_fn: hello
11:43 AM lcvette_fn: logs
11:43 AM logs: lcvette_fn: today's log: http://irc-logs.kcjengr.com/linuxcnc/2019/04/13.html
11:47 AM lcvette_fn: so i found a few ram sticks and had a spare SSD laying around and decided to see if i could salvage any latency improvements from those lowly AMD A6-7400k / ECS A58 FM2+ motherboard pc's I had purchase back in 2016
11:51 AM lcvette_fn: the original test was run and the latency results were staggeringly horrible and unusable neven with mesa hardware if anyone remembers
11:54 AM lcvette_fn: was trying to find those results in the logs but couldn't.. but they were getting up close to a 1,000,0000 in max jitter.. was horrid!
11:55 AM lcvette_fn: so i ran some detailed testing after i saw some improvement from adding in a ssd over the barracuda mechanical HDD and decided to see if some different ram and configuration made a difference as well.
11:56 AM lcvette_fn: it did
11:56 AM lcvette_fn: so i charted it and figured i would share in the event anyone switching to the preempt-RT kernel was also experiencing a jump in latency issues over the RTAI version
11:56 AM lcvette_fn: https://matrix.org/_matrix/media/v1/download/matrix.org/uVORKgYspBLwwVQsiUOgXRZf
11:57 AM pink_vampire: I have a problem with the homing, when I click home he machine move the X axis to the wrong direction
11:58 AM lcvette_fn: pink_vampire you can fix that in the ini
11:58 AM pink_vampire: how?
11:59 AM lcvette_fn: in the homing section for that axis
11:59 AM pink_vampire: what do i need to edit?
11:59 AM gloops: https://www.ebay.co.uk/itm/20mm-linear-slide-guide-shaft-SBR20-1800mm-2rail-4-SBR20UU-bearing-block-CNC-set/113711270690?
11:59 AM gloops: bladdy ell
12:00 PM lcvette_fn: ```
12:00 PM lcvette_fn: HOME_SEARCH_VEL = -3.00 HOME_LATCH_VEL = -.400 HOME_FINAL_VEL = 6.00
12:01 PM lcvette_fn: signs in front of the numbers dictate direction
12:01 PM rmu: lcvette_fn: memory or hard disk should not influence latency jitter in any meaningful way. those numbers look pretty bad IMHO.
12:02 PM lcvette_fn: rmu: these were specifically done to test that and it 100% does
12:02 PM lcvette_fn: and is repeatable
12:02 PM rmu: lcvette_fn: there is something else going on
12:03 PM gloops: if the homing was working, and now doesnt - check power to the switch
12:04 PM Jymmm: ...and loose wires/connections
12:04 PM lcvette_fn: rmu: i have two identical boards and processors and both test the same
12:04 PM rmu: lcvette_fn: what kernel did you use? browser, 2 youtube videos, gimp libreoffice and 10 glxgears seems to be a bit much for a 4gig machine
12:05 PM lcvette_fn: debian stretch latest preempt RT kernel
12:05 PM lcvette_fn: freshly updated and upgraded on a new install
12:06 PM rmu: i suspect thermal issues
12:08 PM lcvette_fn: max jitter jumps to 40k+ just by moving the latency test window before opening anything else
12:08 PM lcvette_fn: on the 4gb sticks
12:09 PM lcvette_fn: it gets up there pretty quick, long before most of those test items have even been opened
12:10 PM rmu: is this running some kind of shared memory graphics?
12:11 PM pcw_home: lcvette_fn: you might try a servo thread only test
12:11 PM rmu: that could be influenced by memory bandwidth. that could be helped with real graphics card
12:12 PM lcvette_fn: chris@lathe1:~$ lspci | grep VGA 00:01.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Kaveri [Radeon R5 Graphics]
12:12 PM lcvette_fn: im not concerned with fixing these, i have since replaced them
12:12 PM lcvette_fn: i was just playing since they were laying around
12:12 PM lcvette_fn: i was impressed the memory and ssd improved everything so much
12:13 PM lcvette_fn: i have since replaced them with the H97M Pro4 and Z97e-itx/ac and G3258 processors
12:13 PM lcvette_fn: which have stupid good latency
12:14 PM lcvette_fn: shout out! thanks to pcw_home
12:15 PM lcvette_fn: pcw_home: how would one test that?
12:15 PM rmu: FWIW numbers i get on an athlon 4450e with on board radeon hd 3200 are much better, system is more than 10 years old
12:16 PM lcvette_fn: completely understand, i think it may just be the mobo and processor combo isn't the best for this application
12:21 PM pcw_home: on the latency test its "latency-test 1ms 1ms"
12:22 PM lcvette_fn: just run that in terminal?
12:22 PM lcvette_fn: i always have run from the menu
12:22 PM pcw_home: on the histogram: "latency-histogram --nobase --sbinsize 1000"
12:23 PM pcw_home: yeah from a terminal
12:30 PM _unreal_: rmu, hello
12:33 PM lcvette_fn: pcw_home: bit confused
12:33 PM lcvette_fn: i got the histogram and can run the latency test, but not sure if i did it right?
12:35 PM lcvette: opened irc on the computer being tested
12:39 PM pcw_home: here's a servo thread only histogram: http://freeby.mesanet.com/dc7800_5.03.png
12:39 PM lcvette: http://i.imgur.com/tCr2PiP.png
12:40 PM lcvette: did i do it right?
12:41 PM lcvette: pcw_home: looks like i have it
12:41 PM lcvette: certainly looks different
12:41 PM pcw_home: Yep, for an idea how infrequent the latency excursions really are, try the linear scale
12:43 PM lcvette: how do i do that?
12:43 PM lcvette: uncheck ylogscale?
12:43 PM pcw_home: yep
12:44 PM lcvette: http://i.imgur.com/hpuURVx.png
12:44 PM lcvette: hows it look in servo thread only?
12:47 PM lcvette: this is with the single 8gb stick of samsung ram
12:47 PM pcw_home: Yeah, that still looks pretty bad but maybe probably usable
12:47 PM lcvette: im gonna throw in the two sticks of 4gb that run in dual channel
12:48 PM lcvette: that was 34% better results before
12:48 PM lcvette: see if it looks the same on the servo thread only histogram
12:48 PM lcvette: brb
12:50 PM pcw_home: Do you have all power management options turned off in the BIOS? ("Cool and Quiet" etc)
12:50 PM lcvette_fn: probably not
12:50 PM lcvette_fn: i can look but i get confused when i go in there
12:53 PM lcvette: back
12:53 PM lcvette: hyperx 4gbx2 installed
01:00 PM lcvette: http://i.imgur.com/lPU115M.png
01:02 PM lcvette: pcw_home: still running, but its a significant difference
01:03 PM lcvette: people say ram should make a difference, but this shows differently does it not?
01:03 PM lcvette: shouldn't*
01:07 PM lcvette: im going to try going into the bios and turning off things
01:07 PM lcvette: but i think last time i tried this on the big machine the results weren't fruitful
01:09 PM pcw_home: Thats a huge change, wonder if the AMD memory controller has some pathology with non-interleaved RAM...
01:10 PM lcvette_fn: well on the motherboard it says dual channel ddr3
01:10 PM lcvette_fn: i read that in dual channel the memory is much more efficient
01:10 PM lcvette_fn: thats why i decided to try it
01:11 PM lcvette_fn: that is a comparison between single 8gb stick and 2 4gb sticks
01:15 PM lcvette: what is the important part of the graph here?
01:15 PM lcvette: the height or the width?
01:16 PM rmu: it may have to do with integrated graphics processor and borderline memory bandwidth for graphics
01:17 PM rmu: lcvette: you could try changing resolution to something lower and/or lowering refresh rate to 30hz
01:17 PM rmu: and look if that makes any difference
01:18 PM lcvette: i was just curious to test based on my reading
01:20 PM lcvette: im not gonna spend much time trying to improve these machines, i will package them and sell them
01:20 PM lcvette: get something better suited for suture projects
01:20 PM lcvette: future*
01:20 PM lcvette: this was my first brush with AMD
01:20 PM lcvette: im sure they make awesome stuff, but i like my intel
01:20 PM lcvette: lol
01:21 PM lcvette: think im gonna sell these and buy a newer system to test
01:21 PM lcvette: i wanna try a h370m-itx/ac + Pentium Gold G5400
01:22 PM lcvette: its a relatively inexpensive mini-itx setup that is newer with onboard m.2 ultra socket
01:22 PM lcvette: makes for compact packaging
01:22 PM lcvette: readikly available and parts are fairly inexpensive and available ciurrently
01:25 PM Tom_L: pcs get 'old' fast
01:25 PM lcvette: yes they do
01:25 PM lcvette: and when they do the parts for them become priced like platinum
01:26 PM Tom_L: or non exhistent
01:26 PM lcvette: yup
01:26 PM lcvette: so hoping to find a new generation build that can be proven
01:26 PM lcvette: i know the 8300 elite usdt's are awesome
01:26 PM lcvette: but they won't be long for this world either
01:27 PM Jymmm: Computers (and most electronics) ar ethe only thing that are worth thousands today, and hundreds in 6 months
01:27 PM Tom_L: find something preempt-rt works good on cause that looks like where it's headed
01:27 PM Tom_L: Jymmm, cars
01:27 PM lcvette: thats the goal
01:28 PM Tom_L: lcvette i'm not using them on a mill but i'm pretty happy with gigabyte boards
01:28 PM Jymmm: Tom_L: worht hundreds in 6 months, where?!?!?!?!
01:28 PM Tom_L: seems that way
01:29 PM lcvette: yeah but the gigabyte 300 series boards are getting some bad reviews
01:29 PM lcvette: im switvhing over to servos on everything now
01:30 PM lcvette: so i want to make sure i have the specs pcw_home says servos should have
01:30 PM lcvette: not that i fully understand what they mean, but i know if he says i need them then i better make sure i have them
01:30 PM lcvette: lol
01:40 PM _unreal_: how would I go about getting my whb04b-6 working in lcnc?
01:43 PM Tom_L: _unreal_, http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Using_A_XHC-HB04_Wireless_MPG_Pendant
01:43 PM Tom_L: for a start
01:43 PM Tom_L: http://wiki.linuxcnc.org/cgi-bin/wiki.pl
01:44 PM Tom_L: some of the information on wiki is old
02:02 PM _unreal_: looking
02:11 PM _unreal_: https://github.com/rubienr/machinekit/tree/feature-xhc-whb04b-6/src/hal/user_comps/xhc-whb04b-6
02:27 PM gregcnc: you can probably buy a skid of HP 8300 for the price of something new
03:34 PM lcvette_fn: gregcnc: i have a 8300 on the way for a small machine also
03:34 PM lcvette_fn: :D
03:35 PM gregcnc: rhino?
03:35 PM lcvette_fn: yeah
03:35 PM lcvette_fn: la banana
03:35 PM lcvette_fn: lol
03:35 PM gregcnc: lol
03:35 PM gregcnc: I may end up with a new retrofit, like I need a hole in the head
03:36 PM gregcnc: A mill turned up locally that probably runs, but I sent a guy with the same machine an email asking about tooling
03:36 PM lcvette_fn: im ready to get mine finished
03:36 PM lcvette_fn: yeah?
03:36 PM lcvette_fn: what kind of machine?
03:36 PM gregcnc: I knew he was having trouble with the control and it sounds like he wants to sell it now
03:38 PM gregcnc: So decisions.
03:38 PM lcvette_fn: what mill?
03:40 PM gregcnc: Emco Mill 155, small but bigger than the puny 125 I have now
03:43 PM lcvette_fn: i almost bought one of the head and changer assemblies from one of those off ebay a few years ago
03:43 PM lcvette_fn: but like you said, the tooling...
03:43 PM lcvette_fn: probably have it made in china
03:43 PM lcvette_fn: $15-20 holder
03:44 PM gregcnc: I figured out how to modify BT30 for the 100/125, but 155 is very different and I haven't had one in hand to measure.
03:45 PM gregcnc: will they do custom tooling for that money? They make piles of std BT30, so it's cheap
03:45 PM lcvette_fn: yeah, unless its a big departure from a similar tool
03:46 PM gregcnc: something to consider. Maritool wouldn't even consider making 20 for the smaller mill
03:46 PM lcvette_fn: of course not
03:46 PM lcvette_fn: its maritool
03:47 PM lcvette_fn: they can't be bothers
03:47 PM lcvette_fn: bothered
03:47 PM lcvette_fn: they are busy making big money on their standard tooling
03:48 PM lcvette_fn: do they just have a bearing on them
03:48 PM gregcnc: they make the holders for PocketNC
03:48 PM lcvette_fn: no dog notches?
03:48 PM gregcnc: the 155 has a bearing and flange with a slot, but the solt run into the taper
03:48 PM lcvette_fn: or with dog notches?
03:54 PM lcvette_fn: geez, can't even google up a picture of one
03:55 PM lcvette_fn: i would get some quotes on tooling before i considered purchasing
03:56 PM gregcnc: https://www.slideshare.net/FahadAlHarbi3/department-of-mechanical-engineering-technology
03:59 PM lcvette_fn: yeah they could make those
04:00 PM lcvette_fn: i don't think i would want them putting junk bearings on them
04:00 PM lcvette_fn: but the machined bodies should be no problem
04:00 PM gregcnc: the bearing floats in use
04:02 PM lcvette_fn: still i think having them add the bearing might push the price up
04:02 PM lcvette_fn: they are in the business of machining parts not installing bearings
04:02 PM gregcnc: sure no need for that. is there a place you've actually used?
04:03 PM lcvette_fn: so if you request them to delve into something that slows their production or stumps them they may pass or quote high
04:03 PM gregcnc: any quote will be less than Emco
04:03 PM Deejay: gn8
04:04 PM gregcnc: one guy posted Emco Spain wanted 250eruo
04:04 PM lcvette_fn: the company i purchased my tooling from makes great tooling and they custom made me a crapload of US sized er collets that are of incredible quality and the grinding is as good as the high end collets i have bought stateside if not better
04:33 PM gloops: clausing lathes - any good?
04:38 PM Rab: gloops, I have an old Clausing mill, so I'm on the Clausing email list/Yahoo group. Most of the posts are from lathe owners with various esoteric-sounding problems. I've formed the opinion that they were commercial-grade machines, but with a lot of complexity and things to go wrong.
04:39 PM Rab: I've been looking for a "classic" lathe, but I wouldn't be interested in a Clausing.
04:39 PM gloops: ahh
04:40 PM gloops: the one im looking at at is vintage though, looks solid enough but i dont know anything abut them
04:40 PM gloops: https://www.ebay.co.uk/itm/engineering-lathe/113716418634?
04:40 PM gloops: only 5 or 6 miles from me
04:41 PM Rab: Specifically, there's some kind of hydraulic clutch system with master and slave cylinders which needs rebuilding. There is some kind of epoxy surface in the power train--maybe the spindle mount?--which deteriorates almost imperceptibly, but enough to affect operation and require some arcane reapplication procedure.
04:41 PM gloops: power cross feed, threading gearbox
04:41 PM gregcnc: RAB specific to 5900, 6900, maybe a few otehrs
04:41 PM Rab: And there are all kinds of adjustments and stuff to get dirty or wear out.
04:41 PM gloops: that does sound like something to avoid then Rab
04:42 PM gregcnc: that ebay lathe has none of that
04:42 PM Rab: gregcnc, as you can probably tell, I'm just ignorantly repeating second-hand info.
04:43 PM gloops: £200 cheap atm, it isnt cheap if its a pain in the ass though
04:43 PM gregcnc: you missed my comment
04:43 PM gloops: yes i did see it gregcnc, so ill keep watching it
04:44 PM gloops: i might message him anyway, being so close i could go and have a look
04:44 PM Rab: So I would say Clausing is/was a great manufacturer of real machine tools, but I would do some very careful research before buying any particular model of lathe.
04:44 PM gregcnc: That looks like the early 100 series
04:45 PM gloops: bit awkward as it isnt badged clausing, it is one though
04:45 PM gregcnc: http://www.lathes.co.uk/clausing/page6.html
04:45 PM gregcnc: well we have colchesters badged clausing
04:47 PM Rab: I like the look of that lathe for a light home shop. That's a tiny chuck though.
04:47 PM gloops: Sold during the late 1940s and early 1950s - and an interesting development of the Clausing 100 Series Mk. 3
04:47 PM gloops: (Fortis)
04:47 PM gregcnc: Check out the jaw ways of that chuck
04:48 PM Rab: gregcnc, they expand at an angle?
04:48 PM gregcnc: never seen that
04:49 PM gloops: english chuck, not be the original
04:50 PM Rab: That looks like a real pain. I guess the workpiece would suck in slightly while tightening. (Or push out with internal grip.)
04:50 PM gloops: trying to think back, im sure ive bought something from him before
04:53 PM Rab: There is a more srs Clausing, although I'm not sure they want to sell it. https://sanantonio.craigslist.org/tls/d/bandera-lathe-clausing-model/6848768994.html
04:54 PM gregcnc: I have the shorter version of that. That's a lot of money.
04:55 PM gloops: that looks a lot more sophisticated
04:55 PM gregcnc: I don't think it is
04:57 PM gregcnc: it has three additional ratios on the feed screw, and 2 more on the input, but that's about it
04:57 PM Rab: It's just got that mid-century mod fascia. My mill is from the same era.
04:57 PM gregcnc: the 4900 are less common and have a 1 3/4"-8 spindle nose
04:58 PM gregcnc: if they don't have a steady or follow rest, they are hard to find
04:59 PM gloops: well if you took the shabby paint off the fortis version, there isnt a lot different really
04:59 PM gloops: ill see how it goes, probably go through the roof near the end of the auction
05:02 PM gregcnc: https://www.ebay.com/itm/133017836092 a newer Emco about the same size
05:03 PM Rab: What's that spindle attachment? A collet chuck with a handwheel for tightening?
05:04 PM gregcnc: yes, i have one of those, they are nice
05:04 PM gloops: id rather have this https://www.ebay.co.uk/itm/173871108669?
05:04 PM gregcnc: Yes so would I
05:04 PM gloops: reserve is over £750 though, hes tried it before
05:05 PM gregcnc: the threading range on the Emco leaves something to be desired
05:05 PM gregcnc: it would have to be cleaned to get it
05:06 PM gloops: tis in yorkshire as well, difficult to get us to part with more than 300 lol
05:07 PM gloops: bedtime here anyway
08:46 PM Tom_L: jthornton around?
08:47 PM XXCoder: yo
08:47 PM XXCoder: dunno its been super quiet here
08:47 PM XXCoder: its so slow I will probably die of old age before tomorrow comes
08:54 PM Tom_L: working on my macro a bit
08:58 PM XXCoder: the mill thread?
08:59 PM Tom_L: yeah
09:00 PM XXCoder: cool :)
09:00 PM XXCoder: I wonder if it would work for wood.
09:00 PM Tom_L: i don't know why not
09:00 PM XXCoder: making wooden bolt and threaded holes
09:00 PM XXCoder: it'd have to be giant tho hm
09:00 PM Tom_L: it's cutting the same path
09:00 PM Tom_L: why?
09:00 PM XXCoder: wood cant be thin threaded it'd break, that i assume anyway
09:01 PM Tom_L: probably can with live tooling
09:02 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/temp/thread.jpg
09:02 PM Tom_L: that's 3 tpi
09:02 PM XXCoder: pretty dang large indeed
09:02 PM XXCoder: partially strippd, looks but old?
09:03 PM XXCoder: *bit
09:03 PM Tom_L: real old
09:03 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/temp/thread1.jpg
09:03 PM Tom_L: replaced it with that
09:03 PM XXCoder: cool :)
09:03 PM XXCoder: i like lathes but im bit too afraid of running em lol
09:04 PM XXCoder: i like cnc routers a lot but not many of companies run one
09:04 PM Tom_L: cutting a 3tpi thread on the lathe was fun
09:04 PM Tom_L: gotta be quick
09:05 PM XXCoder: i bet
09:06 PM Tom_L: https://contest.techbriefs.com/
09:08 PM XXCoder: most invisible glass is interesting
09:08 PM XXCoder: great for solars!
09:08 PM XXCoder: solar panels that is
09:24 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Smartcam/arc_leadin.jpg
09:26 PM Rab: Tom_L, what was that wooden screw from?
09:26 PM XXCoder: looks like 3 cuts of increasing depth
09:26 PM XXCoder: nice
09:26 PM Tom_L: an old old square grand piano
09:27 PM Tom_L: XXCoder i fixed the lead in out from line to arc
09:27 PM Tom_L: the rest was done already
09:34 PM cryptg_: life check
09:35 PM Tom_L: empty
09:35 PM cryptg_: i see
09:35 PM cryptg_: question for the general chat
09:35 PM cryptg_: belt or ball screw
09:36 PM cryptg_: what is recommended for a cnc
09:36 PM Tom_L: question for the curious: application?
09:36 PM Tom_L: i'd go ball screw
09:36 PM cryptg_: cnc mill build
09:36 PM Tom_L: unless it's 100' long
09:36 PM XXCoder: materials to be cut?
09:36 PM Tom_L: then the screw would whip
09:36 PM cryptg_: dimentions 3'x3'
09:36 PM Tom_L: ballscrew
09:37 PM cryptg_: mainly wood/light use of aluminum
09:37 PM cryptg_: alright i was thinking ball screw too, just wondering what was the normal here
09:38 PM cryptg_: i was orignally going to make the mpcnc model, but upon thinking about it i am going to aim for something more solid
09:39 PM cryptg: somehow i timed out
09:39 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Column_Mill_VMC.jpg
09:40 PM Tom_L: that's what i did
09:40 PM cryptg: hmm
09:40 PM cryptg: nice
09:40 PM XXCoder: mpcnc is mostly toy
09:40 PM XXCoder: can do plastics and wood maube
09:41 PM cryptg: im debating whether i should make the y axis like yours or make it so the gantry is the one that moves instead
09:41 PM XXCoder: both have benefits and negatives
09:42 PM XXCoder: since you want squarish workspace router type is good
09:42 PM XXCoder: now fixed gantry or not..
09:42 PM cryptg: haha
09:43 PM cryptg: well fixed gantry i could potentially get less gantry deflection
09:46 PM ryptggg: Freenode not cooperating
09:47 PM ryptggg: Wow
09:48 PM cryptg: Dam
09:48 PM Tom_L: are you registered?
09:48 PM cryptg: Am I back?
09:48 PM Tom_L: yes
09:48 PM cryptg: Yes
09:48 PM Tom_L: hmm
09:48 PM Tom_L: bad connection
09:48 PM cryptg: I had to switch to my phone
09:48 PM cryptg: That's odd
09:49 PM cryptg: Speed test =17ms ping
09:49 PM cryptg: 100m down
09:49 PM cryptg: Oh well
09:50 PM cryptg: Tom what steppers are you running?
09:52 PM Tom_L: https://www.ebay.com/itm/3pcs-NEMA23-570OZ-IN-5A-3-8-DUAL-SHAFT-STEPPER-MOTOR/253223632606?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2060353.m1438.l2649
09:52 PM Tom_L: gecko drivers
09:52 PM cryptg: I see
09:53 PM cryptg: I'm still going for closed loop, I was wondering about holding torque
09:53 PM cryptg: Answered my question nicely though
09:59 PM cryptg: Hmm
09:59 PM cryptg: Looking at that deal, im almost tempted to get them and add my own encoders....
10:00 PM Tom_L: those come in .250" and .375" shaft if you look around
10:00 PM Tom_L: i'd rather have the .375" shaft
10:00 PM Tom_L: which is what those are
10:00 PM cryptg: I do prefer the bigger ones as well
10:00 PM Tom_L: the other end is still .250"
10:01 PM cryptg: I can't find any with encoders built in though
10:01 PM cryptg: Looks like open loop only unless some tomfoolery is had
10:01 PM cryptg: I'm up for it...
10:01 PM Tom_L: i'd be hard pressed to stall them
10:02 PM Rab: Stepper budget + encoder budget = bigger stepper budget, closed loop not needed.
10:03 PM Tom_L: those were the largest nema23 i could find
10:05 PM cryptg: Alright I'm sold :D
10:06 PM cryptg: I have these from a previous project
10:06 PM cryptg: Signswise 600p/r Incremental Rotary Encoder Dc5-24v Wide Voltage Power Supply 6mm Shaft https://www.amazon.com/dp/B00UTIFCVA/ref=cm_sw_r_cp_apa_i_9kQSCb8F4V6XD
10:06 PM cryptg: I was thinking or gearing them 1:1 and finding a closed loop controller,
10:07 PM cryptg: But I hadn't noticed the 600 pulses per rev
10:07 PM Tom_L: i'm only running mine at 2.5A too
10:07 PM Tom_L: that's where the drivers were set for the previous steppers i had
10:07 PM cryptg: All I see are drives that use 1000 pulses so I'm unsure if they'll play nice
10:07 PM cryptg: Nice!
10:08 PM cryptg: So, plenty of torque then.
10:08 PM Tom_L: i'm using gecko 203v
10:08 PM cryptg: OK
10:08 PM Tom_L: https://www.geckodrive.com/g203v-digital-step-drive.html
10:08 PM Tom_L: you can get cheaper chinese ones
10:10 PM cryptg: Ok
10:10 PM Tom_L: the geckos are good though
10:10 PM cryptg: Since this is hobby project for now I'll probably go with the chinesium until further notice
10:11 PM Connor: Any one know if the coolant pumps found in Band Saws have some sort of float switch? I picked up a used horizontal saw.. it has a coolant pump.. but doesn't work. I've not put any coolant in it yet either.. so, I wasn't sure if that was the issue or not.
10:12 PM cryptg: Band saws use a flood coolant?
10:12 PM cryptg: We use mist tipically
10:12 PM cryptg: Cutting oil*
10:12 PM Connor: It's what it came with.
10:13 PM cryptg: Any way you can shoot me a picture of the pump?
10:13 PM cryptg: And the band saw in general
10:14 PM Connor: https://www.ebay.com/itm/DA-Kang-Coolant-Immersion-Pump-1-8-Hp-110V-1PH-/362572539619?_trksid=p2047675.m43663.l44720&nordt=true&rt=nc&orig_cvip=true
10:14 PM Connor: One sold on ebay that looks very much like it..
10:15 PM Connor: maybe not the exact model.. It's still in the saw.. I've not pulled it yet.
10:15 PM Connor: I don't see a float switch.
10:15 PM Connor: and I don't hear anything when I turn it on.
10:16 PM cryptg: It shouldn't have one
10:16 PM cryptg: Open the panel on the side and check to see if the power is direct to the motor
10:18 PM cryptg: Assuming the motor doesn't turn, grab a dmm and check for resistance of the coil
10:18 PM Connor: Has it's own switch.. coming off the saw switch. It looked okay connection wise.. I'll double check that it's functional this week.
10:19 PM cryptg: Could be a bad switch too, the contacts can get gunned up depending how often it's switched on
10:20 PM cryptg: A multimeter will be your friend
10:21 PM cryptg: Step 1 make sure there is power input and output at the switch
10:22 PM cryptg: Good luck
10:22 PM cryptg: OK guys thanks for the info!
10:39 PM cryptg: Hey Tom
10:39 PM cryptg: https://rover.ebay.com/rover/0/0/0?mpre=https%3A%2F%2Fwww.ebay.com%2Fulk%2Fitm%2F163524784676
10:40 PM cryptg: 3.5x the price 😖
10:41 PM XXCoder: wow! you guys know about youtube channel named 11foot8? its about bridge that trucks keep smashing into
10:41 PM XXCoder: apparently it captured factory explosion yeserday
10:42 PM cryptg: Wow
10:42 PM XXCoder: most recent is https://www.youtube.com/watch?v=TvxDhgDjVCA
10:42 PM XXCoder: I dont think guy will upload explosion one
10:43 PM cryptg: No wonder I couldn't find it
10:44 PM XXCoder: one guy died in that explosion I dont think he will upload it. i wouldnt
10:44 PM cryptg: Ah makes sense
10:45 PM XXCoder: i dont know why people keep getting smashed into that bridge
10:45 PM XXCoder: it happens so often that after they added very strong iron bar to protect bridge, they had to replace it after a year.
10:46 PM cryptg: It's an odd height for a bridge with so much traffic
10:46 PM XXCoder: its old
10:47 PM cryptg: The should dig the road deeper
10:47 PM XXCoder: 4 inches higher bridge, or land 4 inches lower yeah
10:47 PM cryptg: Would solve many headaches in the long run imho
10:48 PM XXCoder: yeah.
10:48 PM cryptg: Hey what's your opinion on the chinesium stepper drivers
10:48 PM XXCoder: either of those solutions is quite expensive
10:48 PM XXCoder: which one?
10:48 PM cryptg: Like the TB660
10:48 PM XXCoder: tb660 is ok
10:48 PM cryptg: Tb6600*
10:48 PM XXCoder: my first one was tb6560. it sucked lol
10:49 PM XXCoder: tb6600 runs way better but its still chinese
10:49 PM XXCoder: man that semi trailer got destroyed
10:49 PM XXCoder: $10,000 gone
10:50 PM cryptg: 10k? I've seen dual unit reefers get hit there😖
10:50 PM XXCoder: i shudder to think when someone does same but with standard transport box
10:50 PM XXCoder: 10k was wild guess really
10:51 PM cryptg: A flat bed cost us around that price at work
10:51 PM XXCoder: them transport boxes is made from hard metal. it'd wheck protective bar, and probably jkill bridge too if guy was speeding
10:51 PM cryptg: No enclosure
10:51 PM cryptg: Haha
10:52 PM XXCoder: https://www.youtube.com/watch?v=i7IlRbECGf8 7 years old. it predates bar
10:55 PM pink_vampire: hi
10:55 PM XXCoder: hey pink
10:55 PM pink_vampire: hi XXCoder
10:56 PM cryptg: Hi
11:08 PM cryptg: OK guys I'm out thanks for all the tips
11:09 PM XXCoder: later