#linuxcnc Logs

Apr 23 2021

12:04 AM enleth: veegee: but this statement might very well be provably true
12:11 AM unreal is now known as Guest5925
12:11 AM enleth: veegee: you don't have to do this to get perfectly acceptable results, but you will not get the *maximum* performance otherwise
12:11 AM enleth: question is, how much difference does this make and is the difference meaningful to you
12:58 AM miss0r: morning
01:13 AM Deejay: moin
04:09 AM Tom_L: morning
04:10 AM XXCoder: yoooo
05:05 AM -!- #linuxcnc mode set to +v by ChanServ
05:42 AM Loetmichel: *snicker* Good deeds are punished immedeatly. Coworker just asked how Electromagnetic interference works. Cue me being the "explain bear"... half an hour later: "I think i understand. Now whats the difference between the special / general theory of relativity?" $me: [sigh!]
07:15 AM veegee: enleth yeah for sure. Changing your oil after every drive is _technically_ better for your engine
07:15 AM veegee: Loetmichel cool coworker who genuinely wants to learn?
07:16 AM veegee: Or one of those people who wants to know _just enough_ to pretend like they know what they're talking about?
07:17 AM veegee: enleth but yeah I did neglect the oil change on the cheap vacuum pump I have and there definitely is a lot of water in there. I'm trying to purify it now in the vacuum chamber. Advantage of having two pumps I guess
07:17 AM veegee: Also can use one of them to suck the oil out of the other's drain port to get that last little bit out
07:24 AM JavaBean: can you use either of those vacuum pumps to do a fluids change on your car? example being i am supposed to bring my radiator to "vacuum" to fill it
07:25 AM unterhaus: as cheap as radiators are, I wonder about the wisdom of having it under vacuum
07:28 AM JavaBean: i have had 2 cars now where "the book" says to do it. forget wisdom, i am going to let a fluids place do it. that way if they fsck it its on them
08:23 AM veegee: JavaBean easily
08:23 AM veegee: JavaBean even the shitty sum-ting-wong single stage pump pulls a -0.995 bar vacuum
08:24 AM veegee: unterhaus absolutely zero risk of having a radiator under vacuum pressure
08:25 AM veegee: unterhaus those things see pressures above 2 bar
08:25 AM unterhaus: GM requires a vacuum right?
08:25 AM veegee: But I've never heard needing to vacuum purge a radiator
08:26 AM unterhaus: vacuum means that the pressure is going the other way, but you are probably right about the risk
08:26 AM veegee: yes I know
08:26 AM veegee: if a rigid pipe can handle 2 bar, it can easily handle -1 bar
08:26 AM veegee: 1" schedule 10 pipe is rated for 20 bar working pressure
08:26 AM unterhaus: an aluminum balloon can take positive pressure but put it in a vacuum and it's flattened
08:27 AM veegee: and schedule 10 is so thin that you can't thread it. You have to use special fittings that clamp on
08:27 AM veegee: yeah but a radiator is made from standard metal pipe and its characteristics are well understood
08:27 AM unterhaus: not any car radiator I have seen, it's made from aluminum foil
08:28 AM veegee: ...
08:28 AM veegee: those are the fins
08:28 AM veegee: the pipe itself is quite strong
08:28 AM veegee: the fins aren't subject to any static pressure
08:28 AM unterhaus: you have been working on too many forklifts
08:29 AM veegee: their radiators are the same
08:29 AM veegee: also I just finished changing coolant on my 2007 jeep wrangler unlimited
08:29 AM veegee: and my R nineT motorcycle
08:29 AM veegee: sorry oil on the motorcycle, not coolant
08:30 AM veegee: dumb thing is air cooled
08:30 AM JavaBean: radiators are aircooled
08:30 AM veegee: JavaBean you have successfully attained enlightenment
08:31 AM JavaBean: ohm
08:34 AM veegee: https://www.youtube.com/watch?v=A65613TaWYw
08:34 AM veegee: See the "pipes"?
08:34 AM veegee: they're weaker than a regular cylindrical pipe, but -1 bar is no problem
08:43 AM Loetmichel: veegee: i think he was jokingly telling me that i am a "know it all"
08:54 AM veegee: One of the most annoying things about people I try and teach is their own insecurity
09:49 AM veegee: Going to buy an electric oven today. Seeing lots of good ones for $250
09:49 AM veegee: For the purpose of powder coating and heat treating
09:56 AM veegee: "Each number is the maximum stress that the chain can take per millimeter squared. So, for instance, Grade 30 chain’s maximum stress is 300 newtons per millimeter squared" DID NOT KNOW THAT!
10:46 AM unterhaus: Loetmichel, he should have asked you to explain magnets.
10:47 AM unterhaus: but then you could answer like Feynman. He told a reporter that the reporter didn't know enough for him to answer the question
10:47 AM unterhaus: although, "you don't know enough for me to answer your question" is a good answer
11:05 AM Loetmichel: its a copout though
11:06 AM Loetmichel: the right answer would be "depending on the knowledge you already have it will take anywhere from an hour to a lifetime to explain that to you."
11:06 AM Loetmichel: @unterhaus
11:11 AM unterhaus: well, Feynman went on at length about how the reporter lacked the background to understand it
11:12 AM unterhaus: I will admit I don't have the background to fully understand magnetism, but I could get there in a week, most likely
11:29 AM Loetmichel: yeah, my point was: without assesing the background its a reasonable assumption that he wouldnt understand but STILL a copout.
11:36 AM enleth: Loetmichel: I'm pretty sure he was, at that point, almost qualified to follow this up with "I personally know every person who knows enough and you're not one of them"
11:40 AM Loetmichel: possible
11:41 AM enleth: well, ok, that was 1980s, so "personally" was definitely no longer true
11:41 AM enleth: but there was a point in his career where it was, and that number started at zero
11:56 AM Loetmichel: well, i have heard of the feyman reporter incident
11:57 AM Loetmichel: i still think it would have been funny if the reporter would have said "try me" and would have show him a hole in his theory.
11:57 AM Loetmichel: thats just me being wary of people that think they are "high and mighty" though.
11:57 AM Loetmichel: @ enleth
12:31 PM -!- #linuxcnc mode set to +v by ChanServ
12:43 PM gigasu_shida is now known as gigas_cedar
02:08 PM htasta: unterhaus: that's kinda misrepresenting what Feynman did. He said the reporter has to tell him on what level he wants to know how magnets work
02:10 PM htasta: Loetmichel: he wasnt "high and mighty" he merely showed how it's important to know what you take as a given in order to answer a question, it's a really good interview imo and you should watch it if you find the time, it's on youtube
02:19 PM Connor: Labeling stuff in my panel. I have 48v+ and GND, but, that's not a common ground (because the 5v has one too) How should hat be denoted? 48v- seems wrong, because that implies NEGATE 48v ??
02:22 PM Rab: '48V return' too long?
02:25 PM Connor: Yea. Heat Shrink Label maker..
02:25 PM Connor: DC48vGND too long too.
02:26 PM Rab: 48V+ 48Vg ?
02:26 PM Rab: Or 48v+ 48vG
02:27 PM Connor: I thought a guide existed for this stuff
03:44 PM veegee: Connor then that's not GND
03:45 PM veegee: Connor ground should always mean "the most common rail"
03:45 PM veegee: But if you must, you should clearly mark which circuits are isolated
03:46 PM veegee: If your 5V circuit and 48V circuit are separate, denote them as circuit A or B or whatever, but name them consistently
03:47 PM veegee: But don't label it GND if it's not actually connected to the earth, especially on machine tools
03:47 PM veegee: that's a disaster waiting to happen
03:50 PM veegee: It's super important. In LTSPICE, if you name lines, every line with the same name will be connected. That is by design.
04:27 PM JT-Cave: I label all DC grounds 0vdc as they should all be tied together
04:58 PM -!- #linuxcnc mode set to +v by ChanServ
05:05 PM roycroft: yes, what which dc terminal is ground? positive or negative?
05:08 PM Connor: JT-Cave: I'm not inclined to tie stepper PSU to the logic CPU...
05:14 PM JT-Shop: I guess you have some reason to have floating grounds... enjoy
05:17 PM Guest5925 is now known as _unreal_
05:17 PM Connor: Well, the stepper drivers inputs are opto-isolated. Why would you tie the GND for the stepper psu to the gnd of the logic? Does that not defeat the purpose of opto-isolation?
05:20 PM unterhaus: I don't tie together grounds unless there is a good reason
05:20 PM JT-Shop: floating grounds can build up high voltages from static
05:21 PM unterhaus: unfortunately in my system there is a good reason to tie most of them together and I don't really like it
05:21 PM JT-Shop: and other stuff, not a happy place if the 0vdc can be clean
05:23 PM unterhaus: I think I would mark it 48vRet
05:24 PM unterhaus: but you can have all the grounds tied together back at the star ground and still need optos
05:40 PM Tom_L: all GNDs should be tied together in a star configuration
05:47 PM JT-Shop: aye, and I always do that when building machines so I don't have to try and find ghosts later on
05:47 PM JT-Shop: another tid bit from a machine builder never ground both ends of a shielded cable!
05:47 PM andypugh: Hmmm… It can be a good idea to separate power and logic GND, but it is often not possible. If a single component in the system ties them together internally, tgen you have to tie them together actively.
05:48 PM andypugh: Some stepper drivers do it. And you do not want the link between logic and power GND to be a single PCB trace inside a driver.
05:57 PM Tom_L: now i just need a nice size aluminum project to test this cutter on
05:58 PM XXCoder: :)
05:58 PM Tom_L: i did add it to my cad and lcnc tool list
05:59 PM Tom_L: with moderate feeds to start with
06:59 PM unterhaus: interestingly the manual for my servo drives shows both ends of the shield grounded on the encoder cable
07:00 PM unterhaus: I think that might be redundant, but it's hard to measure through the coatings on the motor end
07:09 PM -!- #linuxcnc mode set to +v by ChanServ
07:11 PM andypugh: Possibly the encoder uses the shield as ground? Does it connect to motor ground? I can see an argument for continuing the shield ground to the ground plane of the encoder PCB. But I suspect that you would not want that to be connected to the motor frame.
07:21 PM buzzmarshall: shield wire in most sheilded cable is to help reject emf and other noise and to do so should only be attached at one end
07:21 PM buzzmarshall: sometimes with single or only a couple of wires it dont really matter but on snake type cable your only asking for random noise issues
07:23 PM buzzmarshall: the shield i the wire isn't meant to conduct any current or vcc... just bleed off noise so it should normally only be connected at one end of the cable
08:07 PM andypugh: What I am saying is that it might make sense to extend the shield _inside_ the encoder.
08:08 PM andypugh: So, it’s not a ground so much as extending the shield into a plane on the PCB,
09:03 PM -!- #linuxcnc mode set to +v by ChanServ
10:47 PM Tom_itx: https://www.lakeshorecarbide.com/videos.aspx
10:47 PM Tom_itx: good videos on feed/speed/chipload etc
10:48 PM -!- #linuxcnc mode set to +v by ChanServ
10:54 PM XXCoder: saved Tom_L