#linuxcnc Logs

Oct 23 2021

#linuxcnc Calendar

01:43 AM Deejay: moin
03:14 AM Berg: raspberry pi cnc program?
03:14 AM Berg: just starting cnc
04:06 AM Tom_L: ok
04:51 AM JT-Cave: morning
07:14 AM Tom_L: morning
07:23 AM tiwake[m]: hi
07:37 AM -!- #linuxcnc mode set to +v by ChanServ
07:43 AM JT-Shop: lost power at 5:30 5% of county is out of power and linux pc in cave won't boot up now
07:43 AM JT-Shop: running on standby generator
07:44 AM XXCoder: hope it didnt fry
07:45 AM JT-Shop: dunno, going to put a livecd in and boot from that if possible
07:51 AM unterhausen: we had no internets all day yesterday, both comcast and at&t out at our house. One of the pipes must have gotten clogged
07:51 AM tiwake[m]: thinking about how I want to linuxCNC up this little tabletop mill thing
07:52 AM tiwake[m]: probably want to get new stepper drivers that arn't ass
08:03 AM Tom_L: tiwake[m], i got gecko 203v on mine
08:04 AM Tom_L: and i beefed up the steppers
08:04 AM Tom_L: when i built the new mill
08:05 AM tiwake[m]: prolly what I'll end up doing... making a bigger one
08:05 AM Tom_L: https://www.ebay.com/itm/253223632606
08:05 AM Tom_L: those are the heftiest nema23 i could find
08:06 AM tiwake[m]: wonder what I have
08:06 AM Tom_L: i swapped a dual stack 282inoz for those and prior to that they were single stack
08:07 AM XXCoder: 3 for 119 thats cheap lol
08:08 AM Tom_L: you can also find that same stepper in 1/4" shaft but i wouldn't recomend it
08:08 AM XXCoder: lol
08:08 AM tiwake[m]: whut
08:08 AM tiwake[m]: it looks like its a different stepper for each axis
08:16 AM Tom_L: on yours?
08:16 AM tiwake[m]: yeah
08:16 AM tiwake[m]: its a BoxZY
08:16 AM Tom_L: i used parts off my sherline to build a newer bigger one
08:16 AM Tom_L: i'm still using the spindle because i had so many holders but that's about it
08:16 AM -!- #linuxcnc mode set to +v by ChanServ
08:16 AM Tom_L: JT-Shop, wb
08:19 AM tiwake[m]: I'm not really seeing any low level specs of this
08:24 AM Tom_L: XXCoder, they were 99 when i got em
08:25 AM tiwake[m]: hmm
08:28 AM tiwake[m]: apparently https://www.pololu.com/product/2981 is an upgrade to what the BoxZY ships with
08:29 AM Tom_L: https://www.ebay.com/itm/112802695589
08:29 AM Tom_L: 1/4" shaft version
08:42 AM tiwake[m]: apparently the boxzy is a 19v 10amp power supply
08:43 AM tiwake[m]: https://www.casunsteppermotor.com/product/%E3%80%9057shd4407-28h%E3%80%91nema-23-bipolar-1-8deg-1-2nm-170oz-in-57x57x54mm-4-wires-stepper-motor-for-xyz-axis-cnc/
08:43 AM tiwake[m]: thats for the Z axis
08:51 AM tiwake[m]: hmm, I can't seem to find 57SHD4407-23H
08:56 AM tiwake[m]: it was an ultramaker driver board
08:57 AM tiwake[m]: erm, ultimaker rather
09:03 AM tiwake[m]: https://reprap.org/wiki/Ultimaker%27s_v1.2_PCB I think is what I have
09:04 AM tiwake[m]: erm, no
09:05 AM tiwake[m]: https://reprap.org/wiki/Ultimaker%27s_v1.5.7_PCB
09:10 AM tiwake[m]: maybe I can use this
09:17 AM Tom_L: those are likely 2.5A drivers
09:18 AM unterhausen: I have thought that someone should design a carrier board for the poulu type drivers
09:18 AM Tom_L: 540 gecko
09:18 AM unterhausen: be nice to have isolation, I suppose you could use a cnc shield
09:19 AM tiwake[m]: alright
09:19 AM tiwake[m]: hmm
09:21 AM tiwake[m]: Sunfounder Mega 2560 R3 is plugged into the back of the Ultimaker board
09:21 AM tiwake[m]: had to take things apart further to see
09:30 AM tiwake[m]: can I get linuxCNC to use it?
09:49 AM Guest9096: hi
09:49 AM tiwake[m]: hi
09:50 AM Guest9096: can sb help me with setting up a slave-axis using a 7i76e ?
09:50 AM Guest9096: I configured it in pncpnf and when I click the homing-button only one motor moves
09:50 AM tiwake[m]: no
09:50 AM tiwake[m]: everyone is dead
09:50 AM tiwake[m]: I killed them
09:52 AM Guest9096: nice
09:52 AM roguish[m]: it's a quiet Saturday morning.....
09:52 AM Guest9096: what time is it roguish ?
09:53 AM unterhausen: look at the signals with halscope
09:54 AM roguish[m]: amost 8:00am here on the left coast.
09:54 AM Guest9096: nice. its already 5pm here
09:54 AM pcw--home: Guest9096 I don't think pncconf sets up the homing constants correctly for gantries
09:55 AM pcw--home: http://linuxcnc.org/docs/html/config/ini-homing.html
09:55 AM Guest9096: ok. so what else do I have to do ?
09:55 AM Guest9096: to make both motors home at the same time
09:55 AM roguish[m]: Guest9096: pncconf is only the beginning, especially for any sort of non-totally-typical setup.
09:56 AM roguish[m]: pcw--home: hey Pete waz up?
09:56 AM pcw--home: First step is to read the homing document
09:56 AM pcw--home: Happy to have some rain
09:56 AM roguish[m]: Guest9096: like pcw said..... RTFM
09:57 AM Guest9096: ok... I will read it first
09:57 AM Guest9096: thanks
09:57 AM roguish[m]: pcw--home: get your booster yet?
09:58 AM pcw--home: The main secret is that for tandem axis you use a negative sequence number
09:58 AM pcw--home: No, not yet
09:58 AM roguish[m]: I jumped at mine 4 weeks ago....... full shot of Moderna
09:59 AM pcw--home: Moderna has been pretty good against fading unlike Pfizer
10:00 AM pcw--home: but yeah I guess its available for us senior citizens now...
10:01 AM roguish[m]: pcw--home: recall I am using your spinx and lincurve to get my spindle working.... well, i tried the vfdb_vfd comp. even though I have model E not B, the base commands used the comp are the same according to both manuals. and it actually works!!!!
10:02 AM roguish[m]: I test real rigid tapping on a wood 2x4. with the modbus control. worked good. now going to try on some pvc, the aluminum
10:05 AM pcw--home: I think real time delays in rigid tapping spindle control are not important _unless_ you are using
10:05 AM pcw--home: I mean you are doing blind holes
10:06 AM pcw--home: when the time to do the reversal determines the depth
10:06 AM roguish[m]: I'm going to test quite a bit. I do have a real, 1000ppr, encoder on the spindle for feedback.
10:07 AM pcw--home: for through tapping, a minor overrun because the reverse command was delayed 100 ms is not important
10:08 AM roguish[m]: there are 2 gcodes for tapping. G33.1 for real, coordinated tapping, and for tapping with a bottom dwell. i'm using G33.1
10:09 AM roguish[m]: and going pretty slow, like 150 rpm. with the accel and decell in the vfd set low
10:11 AM pcw--home: Yeah the slower you go, random delays will add less to any possible depth overrun
10:11 AM roguish[m]: pcw--home: if it's truelly coordinated motion, does that really matter? and is there always a delay in the usb/modbus?
10:12 AM roguish[m]: or just always possible
10:12 AM roguish[m]: or probable
10:13 AM roguish[m]: and right now i'm using a drill chuck, not a compliant collet for the taps. I will get some of those....for insurance....
10:13 AM pcw--home: usb/modbus are a userland processes so not real time (so could have 100s of ms delays)
10:14 AM roguish[m]: ok, now i'm really going to get some springy collets.....
10:15 AM pcw--home: this has no affect on Z axis/ spindle sync, but a delay in the reversal command might be bad for blind holes
10:15 AM pcw--home: so really only a possible issue for blind holes
10:16 AM roguish[m]: blind holes are common. gotta be able to handle 'em.
10:16 AM tiwake[m]: floating tap holders are good
10:16 AM roguish[m]: is there any way to reduce the probable delays?
10:17 AM tiwake[m]: https://www.maritool.com/Tool-Holders-Floating-Tap-Holders/c23_27/index.html
10:18 AM pcw--home: You actually should be able to measure them with halscope ("air" tap a long sequence of holes and check for variations on the minimum Z reached)
10:19 AM pcw--home: that should get you an idea of the repeatability
10:20 AM roguish[m]: ok. i'll try. also think i'll work on getting 'sampler' working to write out a data file.
10:23 AM roguish[m]: tiwake[m]: thanks, yeah, i'm familiar with those. I use a BT40 spring collet holder on a Haas I run occasionally.
10:24 AM tiwake[m]: one of the machines at work is waaaay off in timings in the fanuc controller
10:24 AM tiwake[m]: breaks the tap every time, even though its supposed to be able to ridged tap
10:26 AM roguish[m]: tiwake[m]: on the old VF4 Haas, ya gotta give it a lot of time to synchonize correctly. like at least 1" of motion before the actual tapping.
10:28 AM tiwake[m]: this is a pretty new hyundai mill... the same size as the haas VF3
10:29 AM tiwake[m]: hyundai does not like documenting things
10:29 AM tiwake[m]: fanuc does not like good documentation either
10:30 AM tiwake[m]: several CNC hyundai lathes shipped missing whole feature sets
10:30 AM tiwake[m]: making C-axis work effectively impossible
10:32 AM tiwake[m]: I will never buy such a machine unless it was stupid cheap, because of how crappy they are
10:34 AM tiwake[m]: still trying to figure out how to use this CNC mill I have
11:14 AM -!- #linuxcnc mode set to +v by ChanServ
11:19 AM Tom_L: tiwake[m], that's probably why they bought them
11:19 AM tiwake[m]: ..?
11:20 AM tiwake[m]: oh the hyundai?
11:20 AM Tom_L: yep
11:20 AM tiwake[m]: they bought them brand new
11:20 AM Tom_L: does that make them better?
11:20 AM tiwake[m]: if they were concerned about price, they'd get used machines
11:20 AM tiwake[m]: I did see the price of them new... they are not that much cheaper
11:21 AM Tom_L: you have any with siemens controls there?
11:21 AM tiwake[m]: no
11:23 AM Tom_L: my bud got one machine with a windows software control. he didn't keep it long
11:26 AM tiwake[m]: windows CE?
11:26 AM tiwake[m]: I've seen those before
11:27 AM Tom_L: i don't remember now but that sounds about right
11:27 AM tiwake[m]: it seems horrible
11:27 AM tiwake[m]: windows CE is the embedded version of windows
11:27 AM Tom_L: right
11:27 AM tiwake[m]: because who dosen't want to embed windows in expensive critical CNC machines
11:28 AM tiwake[m]: all businesses do, because microsoft is for businesses
11:35 AM tiwake[m]: I think I can just feed it text files over the USB?
11:38 AM tiwake[m]: https://reprap.org/wiki/CNC_Gcode_controller
11:38 AM tiwake[m]: that... might work?
11:38 AM Tom_L: if that's what you want
11:40 AM tiwake[m]: as near as I can tell, the controler for this BoxZY that I have is the same found in reprap stuff
11:43 AM tiwake[m]: https://reprap.org/wiki/EMCRepRap
11:44 AM Tom_L: very likely
01:03 PM unterhausen: We had some Wince network drives, but then IT had them destroyed
01:03 PM unterhausen: no warning and they didn't image them
01:20 PM snakedLX is now known as snaked
02:20 PM Loetmichel: Oh. Just Clicked "close" on Firefox accidentally. "You really want to close 121 tabs in 3 windows? Y/N" ... Ah, thats why this box here is constantly swapping... May not help that its win7 and has about 160 updays :)
02:30 PM javabean: 121 tabs?? musta just opened a new/clean profile a couple of days ago
02:35 PM -!- #linuxcnc mode set to +v by ChanServ
02:36 PM JT-Cave: the pc is ok but something is wrong with the boot section... wonder how to fix that
02:56 PM roycroft: i just turned a ring to cut up into pads to hold my timing gear in my lathe, and i'm pretty pleased at how it turned out
02:57 PM roycroft: it did not need to be terribly accurate, but i used it as an exercise to see how good i could make it, and in spite of my old, worn out lathe, and old, worn out, relatively inexperienced me, i was able to hold 2 tenths on all dimensions
02:58 PM roycroft: the only problem is that the only round bar i had that was big enough to make this was steel, and the timing gear i need to clamp onto is aluminium
02:58 PM roycroft: i'm wondering if i should wrap some tape around the gear before i clamp it with the steel pads
02:58 PM roycroft: my plan is to shim the outside of the pads to make the part concentric
02:59 PM roycroft: so i don't think a little tape will hurt
03:00 PM roycroft: i don't want to mash the gear teeth, although i should not have to tighten the chuck too much - i'm just enlarging a 6mm hole to 12mm, and the hub is aluminium
03:59 PM -!- #linuxcnc mode set to +v by ChanServ
04:00 PM -!- #linuxcnc mode set to +v by ChanServ
04:04 PM Tom_L: JT-Cave, https://www.debian.org/releases/jessie/amd64/ch08s07.html.en
04:09 PM JT-Cave: thanks
04:11 PM JT-Cave: hmm I don't even get the boot menu just a blinking cursor
04:11 PM JT-Cave: didn't try and type anything though
04:25 PM -!- #linuxcnc mode set to +v by ChanServ
05:02 PM Tom_itx: JT-Cave, might not be as helpful since it includes a windows boot record: https://www.cyberciti.biz/tips/restore-debian-linux-grub-boot-loader.html
05:03 PM Tom_itx is now known as Tom_L
05:17 PM synfinatic_ is now known as synfinatic
06:45 PM roycroft: the transmission gear on for my mill is modified and fitted, so that's awesome
06:46 PM roycroft: the pads i made to hold it worked brilliantly, and i did not damage it at all
06:47 PM roycroft: but i was able to verify today that everything will line up and have clearance, and that the belt is the correct length
06:48 PM roycroft: that transmission gear only has about 0.020" clearance top and bottom - it was a little tricky making the spacers for it and modifying it to fit
08:25 PM Tom_itx is now known as Tom_L
11:33 PM Tom_L: andypugh, you around?
11:34 PM Tom_L: likely not
11:34 PM XXCoder: been a bit since andy last said something
11:34 PM Tom_L: wondering how a rotary axis is treated in linuxcnc
11:35 PM Tom_L: if each angle is a reference from zero or current position
11:35 PM Tom_L: if i go 270deg, then 150 will it rotate in reverse to that position
11:36 PM Tom_L: i've got a rough draft of my rotary post for the cad cam but i need the axis wired up to test it more
11:38 PM Tom_L: also need to know how the canned cycles work on rotary. it doesn't look like they directly support the ABC axis
11:40 PM Tom_L: Rotary axis words are not allowed in canned cycles.
11:40 PM Tom_L: well that explains that..
11:45 PM roycroft: it's also 4:45am on a sunday morning
11:46 PM Tom_L: yeah i noticed after the fact
11:51 PM t4nk_freenode: it's 6:52am, I say: Good Morning.
11:58 PM Tom_L: i just knew andy had done things with a rotary