#linuxcnc | Logs for 2018-09-19

Back
[00:10:23] -!- DaPeace1 has joined #linuxcnc
[00:13:30] -!- DaPeace1 has quit [Client Quit]
[00:15:13] -!- m0n5t3r has quit [Ping timeout: 246 seconds]
[00:15:24] -!- ferdna has quit [Quit: Leaving]
[00:32:17] -!- m0n5t3r has joined #linuxcnc
[00:47:35] -!- \01 has joined #linuxcnc
[00:49:32] -!- \01 has quit [Killed (Sigyn (Spam is off topic on freenode.))]
[01:01:20] <miss0r> trentser: you still around?
[01:06:04] * miss0r is out
[01:06:14] -!- veek has joined #linuxcnc
[01:19:42] -!- Inline has quit [Quit: Leaving]
[01:52:43] -!- ziper has quit [Read error: Connection reset by peer]
[02:17:29] -!- P1ersson has joined #linuxcnc
[02:30:00] -!- rob_h has joined #linuxcnc
[02:38:51] -!- Mian has joined #linuxcnc
[02:43:44] -!- Mian has quit [Remote host closed the connection]
[03:03:45] -!- P1ersson has quit [Ping timeout: 252 seconds]
[03:08:52] -!- Deejay has joined #linuxcnc
[03:09:07] <Deejay> moin
[03:14:40] -!- asdfasd has joined #linuxcnc
[03:19:06] -!- gregcnc has quit [Ping timeout: 264 seconds]
[03:19:32] -!- gregcnc has joined #linuxcnc
[03:29:09] <pink_vampire> what is the "P" value in the g64 command?
[03:29:12] <pink_vampire> G64 P0.05
[03:29:20] <pink_vampire> what is the 0.05?
[03:31:02] <Deejay> http://linuxcnc.org
[03:31:05] <Deejay> tolerance
[03:31:34] <Deejay> example: G64 P0.015 (set path following to be within 0.015 of the actual path)
[03:32:30] <Deejay> hi pink btw ;)
[03:38:37] <pink_vampire> thanks!
[03:38:56] <pink_vampire> I'm faceing wood in 60K
[03:39:01] <pink_vampire> so fun
[03:51:30] <Deejay> wow
[03:52:18] <pink_vampire> it become dust
[03:54:31] -!- veek has quit [Quit: Leaving]
[03:55:07] <pink_vampire> Deejay: https://i.imgur.com
[03:59:33] -!- JesusAlos has joined #linuxcnc
[04:00:30] <MarcelineVQ> particle board? careful not to breathe that in
[04:00:55] <pink_vampire> i know i have a mask
[04:01:55] Jin|away is now known as Jin^eLD
[04:07:43] -!- nallar has joined #linuxcnc
[04:08:16] -!- Ross has quit [Ping timeout: 272 seconds]
[04:08:17] nallar is now known as Ross
[04:24:49] -!- MacGalempsy has joined #linuxcnc
[04:24:52] <MacGalempsy> yo
[04:27:29] <pink_vampire> hi
[04:29:21] <MacGalempsy> making chips PV?
[04:29:30] <pink_vampire> yeah
[04:30:26] <MacGalempsy> what are you making?
[04:37:17] <MacGalempsy> any nacho cheese to go with the chips?
[05:08:04] -!- rob_h has quit [Ping timeout: 240 seconds]
[05:12:13] <pink_vampire> i'm working on my servo spindle
[05:13:41] -!- Valen has quit [Quit: No Ping reply in 180 seconds.]
[05:15:28] -!- Valen has joined #linuxcnc
[05:17:49] <XXCoder> yo
[05:22:27] <pink_vampire> tinkerbell making some magic dust
[05:22:59] <XXCoder> is that codeword for cocoine? lol
[05:24:04] <pink_vampire> aluminum "chips"
[05:24:44] <XXCoder> not exactly snortable lol
[05:25:03] <pink_vampire> I just LOVE the super low load that the spindle apply to the material
[05:28:49] <pink_vampire> WTF was that?
[05:29:03] <pink_vampire> the end mill just gone in a bang
[05:29:14] <XXCoder> ow
[05:29:31] <XXCoder> if it was making owder it would be because it rubbed itself dull and broke by sideload
[05:29:41] <pink_vampire> no, my mistake
[05:30:11] <XXCoder> oh what was it?
[05:31:58] <pink_vampire> I bore the hole and left the retract on center..
[05:32:11] -!- jonathanwash4 has joined #linuxcnc
[05:32:14] -!- jonathanwash4 has quit [Remote host closed the connection]
[05:32:32] <pink_vampire> XXCoder: https://i.imgur.com
[05:32:32] <XXCoder> so it didn't clear hole?
[05:32:48] <pink_vampire> it clear the hole, and then cleat the end mill
[05:33:33] <XXCoder> no, clear as in went higher than hole so it wouldnt hit side
[05:34:48] <pink_vampire> I just found the end mill (half of it)
[05:35:13] <XXCoder> anyway it happens. I had to modify program at work today, because it would try to move X way too far to park table for operator to change part
[05:35:24] <pink_vampire> OMG
[05:35:27] <XXCoder> and for some weird reason one line was messed up
[05:35:37] <XXCoder> so fixed those and uploaded
[05:35:40] <pink_vampire> this is soo bad
[05:36:03] <XXCoder> is part fine or scrapped?
[05:36:12] <pink_vampire> the part is fine
[05:36:23] <pink_vampire> but the endmill is no joke
[05:36:57] <pink_vampire> https://i.imgur.com
[05:37:12] <XXCoder> lol I have no idea how many tools I worn out or destoyed at work
[05:37:15] <XXCoder> hundreds easily
[05:37:17] <pink_vampire> https://i.imgur.com
[05:37:57] <XXCoder> so it center then slope down run to edge, then spirl cut down
[05:37:59] <gregcnc> oops, probably why enclosures are steel
[05:38:13] <XXCoder> then it centers while puck is still there?
[05:38:40] <XXCoder> okay I finally get the reason. ya some programs at work has m0 where I gonna clear out cutoffs
[05:39:06] <pink_vampire> but HF spindles are sooo fun
[05:39:25] <XXCoder> gregcnc: a61 at work has 2 panel bulletproof windows
[05:39:31] <XXCoder> both is cracked, and one has a hole
[05:39:51] <gregcnc> dang
[05:40:09] <XXCoder> programmer had a dumbs on one part, it went 10 times faster than supposed to and endmill one inch thick and 8 inches flute length broke and flew
[05:40:39] <pink_vampire> the best part is my wife not going to ask me anything
[05:41:22] <XXCoder> lol no worries for me, im single for life
[05:41:53] <gregcnc> pink did you move or wake everyone at home up?
[05:42:38] <pink_vampire> gregcnc: I'm the only one here.
[05:43:27] <pink_vampire> and TinkerBell of course.
[05:45:05] -!- benoliver99926 has joined #linuxcnc
[05:48:49] -!- benoliver99926 has quit [Remote host closed the connection]
[05:54:53] -!- fragalot has joined #linuxcnc
[05:57:05] <jthornton> morning
[05:57:17] <XXCoder> yo
[05:57:35] <XXCoder> friday will be hot here
[05:57:40] <XXCoder> it will actually reach 66f
[05:58:23] <gregcnc> probably our last 87°F day tomorrow
[06:02:01] <jthornton> high 92 for today and tomorrow then it falls off to mid 70's :)
[06:03:32] -!- nos has quit [Ping timeout: 272 seconds]
[06:03:47] -!- nos has joined #linuxcnc
[06:03:58] <XXCoder> nice
[06:04:44] <jthornton> it's been cool enough to ride the mountain bike at sunrise
[06:05:19] <gregcnc> so yesterday I put a bunch of 6061 parts off the lathe into a cup with water to rinse them and walked away for lunch
[06:05:48] <gregcnc> some time later they were stained
[06:05:52] <jthornton> https://www.mtbproject.com
[06:06:41] <gregcnc> that looks nice
[06:06:47] <jthornton> except the outer loop is 12 miles
[06:08:47] <gregcnc> our county has thousands of acres in preserves and you are not really allowed to use much for biking
[06:08:59] <gregcnc> as teenagers we did anyway
[06:09:17] <XXCoder> rusted?
[06:09:42] -!- edcragg0 has joined #linuxcnc
[06:09:57] <jthornton> https://www.youtube.com
[06:10:08] <jthornton> that's me riding the trail
[06:11:32] <gregcnc> aluminum, brown staining, parts are fine if I blow them off with or without washing. i just forgot
[06:12:30] <gregcnc> I have to try to see if a dip in somethign will brighten them or if it's going to be scotchbright
[06:14:07] <jthornton> https://www.youtube.com
[06:14:30] <jthornton> they must ride the whole trail in 1:31
[06:14:33] -!- edcragg0 has quit [Remote host closed the connection]
[06:14:43] <XXCoder> one min 31 seconds? ;)
[06:14:53] <jthornton> 1h:31m
[06:15:00] <XXCoder> yeah figured that
[06:15:47] <gregcnc> what distance?
[06:15:58] <jthornton> the forest service had it fixed up real nice
[06:16:21] <jthornton> the outer loop is 12 miles, if you ride all the trails it's 26 miles
[06:20:46] <jthornton> https://www.fs.usda.gov
[06:20:53] <jthornton> finally found the project map
[06:23:24] <jthornton> XXCoder: https://www.youtube.com
[06:26:17] <XXCoder> looking
[06:26:57] <XXCoder> weird it didnt show up in my gmail
[06:30:04] <Tom_L> 74�F Hi 93
[06:30:18] <gregcnc> when a new guy starts at the shop? https://www.instagram.com
[06:30:23] <jthornton> you get email notification of myfordboy videos?
[06:30:55] <XXCoder> jthornton: no wonder it was 2 months ago!
[06:30:56] <XXCoder> yeah
[06:31:00] <XXCoder> sub then click bell
[06:31:17] <jthornton> ah
[06:31:23] <XXCoder> I click bell on ALL subbed channels because summary skips many of those
[06:31:44] <XXCoder> yes it means i get 10 to 20 emails a week but great list of videos to watch usually.
[06:32:07] <XXCoder> I remove channels that make daily videos because thats just too much and tend to be content-free
[06:32:32] <XXCoder> oh yeah that reminded me that he fgot facemill really cheap than ones at work
[06:35:29] -!- P1ersson has joined #linuxcnc
[06:54:20] -!- Blumax has joined #linuxcnc
[06:56:04] -!- diverdude has joined #linuxcnc
[06:56:06] <diverdude> Hi there - So my arduino uno runs a very imple loop where it just sends a number every second. I am trying to read this using boost::asio with a callback on my computer. For some reason it reads a number and then it reads 0 bytes and exits unexpectedly although it should just call the callback funtion whenever data is read from port. If I test same program when i emulate serialport w. socat instead - no problem? Why does it not wor
[07:01:41] -!- emsjessec has joined #linuxcnc
[07:08:06] -!- rob_h has joined #linuxcnc
[07:15:39] -!- c-log has quit [Ping timeout: 252 seconds]
[07:17:01] -!- c-log has joined #linuxcnc
[07:17:02] -!- c-log has quit [Changing host]
[07:17:02] -!- c-log has joined #linuxcnc
[07:18:27] <miss0r> G'day
[07:21:35] <diverdude> miss0r: hey there
[07:21:43] <miss0r> hello :]
[07:22:04] <diverdude> miss0r: ever did serialport communication btwn arduino and a linux machine? :)
[07:22:13] <miss0r> nope
[07:22:18] <diverdude> dang :D
[07:22:24] <miss0r> only to a windoze
[07:22:31] <diverdude> ah yes ok
[07:22:33] <miss0r> What is the issue
[07:22:55] <diverdude> no for some reason the connection gets disconnected after reading a few bytes
[07:24:04] <miss0r> That is odd. What are you using to read the serial communication with?
[07:26:08] <miss0r> Have you tried this description: https://playground.arduino.cc
[07:26:18] -!- P1ersson has quit [Ping timeout: 245 seconds]
[07:40:53] -!- selroc has joined #linuxcnc
[07:41:19] -!- P1ersson has joined #linuxcnc
[07:41:43] -!- Zathrus23 has joined #linuxcnc
[07:43:32] -!- selroc has quit [Client Quit]
[07:44:36] -!- Zathrus23 has quit [Remote host closed the connection]
[08:19:48] <miss0r> I'm seeking inspiration to help sold a problem: I have a 5mm thick flange mounted with 4x 8mm screws. The screws are welded to the surface the flange is attached to. Each screw protrudes 15mm. I have to lift the flange 11.5mm off the surface. Any good ideas as to how to do this?
[08:19:56] <miss0r> sold=solve*
[08:20:36] <miss0r> fragalot: This is something you'd come up with a solution to in a heartbeat ;)
[08:20:46] <fragalot> who what where when why?
[08:20:59] <miss0r> what I just wrote above :D
[08:21:24] <miss0r> The flange is a quite complicated one, that amongst other things holds a pneumatic cylinder. So the fixture needs to still be somewhat strong
[08:21:59] <fragalot> and an 8mm plate with hole pattern
[08:22:04] <fragalot> is not a solution, because.. ?
[08:22:22] <miss0r> 8mm ? :S
[08:22:35] <fragalot> 15mm screw minus thickness for an M8 nut :P
[08:23:02] <miss0r> Well. It only needs to be moved away from the surface, not in any other direction.
[08:23:28] <miss0r> The way i see it, when the flange is lifted 11.5mm, the m8 screw is 1.5mm short of sticking out the top of the flange
[08:24:01] <fragalot> ah the screws are welded from the opposide side, rather than being 15mm studs
[08:24:04] <miss0r> the screws are welded from inside the frame, before the machine was put together
[08:25:01] <miss0r> yeah... So its not *that* straight forward :D
[08:25:25] <fragalot> there's no way around this apart from either going chicaco screw style, or doing it properly and extending said bolts
[08:25:32] <fragalot> or cutting & threading female
[08:27:02] <miss0r> hmm.. No real way for me to extend the bolts
[08:27:08] -!- P1ersson has quit [Ping timeout: 245 seconds]
[08:27:15] <miss0r> so I guess, I'll have to go with the chicago style screws, and maker seprate spacers
[08:27:34] <fragalot> drilling out said bolts and tapping M8 is not an option either I guess?
[08:27:37] <miss0r> I should be able to drill the holes a bit larger in the flanges
[08:27:58] <miss0r> I would be afraid of doing so
[08:28:06] <fragalot> chicago it is.
[08:28:17] <miss0r> The box section they are protruding from is the main cable channel
[08:28:26] <fragalot> ha :D
[08:28:33] -!- KimK has quit [Ping timeout: 252 seconds]
[08:28:35] <gregcnc> hmmm I had to look that up, but wiki calls them sex bolt
[08:28:51] <miss0r> realy? I've always known under that name
[08:28:51] <fragalot> this is why I love having a selection of "sleeves" to put over drills to stop them poking through when drilling in plate
[08:29:15] <gregcnc> there is a list of names https://en.wikipedia.org
[08:29:17] <miss0r> yeah... sleeves or not. I would still have a small heart attack each time
[08:29:19] <fragalot> (still wouldn't risk drilling into acable channel :D)
[08:29:45] <miss0r> its just alot of machining, making spacers & bolts.
[08:30:05] <miss0r> If only I had to move it out further, I would've just made a spacer with an M8 thread all the way through
[08:39:20] <fragalot> cut threads to 8mm long, make spacer with M8F on one end, and M8M on the other
[08:39:21] <fragalot> :P
[08:40:56] -!- KimK has joined #linuxcnc
[08:41:11] <miss0r> That sounds like somethat that could make you alot of money, should it work :D
[08:41:49] <fragalot> they exist, but the part is typically longer
[08:42:03] <fragalot> (eg. my schaublin uses those for the MC ->ISO30 adapters)
[08:43:19] <miss0r> sure... but the "thick part" would need to be -3.5mm long
[08:43:49] <fragalot> removing material is easy
[08:43:53] <fragalot> I don't see the problem
[08:44:00] <fragalot> :D
[08:45:30] -!- rob_h has quit [Ping timeout: 264 seconds]
[08:58:56] <diverdude> I have an arduino program which sends a numeric value (integer) every second via serialport. I have a c++ program using boost::asio which should read this data. It disconnects from arduino stream after having read 1 integer. If i use same c++ program om socat serial simulation it works. Here is hexoutput( made from xxd /dev/ttyACM0 - over a period: https://paste.ubuntu.com) that no EOF or other strange meta is sent. W
[09:09:19] -!- FinboySlick has joined #linuxcnc
[09:13:43] -!- Tecan has joined #linuxcnc
[09:13:44] -!- Tecan has quit [Changing host]
[09:13:44] -!- Tecan has joined #linuxcnc
[09:18:31] -!- P1ersson has joined #linuxcnc
[09:19:47] -!- dkrm15 has joined #linuxcnc
[09:22:41] -!- dkrm15 has quit [Remote host closed the connection]
[09:30:36] -!- MacGalempsy has quit [Quit: Yaaic - Yet another Android IRC client - http://www.yaaic.org]
[09:30:50] -!- syyl has joined #linuxcnc
[09:46:05] <fragalot> diverdude: might want to ask over in #arduino or similar
[09:46:26] <diverdude> i solved it
[09:47:56] -!- HSD has joined #linuxcnc
[09:48:11] <pink_vampire> dry fit - DONE
[09:52:23] <miss0r> This project is turning to shit right fast in a hurry :-/
[09:52:40] <miss0r> haha. I need to run out and find some more suitable stock to get this done.
[09:59:57] -!- Inline has joined #linuxcnc
[10:00:48] -!- Inline has quit [Read error: Connection reset by peer]
[10:01:15] -!- Inline has joined #linuxcnc
[10:01:27] <pink_vampire> https://i.imgur.com
[10:09:13] <fragalot> does it run yet? :D
[10:09:38] <pink_vampire> not yet, just first dry fit.
[10:09:59] <fragalot> miss0r: at least you can now use those ER32 hex blocks you haven't bought yet? :)
[10:10:54] <pink_vampire> I'm not 100% happy with how the pulley connect to the R8 cartridge
[10:18:30] -!- P1ersson has quit [Ping timeout: 264 seconds]
[10:46:25] -!- Tecan has quit [Quit: Live Long And Phosphor!]
[10:56:20] -!- drbean15 has joined #linuxcnc
[10:58:22] -!- drbean15 has quit [Killed (Unit193 (Spam is not permitted on freenode.))]
[11:02:53] -!- fragalot has quit [Quit: Page closed]
[11:05:30] -!- ziper has joined #linuxcnc
[11:19:04] -!- jerryq has quit [Ping timeout: 240 seconds]
[11:23:48] gimps_ is now known as gimps
[11:57:14] -!- jerryq has joined #linuxcnc
[12:03:21] -!- JesusAlos has quit [Quit: JesusAlos]
[12:12:34] <Simonious> Our plotter is being sunset, because there will be no drivers for it for windows10. :/
[12:12:41] <Simonious> so new plotter!?
[12:16:33] <Loetmichel> Simonious: VM?
[12:16:57] <roycroft> i still use a designjet 650c
[12:17:15] <Loetmichel> i have a designjet 650C A0+ here. built in 1991... still running with win7
[12:17:23] <roycroft> there are no windows 7 drivers for it, but i got a copy of 64-bit windows xp and installed the driver from that
[12:17:53] <MarcelineVQ> I still can't imagine a reason to upgrade past windows 7 if your computer isn't for internet browsing
[12:18:13] <Simonious> Loetmichel: the plotter needs to be usable for several endusers, the company is moving to windows 10 with windows 7 sunset in 2020.
[12:18:19] <roycroft> mine is e size - it takes 36" rolls
[12:18:39] <Loetmichel> Simonious: 2023
[12:18:41] <Loetmichel> but right
[12:18:55] <roycroft> if you need an excuse to upgrade the plotter, this is a great one
[12:18:57] <Simonious> I think there are workarounds (ok), but we don't want to spend (much) time on workarounds.
[12:19:10] <roycroft> i don't know if windows 7 printer drivers work on windows 10
[12:19:22] <roycroft> but you might try installing the windows 7 driver to see if it does work
[12:19:34] * roycroft knows as little as he can get away with about windows
[12:19:40] <Simonious> wait - which did you say was an upgrade? the dsignjet 650c? I thought that was an older plotter too?
[12:19:43] <Loetmichel> but where is the problem to set asidea "print server" box that runs winXP and only sees the internal network, not the internet?
[12:19:57] -!- fragalot has joined #linuxcnc
[12:20:13] <Simonious> Loetmichel: that *might* be doable
[12:21:17] * Loetmichel wouldt let go of his designjet 650C... i might have to soon though because i cant seem to get a new maintaineance station for it
[12:21:45] <Loetmichel> and the old one has cracked gaskets so the expensive ink tank/nozzes are drying out every few weeks
[12:22:11] <roycroft> teh internet suggests you may be able to install a windows 7 printer driver on windows 10
[12:22:59] <roycroft> loetmichel: i don't use my 650c often
[12:23:27] <roycroft> and i don't want to waste time/paper/ink doing a test plot every week or two to keep the ink cartridges "fresh"
[12:24:00] <roycroft> i've found that when they start drying out and don't work any longer, if i hold the print head under warm running water for a couple minutes the cartridge starts working fine again
[12:28:16] <Simonious> I guess if I do have to replace it - who has a newer plotter they love and what is it?
[12:33:23] <JT-Shop> https://semo.craigslist.org
[12:35:34] <roycroft> that's not a bad plotter
[12:35:40] <roycroft> a little pricey for a d size plotter though
[12:46:09] <JT-Shop> it cost $2265
[12:51:44] <cradek> that's only like 3 sets of cartridges for a 650c
[12:52:23] <cradek> and they last right up until you have the nerve to stop plotting for 20 minutes
[12:52:34] <cradek> (I hate those machines)
[12:52:46] <cradek> but the way it grabs the knife to cut the roll off is neat
[12:54:44] -!- ktchk has joined #linuxcnc
[13:04:48] -!- sensille has joined #linuxcnc
[13:06:25] <miss0r> fragalot: Good thing I remembered I had some 17mm hex stock on the shelf before going anywhere :D The bolts are done now. Now I just need to make some spacers
[13:06:30] <sensille> can anyone recommend a laser for engraving? i have no idea what the needed specs are
[13:07:04] <sensille> i want to mount it to a corexy and go fast
[13:08:00] <miss0r> fast as in 'wr00m wr00m' or fast as in speed ajuster is turned to rabbit rather than turtle?
[13:09:29] <sensille> up to 1000mm/s
[13:09:41] <miss0r> What are you planning to do engraving on ? :)
[13:09:43] <sensille> it's not for serious results, more to test the hardware
[13:10:08] <sensille> i want a way to record the path of the head, including ripples, overshoots etc
[13:10:23] <sensille> so i can test several strategies to reduce them
[13:10:24] <miss0r> ha... with a laser?
[13:10:36] <sensille> is that a stupid idea?
[13:10:59] <miss0r> Well.. overshoots & ripples can be hard to detect on an engraved line
[13:11:09] <miss0r> if you want to notice something smaller than 0.5m
[13:11:10] <miss0r> mm*
[13:11:29] <miss0r> even if it is a fine laser line
[13:11:35] <sensille> hm
[13:12:01] -!- phiscribe_ has joined #linuxcnc
[13:12:06] <sensille> is the focus of the laser not good enough, or does it depend on the material to engrave on?
[13:12:12] <miss0r> for that stuff, I normally use dial indicators & alot of setups :) I'm sure theres a smarter way though
[13:12:32] <miss0r> well.. in many cases the focus is ajustable.
[13:12:52] <miss0r> but lets be realistic here; you are going to engrave on wood or plastic
[13:13:13] <miss0r> if you want a laser than can engrave on steel, it will have a sizeable pricetag on it, probally not what you want for a test setup? :)
[13:13:16] <sensille> has to be cheap, because i'll do many runs
[13:13:33] <sensille> (the material to engrave)
[13:13:36] <miss0r> In any case; the lines on wood & plastic has a tendency to 'smear' ever so slightly
[13:13:39] <sensille> like... paper?
[13:14:08] <miss0r> sure - wood fibres. you *can* get pretty good results. But seeing the errors with your eyes will be a real challange
[13:14:24] <miss0r> unless they are quite sizeable
[13:14:38] -!- phiscribe has quit [Ping timeout: 245 seconds]
[13:14:42] <sensille> that's why i want to exaggerate the speed so much
[13:14:49] <sensille> to make the errors very visible
[13:15:30] <sensille> target printing speed will be around 100mm/s, so if i can't see errors at 500mm/s with my eyes, it will probably be fine
[13:15:37] <miss0r> I'm not sure what you are trying to accomplish. Any machine can be forced to produce errors if it is run fast enough. You are in danger of noticing some errors that would not even be there at a more reasonable speed
[13:16:07] <sensille> really? aren't they just much smaller?
[13:16:20] <sensille> isn't it all deflection and stretch?
[13:16:21] <miss0r> It all depends. So many variables
[13:16:39] <miss0r> you *could* skip a step or outrun an encoder ect.
[13:16:59] <sensille> the driver would notice that
[13:17:02] <miss0r> but sure. the machine deflection would be amplified af speed
[13:17:20] <miss0r> I guess, I don't know your setup in detail
[13:17:39] <sensille> also skips are very audible
[13:17:49] <sensille> in a quiet setup
[13:18:00] <miss0r> plural, yes :)
[13:18:05] <sensille> so you think the line is not fine enough
[13:18:19] <miss0r> no. I would not think so.
[13:18:51] <miss0r> What I would do; a repitition test: move the spindle to a dial indicator and zero it. move back and forth at different speeds
[13:19:06] <miss0r> and see if it always zeros out the dial or if it overshootes ect
[13:19:27] <miss0r> preferably do this a few different places on the table
[13:19:33] <gregcnc> has nobody built a core xy with laser?
[13:20:02] <sensille> maybe some more background: i have a corexy 3d printer, and am trying to build a new firmware with very tight jerk control
[13:20:23] <sensille> and am searching for a way to directly visualize the differences
[13:20:24] <CaptHindsight> aqueous ink is cheap and easy to make or purchase by the gal/4L from China, always look for printers that are refillable
[13:21:04] <sensille> actual printing adds many more variables
[13:21:19] <sensille> so i though a laser can show the path more directly
[13:21:35] <gregcnc> how will you measure?
[13:21:37] <miss0r> Sensille: hehe. Well, never say never. You can get a pretty cheap laser engraver. But if you are planning on running it realy fast, I would go for a 15watt one.
[13:21:45] <sensille> especially the ripple after taking a turn interests me
[13:22:04] <miss0r> I don't don't see how you will see it on the line, unless it is realy bad
[13:22:05] <gregcnc> i think you want measuring devices not a laser
[13:22:22] -!- Papagno has joined #linuxcnc
[13:22:42] <sensille> hm. measuring devices for the dynamic movement?
[13:22:43] -!- micges has joined #linuxcnc
[13:22:59] <sensille> i could add linear encoders to the axes
[13:23:33] <sensille> they should show overshoots, too
[13:23:41] <miss0r> that would make sense. Have a data logger draw it up
[13:24:01] -!- rob_h has joined #linuxcnc
[13:24:02] <sensille> but it also changes the setup significantly
[13:24:14] <miss0r> sure
[13:24:16] <sensille> they don't move that easily
[13:24:20] <gregcnc> i'm a bit confused as I'm not sure how the laser comes into play
[13:24:27] <sensille> so they will dampen the ripple
[13:24:38] <CaptHindsight> if you can't see the errors are they really there?
[13:24:42] <gregcnc> you really do want to observe the actual position right?
[13:24:58] <gregcnc> not draw cartoons?
[13:25:06] <sensille> yes
[13:25:44] <gregcnc> linear encoder is probably the easiest though i don't know much about other options
[13:26:30] -!- syyl has quit [Quit: Leaving]
[13:26:35] <sensille> i'm afraid of the friction they add
[13:26:53] <sensille> a 3d printer is not a cnc
[13:26:58] -!- jasen has joined #linuxcnc
[13:27:15] <gregcnc> which encoders did you look at?
[13:28:20] <sensille> one sec
[13:28:49] -!- Robh__ has joined #linuxcnc
[13:29:47] <fragalot> miss0r: pfft, who keeps hex stock around :D
[13:29:59] * fragalot doesn't because it gets used up too quickly
[13:30:42] <sensille> gregcnc: i have one of those https://www.aliexpress.com
[13:31:01] <sensille> and i already removed the dust protection
[13:31:18] -!- rob_h has quit [Ping timeout: 272 seconds]
[13:31:41] -!- rob_h has joined #linuxcnc
[13:32:07] <fragalot> sensille: magnetic read heads don't really add significant friction as it's non-contact
[13:32:59] <sensille> this one is optical. and the friction just comes from the mount
[13:33:07] -!- Robh__ has quit [Ping timeout: 240 seconds]
[13:33:21] <fragalot> those do add significant drag on light setups
[13:34:01] <sensille> i could disassemble it and mount the components separately
[13:34:10] <sensille> without the spring
[13:34:40] <fragalot> if there's no chips and coolant flying around, you could indeed reduce friction quite a bit on those
[13:36:34] <CaptHindsight> what is the actual problem with the laser? Trying to get a poorly working one better?
[13:37:05] <fragalot> CaptHindsight: i'm more guessing he's working on step generation & trying to optimize pathing to have minimal overshoot
[13:37:19] <sensille> yes
[13:37:43] <CaptHindsight> stop the guessing
[13:37:45] <gregcnc> someone was here a couple weeks ago talking about jerk for 3d printer, must be a popular topic
[13:37:58] <sensille> i could use a small laser and 'engrave' on a ccd chip
[13:38:01] <fragalot> CaptHindsight: when I said guessing, I meant "i've deduced that..."
[13:38:04] <CaptHindsight> please mention actual problem :P
[13:38:11] <sensille> gregcnc: probably me :)
[13:38:23] <gregcnc> i gave a link to yt
[13:38:30] <sensille> yes
[13:38:54] <fragalot> CaptHindsight: you mean you don't like the xy problem? (or whatever it's called)
[13:38:54] <sensille> but he only started with jerk control
[13:39:16] <CaptHindsight> I have no feelings on the matter
[13:39:59] <CaptHindsight> define problem, solve problem
[13:40:00] <sensille> so, magnetic linear encoders?
[13:40:15] <CaptHindsight> feelings are for Twitter
[13:40:23] <fragalot> CaptHindsight: that is exactly what the XY problem is..
[13:41:16] <gregcnc> if you put encoder on a flimsy frame will it give the answers you need?
[13:41:57] <fragalot> it depends on what you're after, really
[13:42:18] <fragalot> if you take only the actuator with a load on it, and measure the movement of said load, that is one thing
[13:42:20] <gregcnc> what about accelerometer on the axis?
[13:42:30] <sensille> too slow
[13:42:46] <fragalot> if you counterlever said load off-axis, the whole game changes
[13:42:56] <fragalot> sensille: accelerators do exist in high speed models
[13:43:48] <sensille> hm. they _should_ be able to show the overshoot
[13:44:33] <fragalot> for example the ADXL1005 should give you a 23kHz range
[13:45:41] <CaptHindsight> lets just say that we now now the exact amount of "overshoot" now what? Write research paper?
[13:45:49] <CaptHindsight> now know
[13:46:24] <sensille> observe the influence of jerk and acceleration on it
[13:46:42] <sensille> or even compensate it by modifying step generation
[13:46:45] <CaptHindsight> sounds like a research paper
[13:46:55] <sensille> without the paper part
[13:47:06] <fragalot> changing acceleration and jerk ARE step generation modifications :P
[13:47:17] <CaptHindsight> well if the point is smoother engravings then you can see the difference
[13:47:20] <sensille> ok, change the algorithm
[13:47:43] <fragalot> but I do believe that it can be possible to have higher decelerations than standard and compensate for overshoot some other way
[13:48:06] <fragalot> we do very strange things with our drive signals some times to achieve higher accuracies
[13:48:12] <CaptHindsight> modify settings, look at engravings, repeat
[13:48:32] <sensille> fragalot: that's what i have in mind, too
[13:48:39] <sensille> in the end it's just physics
[13:48:53] <sensille> but i need a way to measure the result
[13:48:57] <fragalot> sensille: agreed.
[13:49:01] <pcw_home> Don't 3D printers usually aim for constant velocity? on a constant velocity path so jerk is determined by the Gcode
[13:49:16] <CaptHindsight> only school would have you waste time and make you show your math
[13:49:37] <fragalot> CaptHindsight: or industry trying to push the envelope
[13:49:50] <CaptHindsight> fragalot: nah
[13:49:55] <fragalot> in fact i've got whole papers written about certain tricks to make our controller faster
[13:49:56] <sensille> the gcode always dictates infinite jerk, at least the gcode that gets fed to 3d printers
[13:49:56] <CaptHindsight> that a research paper
[13:49:59] <sensille> only lines
[13:50:05] <fragalot> and the code & drive to prove it
[13:50:18] <CaptHindsight> again a research paper
[13:50:29] <fragalot> CaptHindsight: WITH a real-world application and implementation
[13:50:37] <pcw_home> No, the path determines the jerk on a constant velocity path
[13:50:50] <CaptHindsight> what did school do to you? :P
[13:51:10] <sensille> pcw_home: what, with instant direction changes? that's infinite jerk and acceleration
[13:51:11] <fragalot> CaptHindsight: nothing - on the job training did this to me :P
[13:51:29] <CaptHindsight> he just wants to plot faster
[13:51:52] <sensille> pcw_home: you'd have to stop completely and start again, but that's not constant |v|
[13:52:07] <sensille> CaptHindsight: you guessed wrong
[13:52:22] <pcw_home> the controller can determine the tangential jerk but not the path (unless its allowed to fudge the path)
[13:52:46] <fragalot> I'm all for sensille improving the standard S-curve used today
[13:53:00] <pcw_home> So at a constant tangential velocity you really dont control the jerk
[13:53:12] <pcw_home> your work does
[13:53:13] <sensille> fragalot: thanks :) it's hard to argue against "but we don't do it this way"
[13:53:20] <CaptHindsight> time wasting
[13:53:39] <sensille> CaptHindsight: not your time, though
[13:53:40] <fragalot> CaptHindsight: agree to disagree?
[13:54:02] <gregcnc> don't most cnc controls use some jerk control?
[13:54:04] <CaptHindsight> fragalot: see ^^
[13:54:50] <pcw_home> many do but its well to remember that it doesn't provide much benefit on constant velocity paths
[13:55:11] <pcw_home> since the path determines the jerk
[13:55:13] <sensille> so that accelerometer might give me enough resolution
[13:55:36] <sensille> <0.01mm
[13:55:56] <sensille> that's one way
[13:56:37] -!- ktchk has quit [Quit: Konversation terminated!]
[13:56:42] <sensille> looks like the laser idea is not feasable, because the engraved line would be too smeared to see details
[13:57:21] <sensille> and 15W would be a lot for that
[13:58:05] <fragalot> laser is very impractical
[13:58:37] <fragalot> even if you aim it at a ccd and use a clever algorithm to find the exact center of the smear
[13:59:26] <sensille> ok, thanks for helping with my brainstorming
[14:00:45] -!- pcw_home has quit [Remote host closed the connection]
[14:04:41] -!- jasen has quit [Quit: Page closed]
[14:14:41] -!- merkatorix has joined #linuxcnc
[14:15:09] -!- swarfer has joined #linuxcnc
[14:16:22] -!- ve7it has joined #linuxcnc
[14:18:54] -!- gloops has joined #linuxcnc
[14:22:39] -!- swarfer has quit [Quit: swarfer]
[14:23:40] -!- micges has quit [Quit: Leaving]
[14:24:54] -!- emc has joined #linuxcnc
[14:27:34] -!- rob_h has quit [Ping timeout: 240 seconds]
[14:28:01] -!- Papagno has quit [Quit: Sto andando via]
[14:29:13] <CaptHindsight> don't need the exact center of the smear since it's a smear
[14:29:30] <CaptHindsight> if smear = fuzzy
[14:30:45] <fragalot> CaptHindsight: you do if you are more interested in it's target position than you are in the spot size
[14:30:54] <fragalot> there are common techniques for this used in astrophotography
[14:31:06] <fragalot> to get sub-pixel accuracy of position
[14:32:41] <sensille> i was about to say a ccd chip is too slow, but if i really only want to see the path, i don't need time resolution
[14:33:10] <fragalot> sensille: think 2D, move X axis at constant speed, and do your jerk on the Y
[14:33:26] <fragalot> tadaah, crude oscilloscope view of the path :D
[14:33:28] <sensille> ah :)
[14:33:31] <miss0r> fragalot: At last I am done with all the components for tomorrow morning
[14:34:17] <miss0r> Its always exciting to say yes to do a replace & upgrade on a robot access door on a cnc mill.
[14:34:26] <sensille> maybe there are other light sources that are easier to focus
[14:34:26] <miss0r> (especially when you have to finish in two days)
[14:34:49] <fragalot> sensille: doubt that.
[14:35:04] <fragalot> especially because a laser's whole game is a tightly focussed parallel beam of light :P
[14:35:30] <fragalot> miss0r: do the 2 days include lead-time for any spares? :P
[14:35:51] <CaptHindsight> but.... he's not using a star to engrave, unless the engraving story is just that a story
[14:35:54] <miss0r> fragalot: You a forgetting about 'mkwamg' light focus. It can be a bit random, but it always focus in a hurry when theres a bug near by
[14:36:02] <miss0r> Mean Kid With A Magnifying Glass
[14:36:33] <miss0r> fragalot: yes, why the hell else would I be standing here for HOURS doing parts I can get for next to nothing, if I had a few days :D
[14:36:48] <fragalot> CaptHindsight: the engraving story is literally just a story.. a possible (poor) scenario to measure overshoot
[14:36:59] <CaptHindsight> ah hah!
[14:36:59] <sensille> maybe just rip a laser out of a cd player. they should be quite good at focussing
[14:39:01] <gloops> use the dambusters method
[14:39:12] <miss0r> sensille: Sure, but you'd want a bit more power if you want to engrave with it at a high speed
[14:39:26] <fragalot> miss0r: he's NOT ENGRAVING :D
[14:39:35] <miss0r> hey.. I was gone a few hours :D
[14:39:35] <sensille> miss0r: i'm at the ccd-idea at the moment
[14:39:45] <CaptHindsight> gloops: you want him to fly a Lancaster bomber a drop a rolling skipping bomb onto a laser?
[14:40:14] <CaptHindsight> can i watch?
[14:40:35] <gloops> CaptHindsight well the laser part isnt so important, but the lancaster bomber is
[14:40:52] <miss0r> fragalot: I spoke to GLS today. They now accept the parces is lost. I guess I will have to wait for them to reach out again then :S
[14:41:16] <miss0r> whats this with a lancaster bomber? :D
[14:41:40] <fragalot> miss0r: fun stuff.
[14:41:48] <CaptHindsight> miss0r: gloops idea to test his accuracy
[14:42:07] <miss0r> fragalot: not realy. It realy sucks to get something at a great price, only to have it lost in transit!
[14:42:11] * miss0r is realy pissed
[14:42:17] <miss0r> I was hoping to the last moment!
[14:42:26] <miss0r> CaptHindsight: wha o_0 ?
[14:42:35] <CaptHindsight> lancaster bomber drones for package delivery
[14:42:36] -!- P1ersson has joined #linuxcnc
[14:42:52] <fragalot> more reliable than any shipping service i've used the past few months
[14:43:03] <gloops> to cut the story short, the RAF commander of the dambusters mission had a brainwave while at the theatre
[14:43:06] <miss0r> even if they miss the target by a few miles
[14:43:19] <miss0r> gloops: How is that cutting it short?!
[14:43:27] <gloops> think stage lights - spotlights
[14:43:28] <CaptHindsight> the airdrops worked in Berlin, why not just drop packages every day over villages at the same time and then let the villagers sort it out?
[14:43:47] <CaptHindsight> probably fewer lost packages
[14:43:53] <miss0r> I'd say
[14:44:23] <gloops> spotlight on each wing aimed at the surface at the same angle - when you hit the right height, the spots become 1
[14:44:25] <miss0r> I don't know what pisses me off more about them loosing the package; the fact that it is gone or the fact that fragalot will never get to undo the crazy amount of screws I put into that box...
[14:44:29] <gloops> do that with your lasers
[14:44:48] <fragalot> gloops: that does not in any way shape or form help with what sensille is trying to do
[14:44:55] <MarcelineVQ> miss0r: a service lost a package large enough to have screws in it?
[14:45:19] <miss0r> MarcelineVQ: Yeah. well.. I probally overdid it. but the package was 300x100x100mm
[14:45:21] <fragalot> MarcelineVQ: 22 screws.
[14:45:34] <miss0r> fragalot: That was only the lid, mind you :D
[14:45:43] <fragalot> miss0r: it's probably because yor box looked like garbage
[14:45:47] <CaptHindsight> thats the size package that every deliver service loses
[14:46:11] <miss0r> fragalot: I find that offensive, sir! I used a flush trimmer on the wood router this time, to spiff it up a bit :D
[14:46:12] <CaptHindsight> ~8 x 10 x 5"
[14:46:12] <MarcelineVQ> What's the point of living in first world countries if you can't even get mail to people. Let's make an assumption here that those are related things.
[14:46:15] <miss0r> it made a hellova mess
[14:46:18] -!- rob_h has joined #linuxcnc
[14:47:01] <fragalot> miss0r: pics or it didn't happen and you never actually shipped it
[14:47:03] <fragalot> :P
[14:47:04] <miss0r> MarcelineVQ: Perhaps because most GLS drivers are not from first world contries... :P
[14:47:09] <miss0r> fragalot !
[14:47:12] <fragalot> :D
[14:47:18] <CaptHindsight> miss0r: found your box https://youtu.be
[14:47:21] <miss0r> I'm not even sure I took pictures this time... checking
[14:47:33] <miss0r> CaptHindsight: haha yeah
[14:47:36] <MarcelineVQ> CaptHindsight: it's, beautiful
[14:47:50] <miss0r> I made if from 15mm 11layer plywood. Used glue'n everything
[14:47:56] <miss0r> (not on the lid, though, I have a heart)
[14:48:07] <fragalot> pfft, I used 18mm for yours!
[14:48:15] <fragalot> granted, that's mainly because I had no 12mm
[14:48:18] <gregcnc> some paper about jerking http://lbms03.cityu.edu.hk
[14:48:27] <fragalot> gregcnc: might want to rephrase that
[14:48:29] <miss0r> fragalot: haha yeah
[14:48:37] <gregcnc> it was just for you
[14:48:37] <miss0r> fragalot: LOL!
[14:49:25] <fragalot> CaptHindsight: I wonder how long it's been going, and how long it would take for the corners to round over enough until it just rolls down
[14:49:25] <miss0r> I guess searching google for 'jerking issues' will come up with some less than relevant stuff, to the problem sensille has...
[14:49:28] <miss0r> or maybe not :P
[14:50:14] <sensille> i have more than one problem
[14:51:16] <gregcnc> basically in that case there was no difference in accuracy with jerk control
[14:51:41] <gregcnc> the machine looked pretty stout
[14:51:41] <CaptHindsight> jerks tend to be difficult to control
[14:52:23] <CaptHindsight> laser on delta robot
[14:52:24] <gloops> my favourite line from barnes wallis was when the prime minister was questioning his bouncing bomb idea - he said we havent got a plane that can carry this bomb, yes we have - the lancaster bomber wallis said, the RAF tell me it cant, how would you know better than them? said PM, because i designed the lancaster bomber said wallis hehe
[14:52:58] <CaptHindsight> interesting story
[14:54:04] <CaptHindsight> laser and "mirrors" mounted on delta robot
[14:55:50] <CaptHindsight> https://www.hindawi.com
[14:56:30] <CaptHindsight> https://etd.ohiolink.edu!etd.send_file
[14:57:08] <CaptHindsight> https://etd.ohiolink.edu!etd.send_file?accession=ohiou1460045979&disposition=inline
[14:57:50] <CaptHindsight> method of control is an improvement over traditional methods because it prevents the
[14:57:50] <CaptHindsight> occurrence of infinite jerk spikes, a problem that occurs with traditional robot and 3d printers
[14:57:51] <gregcnc> it's all been done before?
[14:58:47] <CaptHindsight> not everything
[14:59:07] <gloops> theres only 7 stories at the flicks
[14:59:46] <MarcelineVQ> At the cinerama
[15:00:09] <CaptHindsight> https://pdfs.semanticscholar.org
[15:00:24] <CaptHindsight> An Application of Screw Theory to the Jerk Analysis of a
[15:00:24] <CaptHindsight> Two-Degrees-of-Freedom Parallel Wrist
[15:00:42] <MarcelineVQ> That just makes me want to watch The Jerk again.
[15:00:51] <gloops> the angle of the dangle is directly proportional to the ...
[15:00:53] <MarcelineVQ> Amazing entendre aside
[15:01:20] <CaptHindsight> https://openscholarship.wustl.edu
[15:01:31] <CaptHindsight> Minimum Jerk Trajectory Planning for Trajectory
[15:01:31] <CaptHindsight> Constrained Redundant Robots
[15:02:07] <sensille> oh damnit, he's also doing 5th order trajectory generation
[15:02:14] <sensille> it's all been done before
[15:03:54] <CaptHindsight> maybe 6th order
[15:04:00] <CaptHindsight> you'll show em
[15:05:30] <sensille> 5th order is a natural choice
[15:05:38] <MarcelineVQ> there's only 7 orders at the jerk analysys
[15:06:20] <gloops> if we stay in the box that is
[15:06:38] <CaptHindsight> so think outside the box...
[15:06:44] <CaptHindsight> hyperspaces?
[15:06:47] <gloops> what box?
[15:06:59] <CaptHindsight> the box is no longer a box
[15:07:54] <MarcelineVQ> it's still a box, it's just got too much box to hold onto
[15:08:12] <MarcelineVQ> You stick your hand in there and nothing makes sense anymore
[15:08:59] <gloops> now we're getting someplace
[15:09:40] <CaptHindsight> like teenagers
[15:11:05] <CaptHindsight> http://www.et.byu.edu
[15:11:38] <CaptHindsight> Constant Jerk Trajectory Generator
[15:12:02] <CaptHindsight> a novel approach
[15:12:08] <sensille> well, at least that thesis confirms that my approach is not completely crazy
[15:12:12] <CaptHindsight> every move is a jerk so the jerks are all moves
[15:12:41] <CaptHindsight> can't move the mountain so
[15:18:10] <pcw_mesa> You can use infinite order (sine)
[15:19:07] <sensille> sadly impossible to calculate
[15:19:12] <sensille> in finite time
[15:19:33] -!- phiscribe_ has quit [Quit: nick....nick.....Nick....SWAMP!]
[15:19:58] -!- phiscribe has joined #linuxcnc
[15:20:25] <sensille> but he executed the polynomial completely differently. well, interesting link
[15:21:13] <CaptHindsight> anything else I can Google for you? :)
[15:21:57] <pcw_mesa> Pretty sure GRBL is using sine interpolation now
[15:22:14] <sensille> probably in 5th order approximation
[15:22:40] -!- rhamby has joined #linuxcnc
[15:22:50] <pcw_mesa> or one or those u-controller motion widgits
[15:23:08] <pcw_mesa> they make a big deal about infinite order
[15:23:21] <CaptHindsight> it's always something
[15:23:42] <sensille> can't do infinite order
[15:23:57] <sensille> limited jerk is enough
[15:23:58] <pcw_mesa> Sine interpolation has been done for many years in motion control
[15:24:07] <CaptHindsight> throw some non-linear in there
[15:24:14] -!- rhamby has quit [Killed (Sigyn (Spam is off topic on freenode.))]
[15:24:28] <sensille> pcw_mesa: you can't calculate sin exactly
[15:24:45] <CaptHindsight> or irrational numbers
[15:24:45] <sensille> normally you approximate it with polynomials to the desired accuracy
[15:25:19] <pcw_mesa> not if you have a real CPU with FP trig
[15:25:38] <sensille> how so?
[15:27:00] <pcw_mesa> google sine motion interpolation
[15:28:14] <sensille> i think CaptHindsight has to do this for me
[15:28:38] <CaptHindsight> ok ok one sec
[15:29:00] <CaptHindsight> busy reading https://www.nature.com
[15:31:11] <sensille> meanwhile i calculate some taylor polynoms to infinity
[15:33:35] -!- Blumax has quit [Quit: Leaving]
[15:34:30] * gloops continues to calibrate the diy tachyon generator
[15:34:51] <djdelorie> once you get it working, it will be done yesterday?
[15:35:17] -!- emsjessec has quit [Read error: Connection reset by peer]
[15:35:21] <gloops> very quick there djdelorie
[15:35:36] <djdelorie> yeah, well I had since tomorrow to come up with it
[15:39:57] -!- Tecan has joined #linuxcnc
[15:39:57] -!- Tecan has quit [Changing host]
[15:39:57] -!- Tecan has joined #linuxcnc
[15:41:51] -!- fragalot has quit [Quit: Lost terminal]
[15:42:58] <jesseg> hey are these power supplies any good? lol https://www.acopian.com
[15:43:26] <jesseg> they seem pretty proud of themselves
[15:43:46] <JT-Shop> that must be for some high end thing
[15:43:53] <gloops> they want to be for $300
[15:43:57] <JT-Shop> are you looking for stepper drive power supply?
[15:44:02] <djdelorie> I clicked through to the product overview and it says "3-day power supply". I'd believe that.
[15:44:39] <JT-Shop> oh is that a battery power supply?
[15:45:07] <jesseg> no it's just an aerospace quality power supply that can be configured and delivered in 3 days
[15:45:09] <CaptHindsight> Acopian has been around for decades
[15:45:38] <jesseg> I'd never heard of them, but a customer wanted a good quality power supply and said he didn't care what it cost so I googled made in USA power supply
[15:46:04] <CaptHindsight> good stuff
[15:46:11] <jesseg> JT-Shop, nah it's not for a stepper drive.. it's umm. ok laugh now then I'll tell you what it's for :P
[15:47:13] <jesseg> This guy bought this really expensive 2.5 story house, which has an automated lighting system. All the lights are controlled from dimmer boxes in the basement, and instead of light switches, there are RS485 connected touch panels.
[15:47:52] <jesseg> so the 38v 1.5a power supply that runs the logic board died during a thunder storm, so none of the lights in the house work.
[15:48:30] <jesseg> so rather than putting in another cheap chinese power brick, we're going for something hopefully that will last better... and put a surge arrestor in line..
[15:50:18] <CaptHindsight> 120VAC 60Hz input?
[15:50:59] <CaptHindsight> closed? and special form factor?
[15:51:04] <CaptHindsight> and / any
[15:51:34] <gregcnc> yeah for progress
[15:51:35] <XXCoder> must be seperate from this plane of reality
[15:54:52] <CaptHindsight> jesseg: what did they quote you?
[15:57:30] <jesseg> CaptHindsight, I don't understand your questions.
[15:58:10] <CaptHindsight> jesseg: did you get a price from Acopian?
[15:58:29] <jesseg> The power supply is supposed to be PFC 90-265VAC 49-420hz or 110-350DC,
[15:58:45] <jesseg> Oh, I did not ask for a discount. They list $300 on their website.
[15:59:49] <djdelorie> you could just buy three cheap ones and put two on the shelf for later...
[15:59:57] <jesseg> LOL
[16:00:26] <jesseg> customer doesnt care about price but hates hassle
[16:00:42] <djdelorie> buy three cheap ones and diode-OR them together? ;-)
[16:00:49] <jesseg> lol
[16:00:57] <CaptHindsight> should be fine, they don't make junk
[16:01:06] <jesseg> good thanks
[16:07:41] <JT-Shop> jesseg: take a look at Antek power supplies (not Antec) for stepper drives
[16:11:06] <CaptHindsight> Acopian reminds me of another brand that looks like they are gone, they made modular enclosed/sealed bricks into the KW's
[16:11:23] <CaptHindsight> were all black anodized
[16:11:30] <CaptHindsight> low profile
[16:13:04] -!- jesseg has quit [Ping timeout: 240 seconds]
[16:13:18] -!- jesseg has joined #linuxcnc
[16:17:20] <jesseg> JT-Shop, thanks!
[16:17:39] <CaptHindsight> jesseg: just as good http://www.vicorpower.com as Acopian
[16:18:11] <jesseg> cool
[16:18:15] <jesseg> but I have a secret to tell you
[16:18:29] <jesseg> I hate steppers :P I'm a closed loop / servo nut
[16:18:39] <CaptHindsight> me too!
[16:18:59] <jesseg> ok great but these would make fine servo power supplies too :D
[16:19:28] <CaptHindsight> my servo drives input AC
[16:19:35] -!- blkperl4 has joined #linuxcnc
[16:20:14] <jesseg> cool
[16:21:16] <JT-Shop> you can get Antek power supplies with both 5vdc and 24vdc regulated outputs for logic and field power
[16:21:43] <CaptHindsight> Vicor prices are even higher than Acopian
[16:24:00] -!- blkperl4 has quit [Remote host closed the connection]
[16:26:50] -!- syyl has joined #linuxcnc
[16:26:57] <Rab> Mean Well, The Brand You're Willing To Pay For!
[16:27:22] <XXCoder> there should be company called "customer doesnt care about price"
[16:27:53] <CaptHindsight> Mean Well vs Built Well
[16:28:19] <CaptHindsight> Good Intentions -------
[16:28:33] <Rab> XXCoder, TDK-Lambda
[16:29:07] <CaptHindsight> you often get what you pay for
[16:29:48] <Rab> I have found Mean Well to be generally reliable, against all my expectations.
[16:30:42] <Rab> But I don't know that I would put thousands of dollars of equipment behind one.
[16:30:43] <CaptHindsight> when derated
[16:31:07] <Rab> Sure, usually 50%. I apply that to all switching supplies though.
[16:35:12] -!- Roguish has quit [Remote host closed the connection]
[16:38:15] <jesseg> I was running a Power-One 300A 5V power supply at full current at around 3 volts.. and it was pushing 295 amps according to my cheap amp clamp
[16:38:59] <jesseg> its actually a dual 150A output with paralleling bars
[16:40:30] * Loetmichel once ran 3 5V50A PSUs on the same bus for a 640*32 "pixel" text display with 3mm green leds as pixels
[16:41:07] -!- phiscribe_ has joined #linuxcnc
[16:41:08] <jesseg> that sounds like fun Loetmichel
[16:41:16] <Loetmichel> and yes, i had to derate the pixel current in software wen more than 50% of the pixels were lit or they would have switched off
[16:41:24] <Loetmichel> soldering that beast was "fun"
[16:41:39] <jesseg> that's like over 20k leds?
[16:41:52] <CaptHindsight> I once had to design a power supply for a cluster that had 480V 1,000A 3phase input
[16:41:52] <jesseg> so how many LEDs a second did you have to solder to get done in a lifetime :P
[16:42:07] <Loetmichel> 20k+ 3mm leds, 641 Atmega8, double that in shift registers, and a bunch resistors ;)
[16:42:21] <Loetmichel> it took a bit more than a month
[16:42:23] <Loetmichel> 8 hours a day
[16:42:28] <jesseg> wow... just wow...
[16:42:45] <Loetmichel> all THT i may add
[16:42:50] <Loetmichel> not SMT
[16:43:01] <jesseg> so was that for "fun" or did you receive some other solid form of gratitude?
[16:43:11] <Loetmichel> work
[16:43:16] <jesseg> gotcha
[16:43:23] <jesseg> that was one expensive reader board
[16:43:29] <jesseg> is it still in service today?
[16:43:48] <Loetmichel> customer wanted to have a text display thats visible and READABLE from the other wall of a 200ft long beer tent
[16:43:50] -!- phiscribe has quit [Ping timeout: 272 seconds]
[16:44:09] <Loetmichel> nope, it was for a single 2 day event
[16:44:30] <jesseg> haha.. wow... uhm maybe I should stop asking questions now
[16:44:51] <Loetmichel> for the "drivers tent" where all the "mister and misses rich" drivers were "parked" until the riches wanted to drive home
[16:44:59] <jesseg> Didn't they have another 2-day event a year later and use it again?
[16:45:27] <Loetmichel> nope, we actually kept it. i was working for a stage rental company at that time
[16:45:54] <Loetmichel> customer wanted it, costomer got it, paid for it full and we had the display to use it in some more events later
[16:46:07] <jesseg> well it sounds like quite a project all in all
[16:46:23] <jesseg> how fast could you update it?
[16:46:31] <jesseg> could you show video on it?
[16:47:13] <Loetmichel> it could actually take that whole 200ft by 80 ft tent and paint it bright green at night ;)
[16:47:24] <Loetmichel> nope, no video
[16:47:32] <Loetmichel> just scrolling text
[16:47:55] <Loetmichel> i used 19k2 serial daisi chain link between the AVRs
[16:48:22] <Loetmichel> so about a second for a full update
[16:49:47] <Loetmichel> (recalculating: it was more like a 160 ft by 70ft tent, 20 by 50 meters "oktoberfest beer tent")
[16:51:51] -!- Roguish has joined #linuxcnc
[16:59:25] -!- phiscribe has joined #linuxcnc
[17:01:47] -!- phiscribe_ has quit [Ping timeout: 240 seconds]
[17:03:16] <jesseg> Loetmichel, pretty cool alright! and yes, 150W of LED light is going to be pretty bright
[17:04:18] <Loetmichel> jesseg: it was high brightness 3mm green leds
[17:04:45] <Loetmichel> the kind you have orange dots in your retina for a while if you look at them at 20mA
[17:04:54] <Loetmichel> just 20k of them ;)
[17:05:57] <Loetmichel> its a LOT brighter than 150W blue with yellow converter paint (a.k.a white) leds with just 100 led chips on a single module
[17:06:17] <jesseg> yeah
[17:06:39] -!- phiscribe has quit [Ping timeout: 252 seconds]
[17:06:44] <Loetmichel> its like mercury lamps compared to natrium lamps of same wattage
[17:07:32] <Loetmichel> and it was this sick green thats almost cyan
[17:08:00] <Loetmichel> (fitting the colour of the events sponsor logo) ;)
[17:08:58] <jesseg> LOL
[17:09:08] <jesseg> I like a more natural green myself
[17:09:12] <Loetmichel> The event was the marriage of a big german beer company heiress... an audiance like the oscars
[17:09:13] <Loetmichel> :-)
[17:09:34] <jesseg> oh that's why it was a one time event
[17:09:39] <Loetmichel> indeed
[17:09:49] <jesseg> I thought maybe it was an annual beer fest
[17:10:12] -!- phiscribe has joined #linuxcnc
[17:10:25] <jesseg> but yeah most people don't want to get married every year
[17:10:52] <Loetmichel> its fun though when you try to get a signal thru 300m of still working bottling plant on a xlr mic wire.
[17:11:17] Jin^eLD is now known as Jin|away
[17:11:25] <jesseg> lol
[17:11:33] <Loetmichel> had to pump in 50VSS at one end from a 2kW amplifier and set a DI box on the other end to get the noise down to acceptable levels ;)
[17:12:52] <Deejay> gn8
[17:12:53] <Loetmichel> you know: if you cant get rid of the noise: up the SNR until it doesent matter :-)
[17:13:09] <jesseg> yeah if you can't bring the noise down, bring the signal up :P
[17:16:26] -!- Deejay has quit [Quit: bye]
[17:17:32] <Loetmichel> it was a fun jon for 10 years you get old quick in stage rental though
[17:17:42] <Loetmichel> job
[17:18:30] <Loetmichel> not 2 days the same cathastropes. and you got to see the world... ever built a stage in the middle of the maroccan desert?
[17:18:40] <Loetmichel> :-)
[17:24:27] <jesseg> what film was that for? lol or are you under strict NDAs
[17:25:06] <jesseg> or was it a life performance of the Thief of Baghdad ?
[17:26:05] -!- ziper1 has joined #linuxcnc
[17:27:44] -!- ziper has quit [Ping timeout: 250 seconds]
[17:31:52] -!- infornography has joined #linuxcnc
[17:44:40] -!- andypugh has joined #linuxcnc
[17:55:51] -!- FinboySlick has quit [Quit: Leaving.]
[17:57:13] -!- infornography has quit [Remote host closed the connection]
[18:07:33] -!- infornography has joined #linuxcnc
[18:13:17] <infornography> what are we building today?
[18:15:35] <gloops> looks like Ichsguckslives youtube channel has gone
[18:16:01] <gloops> nothing atm infonography
[18:19:11] <gloops> next build is a primitive kind of 4 axis for wood
[18:19:26] <_unreal_> This is going to drive me crazy.....
[18:19:32] <gloops> thing is, its no challenge now, i know its going to work
[18:19:51] <_unreal_> I had a few PCI expansion slot parallel port adapters....... I have no clue where any of them are.
[18:20:22] <gloops> yes madness usually sets in around this stage unreal
[18:20:28] <_unreal_> I just need one for the PARALLEL port shape so I can mill it out in the lid of my cnc control box.
[18:20:40] <_unreal_> Cant find a single one
[18:20:47] <infornography> I can identify with this
[18:21:23] <infornography> Put it where I can find it and now can't find it
[18:21:35] <_unreal_> Soon as I find one. I'll just trace it out and take it to my drill press with the stupid harbor freight XY table and mill it out
[18:22:42] <infornography> I got two of these from old hardware store https://www.amazon.com
[18:22:45] <_unreal_> Also discovered my drill press Dept setting collar is made out of plastic and cracked AND its only designed to stop the press in one direction
[18:22:58] <infornography> but only two bins
[18:24:29] <infornography> you gonna make another stop then?
[18:24:54] <_unreal_> when I finish my new cnc machine YES
[18:24:58] <_unreal_> I'll mill a new one
[18:25:03] <_unreal_> aluminum
[18:28:39] <CaptHindsight> where is a cheap supplier in the USA for machinable wax?
[18:30:11] <CaptHindsight> https://www.machinablewax.com
[18:30:27] <andypugh> _unreal_: https://pdl.designspark.com
[18:31:17] <andypugh> CaptHindsight: https://www.instructables.com
[18:32:19] <CaptHindsight> andypugh: it's not easy to find paraffin these days
[18:32:28] <_unreal_> andypugh, thanks but I'm not going to redraw it...
[18:32:35] <_unreal_> I'll come across it sooner or later
[18:32:44] <CaptHindsight> it's has some sort of magic qualities for hippies
[18:33:26] <_unreal_> CaptHindsight, paraffin? what are you talking about. its whats used in out door lamps
[18:33:39] <CaptHindsight> https://www.amazon.com
[18:33:56] <CaptHindsight> _unreal_: no those use LED's :)
[18:34:15] <_unreal_> What?
[18:34:21] <_unreal_> https://www.walmart.com
[18:34:26] <CaptHindsight> are you from the 50's?
[18:35:17] <gloops> get casting wax
[18:35:24] <CaptHindsight> yeah maybe a try a walmart in Hootersville
[18:36:24] <CaptHindsight> https://www.machinablewax.com
[18:37:53] <hazzy-m> around here you can get paraffin just about anywhere, but its mostly sold for canning
[18:37:55] <CaptHindsight> I was in a Walmart in a city. They laughed at me when I asked where the fabric aisle was
[18:38:48] <gloops> or..go round the flea markets and buy all the big candles
[18:39:02] <CaptHindsight> by tomorrow?
[18:39:31] <gloops> they always have those big 4 inch diameter candles that nobody really wants
[18:39:31] <CaptHindsight> damn maybe England is a much better place
[18:40:09] <CaptHindsight> Fish and Chips, Scotland, Ireland.... hey your on to something
[18:40:14] <CaptHindsight> you're
[18:40:26] -!- asdfasd has parted #linuxcnc
[18:40:35] -!- asdfasd1 has joined #linuxcnc
[18:42:19] <CaptHindsight> andypugh: you missed the discussion earlier about measuring and reducing jerk
[18:46:54] <_unreal_> Strange turn of events
[18:48:10] <_unreal_> So I ordered the right sized fan grill PC fan.... only to discover it was to large??? I ordered the worng size FAN which was the right size. only to discover it was the right size for the hole..... now I need to order a NEW grill that is 70mm not 90mm :)
[18:48:15] <_unreal_> how the hell I did that I'm not sure
[18:54:08] -!- infornography has quit [Remote host closed the connection]
[18:54:33] -!- infornography has joined #linuxcnc
[18:55:03] -!- jerryq has quit [Ping timeout: 245 seconds]
[19:05:31] <CaptHindsight> Quantum physics is in an undefined state between valid and invalid and we must wait for a cat to resolve the state?
[19:05:46] <CaptHindsight> Of course it's right. Clearly we can't rely on a physicist is resolve it, and cats are well known for having a definite opinion on everything.
[19:06:45] <_unreal_> arg....
[19:13:07] -!- asdfasd1 has quit [Ping timeout: 240 seconds]
[19:14:45] -!- gloops has quit [Quit: Leaving]
[19:17:10] -!- HSD has quit [Quit: Leaving]
[19:24:24] -!- merkatorix has quit [Ping timeout: 252 seconds]
[19:29:20] -!- ziper1 has quit [Quit: Leaving]
[19:29:40] -!- ziper has joined #linuxcnc
[19:34:57] -!- jerryq has joined #linuxcnc
[19:38:40] -!- infornography has quit [Quit: Leaving]
[19:44:05] -!- lupine2 has joined #linuxcnc
[19:45:49] -!- lupine2 has quit [Remote host closed the connection]
[19:48:09] -!- njh has quit [Ping timeout: 252 seconds]
[19:48:42] -!- Ross has quit [Ping timeout: 252 seconds]
[19:49:49] -!- Timbo has quit [Ping timeout: 246 seconds]
[19:50:00] -!- MattyMatt has quit [Ping timeout: 252 seconds]
[19:51:55] -!- MattyMatt has joined #linuxcnc
[19:52:28] -!- njh has joined #linuxcnc
[19:56:59] -!- Ross has joined #linuxcnc
[19:58:10] -!- Timbo has joined #linuxcnc
[20:20:34] -!- rob_h has quit [Ping timeout: 240 seconds]
[20:40:06] -!- m4rcu525 has joined #linuxcnc
[20:41:05] -!- m4rcu525 has quit [Remote host closed the connection]
[20:41:58] -!- rmu has quit [Ping timeout: 272 seconds]
[20:48:15] -!- rmu has joined #linuxcnc
[20:59:13] -!- Timbo has quit [Ping timeout: 245 seconds]
[20:59:39] -!- Ross has quit [Ping timeout: 252 seconds]
[21:00:01] -!- njh has quit [Ping timeout: 260 seconds]
[21:00:20] -!- MattyMatt has quit [Ping timeout: 272 seconds]
[21:00:40] -!- kengu has quit [*.net *.split]
[21:00:40] -!- connor_goodwolf has quit [*.net *.split]
[21:00:40] -!- TekniQue has quit [*.net *.split]
[21:00:49] -!- connor_goodwolf has joined #linuxcnc
[21:01:15] -!- kengu has joined #linuxcnc
[21:03:07] -!- MattyMatt has joined #linuxcnc
[21:04:50] -!- njh has joined #linuxcnc
[21:05:55] -!- TekniQue has joined #linuxcnc
[21:06:11] -!- Ross has joined #linuxcnc
[21:06:31] -!- Timbo has joined #linuxcnc
[21:13:05] -!- MrHindsight has joined #linuxcnc
[21:14:14] -!- ve7it has quit [Remote host closed the connection]
[21:16:37] -!- CaptHindsight has quit [Ping timeout: 246 seconds]
[21:16:58] MrHindsight is now known as CaptHindsight
[21:32:31] -!- diverdude has quit [*.net *.split]
[21:35:22] -!- Test has joined #linuxcnc
[21:35:41] -!- Test has quit [Client Quit]
[21:45:32] -!- JMoVS3 has joined #linuxcnc
[21:49:09] -!- njh has quit [Ping timeout: 252 seconds]
[21:49:36] -!- Timbo has quit [Ping timeout: 260 seconds]
[21:49:41] -!- JMoVS3 has quit [Remote host closed the connection]
[21:49:42] -!- MattyMatt has quit [Ping timeout: 252 seconds]
[21:49:43] -!- Timbo has joined #linuxcnc
[21:49:51] -!- nallar has joined #linuxcnc
[21:50:06] -!- MattyMatt has joined #linuxcnc
[21:50:11] -!- Ross has quit [Ping timeout: 260 seconds]
[21:50:11] nallar is now known as Ross
[21:51:23] -!- njh has joined #linuxcnc
[22:14:04] -!- Akex_ has quit [Ping timeout: 240 seconds]
[22:15:43] -!- Akex_ has joined #linuxcnc
[22:45:27] -!- hazzy has quit [Ping timeout: 252 seconds]
[22:52:53] -!- hazzy has joined #linuxcnc
[23:00:07] -!- syyl has quit [Quit: Leaving]
[23:23:34] -!- pcw_home has joined #linuxcnc
[23:33:13] -!- Roguish has quit [Quit: ChatZilla 0.9.92-rdmsoft [XULRunner 35.0.1/20150122214805]]