Back
[00:00:51] <Groguard> Does anyone have any experience with the dc relay coils for a gecko g540?
[00:02:06] <marmite> nop
[00:02:34] <marmite> gecko looks badass thoe
[00:02:37] <marmite> :D
[00:03:23] <Groguard> Im going to be getting one soon, and I have a couple questions
[00:04:48] <marmite> sweeeeeeet
[00:05:12] <marmite> drool
[00:05:22] <Groguard> Yea, I really like it. I had a lot of people recommend it.
[00:07:45] <marmite> alot better then my
http://www.jbcnc.se/product_info.php?cPath=4_37&products_id=98 :D
[00:09:21] <R2e4_> i have a g 540
[00:09:30] <marmite> with 3
http://www.jbcnc.se/product_info.php?cPath=34&products_id=101 , was dirt cheep thoe bougt a kit for 100 usd
[00:17:00] <Groguard> R2E4: I had a couple questions about the relay coils
[00:18:46] <R2e4_> yeah?
[00:22:08] <zeeshan> can linuxcnc understand full circle circular moves
[00:22:12] <zeeshan> or only quadrant circular moves?
[00:22:32] <zeeshan> need to setup my NX post processor accordingly ;p
[00:34:50] <marmite> http://www.lathes.co.uk/emco/page3.html fuck i want this
[00:52:12] <marmite> i did just download a ileagel torrent for software that i have bougth irl :D due to its alot easier to get the install files
[00:52:26] <marmite> is kinda reetarded
[00:52:45] <marmite> how software develepers makes it hard for there users to get there software
[00:58:51] <roycroft> software developers have nothing to do with it
[00:59:15] <roycroft> legal and marketing departments dictate to the release engineering departments how to package and distribute it
[00:59:32] <roycroft> developers usually hate what goes out the door
[01:38:56] <postaL> can anyone recommend a good cam software?
[01:41:37] <zeeshan> yes
[01:41:43] <zeeshan> only 2 that i would ever use in my life
[01:41:49] <zeeshan> Siemens NX
[01:41:55] <zeeshan> and Mastercam
[01:42:25] <zeeshan> siemens nx is highly customizable, including making custom post processors
[01:42:39] <zeeshan> i like mastercam for 5 axis machining due to is amazing tool paths.
[01:43:23] <postaL> i just have a small 12" x 18" flat bed 3 axis cnc
[01:43:26] <postaL> (fireball v90)
[01:43:40] <zeeshan> what cad software do you use?
[01:43:47] <postaL> solidworks
[01:43:59] <zeeshan> get the solidworks mastercam plugin
[01:44:19] <zeeshan> you can then pick up features you've made in it ;)
[01:44:21] <zeeshan> like holes!
[01:44:59] <postaL> i shall give it a try, thanks for the input
[01:45:22] <zeeshan> http://www.youtube.com/watch?v=IAwHBItQJXQ
[01:45:23] <zeeshan> sums it up
[01:45:24] <zeeshan> ;D
[02:15:13] <Deejay> moinsen
[02:37:27] <chrism> Testing
[02:37:55] <archivist> 123
[02:38:32] <chrism> Now to get dev working...thanks
[02:38:57] <Deejay> copied that
[04:59:26] <Matze_> Hi, someone german here ? :)
[05:12:47] <Matze_> Can someone help me to describe the motion module ? I'm writing a documentation while working with linuxcnc, but dont know how to describe it good :)
[05:16:58] <archivist_herron> like what information ?
http://wiki.linuxcnc.org/cgi-bin/wiki.pl?TrajectoryControl
[05:17:32] <archivist_herron> and this
http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Simple_Tp_Notes
[05:18:45] <archivist_herron> and
http://www.linuxcnc.org/docs/html/man/man9/motion.9.html
[05:21:48] <archivist_herron> I imagine you need to read the source to fix "This manual page is horribly incomplete."
[05:23:58] <Matze_> ok thx :)
[05:32:17] <archivist_herron> note the planner is being amended to to improve look ahead
[05:37:15] <Matze_> ok wait, almost done reading
[05:40:49] <Matze_> ok finished :)
[05:42:59] <Matze_> Trajectory Control is quite good, but doesnt help too much :)
[05:54:09] <archivist_herron> not sure what you are doing sometimes it is just finding the information
[06:09:22] <Matze_> im just trying to find the right words to describe motion, because the description in the manpage sucks... nothing more :)
[06:54:33] <archivist> the manpage is automatically made from the source if I remember correctly
[07:29:58] <PetefromTn> Mornin' folks.
[07:30:06] <R2E4> morning
[07:30:46] <R2E4> Not able to get them drives working p[roperly.
[07:31:12] <PetefromTn> damn that sucks..
[07:31:28] <R2E4> I zero out the Vcmd, and it drifts, I adjust zero adjustment, it stops or a moment and then drifts the other way.
[07:31:53] <R2E4> Now when I put the battery on it, it doesnt shake, but doesnt move either.
[07:32:40] <R2E4> Found osme drives, but they want 1800.00 each for them.
[07:32:50] <R2E4> I think I am screwed.
[07:33:16] <archivist> I think a small amount of drift wont matter as the loop will lock them up
[07:33:37] <archivist> I dont think you are screwed
[07:33:50] <skunkworks> if they worked before...
[07:35:12] <PetefromTn> I am unfamiliar with your drive setup so I cannot speak intelligently but if they are moving at all you are most definitely not screwed.
[07:35:30] <PetefromTn> It is just a matter of finding out what they want from the control and giving it to them.
[07:35:35] <R2E4> When I connect the 7i77 to them, they drift a bit faster.
[07:35:56] <PetefromTn> I got brand new drives in my mahince and I still had some frustration getting them working at first.
[07:36:15] <R2E4> Well, that gives me hope
[07:36:22] <skunkworks> the drives should not be anabled until linuxcnc is in control
[07:36:24] <archivist> connect the feedback
[07:36:37] <PetefromTn> Servo tuning is a beotch honestly.
[07:36:49] <R2E4> Linuxcnc is enabling them.
[07:36:59] <skunkworks> when?
[07:37:12] <R2E4> When I turn on.
[07:37:25] <R2E4> The button right next to the estop red X
[07:37:29] <skunkworks> great!
[07:37:31] <PetefromTn> Are you still working with the single Y motor and driver?
[07:37:35] <R2E4> yes
[07:37:51] <R2E4> I removed the enable relay for the x and z
[07:37:56] <PetefromTn> Good that makes things much less confusing.
[07:40:51] <PetefromTn> You said that when you put the battery on it you get no movement, is this still with linuxCNC enabled?
[07:42:01] <R2E4> yes
[07:42:17] <R2E4> The enable light is on on the driver
[07:42:23] <PetefromTn> Maybe post some pictures of the drive board you are adjusting and the wiring so some of the pro's here can look at it.
[07:42:26] <R2E4> Or else it wouldnt move
[07:43:23] <PetefromTn> Those pots are confusing and you need to know what you are adjusting before messing with them or you could get it way out of whack. Maybe put a mark on the pots to know where you started and count turns etc..
[07:44:19] <PetefromTn> I would be hesitant to turn them at all if they were working before on the old control and would probably mess with the PID tuning in LinuxCNC to see if I could get something working. Is the X and Y exactly the same drive and setup?
[07:44:25] <R2E4> I did that. I know where they should be before I started
[07:44:34] <PetefromTn> That's good.
[07:44:49] <R2E4> Yes, they are the same
[07:45:38] <PetefromTn> For instance my drives came from Machmotion and they were setup to work on a common installation parameter so when I got them they locked down right off the bat and all I had to do was tune the PID loop in linuxCNC.
[07:46:19] <R2E4> PCW was saying, if it doesnt work with the battery it wont work with linuxcnc
[07:46:49] <PetefromTn> He is right and to be honest from my view he is the expert here.
[07:47:02] <R2E4> I emailed sany-denki to see if i can get an english manual.
[07:47:04] <PetefromTn> He seems to have a huge knowledge base.
[07:47:36] <PetefromTn> Did you not have movement with the battery before? I thought you did.
[07:47:58] <R2E4> The amp is expecting an AC voltage of +10 -10, i dont know how +1.5vdc is going to make it work, but he knows what he is talking about
[07:48:18] <R2E4> No, it shaked and went crazy before.
[07:48:46] <jdh> 1.5vdc shoudl make it go 15% forward
[07:48:47] <PetefromTn> Actually the way it works if I am not mistakes is that the way linuxCNC controls it is to vary the DC voltage so any voltage will cause movement and the more voltage the faster it goes...
[07:49:21] <PetefromTn> That is I suppose why they recommend using a small battery so you get slow movement to test with quick and easy.
[07:49:22] <R2E4> its not moving at all now, except a drift.
[07:49:46] <PetefromTn> Did you put the pots back to the original settings?
[07:49:51] <R2E4> no
[07:50:10] <PetefromTn> Maybe try to do that and remember where you are now to go back if it works?
[07:50:23] <PetefromTn> doesn't work rather.
[07:51:07] <PetefromTn> I remember you said it shook the whole machine LOL... believe me I can relate to that. Until I got the PID right my machine was making some rather horrible noises from servo oscillation I guess.
[07:53:50] <PetefromTn> I dunno anything about how yours is wired so I dunno I am just trying to help you think thru what you have done and what you may have missed that caused the movement to cease.
[07:54:23] <PetefromTn> If the drive is enabled and it is wired correctly the battery on the right terminals should cause movement.
[07:59:20] <PetefromTn> http://linuxcnc.org/docs/html/motion/pid_theory.html Here's the section on PID tuning I am sure someone may have linked it to you already but if not there it is. There is also a decent one that JT made I think somewhere.
[07:59:20] <R2E4> unless something is fried in the amp
[07:59:20] <PetefromTn> Damn I hope not for your sake.
[08:00:37] <PetefromTn> I gotta take care of something out in the shop be back in a bit. Good luck man. Pete is of course in cali so it will be a few hours before he is awake and online.
[08:03:39] <R2E4> Is your P higher number than 1?
[08:12:57] <jthornton> normally you increase P until you get oscillations then move on the next step
[08:13:05] <jthornton> have you seen my tuning tutorial?
[08:13:09] <R2E4> yes
[08:13:14] <R2E4> Read it ten times
[08:13:21] <R2E4> ok, maybe 6
[08:13:25] <jthornton> did it make any sense?
[08:14:01] <R2E4> yes, I was just wondering what the norm would be, if it will end up being say 5 or 10 or close to 1
[08:14:32] <jthornton> just depends on a whole bunch of things what it might end up as
[08:15:03] <R2E4> It may just stay at 1, if I cant get the drives adjusted....lol
[08:15:35] <jthornton> this is my BP knee mill P_GAIN = 550
[08:15:50] <jthornton> FF1 = 1.8
[08:15:56] <jthornton> FF2 = 0.017
[08:16:01] <jthornton> the rest are 0
[08:16:16] <R2E4> AH,you just kept increasing it until you hit 550?
[08:16:38] <R2E4> Change it, restart linuxcnc....
[08:16:44] <jthornton> but with FF1 not being near 1 I don't have the pid normalized
[08:16:56] <jthornton> no, use the tuning from axis
[08:19:08] <R2E4> AH, ok, I was staring at the P in the file thinking, do I change it to 1.5, or 2 or maybe 10....lol
[08:19:15] <jthornton> Machine > Calibration
[08:19:38] <jthornton> I would go slow
[08:20:01] <R2E4> I have to get the drvie to stop drifting first.
[08:20:28] <jthornton> a little drift is ok as LinuxCNC will hold it still
[08:20:47] <R2E4> It drifts more when I connect the 7i77
[08:20:50] <jthornton> my CHNC drifts and it shocked me at first
[08:21:22] <jthornton> the drive should not be enabled until LinuxCNC is in control
[08:21:35] <R2E4> Theres something I am not doing right.
[08:21:44] <R2E4> I am only enabling it from linuxcnc
[08:22:05] <jthornton> with the power button?
[08:22:09] <R2E4> yes
[08:22:15] <cpresser> R2E4: i guess drift is due to the fact that the servo-drive-input never really interprets an input as zero. but as soon as you close the loop, linuxcnc will read the encoder-count, calculate PID and send signals in such a way that the axis wont move
[08:22:42] * jthornton has to start his day
[08:22:47] <jthornton> see you guys in the shop
[08:23:02] <cpresser> i suggest you look at hal-scope and see what the pid has as input and outputs
[08:23:35] <R2E4> yes, i understand that.... but when I connect linuxcnc 7i77 Vcmd and Vgnd to the drive, and enable it in linux, it drifts and the ballscrew is turning real real slow
[08:25:40] <cpresser> R2E4: what can you see with halscope at this point in time?
[08:26:07] <R2E4> I didnt check. pcw said get it working with a battery first.
[08:26:13] <cpresser> what does pos-fb show?
[08:26:30] <R2E4> I am at work, cant check it, but will later.
[08:26:31] <cpresser> R2E4: in my option: it is working. the drive is moving the axis
[08:26:45] <cpresser> now you need to tune it. make sure the PID works correct
[08:27:17] <cpresser> thats the trick with PID, it will cancel such 'steady errors' like a voltage offset
[08:28:16] <R2E4> Should I start with the motor disconnected from the ballscrew?
[08:28:49] <cpresser> good idea. then you can turn the ballscrew by hand and see how the motor reacts
[08:33:01] <cpresser> R2E4: you can also test the 7i77 without the PID-loop. using setp you can input values into hal. and see if the motor turns the way you expect it to
[08:35:18] <R2E4> cpresser: hmmmm.... Iwouldnt attempt that at my knowledge level....
[08:41:38] <cpresser> R2E4: thats no magic!
[08:42:00] <cpresser> R2E4: fire up a commandline. list all hal-pins; assign values to them.
[08:42:29] <cpresser> you cant damage anything if you got the motor mechanically disconnected
[08:47:44] <R2E4> I have to refresh. I am a bit stressed at the moment about this machine and those drives.
[08:51:32] <PetefromTn> R2E4: Yeah man I would start with it disconnected.
[08:59:58] <JT-Shop> R2E4, after you replaced the cable does the DRO read correctly when you move the axis by hand?
[09:00:06] <R2E4> yes
[09:00:24] <R2E4> I only replaced the Vcmd and Vgnd cable, not the encoder cable
[09:00:47] <JT-Shop> ok, does the DRO read correctly?
[09:00:52] <R2E4> yes
[09:01:08] <JT-Shop> can you pastebin your hal and ini file?
[09:03:07] <R2E4> I am not there, I can paste it later on though.
[09:03:16] <JT-Shop> ok
[09:03:58] <R2E4> I havent changed anything on the PID. I was concentrating on trying to get the battery test to work.
[09:06:53] <JT-Shop> the battery should drive the axis at 15% of full speed
[09:09:01] <R2E4> well, thats not happening.
[09:09:32] <R2E4> I put the battery on it now and nothing happens.
[09:13:52] <skunkworks> did you play with the gain pots?
[09:14:05] <skunkworks> (maybe input gain or loop gain)
[09:24:33] <R2E4> I was turning them down because it was shaking when I applied the battery.
[09:25:10] <R2E4> I didnt turn them down much.
[09:25:31] <R2E4> I returned them to where they were before, and it is not shaking.
[09:29:17] <skunkworks> does it move?
[09:32:47] <R2E4> no, not by command, it drifts.
[09:33:02] <R2E4> I am not there but will be in a short while and will start working on it again.
[09:36:31] <pcw_home> velocity mode drives pretty much always drift when the PID is not running
[09:36:58] <R2E4> hehe, thats a nice one to know.......
[09:37:00] <pcw_home> (even a few mV of offset will cause a slow drift)
[09:37:44] <pcw_home> thats one reason the drives are disabled when the PID is not running
[09:38:01] <pcw_home> (in HAL)
[09:38:22] <R2E4> I tied Vcmd to Vgnd to get 0 volts on input and was able to adjust zero adjust to where it just barely would drift.
[09:39:52] <R2E4> yes, so when I turn the power button on to linuxcnc, it enables the drives. My PID is at the default Pncconf put it 1, and it drifts .
[09:40:26] <cradek> that's a good inital setting. After pid is tuned, I like to tweak the offset pot again to slow/stop the invariable one-count dithering
[09:40:47] <pcw_home> thats wrong, the drive should no be enabled at power on
[09:41:18] <pcw_home> it should only drift is the PID loop is no running
[09:41:21] <skunkworks> he is saying - linuxcnc power on - which should be when the pid is activated.
[09:41:39] <pcw_home> OK so PID is not working
[09:41:51] <R2E4> the 7i77 is enabling the drives with ENA
[09:42:23] <R2E4> yeah.
[09:44:25] <pcw_home> a one point you were going to swap VCMD and VGND I hope you didn't do this (7I77 gnd must go to VGND)
[09:44:47] <R2E4> The drive is not moving with the battery, so whatever I do in Linux it shouldnt work?
[09:45:01] <R2E4> I didnt swap Vcmd and Vgnd
[09:45:38] <pcw_home> well if its not moving with a battery nothing you do with linuxcnc is going to help
[09:45:53] <R2E4> from what I understand if the battery doesnt drive the motor linux wont do anything.
[09:46:00] <quitte_> what is a home latch?
[09:46:22] <pcw_home> Thats right (is the drive enabled?)
[09:46:53] <R2E4> yes, I do turn on linux to enable the drive
[09:46:56] <jthornton> quitte_, it is the slow move after home seek
[09:48:26] <quitte_> jthornton: i have never seen a proper home seek
[09:50:02] <quitte_> okay. I just saw a tiny extra move. what is the purpose of that?
[09:54:16] <R2E4> I have four pots for adjusting. one for speed scale, one for zero adjust and tow loop gain
[09:54:43] <R2E4> The two for loop gain should be adjust both in the same direction? one is gain and one is offset?
[09:55:12] <pcw_home> one is probably P gain and one is probably I gain
[09:56:05] <pcw_home> although they may be current and velocity loop gains
[09:56:26] <pcw_home> you need a manual before adjusting
[09:57:05] <pcw_home> (well the zero and speed scale are fairly obvious)
[09:58:55] <pcw_home> speed scale should be set so the axis speed at 10V is maybe 20% higher than rapid speed for that axis
[10:03:30] <R2E4> Nobody is getting manuals from sanyo-denki on these drives, in english.
[10:03:47] <R2E4> I have the original in Japaneese. I guess its time to get it translated.
[10:04:43] <R2E4> its 5,000.00 budget price to replace motors and drives.
[10:05:30] <R2E4> IT would be worth the two to three hundred to get the manual translated.
[10:08:01] <jdh> how much can you get for the motors and drives?
[10:10:08] <R2E4> heres the drive
[10:10:38] <pcw_home> pictures may be good enough
[10:11:18] <R2E4> http://www.ebay.com/itm/20BM040HXTPT4-SANYO-DENKI-SERVO-MOTORS-NEW-/360754234377?pt=LH_DefaultDomain_0&hash=item53fea0c409
[10:11:25] <R2E4> New WOW!
[10:11:58] <R2E4> I got a good deal, I'll sell mine for 2000.00 each...lol
[10:13:31] <pcw_home> The first thing to find out is why the drive is not working anymore, that is, what changed
[10:14:04] <R2E4> Gotta laugh, or else I'll cry.... spent 6000.00 personal money on this machine, it was suppose to be working, keep dumping money into it.
[10:14:50] <R2E4> The only thing i changed was adjust the pots and changed the cable to shielded cable for the Vcmd and Vgnd.
[10:15:40] <jdh> did you check your cabling?
[10:16:06] <jdh> create some other problem when you swapped the wiring?
[10:16:20] <R2E4> I'll look at that when I get there in a bit.
[10:16:31] <R2E4> possibly.
[10:18:16] <pcw_home> I dont seen anything so far thats suggests any real problems other than tuning and connection issues
[10:19:47] <pcw_home> slow creeping is a very good sign (it means the velocity feedback to the drive is working so you have slow controlled movement)
[10:20:37] <pcw_home> since these look like AC servos it also means commutation is working
[10:22:37] <R2E4> I dont think this matters but, I have a different power source from 110 and the 220 3 ph
[10:23:51] <R2E4> I cannot get a good 120 from the existing transformers cause it only has 100v on secondary.
[10:24:37] <R2E4> So I am running all new stuff from 120 off my single phase panel, and 220 coming from 600v entrance powering the machine
[10:24:45] <R2E4> same electrical ground though.
[10:25:27] <tjtr33> R2E4, if you have another drive that used to work... do NOT touch the pots. open it up and ohm out the pots in circuit ( 3 connections each, 3 values each) adjust wacko drivbe to those values, you are now at step 0 :)
[10:26:25] <R2E4> I caould try that.
[10:26:57] <tjtr33> it would restore the amp pots to a 'good' state
[10:27:13] <tjtr33> all power off!
[10:27:38] <R2E4> I didnt turn them morte than half a revolution, so I can get them back by placing them the same as the X axis
[10:28:50] <tjtr33> do these pots have a tiny brass button adjust screw ( multiturn) or a clockface with markings arounf the white plastic screw?
[10:28:55] <R2E4> haha, I throw the breaker on the machine and the 220 3ph switch next to the machine each time I go in to do something.
[10:29:25] <tjtr33> yeah, making sure power off is a little litany done each time
[10:29:41] <R2E4> This is them.
http://www.ebay.com/itm/Sanyo-Denki-BL-Super-27BA050FXT38-Servo-Amplifier-A050-08AB-/360548814953?pt=BI_Control_Systems_PLCs&hash=item53f2625069
[10:29:44] <tjtr33> describe the pots tho ( pic not clear )
[10:30:13] <pcw_home> I suspect these a mistake on the input wiring (if the drives still creep they are very likely OK)
[10:30:27] <pcw_home> s/these/there's/
[10:30:33] <tjtr33> those yellow pots are multiturn 1/2 turn dont mean squat.. you made very little or no difference
[10:30:39] <R2E4> square yellow with brass. multiturn
[10:31:20] <tjtr33> "half a revolution" didnt make a significant diff
[10:31:34] <tjtr33> return to 'good; settings then...
[10:31:52] <tjtr33> listen very closely to the wacko pot, it will click at one end
[10:32:17] <tjtr33> adjust to match recorded values from click, this is the turns needed to get back to 'norm'
[10:32:34] <tjtr33> then COUNT as you change the pots ( cw/ccw)
[10:33:24] <tjtr33> note the full value of the pot, and what 1 turn changes it. so you can estimate if your adjust was 'significant'
[10:33:51] <R2E4> to find out where it is in the scale of the pot?
[10:34:01] <R2E4> ok
[10:34:16] <R2E4> I am going to look at the wiring first.
[10:34:22] <tjtr33> well, to find out where the 'good' value is in reference to the range of the pot
[10:34:55] <tjtr33> the diagrams on your website were confusing, i was not convinced they were for your amp
[10:35:29] <jdh> r2e4: try eharmony.com Specify you are looking for a japanese date, take her to dinner, have her read you the manual.
[10:35:56] <tjtr33> hehe thats a great idea ( except that the jap words used in the manual might throw her )
[10:36:57] <tjtr33> if you can find the manual as pdf, you might try translating parts with google xlate
[10:37:16] <tjtr33> (guaranteed jinglish )
[10:37:42] <R2E4> Which diagrams are you talking about?
[10:38:13] <tjtr33> hmm you posted a link to a pic, i truncated the p;ic off the url, that opens a directory, theres 4 diagrams there
[10:38:53] <tjtr33> yesterday
[10:38:58] <R2E4> schematics1 1A 2 and 3
[10:39:13] <tjtr33> yeah 1a is only for absolute encoder, toss that
[10:39:39] <R2E4> 3 is the one
[10:39:47] <tjtr33> link?
[10:40:02] <R2E4> 2 is the spindle.
[10:40:40] <R2E4> 1 is the correct one. this is out of the machine manual.
http://irmtl.com/vm40/schematics1.pdf
[10:40:56] <tjtr33> what i saw were wiring diagrams ( external ) not schematics ( internal ), so little aid to understanding
[10:41:03] <R2E4> This is correct. I checked th epins
[10:42:26] <tjtr33> that diagram shows no pots, so is nice, but not useful for task at hand
[10:42:46] <R2E4> I dont have schematics of the amps.
[10:43:18] <R2E4> oh, I do....
[10:43:23] <tjtr33> !
[10:44:02] <R2E4> I'll go scan it.
[10:45:11] <tjtr33> you get those together for the group here, i gotta go for a bit. will look later today. schema may diambiguate the 2 pots for 'gain'
[10:46:15] <R2E4> ok uploading
[10:47:32] <R2E4> http://irmtl.com/vm40/SD-amp.pdf
[10:48:26] <tjtr33> IMO vri is Vcmd offset VR2 is Vcmd scale...
[10:48:35] <tjtr33> IMO vr1 is Vcmd offset VR2 is Vcmd scale...
[10:50:16] <tjtr33> whats the link to the troubleshooting chart?
[10:50:37] <tjtr33> ( 2 pots on right edge are unlabled or un-numbered :(
[10:51:38] <tjtr33> got it
http://irmtl.com/vm40/BL-man.pd
[10:52:35] <R2E4> yeah
[10:54:24] <tjtr33> for VR3 and VR6 both are abouit loop gain VR3 is offset (affects creep) VR6 is gain (affects 'violence' of reaction to a given error)
[10:55:40] <R2E4> ok, cool thanls
[10:56:25] <tjtr33> hehe it lists up to VR9 ( dont mess with them tho ) begin with settign the drive pots to a known state. then use this block diagram to understand how the Vcmd pushes & pulls the motor.
[10:56:49] <tjtr33> and make sure the TG feedback is connected ( bldc tacho )
[10:57:10] <tjtr33> hmm if blds tacho, also make sure it has power
[10:57:29] <R2E4> Those are connected. I didnt mess with them, they are directly from the motor directly to the drive.
[10:57:34] <tjtr33> k
[10:57:48] <tjtr33> oh thru a spcl cable, gotcha
[10:58:16] <R2E4> Yeah, that is CN2, I only connected to CN1.
[10:58:31] <tjtr33> looks nice, just not used them ( panasonic and yaskawa , baldo, not these tho )
[10:59:01] <tjtr33> ok slow & methodical, have fun!
[10:59:12] <R2E4> thanks
[11:10:31] <CaptHindsight> http://www.3ders.org/articles/20140218-watch-amanda-boxtel-walk-again-with-first-3d-printed-hybrid-exoskeleton.html I was just imagining using EMC and a joypad to control an exoskeleton like this
[11:19:07] <tjtr33> CaptHindsight, applause! but send her some nicer shoes :)
http://www.griffith.edu.au/visual-creative-arts/design-digital-media/additive-manufacturing-2013/3d-print-fashion-show-malaysia
[11:24:24] <CaptHindsight> these would also be handy for very busy people, you could sleep while walking by just loading your G-code for the trek home :)
[11:27:09] <zeeshan> R2E4 thats against code, because they need to be coming from the same circuit breaker.
[11:27:29] <zeeshan> thats why you have 3 pole breakers for your washing machine and stove.
[12:27:25] <Gigs-> you all think a leather flat belt would stand up to high temperature use?
[12:27:36] <IchGuckLive> hi all B)
[12:28:05] <IchGuckLive> Gigs-: why not using a PZ
[12:28:10] <Gigs-> what's that
[12:28:24] <IchGuckLive> high temperatur belt
[12:28:35] <IchGuckLive> SPZ for better Tourch
[12:28:38] <archivist> Gigs-, use the right material for the temperature, silicon based belting maybe
[12:28:50] <Gigs-> where can you get such things?
[12:28:50] <IchGuckLive> and RipPZ for a realy effectfull run
[12:29:06] <IchGuckLive> Gigs-: where in the world are you
[12:29:09] <Gigs-> US
[12:29:10] <IchGuckLive> im in germany
[12:29:30] <IchGuckLive> what is the expected belt length
[12:30:05] <Gigs-> not sure yet, probably 12 inches conveyance length so ... 28 inches or so?
[12:30:06] <IchGuckLive> so you can go to the car junkjard and do as our Bolivian and south american CNC freks do
[12:30:18] <Gigs-> oh interesting, like an overhead cam belt
[12:30:38] <IchGuckLive> are you going for a 3D printer
[12:31:10] <Gigs-> no, conveying annealed brass
[12:31:34] <IchGuckLive> ah then its useless
[12:31:36] <Gigs-> right now I use a carousel annealer but it's slow, I'd like to make a linear annealer that could roll the pieces through the flame in a line
[12:31:39] <archivist> chain based belt
[12:31:40] <IchGuckLive> forget all about ne
[12:31:44] <archivist> steel
[12:31:44] <IchGuckLive> O.O
[12:31:56] <Gigs-> it needs to be grippy enough to make the brass spin
[12:32:06] <archivist> use rollers then
[12:32:32] <Gigs-> hmm I'm not communicating well
[12:32:33] <archivist> rollers with a gear on the end on a rack
[12:32:54] <archivist> look up roller conveyor
[12:33:10] <Gigs-> I'll look but I think I know what you mean
[12:33:15] <Gigs-> the kind with the freewheel rollers on it right
[12:33:41] <archivist> no not free as you want the drive them
[12:33:50] <Gigs-> I think all of that would be too large
[12:34:07] <tjtr33> pz belts used in conveyors and farm equip
[12:34:08] <archivist> can be 1/2" bars
[12:34:39] <Gigs-> let me start over, the brass is small, rifle brass... I'm thinking a fixed plate with some kind of high temperature grippy surface, vertical or nearly so, a belt on the other side, rolling the cases across the fixed plate
[12:34:51] <Gigs-> all vertical
[12:35:40] <Gigs-> like imagine rolling a pencil between your palms
[12:35:46] <Gigs-> one palm moving the other still
[12:36:16] <archivist> that will fail due to bucking with the heat
[12:36:49] <Gigs-> what do you mean?
[12:36:54] <tjtr33> fixtufre brass cyl between 2 gears, run 2 chain drives over gears, one slow other fast will roll and go 'down thbe line'
[12:37:33] <Gigs-> that's similar to how the carousel annealer rotates them tjtr33, and it's iffy at best
[12:38:00] <Gigs-> the carousel has a fixed bottom plate with a tapered surface so that only one edge of the rim touches it
[12:38:19] <Gigs-> it works, more or less, but I'd prefer a much more solid way to rotate them than relying on friction on the bottom
[12:38:25] <tjtr33> gears wont slip when attached to cylinder
[12:39:00] <tjtr33> cyl and gears are 1 unit, driven between 2 chains
[12:39:27] <Gigs-> oh I think I see what you are getting at
[12:39:46] <Gigs-> a sort of "flighted" system with circulating platforms
[12:39:49] <tjtr33> wiat till i get this mind imaging thingy working ;)
[12:40:21] <archivist> at last he "sees"
[12:41:32] <Gigs-> well I'm not sure I see fully, in that I don't see how that could be accomplished with normal type chains
[12:41:51] <Gigs-> unless you mean for it to reciprocate back and forth
[12:42:17] <tjtr33> i'm just guessing that bike chains should be able to stand a few hundred degrees
[12:42:21] <Gigs-> sure
[12:43:02] <Gigs-> with heat sinking I can currently keep my carousel down to around 70C
[12:43:22] <Gigs-> but there will be exposure to hotter surfaces with local contact
[12:43:32] <tjtr33> not reciprocate inner oval chain at seed #1 outer at speed #1-x, the diff causes parts to spin, the faster rate is the thruput
[12:44:16] <Gigs-> you'd have to use some kind of double chain though because the gears can't jump the drive gear right?
[12:44:20] <tjtr33> two concentric oval drives, fixtured part pinned between them ( well 4 concentric, 2 up 2 down )
[12:44:43] <Gigs-> yeah
[12:44:49] <Gigs-> ok I think I fully see what you mean now
[12:45:09] <tjtr33> mind-o-vison mark II :)
[12:45:42] <Gigs-> so basically it's back to the carousel concept, except little driven islands
[12:45:50] <tjtr33> yah yah
[12:46:16] <Gigs-> what did you mean by bucking archivist
[12:46:28] <tjtr33> ?
[12:46:38] <Gigs-> his earlier comment about my idea
[12:46:45] <archivist> stuff kept hot can buckle
[12:47:02] <archivist> so a flat sheet wont be flat
[12:47:25] <Gigs-> oh, the fixed plate I am thinking would be metal with a thin high temperature grippy surface, some kind of silicone
[12:47:56] <tjtr33> gear tooth in chain is kindy grippy
[12:48:00] <Gigs-> hehe
[12:48:18] <Gigs-> your idea is good tjtr33 but it introduces feeding and unloading complexity
[12:48:25] <tjtr33> k
[12:48:26] <archivist> I wish I had taken pictures of bucked sheet and bars
[12:49:16] <tjtr33> we used to get mold inserts warped all the time, too long too thin...
[12:49:39] <tjtr33> from good heat treateing houses
[12:50:06] <archivist> out clock hand bluing plate had a nice bend
[12:50:10] <archivist> our
[12:51:12] <tjtr33> thats an idea, Gigs- go visit local heat treater
[12:51:33] <JT-Shop> to use the jog positive/negative input in Touchy do you connect the input to all the touchy jog positive pins and the axis you select is the one that will jog?
[12:51:34] <archivist> that was a brass sheet on a cooker hob with brass chips
[12:54:37] <archivist> the result I did take a picture
http://www.collection.archivist.info/showresult.php?prog=1&srcprog=searchv13.php&srcdata=title&Type=PD&Accn_no=525&subject=9813
[12:55:32] <tjtr33> gorgeous, need a penny for scale ( a a scale for scale )
[12:55:49] <tjtr33> or a 6" scale for scale
[12:56:00] <archivist> about 3 and 5" long
[12:56:08] <archivist> was years ago
[13:17:07] <PetefromTn> archivist: Hey man that is some truly beautiful work my CNC friend.
[13:17:25] <IchGuckLive> zeeshan: losing ground in sotchi
[13:18:24] <archivist> PetefromTn, no cnc involved on those, drill, saw, file, and the engraving was dremel
[13:18:55] <PetefromTn> Yes precisely which is why I commented that it was truly beautiful...
[13:20:09] <archivist> some others not all my work
http://www.collection.archivist.info/searchv13.php?searchstr=hands
[13:22:23] <Loetmichel> nice "art"
[13:23:00] <Loetmichel> but to be honest i would be MUCH to lazy to hammer something that detailled into form
[13:23:07] <Loetmichel> i would use a cnc mill ;-)
[13:23:54] <archivist> we started with a bit of spring steel for the first set
[13:34:36] <Gigs-> heh when you said bluing you really meant bluing
[13:34:57] <Gigs-> I guess that's more technically some kind of pickle isn't it?
[13:39:16] <archivist> no just heat to blue temperature
[13:40:20] <archivist> pickle methods are usually cold
[13:44:04] <R2E4_> afternoon
[13:46:12] <R2E4_> I found the issue.
[13:46:35] <R2E4_> or the issue that was causing the issue we were discussing this morning.
[13:46:36] <R2E4_> error 018
[13:50:08] <skunkworks> ?
[13:53:00] <JT-Shop> you gotta love craigslist... the guy that made me a low ball offer on my phase converter shows up to pick it up in a $50k F250
[13:53:37] <archivist> some people know how to get and stay rich
[13:54:17] <R2E4_> 18 inches from the terminal strip.
[13:54:28] <cradek> others buy $50k trucks
[13:54:50] <CaptHindsight> heh, I had a guy come for a visit a week ago that sold his co for $250M and wanted free work
[13:54:54] <R2E4_> I was working on the Y and axis and had the Z axis Vcmd and Vcgnd wires, which was not enabled.
[13:55:21] <archivist> :)
[13:55:25] <R2E4_> I connected it one terminal next to where it should be.
[13:55:39] <archivist> off by one error
[13:55:44] <JT-Shop> he is a retired tool and die maker
[13:56:04] <zeeshan> how much did you sell it for?
[13:56:12] <CaptHindsight> it's more like a compulsive disorder over collecting and holding money
[13:56:19] <R2E4_> I figured that would never work, so I am fixing that error....:-)
[14:02:45] <IchGuckLive> im off BYE
[14:05:34] <R2E4_> Going to return the p[ots to the place they were before, before going further.
[14:06:07] <R2E4_> sorry for that messup.
[14:21:15] <Belerafon> Hi. I want to rebuild linuxcnc kernel to add other modules. Where can I got sources of this kernel (is it special or "vanila"?) or linuxcnc kernel config?
[14:21:57] <cradek> you can get corresponding source with apt-get source
[14:22:24] <cradek> but to add new modules, you often don't rebuild the kernel, you build them against the headers package
[14:22:32] <cradek> depending on exactly what you're trying to do. elaborate?
[14:23:22] <JT-Shop> a friend is using 2.5 and when he touched off the G54 coordinate system the G code file started to run when he clicked OK
[14:23:36] <PetefromTn> R2E4_: Thats great news man maybe you can just get the pots put back where they were now and attempt some linuxCNC PID tuning...
[14:24:03] <cradek> JT-Shop: sounds like he should check his halui setup
[14:24:32] <Belerafon> I try to play with usb-can adapter. It need CONFIG_CAN=m CONFIG_CAN_RAW=m and etc.
[14:24:34] <JT-Shop> ok thanks
[14:24:42] <cradek> JT-Shop: just a guess...
[14:25:46] <CaptHindsight> Belerafon: will a car be a tool you'll be controlling?
[14:25:48] <cradek> Belerafon: you realize that won't be usable in realtime threads, right?
[14:26:42] <R2E4_> PetefromTn:, yeah, first I'll set the zero, then try with battery, make sure it don't shake.
[14:26:52] <PetefromTn> sure..
[14:27:00] <R2E4_> scared to touch it at this moment....lol
[14:27:37] <JT-Shop> cradek, I think I might have found the problem he was using in-not and I bet the button is N/O
[14:27:51] <cradek> eek
[14:27:54] <JT-Shop> so as soon as is-auto is true it runs
[14:28:24] <JT-Shop> halui.mode.is-auto
[14:28:24] <cradek> I'm pretty proud to have guessed this from your description
[14:28:35] <JT-Shop> I'm just as proud
[14:30:08] <cradek> Belerafon: I don't see those options in our kernel (2.6.32-122.35)
[14:31:13] <Belerafon> cradek: I try to install native drivers for usb-can module and I don't now about it realtime use
[14:34:09] <Belerafon> This is readme from driver
http://pastebin.com/mJWqnX9H
[14:37:35] <cradek> oh ok, I do see those settings
[14:37:46] <cradek> the only one not set how they suggest is CONFIG_CAN_CALC_BITTIMING
[14:38:10] <cradek> what are you trying to do?
[14:40:58] <Belerafon> I try to see config too :) And set CONFIG_CAN_CALC_BITTIMING and rebuild
[14:42:37] <Belerafon> Driver module can't load with "Unknown symbol in module" now and I try different ways
[15:11:18] <Belerafon> Can't got sources with apt-get:
http://pastebin.com/3SA6DZjv where is my sources? :)Can't get sources with apt-get:
http://pastebin.com/3SA6DZjv where is my sources? :)
[15:13:06] <Belerafon> I wait archive in /usr/src but nothing
[15:26:49] <CaptHindsight> Belerafon: it ends up in /usr/src/headers since Ubuntu does it wrong
[15:27:19] <CaptHindsight> or do they dare just to be different?
[15:28:29] <cradek> apt-get source linux-image-`uname -r`
[15:29:16] <R2E4_> ok, back to square one. With the onthe Y it shakes and axis moves 3 inches.
[15:29:36] <R2E4_> vibrates really hard
[15:30:01] <R2E4_> probably gain?
[15:30:23] <R2E4_> I turned the two potsdpown 1/2 turn and samething.
[15:32:26] <cradek> these are ac servos?
[15:32:47] <R2E4_> yeah
[15:32:55] <cradek> do you have reason to believe Y worked?
[15:33:17] <cradek> sounds like the commutation feedback is totally wrong or missing
[15:33:28] <R2E4_> Zeroed out Y turns pretty quick.
[15:33:45] <cradek> I don't understand what you mean by that
[15:33:54] <Belerafon> Problem solved after installation linux-ec2-source package. I have got archive
[15:34:00] <R2E4_> When it is not connected to 7i77 board.
[15:34:38] <cradek> ok I don't know what's going on. you lost me at "With the onthe Y it"
[15:34:39] <R2E4_> I apply 0V to Vcmd from Vgnd. The driver input.
[15:34:52] <R2E4_> I only have the Y enabled.
[15:35:51] <R2E4_> OV in the input to driver should not move but a little. This is moving pretty good. I can feel it on the ballscrew.
[15:36:20] <R2E4_> dand see the axis move. I tried turning the zero pot down a half turn and it didnt change.
[15:39:02] <cradek> that sure smells wrong
[15:46:28] <R2E4_> Theres obvious voltage or zero is off in the driver. With the 7i77 connected Vcmd and Vgnd, would hal scope be able to see it on a signal by sensing or it only know what it is sending out? Do I need a scope on the Vcmdand Vgnd?
[15:47:24] <cradek> if it moves crazily it's not just a problem with the offset
[15:50:07] <PetefromTn> Well that sucks... Drove to KnoxVegas today and got a big friggin truck pass in front of me before heading off the offramp and a big rock rolled off it and bounced into my windshield cracking it. My Freakin' Geico insurance only cover it up above our deductible so basically there is no coverage we have to pay for it ourselves..
[15:51:00] <R2E4_> tach?
[15:51:15] <cradek> ac servos usually don't have tachs
[15:51:34] <cradek> does yours?
[15:51:46] <R2E4_> These do, they go directly into the driver. I didnt touch that cable.
[15:52:43] <R2E4_> http://irmtl.com/vm40/schematics1.pdf
[16:02:08] <PCW> does the drive still creep with 0V in?
[16:03:10] <R2E4_> yes, been trying to zero it out by turning zero pot to no avail.
[16:04:26] <PCW> if it slowly creeps (and keeps creeping with low speed and high torque) commutation and tach are OK
[16:06:13] <PCW> so can you drive it in both directions with the battery?
[16:09:12] <R2E4_> terrible connection
[16:11:46] <R2E4_> This is what it is doing.
http://www.youtube.com/watch?v=laFgNCpfYmc&feature=youtu.be
[16:14:36] <PCW> i guess I would not worry too much about creeping at this stage of the game
[16:16:07] <PCW> thats about 10 mills/second
[16:16:26] <PCW> what is the full speed of that axis?
[16:17:32] <R2E4_> Max velocity of that axis is set at 50
[16:18:18] <PCW> I guess what i mean is what is it capable of (not 50 IPS I'm sure)
[16:19:59] <R2E4_> rapids at 590" /min
[16:20:48] <PCW> ok 10 mill per second is 1/1000 of full speed or about a 10 mV offset not a big deal
[16:21:30] <PCW> so I would not worry about it yet
[16:23:26] <PCW> It should be in the range of adjustment of the zero pot but actually the zero post should be
[16:23:28] <PCW> adjusted a different way with the 7I77 analog outputs enabled but set for 0V
[16:23:29] <PCW> But that can wait
[16:24:22] <R2E4_> ok, on to the shalking when the battery is applied. The two pots to lower loop gain.
[16:25:37] <R2E4_> the manual says tolower the loop gain when the motor vibrates when it is stopped or running.
[16:25:59] <R2E4_> VR3 and VR6
[16:26:48] <PCW> It may possibly be a pathological response to the step input in voltage
[16:27:22] <PCW> can you try a lower voltage (a couple resistors maybe)
[16:27:41] <PCW> or if you have a free potentiometer
[16:27:52] <Deejay> gn8
[16:30:51] <PCW> so you don't start out with such a bang!
[16:36:49] <R2E4_> Only have 250 k log
[16:38:26] <CaptHindsight> what are the simple text to g-code programs besides?
http://wiki.linuxcnc.org/uploads/engrave-11.py and
http://wiki.linuxcnc.org/uploads/engrave-lines-v3.zip
[16:43:01] <PCW> R2E4_ better than nothing
[16:45:18] <CaptHindsight> http://www.3ders.org/articles/20140218-mark-one-the-world-first-carbon-fiber-3d-printer-now-available-for-pre-order.html
[17:11:16] <CaptHindsight> I'll take anyones pre-order for only $8699ea :)
[17:12:42] <CaptHindsight> http://www.youtube.com/watch?v=Y5wjjDBdgeE
[17:14:10] <Einar_> R2E4_: A pot is a good idea. I tested my servo drive+motor with applying a voltage. It was fixed to the table. As it was commanded to go from 0 to 3000RPM immediately, it almost flipped the table, throwing components and stuff on the wall!
[17:14:26] <R2E4_> OK, have created a variable voltage devider. Can go from .007 to 1.5
[17:14:51] <Einar_> Wit a pot it ran nicely up in speed & down again.
[17:15:01] <PCW> or -.007 to -1.5
[17:15:32] <R2E4_> no, .007 to 1.5
[17:15:38] <R2E4_> yeah,
[17:15:41] <R2E4_> haha
[17:16:38] <R2E4_> Its not linear so, takes almost a full turn to go to .9, then 1/4 turn to get to 1.5
[17:17:13] <R2E4_> Just want to see it move in either direction, doesnt need to go fast,
[17:17:38] <R2E4_> not going tomove fast anyway, its onl;y set to max velocity 50.
[17:21:18] <tjtr33> how did you limit the max velocity in the amplifier? are you letting linuxcnc get in these tests?
[17:22:09] <MarkusBec> sound like a logaritmik pot
[17:22:18] <tjtr33> yep
[17:27:12] <R2E4_> ok , it jogs slowly in both directions with battery. get jopiont errors, but it does jog in both directions
[17:27:39] <tjtr33> great!
[17:28:32] <PetefromTn> AWESOME!!
[17:28:38] <tjtr33> if you want faster, you can scale the Vcmd, theres a pot for that. use with caution!
[17:30:18] <R2E4_> or connect to LinuxCNC, andjog axis with arrow keys
[17:32:09] <R2E4_> I only inputted about .1v
[17:32:34] <R2E4_> I think I have a linear pot at home, I can bring.
[17:33:18] <PetefromTn> Man you will be joggin axes before you know it now LOL...
[17:34:36] <R2E4_> PetefromTn: does your machine home by itself?
[17:34:50] <PetefromTn> Whaddya mean by itself?
[17:35:08] <PetefromTn> When I first turn it on I hit home all and it auto homes all three axes one at a time..
[17:35:22] <R2E4_> press a button and off it goes and homes itself when it hits its home limit switches
[17:36:37] <PetefromTn> Yeah first the Z goes up and then hits the home kinda medium speed then backs off and goes again slower and then goes to the Home offset location I programmed. Then the X and Y the same way.
[17:38:44] <R2E4_> So theoretically, I should be able to connect the 7i77 to the driver, and jog the machine.
[17:39:19] <PCW> assuming the feedback is in the correct direction
[17:41:03] <R2E4_> I think it is. When the y axis is moving towards me the dro is going negative towards the Y over travel - limit switch
[17:42:00] <R2E4_> How do I stop the creaping though.
[17:53:58] <PCW> once feedback is enabled it will stop creeping
[17:56:56] <Stratrascal> Hey zeeshan! you are her a lot :)
[17:57:08] <Stratrascal> here
[18:02:12] <Stratrascal> So, I have another stupid question, You guys must be getting sick of me.... What is "Rotary Angle"?
[18:02:35] <JT-Shop> for a rotary axis
[18:02:51] <Stratrascal> Thanks.
[18:02:55] <JT-Shop> or rotating the coordinate system about Z
[18:03:13] <Stratrascal> I only have a 3 axis machine so I don't have to worry about it.
[18:03:21] <JT-Shop> yup'
[18:03:37] <Stratrascal> Oh I dont understand your second answer
[18:03:58] <Stratrascal> I think that may be it. It is asking me that in my wizzard
[18:04:42] <Stratrascal> Let me rephrase, ouput rotary angle
[18:05:01] <Tom_shop> ignore it, it doesn't apply to normal 3 axis
[18:05:17] <Stratrascal> Great, thanks Tom and JT.
[18:05:29] <Stratrascal> Back on my adventure. I will check in later
[18:05:44] <Tom_shop> unless you're rotating geometry about an axis in your cad file
[18:06:17] <Stratrascal> And wouldnt you know it, the very next thing I dont understand.
[18:06:35] <Stratrascal> Contour Ramping Output, Line moves and Arc moves
[18:07:13] <Stratrascal> Does that mean that when ramping it moves in an arc or straight line?
[18:07:14] <Tom_shop> will your machine accept arc ramp moves? if not it converts them to short line segments
[18:07:24] <Tom_shop> or what you said
[18:07:45] <Tom_shop> you can straight line ramp or arc
[18:07:55] <Tom_shop> i've always used straight lines
[18:07:56] <Stratrascal> Oh, Well I don't know I use Mach3
[18:08:16] * Tom_shop wipes his hands clean
[18:08:25] <Stratrascal> Sorry tom.
[18:08:30] <Stratrascal> What can I say.
[18:09:12] <Stratrascal> But thanks for the answers anyway.
[18:13:02] * JT-Shop hates to take the vise off the BP but it is in the way I'm afraid
[18:13:30] <Tom_shop> aww
[18:13:37] <Tom_shop> what you workin on now?
[18:13:53] <Stratrascal> Me?
[18:13:59] <Tom_shop> jt
[18:14:04] <Stratrascal> Sorry
[18:14:25] <Tom_itx> i'm just givin ya crap about mach
[18:14:57] <Stratrascal> Hey, I am asking advice in a room called Linuxcnc, I expect it
[18:15:06] <Stratrascal> :)
[18:15:21] <JT-Shop> I have two broken off bolts in the idler...
[18:15:32] <Tom_itx> ick
[18:15:42] <Tom_itx> no easy outs big enough?
[18:15:47] <Tom_itx> and a torch
[18:15:49] <Stratrascal> when taling about arc slowdown does the percentage mean it will go that percent or slow down that much?
[18:16:57] <Stratrascal> So if it give it an answer 90 will it go 90percent of the speed or 10 percent?
[18:17:03] <Tom_itx> i would interpret it as the percentage of slowdown
[18:17:08] <JT-Shop> they are 1/4-20 SHCS that hold the cover over the seals
[18:18:07] <Stratrascal> Oh my bad, the default is 100, so that must mean not to slow down at all
[18:18:41] <tjtr33> hope ctr is soft, then use ezout, else edm ( tap buster not real edm )
[18:27:50] <JT-Shop> dang thing must be hollow... I could lift it up onto the mill
[18:32:25] <CaptHindsight> tjtr33: these are about the cruelest 3d printed shoes I could find
https://tinyurl.com/lu2g2w5
[18:35:00] <tjtr33> hahaha
[18:35:37] <tjtr33> some wimmen would pay big for that ( we'd say the plastic injection machine just puked )
[18:36:04] <tjtr33> dang sprue didnt shut off!
[18:36:11] <CaptHindsight> 3d printed shows $49.99, 3d printed mistakes 499.85
[18:36:20] <CaptHindsight> shows/shoes
[18:40:26] <Stratrascal> Those shoes rock!
[18:40:43] <Stratrascal> and I am a guy who doesent give a rip about shoes
[18:49:48] <CaptHindsight> our hosting co is overloaded today with cleaning malware, who are some good hosts and colocators?
[18:52:49] <CaptHindsight> anyone use
http://www.wiredtree.com/?
[18:57:26] <PetefromTn> Wouldn't ya know it..... I got a lady who wants to buy my Van I posted on Craigslist and now we can't find the title anywhere in the damn house... TYPICAL!!
[18:58:08] <CaptHindsight> heh, probably will turn up after she leaves
[18:59:51] <PetefromTn> YUP...probably. I need to get my shit together man LOL
[19:00:37] <CaptHindsight> unless you're overtired and it's right in front of you
[19:01:10] <CaptHindsight> you know it's time to put the tools down when you can't find the tool right in front of you
[19:29:20] <PetefromTn> yeah well this is a title not a tool. The tool lost the title hehe..
[19:41:37] <zeeshan> Stratrascal: im on reading week so im free a lot ;p
[19:41:41] <zeeshan> hence why im active :P
[19:42:37] <zeeshan> got the ball screws
[19:42:46] <zeeshan> feel no play between the ball screw and the ball nut at various locations
[19:43:07] <zeeshan> won't have a video of the dial gage till i have it mounted on the lathe
[20:00:41] <zeeshan> http://www.youtube.com/watch?v=R6oHptAnqtA
[20:00:52] <zeeshan> is a machine like this using a brake to index the spindle?
[20:00:58] <zeeshan> er
[20:01:22] <zeeshan> using an econder to know the spindle position and a brake to hold it in position?
[21:03:18] <R2E4_> evening!!!!
[21:03:37] <skunkworks> R2E4_: any luck?
[21:04:07] <R2E4_> Well. it isstill creeping. but can jog.
[21:04:16] <R2E4_> http://www.youtube.com/watch?v=KPAFP-rLg0k&feature=youtu.be
[21:04:25] <R2E4_> IS this too much to add offset?
[21:05:31] <skunkworks> that doesn't make sense.. that is a lot of drift...
[21:07:50] <R2E4_> here it is with the input to the driver at 0 volts by shorting Vgnd to Vcmd.
http://www.youtube.com/watch?v=laFgNCpfYmc
[21:10:13] <R2E4_> yeah, it seems to me it is. The manual says "adjust VR1 for zero but, However a drift cannot be adjusted."
[21:11:17] <R2E4_> The wires are shielded and the shield grounded on the 7i77 side, like pcw said todo.
[21:16:41] <R2E4_> It drifts faster when the 7i77 is connected than with Vgnd kand Vcmd are connected.
[21:23:42] <R2E4_> Could this be something wrong with the encoder and the sevo correcting it allways in the same direction? I would think I linuxcnc would report an ecoder error though.
[21:24:11] <R2E4_> The X was driftiung when I had it connected also.
[21:32:12] <skunkworks> odd
[21:33:02] <skunkworks> that is a pretty slow creep when grounded...
[21:33:38] <skunkworks> are you sure that the mesa and computer hardware is grounded good?
[21:34:38] <skunkworks> and - are you sure the feedback is correct? If you had really low tuning and backwards feedback - you would get drift..
[21:35:40] <R2E4_> The tuning is what Pncconf set. P=1 and the rest to 0
[21:36:12] <skunkworks> ok - so really soft tuning.. did you try negating the SCALE in the ini?
[21:36:23] <skunkworks> for grins
[21:36:34] <R2E4_> the Y gantry is moving in the correct direction according to the dro which is fed via encoder.
[21:36:50] <skunkworks> ok - to negate the output scale..
[21:37:03] <skunkworks> *ok - then negate the output scale on y.
[21:37:15] <R2E4_> I haven't started tweaking the PID
[21:37:21] <skunkworks> and see if it either - runs away faster or stops
[21:37:25] <R2E4_> I dont know what I am doing in there yet
[21:38:03] <skunkworks> sure.. but - I think with a P=1 imho - it should not drift faster than grounded amp inputs...
[21:39:55] <Tom_itx> with lcnc, given the encoder counts and scale of the drive you don't need any pid tuning constants to offset the PID values do you?
[21:40:02] <Tom_itx> i hope i said that right...
[21:40:27] <Tom_itx> just the direct PID values
[21:40:42] <Tom_itx> with no offsets for drive / encoder mismatch
[21:40:59] <pcw_home> I think you have the feedback reversed (and output scales set so you cant get more than a fraction of a volt output)
[21:41:22] <skunkworks> ^I agree..
[21:42:06] <pcw_home> so what you see is a slow runaway
[21:42:25] <skunkworks> well - P=1 would give you pretty slow run away - wouldn't it?
[21:42:36] <skunkworks> it seems to be slowly accellerating
[21:42:42] <skunkworks> as much as you can tell
[21:43:06] <pcw_home> not when it gets any distance away from null...
[21:44:02] <pcw_home> if its all in inches and volts then a PID error of 1 inch is 1 volt
[21:44:28] <pcw_home> ( with P of 1)
[21:45:51] <R2E4_> IF A channel of encoder leading pulse is before B then it goes one way, if opposite it goes the other way. If it were reveres wouldn't it drift in the other directioon?
[21:46:13] <pcw_home> no
[21:46:39] <R2E4_> Well, that answers my question.
[21:46:42] <R2E4_> lol
[21:46:44] <pcw_home> that will change the count directions. if its correct leave it alone
[21:47:16] <PetefromTn> Honestly I think that there is nothing wrong with the drives since he said earlier that both the X and Y drift, do they drift in a similar manner?
[21:47:22] <pcw_home> Is the DRO correct?
[21:48:01] <R2E4_> yes, it seems to be. 1onthe DRO euates to 1 inch on the axis. or real close
[21:48:27] <pcw_home> I mean the sign (which direction is plus)
[21:48:48] <pcw_home> thats the first thing to get right
[21:49:01] <R2E4_> yes,that is correct.
[21:49:23] <pcw_home> OK so you are done with the encoders
[21:49:45] <pcw_home> I think you have the feedback backwards
[21:50:08] <pcw_home> so the output polarity needs to be reversed
[21:50:18] <pcw_home> This is done in HAL
[21:51:15] <pcw_home> (as a quick test you can try setting P to -1) but this can only be done as a test
[21:51:34] <R2E4_> ok
[21:52:25] <pcw_home> hand on estop...
[21:52:38] <R2E4_> Should I be leary of anything crazy that could happen?
[21:52:48] <pcw_home> always
[21:53:05] <PetefromTn> I tell ya man I would uncouple the ballscrew until you get that motor locked down.
[21:53:07] <skunkworks> but it still should trip out with a following error.
[21:53:29] <skunkworks> but still - hand on the estop
[21:55:58] <R2E4_> IT worked. YEAH!
[21:59:01] * skunkworks goes off to bed... Good luck!
[22:00:40] <PetefromTn> Drift stopped?
[22:01:07] <PetefromTn> So now he needs to reverse it in the Hal file I guess?
[22:01:46] <Tom_itx> sounds like that's what he did
[22:02:09] <pcw_home> well negative P is not a great idea
[22:03:01] <R2E4_> reverse the min and max output? -10 to 10 and 10 to -10?
[22:03:09] <pcw_home> setting hm2_5i25.0.7i77.N.1.analogout0-scalemax to a negative number is the right way
[22:03:50] <pcw_home> sorry hm2_5i25.0.7i77.0.1.analogoutN-scalemax
[22:05:04] <pcw_home> dont change minlim or maxlim
[22:05:33] <pcw_home> and then P back to 1
[22:19:05] <R2E4_> setp hm2_5i25.0.7i77.0.1.analogoutN-scalemax [AXIS_1] OUTPUT_SCALE - ?
[22:35:48] <Tom_itx> the scale would be the ticks per rev?
[22:36:43] <Tom_itx> mm i'm not sure about OUTPUT_SCALE
[22:39:05] <R2E4_> oops, missed the 10. should read setp hm2_5i25.0.7i77.0.1.analogoutN-scalemax [AXIS_1] OUTPUT_SCALE -10
[22:39:41] <Tom_itx> if it's -10 and you want to reverse the direction it would be 10
[22:39:58] <Tom_itx> make sure to change P to a positive number too
[22:40:22] <R2E4_> it defaults to + I imagine, cause there is nothing, so just adding the -10 should reverse it.
[22:40:46] <Tom_itx> i thought you tested with a negative P just a bit ago
[22:41:41] <Tom_itx> anyway... time for zzzzz's
[22:42:15] <R2E4_> I'll wait for pcw's word....hehe
[22:51:17] <PetefromTn> Sounds like you are finally getting somewhere man Congrats!
[22:53:52] <R2E4_> I dont understand the HAL good enough yet.
[22:56:59] <R2E4_> Thanks
[22:57:12] <R2E4_> Have to do the same with the X and Z
[22:58:04] <PetefromTn> Yeah but that should be easy now that you know what to look for.
[22:58:25] <PetefromTn> I am trying to draw my next part in freecad. UGH I suck at 3d drawing...
[23:01:02] <R2E4_> freecad does surfaces?
[23:02:17] <PetefromTn> yeah I guess so LOL..
[23:03:27] <R2E4_> yeah, cool an open source full blown parametric modeler.
[23:03:29] <zeeshan> are you doing solid modeling?
[23:03:34] <zeeshan> or surface modeling
[23:03:49] <zeeshan> autodesk alias is amazing for surfaces
[23:03:57] <zeeshan> im making a wheel in it right now
[23:05:36] <PetefromTn> yeah it is a real parametric modeler and it is free.... Go figure.
[23:05:52] <zeeshan> why dont you use a real software ;p
[23:06:03] <zeeshan> like solidworks or inventor or nx or soliedge
[23:07:19] <zeeshan> they are trying to teach us alias right now
[23:07:28] <zeeshan> apparently all the car companies do their sketches in it
[23:07:47] <PetefromTn> When you can afford to buy those programs for real let me know okay.
[23:07:56] <zeeshan> theyre fairly cheap!
[23:08:03] <zeeshan> ... for a student :)
[23:08:05] <PetefromTn> yeah man
[23:08:12] <zeeshan> but inventor license is like 2k
[23:10:34] <zeeshan> SolidWorks is sold in three separate versions: Premium ($7,995.00), Professional ($5,490.00), and Standard ($3,995.00) -- not exactly personal use pricing. The costs involved take it out of the range of most common users but Dassault Systems does offer a reduced price ($150.00) Student Version f
[23:10:39] <zeeshan> cost of solidowrks 2013
[23:11:50] <R2E4_> I have inventor student version. IT knows and there are differences when your using student version.
[23:12:05] <zeeshan> i'm using the real version
[23:12:08] * zeeshan hides
[23:12:56] <zeeshan> eaton must have a lot of $ for software
[23:13:12] <R2E4_> I use the real autocad at work.
[23:13:12] <zeeshan> about 1000 computers are the two locations i was at had inventor
[23:13:20] <zeeshan> why
[23:13:28] <zeeshan> for what
[23:13:30] <R2E4_> Why what?
[23:13:38] <R2E4_> Electrical drwings
[23:13:45] <zeeshan> ah makes sense. eaton did too
[23:13:49] <zeeshan> it's good for that stuff
[23:13:59] <zeeshan> i thought you were using it for mechanical stuff
[23:14:21] <zeeshan> my friend who does hvac design uses autocad at his place still..
[23:14:26] <zeeshan> that'd make me go crazy
[23:15:28] <R2E4_> design hvac is fun, I do that and card access for Ingersoll Rand
[23:15:45] <zeeshan> using autocad?
[23:16:05] <R2E4_> yes
[23:16:09] * zeeshan cringes
[23:16:45] <zeeshan> i will never go back to autocad ever again. 2 years of using that was enough
[23:17:18] <zeeshan> that being said, i made some schematics in autocad electrical
[23:17:35] <zeeshan> which interact with inventor's routing features
[23:17:38] <zeeshan> it's pretty damn cool
[23:19:28] <PetefromTn> There is nothing wrong with autocad based drawings for 2d stuff it has its place as does the 3d modelers. I have used Solidworks and have played with alibre design which is supposed to be similar and honestly you would be amazed at just how good freecad is getting.
[23:20:04] <R2E4_> I am going to try freecad.
[23:20:08] <zeeshan> one major flaw that makes me not even want to use autocad for 2d stuff
[23:20:11] <zeeshan> is it's not parametric.
[23:20:18] <zeeshan> and that's a major pitfall
[23:20:20] <R2E4_> Can you export to stll with freecad?
[23:20:39] <R2E4_> yes it is parametric
[23:20:46] <zeeshan> PetefromTn: i haven't tried freecad but i don't think ill be using it in indusutry
[23:21:01] <zeeshan> in 2011 it wasnt
[23:21:02] <R2E4_> FreeCAD is a parametric 3D modeler. Parametric modeling allows you to easily modify your design by going back into your model history and changing its parameters
[23:21:37] <PetefromTn> yeah you can export >stl as well as several other formats.
[23:22:01] <PetefromTn> Import SW drawings etc..
[23:22:13] <zeeshan> yea but you always loose the feature/history tree
[23:22:25] <zeeshan> cause thats always propietry to the software
[23:22:40] <R2E4_> Can you access surfaces though when you import SW.
[23:22:43] <PetefromTn> They are working on it all the time you should check out their forum.
[23:22:52] <R2E4_> Inventor doesnt recognize surfaces from SW
[23:23:05] <zeeshan> standardization requires all cad software to be able to export to a universal format
[23:23:29] <zeeshan> but sometimes it breaks
[23:23:50] <PetefromTn> I have a drawing I once drew in SW and imported it and was able to work with it but he is right SW makes it difficult. I am kinda new to the 3d stuff so I am not the best to answer these questions.
[23:23:50] <zeeshan> like i've exported in iges, step, parasolid formats from solidworks and imported to inventor
[23:24:08] <zeeshan> and the part volume goes from 50m^3 from example, to 50.04232134321332123m^3
[23:24:15] <zeeshan> its so irritating
[23:24:58] <zeeshan> and by the time you bring it into ansys cfx for fluid simulation, the geometry is no longer what it began with
[23:25:16] <zeeshan> it's the same shape, but the nominal dimensions aren't nominal
[23:25:25] <R2E4_> I need the nextengine scanner
[23:26:02] <zeeshan> is it a laser scanner?
[23:26:15] <PetefromTn> http://imagebin.org/294363 Heres my model of a part someone else made so far.
[23:26:33] <zeeshan> PetefromTn: how long did that take
[23:26:47] <zeeshan> looks like a 10 min job in inventor/solidworks
[23:26:52] <PetefromTn> About twenty minutes just sitting here chatting with you guys and watching TV.
[23:27:08] <PetefromTn> Oh I am not saying it is complex whatsoever.
[23:27:18] <zeeshan> you're saying it can be done :)
[23:27:27] <PetefromTn> But it works and is easy to use.
[23:27:29] <zeeshan> http://www.capture3d.com/products-ATOS-TripleScan.html
[23:27:34] <PetefromTn> fully parametric.
[23:27:34] <zeeshan> this is the latest technology in 3d scanning
[23:27:38] <zeeshan> prof is developing software for it
[23:27:44] <zeeshan> its literally blue light (not laser)
[23:28:06] <zeeshan> in one snap shot, and using a complex algorithm, it imports a 3d software of the whole part into the software
[23:28:08] <zeeshan> crazy.
[23:29:10] <PetefromTn> Jeez man I saw 3d scanners like that almost six years ago at the IWF show...
[23:29:18] <zeeshan> basically if they can working to the point its reliable, companies like ford want their manufacturers ie a waterpump manufacturer
[23:29:33] <zeeshan> to povide the "inspection 3d model" for every part that is recieved by forward
[23:29:34] <zeeshan> *ford
[23:29:58] <PetefromTn> R2E4_: Hey man did you get the axis sorted out yet?
[23:30:07] <zeeshan> PetefromTn: laser scanners have existed since even longer than that, so its fairly new ;p
[23:30:11] <R2E4_> http://www.nextengine.com/
[23:30:31] <R2E4_> That one is nice and onl;y 3,000
[23:30:41] <R2E4_> plus the software
[23:30:56] <zeeshan> if jay leno says its good
[23:30:57] <zeeshan> its good!
[23:31:02] <PetefromTn> Software is what kills ya probably.
[23:31:03] <zeeshan> he's the man,.
[23:31:17] <R2E4_> Yeah, the axis is working. Have to figure out the accel, and it takes a while to stop....
[23:31:33] <zeeshan> it comes with a standard software
[23:31:55] <R2E4_> HEy pete, can I study your hal files?
[23:32:02] <PetefromTn> Yeah it takes awhile to get everything working and setup. I spent a couple weeks getting my axes dialed in and they still need work.
[23:32:57] <PetefromTn> R2E4_: Yeah man but I gotta get them outta the machine here and I am about to head to bed. I got some stuff to do in the morning to hopefully sell this van of mine but after that I will get it off and zip it over to you.
[23:35:00] <zeeshan> http://i130.photobucket.com/albums/p252/turbozee84/Concept2_zps58b0fe0b.png
[23:35:10] <zeeshan> doesnt the model look vibrant
[23:38:01] <R2E4_> ok, thanks. I'm leaving also. BYE Bob
[23:42:06] <zeeshan> PetefromTn: looking more into freecad
[23:42:10] <zeeshan> it looks good!
[23:42:46] <PetefromTn> Why? I thought I you wanted more profressional programs LOL.
[23:42:56] <zeeshan> PetefromTn: i'm not against it
[23:43:05] <zeeshan> im just saying ill never use it at work because they'll never use it
[23:43:17] <zeeshan> so whats the point of learning it right?
[23:44:04] <zeeshan> i like the fact you can run python scripts in it
[23:44:33] <PetefromTn> From what I have seen most all the parametric modelers work pretty much the same with subtle differences. However if you have no intention of using it why bother. It takes time to become proficient with any software after all..
[23:46:33] <zeeshan> i agree they're similar but at the same time they're also different. i've used 4 parametric software very thoroughly, solidworks, inventor, nx, and ansys. i personally love to use symmetry in my designs, and if i can get away with make 1/4 of something, ill do that.
[23:46:34] <PetefromTn> Honestly I am a small shop making niche market parts and as long as I can take a customers 3d drawing and do something with it I am good. It works for me for much the same reason that I can get away with using something like CamBam instead of Mastercam or whatever because at this point in the shops progression I am not making anything that requires amazing toolpaths or super speed. The machine just started working for me
[23:46:34] <PetefromTn> about a couple months ago and I am trying to find work for it now.
[23:46:40] <zeeshan> now in solidworks, and inventor i can do that very easily
[23:46:55] <zeeshan> but holy shit, in nx, its SO much more effort, that its easier just to make the whole part instead of 1/4 of it
[23:47:01] <zeeshan> and in ansys -- just forget about it
[23:47:03] <zeeshan> its terrible
[23:47:21] <zeeshan> i understand
[23:47:23] <PetefromTn> Freecad has symmetry constraints and I use them frequently.
[23:47:31] <zeeshan> no i don't mean symmetry constraints
[23:47:42] <zeeshan> if you're trying a box for example
[23:47:54] <zeeshan> and you want to align the mid points of the vertical and horizontal lines with the orign
[23:48:00] <zeeshan> how easy is that to do in freecad?
[23:48:27] <zeeshan> in inventor and solidworks its 3 clicks, in nx its about 14 clicks, in ansys its not possible
[23:48:47] <PetefromTn> pretty easy from the videos I have seen. They are working on a module now that will allow complex multi part drawings
[23:49:02] <PetefromTn> The parts in the drawing will be assembly style setup.
[23:50:00] <PetefromTn> It is not SW or NX or those others but man it is free and works pretty damn well. When I first downloaded it I had crashes a lot now it almost never crashes and I can draw pretty much whatever I want.
[23:50:21] <zeeshan> yes
[23:50:38] <PetefromTn> Best part is if you take the parts into something like Blender and analyze the solids it looks good in other programs too for the most part.
[23:51:34] <zeeshan> you know what sucks?
[23:51:53] <zeeshan> when solidworks is dealing with very large assemblies (500+ parts)
[23:51:57] <PetefromTn> Not being able to afford anything I want?
[23:52:00] <zeeshan> it gets extremely slow on my machine
[23:52:06] <zeeshan> but nx handles it like it is a joke
[23:52:14] <zeeshan> PetefromTn: lol
[23:52:22] <zeeshan> PetefromTn: in 1 year you'll be too cool for irc
[23:52:36] <PetefromTn> Hell I am already too cool for it...
[23:52:45] <PetefromTn> ;)
[23:52:59] <zeeshan> in 2 years you'll be working so much on cad stuff that your clients will be requesting you to see their solidworks designs and you'll cave in to buying it
[23:53:06] <zeeshan> cause you'll be making $$$$$$$$$$$$$$$$$$$$$$44
[23:53:15] <zeeshan> :D
[23:53:53] <zeeshan> my friend whos a machinist started his machine shop when i was starting college
[23:53:56] <PetefromTn> We'll see..
[23:54:03] <zeeshan> after 2 years, he bidded on some big jobs
[23:54:15] <PetefromTn> Right now I would just like to be able to make a decent living with this stuff.
[23:54:23] <zeeshan> 7 years since he started, he's making crazy money
[23:54:31] <zeeshan> cause he's doing work for big companies
[23:55:51] <PetefromTn> http://imagebin.org/294370 got it champfered. Still needs a lot of work to appear real. There is also a bunch of rendering options like you showed in your picture before.
[23:56:01] <zeeshan> what is that
[23:56:11] <zeeshan> looks like an airplane truss
[23:56:15] <PetefromTn> Scope rail for precision airgun.
[23:56:19] <zeeshan> oh
[23:56:20] <zeeshan> haha
[23:56:26] <zeeshan> looks nice man
[23:56:56] <PetefromTn> Its dirt simple, I did not design it just gonna make something similar but better looking for a customer.
[23:57:40] <zeeshan> are the holes functional
[23:57:45] <zeeshan> or speed holes (ie to remove weight)
[23:57:50] <PetefromTn> Both.
[23:57:57] <zeeshan> the bottom 5 holes look like mounting holes
[23:57:59] <zeeshan> the big holes
[23:58:02] <zeeshan> is what im talking about
[23:58:17] <PetefromTn> Big ones are just for fun small ones are for screws to clamp onto the gun.
[23:58:28] <zeeshan> maybe you can make some fancy shape on there? :D
[23:58:30] <zeeshan> like a dragon!
[23:59:04] <PetefromTn> Hell I can make about anything now that the machine is working but it has to be tasteful and still be quick to machine.
[23:59:36] <PetefromTn> I can talk all night but I gotta get to bed here or I will not be able to wake up tomorrow LOL...
[23:59:42] <zeeshan> haha cya man
[23:59:46] <PetefromTn> Did not realize it was so late.
[23:59:53] <zeeshan> i gotta be up for 7am ;[
[23:59:53] <PetefromTn> Gn8
