Back
[00:01:29] <Jymmm> Maybe a bench grinder
[00:08:49] <RyanS> It would probably shift the garage from its foundations when you start .it would need a really soft soft starter
[00:18:18] <Jymmm> The "62 Ohm" resistor gets hot... GO FIGURE!!!!!!!!!!!!!!!!!!!
https://www.youtube.com/watch?v=tyvZGIlHEok
[00:28:29] <MacGalempsy> when running the PID test, the ferror has a baseline of .0004 before the axis even moves. is the total error the produced error minus the baseline?
[01:53:26] <archivist> MacGalempsy, dont think so, the start error is just that an error in position, (but within ones error limits)
[02:21:52] <_DJ_> moin
[02:30:26] <archivist> MacGalempsy, for error limits you can also think dead band
[03:27:49] <Vinci89> hello there
[03:28:32] <Vinci89> I'm still in need of an activation e-mail... I've registered to the forum two days ago but still haven't received anything
[03:29:00] <Vinci89> I'd really appreciate it if someone could look into this
[03:29:49] <archivist> hang around in here and poke jthornton when he is awake (USA time)
[03:30:58] <Vinci89> alright, in the meantime... anyone having experience on getting PCI to parallel port cards run on the latest version of linuxcnc?
[03:31:51] <Vinci89> i bought some crappy adapter board from ebay china but simply cant get it to work
[03:32:31] <Vinci89> mainly because the "drivers" they delivered are god damn .c files with a readme basically saying "build the fuck yourself"
[03:32:48] <archivist> some need some coaxing and probing to find the address they are at
[03:33:18] <archivist> drivers... never needed that for parallel
[03:34:12] <Vinci89> yeah i'm not really sure on that either, some threads i found on stackoverflow suggest that parallel thing is built in as simple IO, but I'm not really an expert on that...
[03:34:46] <Vinci89> checking the addresses with "cat /proc/ioports | grep parport" I got 2x ports showing... so I thought it should already work
[03:35:10] <Vinci89> but if I try to use the second address (apart from x378) shown I get weird errors from linuxcnc
[03:35:30] <Vinci89> and it won't even let me start the gui, let alone trying out the config
[03:35:34] <archivist> does lspci say something sensible
[03:36:09] <Vinci89> sensible like?
[03:36:39] <Vinci89> lspci doesn't even show me the parallel port on the mainboard
[03:36:41] <archivist> it will tell you about devices it finds on the pci bus and the addresses
[03:37:02] <archivist> man lspci for the options
[03:37:33] <Vinci89> ok, let me check, brb
[03:48:05] <Vinci89> alright
[03:48:36] <Vinci89> I just dumped the whole output
[03:48:41] <Vinci89> it lists a pci bridge
[03:49:25] <Vinci89> but thats it, no information on any cards plugged in or whatsoever
[03:56:21] <archivist> did you try lspci -vv
[04:02:35] <Vinci89> you mean lspci -v?
[04:02:36] <Vinci89> yeah i did
[04:02:39] <Vinci89> funny thing though
[04:02:49] <Vinci89> the parallel port got listed as "serial device"
[04:03:12] <Vinci89> i just tried the address listed at the serial device and it worked...
[04:03:24] <archivist> ew
[04:09:25] <Vinci89> thank god that worked... otherwise I would have been screwed
[04:09:58] <Vinci89> my stepper control relies on 5V signals and the port from the motherboard is only 3.3
[04:10:10] <archivist> feedback is faster in here than on a forum :)
[04:10:25] <Vinci89> yeah, thank you btw!
[04:10:44] <archivist> use a buffer chip to level translate
[04:11:27] <archivist> by the way the ports are basicly open collector you may get away with pull ups to 5v
[04:11:44] <archivist> I use a buffer chip though
[04:12:16] <Vinci89> probably next time... I built everything from scratch and basically out of the trash... I'm 1 1/2 years in on building the machine now and I simply want to use it right now
[04:13:30] <archivist> for buffer chip read 74 series whatever I can find or a uln200x
[04:17:07] <Vinci89> where can I configure the second parallel port? just in the .ini file?
[04:19:09] <archivist> once you have an ini most of use hand edit it to add stuff
[04:19:23] <archivist> and the hal file
[04:26:12] <Loetmichel> mornin'
[04:29:13] <Loetmichel> *yawn*... i managed to get out of in time bed this moring, just to fall asleep again in the bath tub :-( (45 min to late at work)
[06:00:40] <Vinci89> anyone can think of any reasons why from one to the next day suddenly the spindle won't run?
[06:00:54] <Meduza> you have tripped over the cord ;)
[06:02:48] <Vinci89> monday it was just working fine... now I switched to linuxcnc and nothing works, I already disconnected all the control signals from the frequency converter and tried to manually run it... but nothing
[06:03:19] <Vinci89> it just makes some stupid high pitch noise and wont start running
[06:03:53] <archivist> check the rest of the wiring :)
[06:14:28] <archivist> I have had VFDs die too
[06:21:09] <jthornton> if you disconnected the control signals then it can't have anything to do with LinuxCNC
[06:44:36] <Vinci89> of course it doens't have to do with linuxcnc, im just curious because I ran the same configuration just two days ago :(
[06:46:04] <archivist> wires and connections often fail when moving stuff around when upgrading/modifying things
[06:47:05] <Vinci89> it must be something else, I just tried to run a minimal configuration with just the spindle and a potentiometer connected... and nothing
[06:47:21] <Vinci89> ill try another motor after a well deserved cup of coffee....
[06:47:28] <archivist> the cable drive to motor
[06:47:54] <Vinci89> well, might be, but I hardly moved the motor around
[06:47:59] <Vinci89> since monday
[06:48:01] <archivist> drop a phase and it will just sit there and whistle
[06:48:28] <Vinci89> ok ill look into that
[06:55:49] <Vinci89> there it is
[06:55:51] <Vinci89> test mail
[06:55:54] <Vinci89> woops
[07:14:40] <MacGalempsy> morning all.
[07:24:59] <MacGalempsy> after 53 days, its finally friday! vacation here I come!
[07:29:23] <Vinci89> cable drive to the motor seems fine, other motor works as well
[07:29:49] <Vinci89> the fucking vfd works like a charm... looks like the spindle is done
[07:30:45] <Vinci89> which would suck because I just spent like 300€ on chucks
[07:32:19] <archivist> dismember and repair if possible
[07:32:54] <Jymmm> archivist: Dr Frankenstein I presume
[07:42:36] <Loetmichel> Jymmm: no, tahts me
[07:44:08] <Loetmichel> just building a 24" tft out of a pannel from a fujitsu, a case and stand from a HP 24", electronics LVDS to DVI/VGA from Pollin.de, a inverter for the 4 ccfls from ebay(china) and a 12V 5A smps i had laying around in the spare parts box
[07:44:13] <Loetmichel> :-)
[07:44:56] <Loetmichel> and a spindle is a niormal ac sync or async motor... mkore or less
[07:45:11] <Loetmichel> .. there can not be much defective parts.
[07:45:27] <Loetmichel> isolation failure in the windings is the one that makes most work but is the cheapest
[08:38:05] <kengu> hum. plasma is moving at 20 000mm/min. ok.
[08:42:40] <jdh> yeah, considering there is zero ambient light, ti's really good
[08:42:43] <jdh> <urk>
[09:47:23] <MacGalempsy> anyone got anything good going on?
[09:51:32] <gonzo_> the kettle is about to go on.
[09:51:42] <MacGalempsy> just rolling out of bed?
[09:58:23] <jmasseo> whats out there in the open source cam world?
[09:58:38] <jmasseo> or has some sort of discount for non profits/educational institutions?
[09:58:56] <archivist> stuff that often disappoints
[09:58:57] <JT_Shop> cam for what?
[09:59:02] <JT_Shop> it makes a difference
[09:59:46] <jdh> it does?
[09:59:56] <jdh> is there good OS CAM for anything?
[10:00:08] <archivist> define anything
[10:00:15] <jdh> no
[10:00:47] <archivist> each cam program has a niche
[10:01:03] <archivist> there is no allcando
[10:01:19] <jdh> is there any OS CAM that can do anything useful?
[10:01:29] <archivist> well 60k pounds I had quoted once
[10:01:31] <JT_Shop> cam for a plasma table is very different from mold making
[10:01:45] <cradek> "cam" can mean almost anything
[10:01:53] <JT_Shop> yep
[10:01:56] <jdh> sure, but when you limit it to Open Source, it gets limited.
[10:02:17] <archivist> a widget in ngcgui could be called cam
[10:02:32] <cradek> REALIZE is probably CAM and it works really great for what it does
[10:02:53] <JT_Shop> my fingers are CAM
[10:03:11] <archivist> this is very much use the right tool for the job area
[10:03:29] <archivist> I use inside rear of skull for my cam
[10:03:57] <archivist> and let the subroutines do the boring bit
[10:07:12] <JT_Shop> so jdh what is your application now that you have narrowed it down to free CAM
[10:09:45] <jdh> I don't, I was narrowing down the "Open Source CAM" and "good"
[10:10:45] <jdh> but, I was not considering ncgui type stuff. image2gcode, ttt, etc are all OS and quite workable, but perhaps not quite definition of CAM.
[10:10:59] <jdh> err... my definition.
[10:11:02] <cradek> I think everyone is asking what TASK are you trying to accomplish
[10:11:14] <CaptHindsight> jmasseo: what are you going to use the CAM for?
[10:12:11] * JT_Shop gets tired of asking the same question over and over
[10:12:15] <jdh> heh
[10:12:26] <archivist> jdh something like pycam maybe but some also moan about it
[10:12:35] <jmasseo> our makerspace may be getting a used haas vf2
[10:12:51] <jmasseo> obviously we don't want to put out $10,000 for mastercam
[10:12:52] <CaptHindsight> http://hackaday.com/2013/10/19/blender-cam-open-source-cam-software/ maybe this? :)
[10:12:54] <jdh> I could not coerce pycam into doing anything useful.
[10:13:06] <jmasseo> i'll look at that
[10:13:14] <jmasseo> we have blender training classes at our space so that's a real good fit if it works
[10:13:23] <jmasseo> we have plasmacam for our plasma
[10:13:28] <jmasseo> retinaengrave for our laser cutter
[10:13:33] <jmasseo> slicer for the 3d printers
[10:13:43] <jdh> I got good results with realize, but I dont' have a autocad license
[10:14:13] <jmasseo> i'm inclined to inquire to autocad about licensing
[10:14:19] <CaptHindsight> jmasseo: have you seen pycam?
http://pycam.sourceforge.net/
[10:14:20] <jmasseo> techshop members count as educational licenses
[10:14:27] <jmasseo> CaptHindsight: i was not impressed but i'm willing to give it a try
[10:14:45] <archivist> freecad is getting more popular slowly
[10:14:59] <jmasseo> anyway, if we want any % of our membership to be able to do anything with the machine, we'll probably have to have some sort of cam solution
[10:15:04] <jmasseo> some of our members are into freecad
[10:15:07] <CaptHindsight> some others here use some <1K CAM software for their routers IIRC
[10:15:08] <jmasseo> it crashed on me the one time i loaded it
[10:15:15] <jmasseo> i know a guy who loves artcam
[10:15:20] <jmasseo> i think its <$1k
[10:15:23] <CaptHindsight> don't recall the names of the apps
[10:15:58] <jdh> vectric stuff for windows is < $1k (for some of them)
[10:19:45] <CaptHindsight> jmasseo: nice addition
http://www.amtechmachine.com/images/Haas%20VF-2%20Vertical%20Machining%20Center.JPG is it this?
[10:35:26] <somenewguy> realize?
[10:36:02] <jdh> http://timeguy.com/cradek/autocad
[10:36:16] <somenewguy> thank you
[10:36:39] <somenewguy> I used freecad (cad, not cam) but it was very buggy although it did work
[10:37:06] <somenewguy> I suspect if you used it thru python and not the gui it would be mroe stable, but the biggest trick I learned is to save and re-start freecad often, and you would have much more stable boolean operations
[10:37:17] * JT_Shop gets a good chuckle out of the oxymoron "open source CAM" and "good" in the same sentence
[10:37:39] <somenewguy> I am in hte middle of a varicad trial, again still CAD not CAM, but it is actually working pretty well
[10:37:40] <jdh> JT: that was my point above.
[10:37:58] <somenewguy> parametric modeling is kinda strange, but I haven't hit a single software bug yet, and it runs in linux/windows/mac
[10:38:26] <somenewguy> and there are several 2.5 milling CAM software bits that are free/open source that can make tool paths from DXFs, so that is something
[10:45:57] <CaptHindsight> maybe you'll just have to wait for someone over at reprap to write a CAM app <ducks and runs>
[10:46:14] <archivist> spitcam
[10:47:02] <somenewguy> dumb question, but what is the real difference between custom m-codes and subroutines?
[10:47:29] <somenewguy> I was surpised when I discoverd you can only pass 2 parameters to a custom m-code, so I suspect maybe they aren't meant for what I thought
[10:47:47] <CaptHindsight> I do have to give them credit, they seem to be able to come up with useless stuff that many love to spend money and waste on on
[10:50:13] <archivist> I walked past a stand at a show last week and muttered gluegun at some wide eyed noobs looking at reprap on a stand :)
[10:50:43] <jmasseo> CaptHindsight: older. i think the 1995 is only 10hp
[10:52:05] <JT_Shop> and if I knew the application I might be able to recommend something...
[10:52:14] <jmasseo> there are reprap CAM products.
[10:52:17] <jmasseo> repetier host
[10:52:18] <jmasseo> etc
[10:52:23] <jmasseo> i don't know if they are open source or not.
[10:52:49] <jmasseo> You could port slicer to be a 3d milling engine.
[10:52:55] <jmasseo> i don't think i'd want to.
[10:53:10] <jmasseo> well, you could with a 4th axis maybe :)
[10:53:14] <CaptHindsight> repetier is open, it was a good attempt at a decent UI
[10:53:23] <jmasseo> hmm no.
[10:53:27] <jmasseo> because 3d printing is bottom up
[10:53:32] <jmasseo> milling is top down
[10:55:07] <archivist> if a standard xyz machine maybe
[10:56:48] <archivist> and I think not all 3d printing is bottom first either
[10:57:29] <CaptHindsight> jmasseo: no it's not open anymore, or no it wasn't a good UI?
[10:57:45] <jmasseo> oh i was talking to myself. :)
[10:57:53] <jmasseo> about using sl1cer for 3d milling toolpath generation
[10:59:33] <CaptHindsight> for most people 3d printing = bottom up glue guns since the term 3D printer was hijacked by the glue gunners
[10:59:56] <jmasseo> even the sls is bottom up isn't it?
[10:59:59] <jmasseo> only the form1 is top down
[11:00:07] <jmasseo> if it's even top down
[11:00:10] <jmasseo> and not just upside down :)
[11:00:51] <CaptHindsight> form1 is bottom up, they use a laser up through the bottom of a vat filled with resin
[11:02:26] <CaptHindsight> form1 is SLA, SLS melts powdered materials
[11:02:59] <jmasseo> right
[11:03:18] <jmasseo> there was a zcorp on sale for like $5k locally
[11:03:28] <somenewguy> most everything is bottom up, becasue gravity
[11:03:36] <somenewguy> I don't know of any that work in reverse
[11:03:38] <jmasseo> gravity is a dick!
[11:03:48] <CaptHindsight> zcorp, inkjet + binder or?
[11:03:51] <somenewguy> at least it doesn't discriminate too much
[11:04:06] <jmasseo> i don't remember which one
[11:04:07] <CaptHindsight> 3d systems bought them last year
[11:04:13] <jmasseo> it used some sort of powdered ABS or something?
[11:05:02] <archivist> inkjet and starch was one
[11:05:35] <archivist> sample was a ball bearing years ago
[11:14:45] <CaptHindsight> I think the MIT patent on the thermal inkjet with binder + powder expires soon
[11:15:50] <jdh> we got a Stratasys Fortus 400mc 3-D printer
[11:57:55] <IchGuckLive> hi all B)
[11:58:17] <IchGuckLive> Loetmichel: the gps and atmel works fine
[12:02:27] <JT_Shop> I still think archivist CAM solution is the only one that fits the "anything" requirement...
[12:02:39] <CaptHindsight> IchGuckLive: what CAM do you use for your routers?
[12:02:54] <IchGuckLive> CaptHindsight: lots of
[12:03:04] <IchGuckLive> XYZ goes on heeks
[12:03:10] <IchGuckLive> plasma on sheetcam
[12:03:24] <IchGuckLive> 5axis on proENc
[12:03:38] <IchGuckLive> foam on heeks
[12:03:47] <jdh> isn't that HAM, not CAM?
[12:04:10] <IchGuckLive> HAM is best for DX
[12:04:18] <CaptHindsight> IchGuckLive: there was a discussion earlier, you answers match what we discussed, it depends on what you are planning to do and on which machine
[12:04:24] <CaptHindsight> you/your
[12:04:28] <IchGuckLive> as i us RX TX its on UART jdh
[12:04:29] <archivist> HAM and Eggs
[12:04:44] <Tom_itx> Green Eggs and Ham
[12:04:55] <Tom_itx> but i am not sam
[12:05:07] <IchGuckLive> CaptHindsight: fully afree
[12:05:09] <CaptHindsight> Suess CAM
[12:06:05] <CaptHindsight> Green Eggs and CAM
[12:06:35] <JT_Shop> LOL
[12:06:46] <Tom_itx> just how far can this go..
[12:07:16] <IchGuckLive> <- Hungry
[12:07:34] <andypugh> jdh: PyCAM is OK, but I find it unbelievably slow for some processes.
[12:08:03] <andypugh> MeshCAM is not OS or Free and only runs on Windows, but seems pretty good and is relatively inexpensive at $250.
[12:08:08] <IchGuckLive> how far is awallin with the freecad cam module
[12:08:22] <CaptHindsight> zultron mentioned that the latest Pycam version was much faster
[12:09:15] <andypugh> I should look into that. (I did suggest some ways to speed things up, I don't know if they were useful suggestions, or got acted on)
[12:09:47] <archivist> step one, dont use an interpreted language
[12:09:52] <jdh> I haven't tried PyCAM in a while, and only in windows
[12:10:16] <jdh> cut2d is $150ish and very easy (and windows)
[12:10:59] <jdh> I'm trying to extrude an encoder wheel in Inventor(tm) with no luck.
[12:11:13] <CaptHindsight> NX CAM and Mastercam is what just about everybody we talk to uses, $10K +
[12:11:44] <Tom_itx> mastercam is used alot around here unless they're using Catia
[12:12:02] <jdh> doesn't Catia make everything else look cheap/free?
[12:12:04] <Tom_itx> most shops around here have gone to Catia
[12:12:09] <Tom_itx> prety much
[12:12:13] <Tom_itx> pretty
[12:12:28] <archivist> is APT hiding under catia
[12:12:29] <Tom_itx> my bud gave about 75k for a seat of it several years back
[12:12:59] <andypugh> jdh: I am a pro with Inventor. What's the problem?
[12:13:01] <Tom_itx> including the sun machine etc
[12:13:42] <andypugh> (I mean that literally, I have been paid to run an Inventor seat)
[12:13:56] <jdh> andypugh: mainly lack of any knowledge of what it expects. I inserted a 2d drawing of the encoder, but can't figure out what to select to extrude.
[12:14:03] <jdh> http://tinyurl.com/n4kjmyb
[12:14:28] <jdh> that took a few minutes to draw (in draftsight) and a few minutes to CAM in Cut-2D
[12:16:17] <Tom_itx> is that copper?
[12:16:24] <Quintox> hello everybody, can anyone give me the configuration of the TB6560?
[12:16:36] <jthornton> there is one on the forum
[12:16:43] <archivist> in solidworks draw disk extrude to thickness, draw slot, cut hole, make an array around axis
[12:17:03] <Quintox> where?
[12:17:06] <jthornton> yep
[12:17:11] <jthornton> the forum
[12:17:19] <Tom_itx> archivist, same with my cad cam
[12:17:22] <jdh> I was hoping I could select the area that wasn't the hole and extrude that
[12:17:24] <Quintox> do you have the link?
[12:17:34] <jthornton> linuxcnc.org
[12:17:40] <andypugh> In Inventor this is how I would model that: New part, then it wants you to click somewhere (possibly on an origin plane in the tree) for the starting sketch location. Then select the circle tool, click somewhere (probably the centre dot) then drag out and type the inner hole diameter. Then repeat for the outer diameter. Then "Return" top right of the Ribbon to exit the sketch.
[12:17:49] <andypugh> The Extrude that to thickness.
[12:18:10] <Quintox> ok thanks
[12:18:22] <jdh> yeah, I assume it would be easier to do it that way than to work with my existing 2d drawing
[12:18:38] <jthornton> just search for 6560 when you get there
[12:19:39] <andypugh> Then, new Sketch, click the flat surface. Draw a slot (line, construction geometry, centre point, out to edge, constrain vertical). Turn off construction, draw two circles on the construction line, constrain to be equal, two lines, constrain to be tangent, then trim the circles to make the slot. Then add dimensions to suit.
[12:19:56] <jdh> tomorrow is my last day of work until Dec.2, I'm hoping I can play with it then
[12:19:56] <andypugh> Return and extrude-cut
[12:20:25] <andypugh> Then make a polar array of the slots.
[12:21:10] <andypugh> There are quiicker ways, but the sequence I described will be more adaptable (Edit the array to change the number of slots, for example)
[12:23:23] <Vinci89> ok, so... back I am... and my spindle still won't run
[12:23:27] <Vinci89> at least not for long
[12:23:58] <IchGuckLive> hi vince
[12:24:09] <Vinci89> I figured I should change the wiring to the motor but that didn't really help... well, at first I'd thought it did, but 10minutes in on the gcode it suddenly stopped again
[12:24:40] <jdh> I dont' see construction line geometry
[12:24:56] <archivist> Vinci89, is the cable flexing?
[12:25:19] <Vinci89> well... kinda, it's a usual 0.75² (metricly speaking)
[12:25:30] <Vinci89> but I really doubt that the wiring is the fault
[12:25:46] <IchGuckLive> what stops the spindel
[12:25:56] <archivist> bend radius needs to be large else it breaks
[12:26:02] <IchGuckLive> crappy china board
[12:26:18] <Vinci89> I get a perfectly measured 0.06R internal resistance if I measure through the phases
[12:26:41] <Vinci89> plus I just changed the cable from the plugs inside the motor to the vfd...
[12:26:51] <IchGuckLive> let me read the full storry at the loggs
[12:27:22] <archivist> worth trying to flex cable while measuring
[12:28:13] <Vinci89> alright then, brb
[12:28:14] <archivist> we had standard cable last less than an hour till we learned about the right way on a printer
[12:29:14] <andypugh> jdh Give me a moment :-)
[12:29:34] <IchGuckLive> Vinci89: are both parport in out mode
[12:30:19] <archivist> IchGuckLive, wrong problem
[12:30:57] <archivist> his spindle motor is failing (whistles and fails to start)
[12:31:21] <jdh> found construction, it is an angle looking thing under format
[12:31:33] <IchGuckLive> i se the logs does say the same but only after second parport came in
[12:35:53] <Vinci89> no luck, i can flex the cable all i want, im still able to measure a pass between the phases
[12:36:31] <Vinci89> might it be some kind of temperature thing? i mean I don't really have any data on the motor, beside a frequency range and the max. voltage
[12:36:38] <Vinci89> maybe I set the vfd parameters wrong?
[12:36:56] <archivist> doubt it
[12:37:21] <cpresser> does the VFD show an error?
[12:37:30] <Vinci89> no
[12:37:41] <Vinci89> the spindle simply stopped turning
[12:37:50] <cpresser> it (the vfd) thinks the motor is still runing?
[12:37:51] <archivist> that is odd it should sense the problem
[12:38:17] <Vinci89> i guess it does, at least it still shows me the frequency from my pwm input
[12:38:39] <Vinci89> plus i can hear the current flow in the windings...
[12:39:15] <andypugh> jdh: I was going to record a video, but something is eating all my CPU...
[12:39:28] <archivist> I would attempt carefully to spin the spindle
[12:40:10] <Vinci89> tried that, i tried to "give it a little" push at pretty much every frequency after it stopped...
[12:40:11] <Vinci89> nothing
[12:40:26] <jdh> I got the extruded disk, and a slot
[12:40:31] <andypugh> It sounds like you have lost a phase.
[12:40:57] <andypugh> But you said you checked continuity
[12:41:07] <Vinci89> i can measure internal resistance between all phases, i doubt it
[12:41:44] <archivist> internal short between phases, but that really should cause a drive fault
[12:42:59] <Vinci89> until ~6h ago i was pretty certain that it was the wiring... and it even ran like 10minutes without any flaws
[12:43:21] <Vinci89> and then, from one moment to the next it just stopped...
[12:43:21] <jdh> could you run a 3 phase motor off 2 phases and use the third as a tach?
[12:45:19] <archivist> jdh two phases summed is actually one phase so no, unless you take the motor apart to actually rewire into two phase like a single phase with capacitor is run
[12:45:41] <jmasseo> heekscam?
[12:46:32] <archivist> jmasseo, lacks support these days
[12:46:47] <IchGuckLive> i am the only devel at this time
[12:46:54] <IchGuckLive> on heeks
[12:47:14] <jmasseo> ah
[12:47:37] <jmasseo> is your name heeks?
[12:48:16] <IchGuckLive> no dan is not more interested in his projekt
[12:48:17] <archivist> dan heeks has not been into the heekscad channel in a long while
[12:48:27] <jmasseo> heh
[12:48:39] <IchGuckLive> he is courentlie doing things in Russia
[12:48:56] <archivist> <the_wench> last seen in 2011-08-01 18:34:53GMT 838:59:59 ago, saying Quit: Miranda IM! Smaller, Faster, Easier.
http://miranda-im.org
[12:52:10] <IchGuckLive> jmasseo: did you find my youtube channel where i interact with heeks
[12:53:29] <tjtr33> IchGuckLive, where do hear about Dan? his last blog is 2012. ( & your youtube stuff is great )
[12:53:55] <jmasseo> IchGuckLive: i did not
[12:54:18] <IchGuckLive> dan mailes me from time to time he is on the blender 3D projekt doing things for his own Heeksart development
[12:54:35] <IchGuckLive> http://www.youtube.com/user/magic33de/videos
[12:54:36] <tjtr33> ah the BlenderCam maybe ( i hope i hope i hope )
[12:54:47] <IchGuckLive> no no cam at all
[12:54:50] <tjtr33> :(
[12:55:07] <IchGuckLive> heekscad has been a fal away from the art projekt
[12:56:25] <IchGuckLive> tjtr33: awallin is the cam guru
[12:56:36] <IchGuckLive> SMILE ;-)
[12:59:00] <tjtr33> awallin is very good, not as easy to use as Blender ( voronoi curves = scary! )
[12:59:33] <tjtr33> and we all know how easy Blender is ;)
[12:59:34] <Vinci89> guys... it's been a pleasure, I really need one or two cold beers...
[12:59:39] <Vinci89> thanks for the help everyone!
[13:34:17] <andypugh> jdh:
http://youtu.be/WeIMZiqkuFY
[13:34:21] <andypugh> It's HD
[13:34:36] <andypugh> (So you may want to enlarge the viewer)
[13:35:23] <andypugh> It shows me a) Applying constrints too late, and having to rescue the geometry and b) Because if that having to manually close the loop.
[14:53:41] <tjtr33> archivist, nice PD mechanism reference scanned by Google
http://www.pdnotebook.com/2013/02/507-mechanical-movements-site/
[14:53:58] <tjtr33> but i think i see the scanner's finger on page 4 :)
[15:01:07] <Jymmm> archivist: All it needs now is the vector artwork for them all =)
[15:01:42] <Jymmm> #234 is different
[15:04:19] <tjtr33> theres some guy wanting to animate them all (on that web page)
[15:05:05] <tjtr33> was looking for an adjustable orbiting mechanism ( adjust circular orbit from 0 to 2mm )
[15:14:47] <andypugh> tjtr33: Like :
http://www.youtube.com/watch?v=VrvwlPxRhKU
[15:20:01] <andypugh> tjtr33: I have the solid model for that if you want
[15:20:57] <tjtr33> andypugh, yes, (sorry, i was away reading that book)
[15:21:24] <tjtr33> veru nice work, do the wedges change the eccentricity?
[15:21:51] <tjtr33> very
[15:27:51] <tjtr33> i suppose it does, maybe steppers on the wedge screws could shift the eccentricty
[15:28:01] <tjtr33> thx!
[15:39:12] <andypugh> The wedges take up the slack in the anti-rotation cross. The accentric adjustment isn't actually in that model, but is in the finished one:
http://www.bodgesoc.org/cycloidal.stp for STEP or
http://www.bodgesoc.org/cycloidal.igs for IGES
[15:41:10] <andypugh> The way the eccentricity is adjusted is by having two coaxial eccentrics of half the required throw. Aligned you get double, and opposed they cancel.
[15:41:46] <andypugh> If you need to adjust the eccentricity while the mechanism is rotating, then it gets a bit more interesting.
[15:48:11] <tjtr33> in the video, i think the wedges would chg the throw but the outer ring pin diameter is fixed :( ( and that would bind)
[15:48:37] <tjtr33> oh, , ok I'll dig out a viewer ( not linux afaik )
[15:48:40] <tjtr33> thx again
[15:50:41] <tjtr33> and i get the idea of the 2 coaxial excentrics being able to null each other, cool
[15:54:20] <andypugh> If you need to electrically adjust the throw while the system is rotating then a DC motor and slip-rings would be one way. A neater way would be a brushless motor magnet ring on the rotating part and a stationary stator with hall sensors. Then an encoder on the totating part. Keep the motor and encoder in phase to keep the throw constant, adjust for lead or lag to adjust.
[15:54:36] <andypugh> This is the system I used on my CNC boring head.
[15:55:18] <andypugh> (err, which I was going to patent but the agents defcided wasn't commercial enough)
[15:56:56] <tjtr33> i thimk im working out a mechanical way... you're much faster than i, but its one external motor, driving 2 ring gears.
[15:57:05] <tjtr33> One ring gear on OD of inner sleeve. One on OD of outer ring.
[15:57:06] <tjtr33> They turn opposed by the single spur. infinite turn go 0 to Max throw.
[15:57:24] <tjtr33> thimk hehe
[15:57:57] <tjtr33> want a commercial app for it?
[15:58:23] <andypugh> The exact mechanism you need is on this machine:
https://picasaweb.google.com/lh/photo/lorjlRbUi9B2VQy0Cd0KcNMTjNZETYmyPJy0liipFm0?feat=directlink
[15:59:08] <andypugh> That has a cross-slide on the spindle that is adjusted by a wheel at the back. No idea how it works :-)
[15:59:11] <tjtr33> hehe wheres waldo.. which buit might that be?
[15:59:20] <tjtr33> ok
[15:59:40] <andypugh> (It's my dad's Kears boring machine)
[15:59:55] <andypugh> Kearns, that is
[16:00:21] <tjtr33> wow nice toys, i'll sketch this up and post it or email it thx very much
[16:01:47] <andypugh> There is one in motion here;
http://www.youtube.com/watch?v=cnaGSZ9lYss
[16:02:15] <andypugh> At 1:20 you can even see it adjusting the throw :-)
[16:13:29] <tjtr33> wow, is it balanced? ( no washing machine syndrome ?)
[16:13:43] <andypugh> No, it's just a big heavy machine :-)
[16:14:28] <tjtr33> hahaha
[16:14:34] <andypugh> I have no idea what I will do when I inherit it. Hopefully it is a long time in the future, but I suspect not :-(
[16:14:46] <tjtr33> i hope its a long time
[16:15:42] <andypugh> So do I, and only partly to avoid building a new workshop.
[16:16:33] <tjtr33> it looks like some auto feed for the rear adjuster, very interesting mechanics
[16:17:43] <andypugh> There is a flat-belt drive from the spindle itself to the spindle slide adjusting handle
[16:48:44] <_DJ_> gn8
[17:09:40] <Tom_itx> andypugh building a new workshop wouldn't be such an issue if you didn't build things to last 500 years
[17:09:57] <Jymmm> Tom_itx++
[17:09:59] <somenewguy> hahaha
[17:10:17] <Jymmm> andypugh: ...or need 40 TONS of rock!
[17:20:08] <Jymmm> What size tank is that on the far right? Looks 40cf to me
http://images.craigslist.org/01616_5EG0TB1E7dO_600x450.jpg
[17:20:59] <Jymmm> The two in front are suppose to be 10 CF acetylene
[17:34:47] <andypugh> I persuaded Renishaw to lend me one of these:
http://www.renishaw.com/en/resolute-rotary-angle-absolute-encoder-options--10939
[17:36:04] <andypugh> It's a pity they want it back, but I did just want it to check the BiSS driver.
[17:36:26] <Tom_itx> for a lathe or what/
[17:36:27] <Tom_itx> ?
[17:36:36] <andypugh> Not _for_ anything.
[17:36:57] <Tom_itx> the ring size just seems rather large
[17:37:17] <andypugh> Though if they had said "just keep it" I would have found a use.
[17:37:32] <andypugh> Round the outside of a rotary table?
[17:38:13] <andypugh> On the actual spindle flange of a lathe? (with that accuracy the torsional wind-up of the lathe spindle would matter)
[17:38:41] <Tom_itx> that might not be a good place for one
[17:38:49] <Tom_itx> as the chips etc might end up all over it
[17:38:58] <andypugh> Non-contact optical. Probably OK
[17:39:41] <Tom_itx> it has high tolerance to dirt, scratches and greasy fingerprints that can cause other encoder systems to miscount
[17:39:42] <Tom_itx> huh
[17:39:46] <andypugh> http://www.renishaw.com/media/pdf/en/5af22d458b614ba3a55d0082ae7ec395.pdf shows examples of dirty scales it is happy with :-)
[17:40:58] <andypugh> They do the rotry version in 52 to 550mm diameters.
[17:41:28] <andypugh> I suspect that the only thing that they aren't is cheap :-)
