#linuxcnc Logs

Apr 22 2022

#linuxcnc Calendar

01:17 AM Deejay: moin
02:08 AM randy: morning
04:52 AM Tom_L: morning
05:03 AM tcurdt: Does RTAI have any significant advantage over PREEMPT-RT with a Mesa card?
06:02 AM JT-Cave: morning
06:02 AM JT-Cave: no
06:04 AM JT-Cave: Tom_L, never seen them before
07:03 AM Tom_L: me either
09:23 AM mrec: chamfer in freecad is nice...
09:23 AM mrec: a bit too many values to enter but that can be fixed easily.
09:54 AM randy: chamfer in path or chamfer in part design?
09:54 AM mrec: chamfer in part design
09:55 AM mrec: path I mean
09:55 AM mrec: not part design.
09:55 AM mrec: the terminology is not fully clear I'd say
09:55 AM mrec: (for the chamfer settings) but it did something and the edges were okay at least
09:56 AM mrec: I'll have a closer look at it during the next days
10:06 AM randy: kk
10:41 AM mrec: I want to have it a little bit more comfortable maybe with a small graphic demonstrating the terminology of the chamfer endmill
10:42 AM mrec: the topological naming issue is a serious issue for the path workbench
10:42 AM mrec: the path workbench can only be applied to ready objects (no modification is allowed anymore otherwise the path will be messed up)
11:39 AM * roycroft shakes his head
11:39 AM roycroft: my boss' daughter has a website hosted with us
11:39 AM roycroft: it's been sending spam
11:40 AM roycroft: i just logged into it to correct the problem and i found it's not even a developed website - all the images are stock images, and all the text is lorem ipsum filler
11:40 AM roycroft: it's been that way for almost a decade
11:40 AM roycroft: i just told my boss that i'm putting the site in maintenance mode until someone develops it properly and he told me not to do so
11:40 AM unterhausen: I looked at Moore Nanotech's website once and it was filled with hidden spam
11:40 AM roycroft: he said it's his kid's site and we can't just disable it
11:41 AM roycroft: i put it in maintenace mode anyway
11:41 AM roycroft: i had to stop the spam
11:41 AM unterhausen: we know the boss, told him, and he did nothing
11:44 AM roycroft: i don't get how people can just stick their heads in the sand while people are whapping on their behind with a huge paddle
12:36 PM XXCoder: roycroft: honestly I dont get "its my childs there its untouchable" thing. If I manage a site, and my kid sub-site is having spam issues, id either fix or suspend till kid fixes it
01:08 PM roycroft: one would think that, since my boss fancies himself a website developer, when a problem with his daughter's website is reported, he would spring into action to fix the problem
01:08 PM roycroft: this site really looks something my boss started developing and then got bored with it
01:09 PM roycroft: he does have a pretty serious case of adhd
01:10 PM roycroft: but the odd thing about that is that once he gets bored with something and moves on, he will eventually become interested in the first thing again, and do a little more work on it
01:24 PM mrec: Tom_L: I hope I can install the motor tomorrow, today I was able to modify the mounting plate but the housing of the mini lathe is blocking the phase/hall wires
01:25 PM mrec: I can't wait to experience the new lathe :-)
01:26 PM mrec: and on monday I'll buy a delta controller + jianken jgl-100/24k rpm spindle
01:35 PM roycroft: how much is that spindle going to cost?
01:37 PM roycroft: i see it for US$1500 on aliexpress
01:37 PM roycroft: and i get huanyang spindles with similar specs for US$225
01:37 PM roycroft: i wonder what the difference is
01:52 PM Wout: Somebody knows where I can find an updated page to this one? http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Hardware_Requirements
01:58 PM XXCoder: need a LOT of hardware and money to update that
02:05 PM CaptHindsight[m]: Wout: the updates are now live on IRC or the forums
02:05 PM CaptHindsight[m]: Wout: https://forum.linuxcnc.org/18-computer
02:07 PM Wout: Yea, I am reading this already for 1 hour, but I find a lot of specific information. Looking more for general information.
02:07 PM Wout: Like what type of 3d card is advised or is intergrated grafix better.
02:09 PM CaptHindsight[m]: hasn't mattered for hardware since around 2010
02:10 PM CaptHindsight[m]: BIOS is the main issue, they are often broken and will leave power management features on that interfere with real time
02:10 PM CaptHindsight[m]: even though you turn them OFF in the menu
02:11 PM XXCoder: rpi4 based system sounds more and more like great idea to me lol
02:11 PM XXCoder: with MESA latency matters less, and theres ethernet based MESA to use with it
02:11 PM Wout: Aa oke. I would like to get a low latency. I studied Control engineering and i really like the PID tuning part.
02:11 PM CaptHindsight[m]: except for RPI4's being out of stock or >$150ea now
02:11 PM XXCoder: yeah
02:11 PM Wout: I think it matters for closing the position loop if i am correct.
02:12 PM Wout: Also i am not using stepper motors. ;)
02:13 PM CaptHindsight[m]: Wout: actually getting hardware can be difficult currently
02:13 PM Wout: I just order the 6i25 and a 7i77. Will be here on Monday.
02:14 PM Wout: So finding an old PC should be doable, I hope....
02:14 PM Tom_L: mrec, you can't rotate the motor 90 deg for wire clearance?
02:15 PM CaptHindsight[m]: Wout: I have only found a few PC's that had too poor latency for use with Mesa FPGA's
02:15 PM Wout: But to summarize. Get a PC that is not too old and has a good BIOS without power saving options.
02:16 PM CaptHindsight[m]: they all come with power management options, you just need to be able to turn them off
02:16 PM Wout: Then I will just start with something. I can always upgrade later if it is not good enough I suppose.
02:16 PM Tom_L: besides you just can't have too many PCs
02:17 PM CaptHindsight[m]: some of the corporate office PC's leave out BIOS options to reduce support calls
02:17 PM Wout: Tom_L Hahaha tell that to my wife.
02:17 PM CaptHindsight[m]: off lease systems for sale
02:17 PM CaptHindsight[m]: some Dells and HP's are fine, some are just too jittery for real time
02:19 PM Wout: Yea, that is the thing. I like the long list on the forum with tests, but I cannot figure out what the most important parameters are that influence the latency. But now I know it is the BIOS
02:19 PM CaptHindsight[m]: Wout: the most recent discoveries of poor systems are in this thread https://forum.linuxcnc.org/18-computer/39371-results-of-latency-test-list-of-computers-tested-for-use-with-linuxcnc?start=300#240319
02:20 PM Wout: Yup that is the tread i am referencing.
02:20 PM CaptHindsight[m]: gaming and tweaker mother boards tend to work best for software stepping
02:20 PM CaptHindsight[m]: since the BIOS options often work as intended
02:21 PM CaptHindsight[m]: some power management buttons are like the "Close Door" button on elevators
02:21 PM CaptHindsight[m]: lies!
02:23 PM Wout: Well with that info I will be going shopping! THX
03:17 PM roycroft: my new gpu is out for delivery, finally
03:18 PM XXCoder: yay
03:18 PM roycroft: hopefully my cad workstation will be even faster this weekend
03:19 PM roycroft: i also just ordered 60 drawer slides
03:20 PM roycroft: the ups driver is not going to be very happy with me
03:22 PM XXCoder: ok I got a question
03:22 PM XXCoder: if I change threads from 1 to 2 start, is it still compitable with threads on single thread hose?
03:23 PM XXCoder: if so, any special concerns on adjustment like peak to peak or whatever
03:23 PM CaptHindsight[m]: roycroft: have you tried different north south positions that the monitor faces for optimal speed? :)
03:24 PM XXCoder: after thinking some time, I dont belive 2 start is compitable
03:25 PM XXCoder: bah. it'd make mirroring 2 sections so much easier
03:39 PM CaptHindsight[m]: JT-Shop: https://forum.linuxcnc.org/media/kunena//attachments/legacy/files/align.zip is this your latest for sheet alignment?
03:41 PM CaptHindsight[m]: I'm working on machine vision widgets and looking at integrating edge, corner and fiducal detection to this or similar
03:47 PM roycroft: i do not want to begin with the feng shui ritual until the hardware configurtion is finalized
03:48 PM roycroft: i'm afraid the qi might be altered by the change of gpus
03:51 PM CaptHindsight[m]: roycroft: https://answers.microsoft.com/en-us/windows/forum/all/monitor-affecting-internet-speed/99ee38ed-4e65-4ea4-95a8-280bdd3528ef
03:53 PM CaptHindsight[m]: MS blames it on hardware :p
03:53 PM roycroft: well they would have to blame it on dell, then
03:53 PM roycroft: the machine is a dell, btw
03:54 PM roycroft: but i think that's irrelevant
03:54 PM roycroft: the old, 1080 display was a dell
03:54 PM roycroft: and windows recognized it as such
03:54 PM roycroft: the new display is a spectre, and that's pretty much a generic brand, and windws sees it as a generic 4k display
03:55 PM roycroft: one generally expects a vendor-specific driver to perform better than a generic driver
03:55 PM roycroft: but the opposite appears to be the case here
03:55 PM roycroft: whatever the reason, it will all be moot when the new gpu arrives
03:55 PM roycroft: because everything will change then
04:35 PM Tom_L: XXCoder, how could a 2 start be compatible?
04:36 PM Tom_L: 2x the pitch
04:36 PM Tom_L: per thread
04:36 PM XXCoder: well im no engineer, though did figure that out evenually
04:37 PM Tom_L: not quite like that reverse thread thingie
04:39 PM XXCoder: whats reverse thread? left tighty type?
04:40 PM Tom_L: the thing with the 2 nuts threaded backward from each other
04:40 PM XXCoder: oh yes. I have printed that, very fun object
04:40 PM Tom_L: i knew you did :)
04:40 PM Tom_L: now machine one
04:41 PM XXCoder: sadly no lathe
05:47 PM CaptHindsight[m]: andypugh: to add image processing/machine vision to LCNC should I create user space Hal components for each filter for anyone to use (even without a GUI) vs as just QTPYVCP widgets?
06:55 PM roycroft: i am pleased to announce that today is the official start of mojito season at bistrĂ³ roy
06:59 PM CaptHindsight[m]: viva los mojitos
06:59 PM JT-Shop: a fellow Missourian saw the first humming bird of the year today here
07:02 PM roycroft: that's always a good sign
07:02 PM CaptHindsight[m]: I got my first flying to small toseem bug bites yesterday
07:03 PM CaptHindsight[m]: I likely not survive the summer
07:03 PM CaptHindsight[m]: first bites are usually in June
07:04 PM roycroft: we're on track to have our 8th wettest april ever
07:04 PM roycroft: after years of drought, though, it's actually kind of nice to have too much rain
07:05 PM roycroft: sadly, though, we did not get a good snow pack last winter again, so we'll probably be in drought again come summer
07:07 PM CaptHindsight[m]: no summer sky waters?
07:14 PM Tom_L: what sort of sheet plastic (roll) is impervious to sodium hydroxide?
07:16 PM Tom_L: looks like most except PC
07:17 PM Tom_L: PE is good
07:17 PM roycroft: i just made the lip that holds labels on my hardware boxes thicker, and filleted the edge
07:17 PM roycroft: hopefully it will print better with the new nozzle
07:17 PM Tom_L: but does it come in rolls
07:18 PM * roycroft waits for the file to sync
07:18 PM Tom_L: bigger nozzle means wider layers
07:18 PM CaptHindsight[m]: Tom_L: https://omnexus.specialchem.com/polymer-properties/chemical-resistance/sodium-hydroxide-10-20-c
07:19 PM roycroft: damn
07:19 PM roycroft: i have another of those 6pm meetings tonight
07:19 PM * roycroft starts making pizza early
07:19 PM Tom_L: ur just so popular
07:20 PM Tom_L: capn i was looking at this one: https://www.tedpella.com/company_html/PlasticsChemResistance.aspx
07:20 PM roycroft: no, i'm just serious about my job is all
07:21 PM roycroft: https://roycroft.us/New-Hot-End/Box-3.jpeg
07:21 PM roycroft: that's from my first go with the new hot end/nozzle
07:21 PM Tom_L: CaptHindsight[m], what is that sheet plastic you get at the big box store
07:21 PM roycroft: the edge of the lip is sharp in that design
07:22 PM roycroft: i just made it 0.050" thick, and hopefully that will keep it from sagging/warping/going wonkly
07:22 PM roycroft: wonky
07:23 PM CaptHindsight[m]: Tom_L: PVC?
07:23 PM Tom_L: probably
07:23 PM CaptHindsight[m]: polyethylene
07:24 PM Tom_L: LDPE?
07:24 PM roycroft: i don't think most plastics are significatnly affected by naoh
07:25 PM CaptHindsight[m]: https://en.wikipedia.org/wiki/Plastic_wrap
07:26 PM unterhausen: amazon has decided I'm the kind of person that would spend $95 for a 3d printer nozzle
07:26 PM Tom_L: were they right?
07:27 PM unterhausen: I'm the kind of person that would spend someone else's $95 to see how it worked, maybe
07:27 PM unterhausen: my printer cost me $80, so probably not
07:29 PM XXCoder: roycroft: nice! tab still bit issues but clean
07:30 PM XXCoder: maybe tab get twice as thick would help as top layer would be clean then
07:31 PM roycroft: that's what i'm thinking, xxcoder
07:31 PM roycroft: if there's more support below the tab it might come out better
07:31 PM XXCoder: yeah
07:34 PM roycroft: i just ran the numbers, and these new cabinets are going to cost about $1200, not counting the cost of all the boxes, but it will be worth it - it will be the biggest storage improvement i've ever made to my shop
07:34 PM XXCoder: nice
07:35 PM XXCoder: how will you secure it to wall?
07:35 PM roycroft: and fortunately i had saved a big chunk of money for this project already
07:35 PM roycroft: i'm not
07:35 PM roycroft: they are going to be base cabinets
07:36 PM roycroft: i'm redoing the drawings again, now that i have the final parameters calculated, and i'll be able to post a rendering of them soon
07:36 PM XXCoder: awesome :)
07:36 PM roycroft: i'm going to wait until the drawers slides arrive first, as i'm using a brand that's new to me, and i'm not 100% certain of the hole spacing yet
07:37 PM roycroft: i don't like switching brands, but i got these for $69.99 for a ten pack, vs. $12.79 each for the brand i usually get
07:37 PM roycroft: and i had to order 60 of them, so that's a *huge* price difference
07:38 PM XXCoder: indeed
07:38 PM roycroft: oh, it looks like it's printing now, and the webcam is down again
07:38 PM * roycroft heads out to inspect the first few layers
07:39 PM XXCoder: I finally did the stone paver upgrade to my printer
07:43 PM roycroft: it seems to be going fairly well so far
07:43 PM roycroft: the real test will be the end of the print, of course, when that lip is printed
07:44 PM roycroft: but my print time is done from a bit under 2 hours to 30 minutes, and i've both added that thickness at the lip and made one of the internal fillets a lot bigger, so i'm extruding a fair amount more material
07:44 PM roycroft: if this works well, i'll be happy with a 30 minute print time
07:46 PM XXCoder: nice
07:47 PM Tom_L: you broke the budget though!
07:48 PM XXCoder: yeah considering your huge need for those, time savings is good
08:08 PM roycroft: i upped the print speed from 80mm/s to 150mm/s
08:10 PM roycroft: i also lowered the temperature by 5 degrees, and i'm not sure that's the best idea
08:11 PM roycroft: but i can print so fast now that it's no big deal to do another go after this
08:13 PM Tom_L: maybe you need a filament extruder now
08:15 PM roycroft: if i get this new hot end/nozzle working well i'll be pretty happy
08:32 PM roycroft: and lowering the temperature was not a good idea
08:37 PM XXCoder: yay oprinting something lol
08:37 PM XXCoder: remote holder
09:35 PM roycroft: the tabs are still not printing well :(
09:39 PM XXCoder: sucks
09:44 PM roycroft: i'll just keep tweaking things until it turns out decently
09:44 PM roycroft: i slowed the print speed down for this go
09:44 PM roycroft: i should probably adjust the design so it is not as tall - that will save a lot of time and materials while i'm doing the testing
09:56 PM XXCoder: yeah
09:56 PM XXCoder: maybe dont bother with bottom too
09:56 PM XXCoder: just walls then bridge
10:21 PM XXCoder: roycroft: nice, it worked very well
10:22 PM XXCoder: had one design error, bores wasnt big enough for magnets. was easy fix, I just used hot air then push it in
10:32 PM XXCoder: updated design to have 5 magnets too if this one fails ill print that one
10:33 PM roycroft: cool
10:33 PM roycroft: this print is almost done, and i think it will be better than the last two
10:34 PM roycroft: slowing it down is only adding 10 minutes to the job, which is acceptable if it works
10:35 PM XXCoder: nice
10:35 PM XXCoder: updated one has fillets all over lol
10:37 PM XXCoder: https://imgur.com/iiibNvs.png
10:38 PM roycroft: this last go is almost acceptable
10:38 PM roycroft: i've not been adding supports for the lip
10:38 PM XXCoder: no wonder. magnets is 8 mm. I assumed it was 7 lol
10:38 PM roycroft: perhaps i'll try that
10:39 PM XXCoder: yeah
10:39 PM XXCoder: add thin bar under end of lip
10:39 PM roycroft: so you have a good interference fit
10:39 PM XXCoder: it'll bridge, build one layer, then bridge tab to it
10:39 PM roycroft: the lip profile is an arc
10:39 PM XXCoder: much stronger bridge
10:40 PM roycroft: i could make it just a 45 degree triangular profile, and it would probably print better, but the lip is there not just for holding a label - it's for lifting as well
10:40 PM XXCoder: hold on a sec
10:40 PM XXCoder: lemme draft idea very fast
10:42 PM roycroft: i'm not concerned about wasting material if i add supports, but that would add time to the prints, and require some cleanup after
10:43 PM XXCoder: https://imgur.com/jgAw8QR.png
10:43 PM roycroft: i suppose i could be really clever and design these boxes as two pieces, the tab and the box, and snap them together
10:43 PM XXCoder: the top is label, side profile
10:43 PM XXCoder: notice how there is 0.4mm at bottom. it will cause slicer to bridge that spot first, then build a layer
10:43 PM XXCoder: then it will bridge from edge of wall to that bar
10:43 PM roycroft: yeah, but i think that would print even worse than what i'm doing
10:44 PM roycroft: let me see if i can grab a pic of my profile
10:44 PM XXCoder: not really, its trick I used before. it works great even on my crap printwre when I had it
10:45 PM roycroft: adding that bit at the bottom of the lip might help, yes
10:45 PM roycroft: it's the flat profile that i think would cause problems
10:48 PM XXCoder: well if you make it slope 45 degrees it'll be printable but dunno
10:49 PM XXCoder: not really good, takes more space inside
10:49 PM roycroft: s://roycroft.us/LipProfile.png
10:49 PM roycroft: ack
10:49 PM roycroft: https://roycroft.us/LipProfile.png
10:49 PM roycroft: try that :)
10:49 PM XXCoder: works fine lol
10:49 PM roycroft: anyway, that's what i have now
10:49 PM XXCoder: topmost is eh since it runs into angle issue
10:50 PM roycroft: i can always add a tab at the end like yours
10:50 PM roycroft: that's sideways, btw
10:50 PM roycroft: the top is on the right
10:50 PM XXCoder: slightly more gentle but yeah
10:50 PM XXCoder: a second
10:51 PM roycroft: dropping the edge down a bit more would make for a better grip anyway
10:54 PM XXCoder: you probably could do arc then strighten it
10:54 PM XXCoder: but yeah
10:54 PM XXCoder: trying to draft it, it dont work well lol
10:54 PM roycroft: i got my new gpu today and installed it during my meeting tonight
10:54 PM roycroft: sw is actually fun to use now
10:54 PM XXCoder: nice lol
10:54 PM roycroft: but i'm taking a break for a bit right now
10:55 PM roycroft: i have been in my office since 6am this morning, other than when i made dinner
10:55 PM roycroft: so i'm watching a star trek episode and chilling on the sofa for a bit
10:55 PM roycroft: i'll go do another edit and print in a bit
10:56 PM XXCoder: :)
11:43 PM roycroft: https://roycroft.us/LipProfileNew.png
11:43 PM roycroft: that's what i'm printing now
11:45 PM XXCoder: hope it works
11:45 PM roycroft: it's going to add another 10 minutes
11:45 PM roycroft: but i'm still printing in less than half the time as with the old nozzle
11:45 PM XXCoder: indeed
11:46 PM roycroft: i think that encompasses your suggestion on the end of the lip while letting me keep the curved support
11:46 PM roycroft: although i did change that curve from an arc to a spline
11:46 PM XXCoder: indeed
11:46 PM XXCoder: lip makes it more grippable
11:46 PM roycroft: yes
11:47 PM roycroft: this will probably be my last test print of the night
11:47 PM roycroft: so i hope it's the good one
11:48 PM roycroft: i forgot to add supports
11:48 PM roycroft: dammmit
11:48 PM roycroft: oh well, i'll see how it goes without them
11:48 PM roycroft: i can't see how it will print that lip on the bottom without supports, though
11:49 PM roycroft: i think i need to abort, and add supports and reprint
11:55 PM roycroft: and that adds another 20 minutes
11:55 PM roycroft: it's getting to the point where it's not going to be much faster than with the old nozzle and the original design, and will use more material