#emc | Logs for 2011-02-04

[00:03:59] -!- theorb [theorb!~theorb@] has joined #emc
[00:04:15] -!- theorbtwo has quit [Ping timeout: 260 seconds]
[00:04:22] theorb is now known as theorbtwo
[00:10:03] -!- servos4ever has quit [Quit: ChatZilla 0.9.85 [SeaMonkey 2.0.11/20101206162726]]
[00:31:46] -!- Emcrules_Laptop has quit [Ping timeout: 272 seconds]
[00:39:17] -!- skunkworks [skunkworks!~chatzilla@str-bb-cable-south2-static-6-412.dsl.airstreamcomm.net] has joined #emc
[00:46:56] <danimal_garage> so when setting up max and default velocity and acceleration in the traj section of the ini, should they be higher than the values for each axis
[00:46:57] <danimal_garage> ?
[00:47:12] <danimal_garage> i set them to the same and it ferrors
[00:53:05] <danimal_garage> ah i accidentally rasied the servo thread period instead of lowering it
[00:53:09] <danimal_garage> whoops
[00:56:41] <danimal_garage> weird, it doesnt throw a ferror when it's jogging at max vel, but it does when a program is runnng now for some reason
[00:56:44] <danimal_garage> in the z axis
[00:58:55] <danimal_garage> hmm i i think it was the power line for the spindle motor that was too close to the feedback cable for the axis
[00:59:05] <danimal_garage> i moved it and it didnt ferror
[00:59:51] <danimal_garage> weird
[01:00:15] <danimal_garage> easy fix at least
[01:02:18] -!- Emcrules_Laptop [Emcrules_Laptop!~Jean-Paul@CPE001310e6c94f-CM001bd71cb794.cpe.net.cable.rogers.com] has joined #emc
[01:05:42] -!- andypugh has quit [Quit: andypugh]
[01:08:43] -!- Dannyboy has quit [Remote host closed the connection]
[01:42:14] -!- dimas_ has quit [Ping timeout: 255 seconds]
[01:45:14] -!- adb has quit [Ping timeout: 272 seconds]
[01:51:04] -!- adb [adb!~Moldovean@178-211-232-101.dhcp.voenergies.net] has joined #emc
[02:21:31] <Emcrules_Laptop> pcw_home, How hard is it to modify the 7i49 to accept DC excited resolvers? Apparently the hiperface encoders on my servo motors also output a differential analog sin/cos. I would like to try to use these motors with a 7i49 and the 8i20
[02:28:28] -!- ries has quit [Read error: Connection reset by peer]
[02:42:24] -!- rooks has quit [Ping timeout: 276 seconds]
[02:42:28] <KimK> Emcrules_Laptop: DC excited resolvers? Does that even work? Tell me more.
[02:44:50] -!- adb has quit [Ping timeout: 246 seconds]
[02:44:54] <Emcrules_Laptop> Its actually a absolute hiperface serial encoder but also has analog sine cosine output
[02:45:34] -!- adb [adb!~Moldovean@178-211-232-101.dhcp.voenergies.net] has joined #emc
[02:46:43] -!- ries [ries!~ries@] has joined #emc
[02:47:09] <KimK> Interesting, do you have a link?
[02:48:14] <skunkworks> KimK: how is it going?
[02:49:07] <skunkworks> heh - scary - I just copied and pasted some text from a web page and it added the link to the paste buffer.
[02:49:24] <Emcrules_Laptop> http://www.stegmann.com/product/servo/servo_all.html look at the sks/m 36
[02:49:29] <atmega> I can get in another 10 or 12 hours of caving before you think about diving again
[02:49:40] <KimK> Hi Sam. No big news here. How are things there? Ha, are you running one of those pastebin plugins?
[02:49:41] <atmega> <urk>
[02:50:33] <skunkworks> no that I know of
[02:50:42] <skunkworks> http://www.engineeringtalk.com/news/rea/rea165.html#ixzz1CxEzMnbs
[02:50:51] <skunkworks> copy something off of that page
[02:50:56] <PCW> I dont think the 7I49 is terribly suitable because you would need a much higher sample rate for the sine/cosine encoders
[02:50:57] <PCW> I think the 7I49 can only sample at 40 KHz or so (6 channels) but maybe could be sped up as you dont have or need as much interpolation accuracy
[02:51:12] <skunkworks> (maybe it is just me)
[02:51:39] <kb8wmc> when wiring home switches is it preferable to wire as normally open or normally closed?
[02:51:40] -!- rooks [rooks!~rooks@102-bem-18.acn.waw.pl] has joined #emc
[02:52:04] <Emcrules_Laptop> normaly closed IMO
[02:52:35] <kb8wmc> Emcrules_Laptop: rgr....
[02:52:46] <atmega> NC for limits, can't see that it matters for dedicated homes?
[02:52:51] <Emcrules_Laptop> That way contacts are forced open logic can be inverted
[02:53:20] <kb8wmc> ok
[02:53:32] <KimK> skunkworks: Yes, I copied a paragraph and it added the page url onto the end of the copied text.
[02:53:43] <skunkworks> ok - so it isn't just me :)
[02:54:32] <KimK> No, must be a javascript thing, but I'm like you, I consider it unwanted help.
[02:54:46] <skunkworks> exactly
[02:54:53] <skunkworks> KimK: nothing new lately
[02:55:43] <KimK> I could put this up, let me paste a very small image, brb...
[03:00:12] <KimK> http://imagebin.org/136055
[03:00:22] <Emcrules_Laptop> PCW: do you have time to discuss 8i20 inverter circuit
[03:06:55] <KimK> skunkworks: How often are the temps the same from Dallas to Minneapolis? I'll be glad when we're done with this. Hope you're staying warm there. (And that your K&T doesn't freeze solid.)
[03:07:42] <KimK> skunkworks: So are you going to build a strain gauge probe?
[03:07:47] <Emcrules_Laptop> KimK: yeah this weather is getting old allready
[03:10:30] <KimK> Yes, I'm tired of it for sure. And a month and a half to go yet! (To me Spring arrives in mid-March. Your um, Springing may vary.)
[03:11:07] <Emcrules_Laptop> Nope same as you. Im just north of toronto
[03:11:54] <KimK> So are you running an 8i20?
[03:12:11] <Emcrules_Laptop> just got one and want to play with it
[03:12:59] <KimK> Nice, have you got a motor to match? (If so, that would be a big motor.)
[03:13:20] <Emcrules_Laptop> just wondering if some how it could figure out to do sensorless commutation
[03:13:25] <Jymmm> KimK: Yeah, I hate the bitter cold weather too... http://i55.tinypic.com/2ue1idl.jpg
[03:13:36] <Emcrules_Laptop> yes i have a range of motors i can use
[03:14:06] <Emcrules_Laptop> 100W, 1KW, 2KW
[03:14:57] <Emcrules_Laptop> just kind of a pain in the ass as they all have absolute hiperface encoders on them
[03:15:09] <KimK> Emcrules_Laptop: Possibly, if you have an absolute position sensor on it. Like a resolver. And so that's why you're asking about absolute position sensors?
[03:15:36] <Jymmm> KimK: Having to wear a tshirt, jeans, and sneakers is dreadful
[03:16:50] <KimK> Jymmm: Haha, there's Jymmm making us all jealous in California there. Well, OK, but your state is broke! (Come to think of it so is the rest of the country...)
[03:16:55] <elmo40> Jymmm: that is bitter cold weather?
[03:17:17] <Emcrules_Laptop> sensorless does not require a feedback device to commutate
[03:17:24] <elmo40> KimK: with no such thing as 'money' who is to call anyone 'broke' ? ;)
[03:17:25] <Jymmm> elmo40: Not the worse.
[03:18:30] <elmo40> Emcrules_Laptop: evening neighbour. it seems that absolute encoders are still not supported by EMC (yet)
[03:18:31] <skunkworks> KimK: - no - probably a 6 ball 3 pin type probe
[03:19:11] <Emcrules_Laptop> well they kind of are hostmot2 will interface with ssi encoders
[03:19:11] <KimK> Emcrules_Laptop: You'll need a postion sensor anyway though. And if it's absolute you might be able to run sensorless.
[03:20:20] <elmo40> I have a lowly ol' TTL encoder. I still need an interface board for it. I was told a 5i22 would work. I have never used one of them, it will be interesting :)
[03:20:25] <KimK> elmo40: Does that include the new Mesa resolver input cards? I thought they were ready.
[03:20:58] <Emcrules_Laptop> KimK: yes that is true. i know andy has done alot of work on the BLDC comp so I was thinking sensorless might not be possible with the 8i20 hardware
[03:21:56] <Emcrules_Laptop> elmo40: TTL is fine. Thats what I use. But i am using mesa interface hardware
[03:23:04] <Emcrules_Laptop> Elmo40: the 5i22 will work fine but a daughter board would be recommended for interfacing
[03:23:23] <KimK> elmo40: Yes, I've used 5i22, both sizes, but always with 7i33(?) 4-input-servo.
[03:23:50] <elmo40> hrmm... all I have right now is the computer and ttl encoder. now I need two more boards?
[03:23:51] <Emcrules_Laptop> Elmo: i am the same as kim except i use a 5i23
[03:24:33] <Emcrules_Laptop> And a 7i48 (six axis version of the 7i33)
[03:24:39] -!- kb8wmc has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101206122310]]
[03:25:39] <Emcrules_Laptop> Elmo40 if you need help just let me know i could stop by anytime
[03:26:14] <KimK> elmo40: Yes, you need a servo interface (or two) and (maybe) a DC interface 7i38(?) or two. You can spend more money to get more interfaces on one card. And you need ribbon cables. I was told 5 ft max for servo, 10 ft max for DCIO.
[03:27:17] <KimK> I like to have extra servo I/O for spindle encoders, 4th/5th axes, handwheels, knee DROs, lots of "extra" stuff.
[03:27:30] <pcw_home> Emcrules_Laptop Yes I can discuss 8I20
[03:28:20] <Emcrules_Laptop> pcw_home: is it possible to derive rotor position sensorless with the 8i20?
[03:28:53] <Jymmm> KimK: Sorry man, I just couldn't resist the temptation of trying to figure out why in the hell all you people live in a freezer for!
[03:29:00] -!- OoBIGeye has quit [Ping timeout: 260 seconds]
[03:29:11] -!- OoBIGeye [OoBIGeye!~kent@194-237-166-4-no48.tbcn.telia.com] has joined #emc
[03:29:11] <Jymmm> We got lots of nice warm desert too!
[03:29:21] <Emcrules_Laptop> Jymmm: brain cant thaw out to decide to move!!!!
[03:29:38] <Jymmm> lol
[03:29:45] <Emcrules_Laptop> When it does the weather is nice again!!
[03:30:08] <KimK> Jymmm: It's because we're closer to the food, and further from the food riots.
[03:30:43] <KimK> Sorry, I'll pipe down so PCW can chat.
[03:31:29] <pcw_home> Well you can use magnetic align
[03:32:00] <Emcrules_Laptop> Im thinking more clarke and back EMF
[03:32:01] <pcw_home> (Andys BLDC comp does that)
[03:33:03] <elmo40> well, lets put this all together then. I am converting a lathe. I want to use the TTL on the spindle for threading capability. I would also like an MPG. X&Z will be steppers. I have the PC, the encoder, the steppers, the stepper control board and an analog speed controlled DC motor. From reading up it seems as though I need a 5i20(22?), two daughter boards and the cabling? What is a 7i33 used for?
[03:33:09] <pcw_home> No we dont have any back emf sensing (plus you have to mode to measure backemf so not much advantage)
[03:33:23] <pcw_home> s/mode/move/
[03:34:51] <pcw_home> If you have to move to a zero crossing, you might as well use magnetic/Halls or index
[03:35:40] <KimK> elmo40: A 7i33 gives you 4 encoders in and 4 +/- 10V out (or PWM out), in exchange for one "anything I/O" port/ribbon.
[03:38:10] <KimK> elmo40: A 7i42(was it?) gives you 6 encoders in and 6 step/dir or CW/CCW out (or other signals), in exchange for one "anything I/O" port/ribbon.
[03:39:32] <KimK> elmo40: A 7i38(was it?) gives you 16 DC inputs and 8 DC outputs (or other, see manual), in exchange for one "anything I/O" port/ribbon.
[03:40:46] -!- danimal_garage has quit [Ping timeout: 250 seconds]
[03:40:54] -!- danimal_garage [danimal_garage!danimal_ga@ip98-176-169-90.sd.sd.cox.net] has joined #emc
[03:41:12] <KimK> elmo40: Does that help? ( now back to pcw_home and Emcrules_Laptop )
[03:42:37] <pcw_home> If you only have one analog drive (and its 0 +10 and dir) I would use a 7I47S, This will give you 4 encoder inputs, one analog out and 4 step+dir outputs
[03:42:39] <pcw_home> (its differential input but this can be worked around)
[03:42:45] <Emcrules_Laptop> pcw, yes i was just reading up on a couple of freescale app sheets using sensorless vector contol on PMSM and BLDC
[03:43:50] <pcw_home> I always liked how JMK called thenm senseless controls...
[03:43:55] <Emcrules_Laptop> elmo40: i have all of those board here if you want to review what you need
[03:44:18] <Emcrules_Laptop> pcw, ?
[03:46:29] <pcw_home> I think it may be possible to refine the thump align so it just buzzes a little to zero in on the angle that generates no torque
[03:47:27] <Emcrules_Laptop> I would think it would be but I just dont know of how you control and measure the inverter
[03:48:39] <Emcrules_Laptop> Well control i have a good idea just dont know what measurements you are making in the inverter
[03:49:13] <pcw_home> I will play around with when I get a little free time
[03:49:15] <pcw_home> (all control would be in EMC the 8I20 just supplies the requested current at the requested angle)
[03:49:53] <Emcrules_Laptop> Ah and if the 8i20c is used then control is in it?
[03:50:34] <pcw_home> Yes but for EMC the 8C20 is not needed or really useful
[03:51:16] <Emcrules_Laptop> PCW is any of the developmet work on the 8i20 in master or is it offline development
[03:51:45] -!- OoBIGeye has quit [Ping timeout: 260 seconds]
[03:51:56] -!- OoBIGeye [OoBIGeye!~kent@194-237-166-4-no48.tbcn.telia.com] has joined #emc
[03:52:35] <pcw_home> as EMC can close the position loop and does the initialization
[03:52:37] <pcw_home> Pretty sure its all in master (what needed it the BLDC comp and the sserial support in HostMot2)
[03:53:23] <Emcrules_Laptop> Thanks i got my 8i20 but have to wait for a serial interface board
[03:53:35] <pcw_home> ran a parker linear motor today from the 8I20/EMC
[03:53:53] <Emcrules_Laptop> A trilogy motor?
[03:54:27] <pcw_home> You can run it it from a 7I47
[03:54:29] <pcw_home> Parker daensomething
[03:54:39] <Emcrules_Laptop> last time i bought from parker it was a trilogy. Thay are nice
[03:54:50] <Emcrules_Laptop> how did it go
[03:56:25] -!- OoBIGeye has quit [Ping timeout: 260 seconds]
[03:57:08] <Emcrules_Laptop> pcw, do you think it's worth me getting a 7i49 to try out these hiperface encoders. I dont mind buying a board just for testing!
[03:57:10] <pcw_home> worked fine but there are some issues with thump start that I need to check with Andy
[03:57:12] <pcw_home> (Ive had no problem with normal motors but got a jerk on startup with the Parker)
[03:57:13] <pcw_home> Other than that really neat and fast
[03:58:50] <Emcrules_Laptop> I really want to make a 3 axis machine using the 8i20's i think it would be the cats ass. In terms of the whole project and overall cost vs performance
[04:00:08] -!- OoBIGeye [OoBIGeye!~kent@194-237-166-4-no48.tbcn.telia.com] has joined #emc
[04:00:35] <pcw_home> Basically the 7I49 would just be an A-D card, with maybe 120 KHz max input sampling rate (of all 12 input channels)
[04:00:36] <pcw_home> dont know if thats fast enough for your encoders (how many cycles per turn?) You probably want a faster,lower res A-D
[04:00:48] <KimK> Emcrules_Laptop: If you do, we want pictures!
[04:00:54] <Emcrules_Laptop> 128
[04:01:44] <Emcrules_Laptop> KimK no problem. It will probbably be a 6' x 4' x 10" plasma or router
[04:02:15] <Emcrules_Laptop> With nice and fast rapids
[04:03:09] <pcw_home> well 128 is not too bad (6.4 KHz at 3000 RPM) maybe Ill give it some thought (I would have to patch the firmware to run faster and not average it A-D readings as it does n ow)
[04:06:07] <Emcrules_Laptop> pcw: So if i ordered one i would have to arange it with you to have it patched?
[04:09:40] <pcw_home> I have to see how much it looks like work...
[04:09:51] <Emcrules_Laptop> LOL
[04:10:40] <Emcrules_Laptop> Would you mind looking? I have a few of these motors and replacing the encoders is a massive pain.
[04:11:15] -!- Valen has quit [Quit: Leaving.]
[04:14:25] <pcw_home> I will take a look
[04:17:07] <Emcrules_Laptop> pcw. thanks
[04:17:26] <Emcrules_Laptop> kimK http://imagebin.org/136058
[04:23:26] <KimK> Emcrules_Laptop: Cool! What am I looking at?
[04:26:45] <Emcrules_Laptop> a robotic deburring app using a 6 axis DOF force sensor. the force sensor allows the robot to apply a constant pressure on the surface regardless of programmed path
[04:27:50] <Emcrules_Laptop> Cant show any pictures of that actuall process. Millitary parts
[04:38:38] -!- dimas_ [dimas_!~dimas@] has joined #emc
[05:27:38] -!- Emcrules_Laptop has quit [Ping timeout: 276 seconds]
[05:31:15] -!- ve7it has quit [Remote host closed the connection]
[05:41:25] -!- adb has quit [Ping timeout: 260 seconds]
[06:07:08] -!- kb8wmc [kb8wmc!~chatzilla@] has joined #emc
[06:57:30] -!- Dannyboy [Dannyboy!~Dan@ip68-101-102-133.oc.oc.cox.net] has joined #emc
[07:01:18] -!- psha [psha!~psha@] has joined #emc
[07:04:50] -!- dimas_ has quit [Read error: No route to host]
[07:06:43] -!- dimas_ [dimas_!~dimas@] has joined #emc
[07:11:47] -!- micges_work [micges_work!~x@aeji46.neoplus.adsl.tpnet.pl] has joined #emc
[07:54:28] -!- toastydeath has quit [Ping timeout: 255 seconds]
[07:56:20] -!- toastydeath [toastydeath!~toast@c-69-140-223-139.hsd1.md.comcast.net] has joined #emc
[07:57:35] -!- toastyde1th [toastyde1th!~toast@c-69-140-223-139.hsd1.md.comcast.net] has joined #emc
[08:00:46] -!- toastydeath has quit [Ping timeout: 255 seconds]
[08:06:42] -!- Dannyboy has quit [Remote host closed the connection]
[08:07:37] -!- micges_cnc [micges_cnc!~x@aeji46.neoplus.adsl.tpnet.pl] has joined #emc
[08:22:26] -!- micges_work [micges_work!~x@aeji46.neoplus.adsl.tpnet.pl] has parted #emc
[08:25:42] -!- cnc [cnc!~Valen_cnc@60-240-68-28.tpgi.com.au] has joined #emc
[08:26:13] <cnc> i have .001mm resolution scales on our mill
[08:27:07] <cnc> when i am adjusting it though the ferror is floating between -005121 and +.0004879
[08:27:20] <cnc> .0005121
[08:27:32] <cnc> which is a difference of .001
[08:28:02] <cnc> why would it be using those additional digits of resolution which aren't there
[08:28:03] <archivist> do you have separate encoders on the servos too
[08:28:04] <cnc> ahh
[08:28:07] <cnc> no
[08:28:15] <cnc> its the jogging
[08:28:29] <archivist> floats
[08:28:48] <cnc> i wager jog has set it to a position in between a .001 mesurement
[08:28:59] <cnc> so the ferror is indeed that
[08:30:10] <cnc> nope thats not it
[08:30:43] <cnc> just told it to do a g0 to a .000 and ferror is still off
[08:31:42] <cnc> problem is its causing the axis to hunt as it never sits t 000
[08:31:57] <cnc> it just keeps moving by one count
[08:32:18] <cnc> i wonder if its perhaps an offset thing?
[08:32:38] <archivist> have you set too close a limit
[08:32:47] <awallin> use deadband to get rid of hunting?
[08:33:59] <cnc> deadband is the wrong solution
[08:34:11] <cnc> too close a limit archivist?
[08:34:45] <cnc> if my scales are in .001 I shouldn't see a measurement of .0003453
[08:36:39] <archivist> what ferror have you set/allowed for, its trying to do what you are commanding, this is resolution v accuracy
[08:37:46] <cnc> where does one "set" ferror allowed?
[08:37:50] <archivist> so a deadband will allow it to stop
[08:38:12] <cnc> yes but it would also cause a .002 region in which it does nothing
[08:38:13] -!- tris has quit [Quit: Leaving]
[08:38:28] <awallin> you need to home accurately to an edge of the encoder to see 'even' counts of 0 0.001 0.002 and so on ?
[08:39:08] <cnc> where does it get these non even counts from?
[08:39:35] -!- tris [tris!~tristan@173-164-188-122-SFBA.hfc.comcastbusiness.net] has joined #emc
[08:39:44] <archivist> you only have counts/edges at your scales rate
[08:39:50] <awallin> you said earler "floating between -005121 and +.0004879" that makes sens, the difference is 0.001
[08:39:55] <cnc> I just did a G53 Z0 and it went to an even offset
[08:40:10] <cnc> yes and so why is it .0005 not .001
[08:40:12] <awallin> you are commanding the machine to be at 0.000 but there are no encoder edges there
[08:40:34] <cnc> thats what i have been saying
[08:40:36] <archivist> maybe you should program it in mm to get the best out of it
[08:40:44] <cnc> it is in mm
[08:41:04] <cnc> how do i tell it that there are 3 digits of resolution on the encoders
[08:41:42] <cnc> (it went from -.001 to 5.329071E-14, thats a floating point issue)
[08:42:12] <cnc> (after the g53)
[08:42:40] <awallin> you can round off the numbers for display in AXIS, otherwise I don't think theres a place inside emc2 to adjust accuracy/significant-digits
[08:42:52] <archivist> should be using doubles though, but yes classic float problem
[08:43:19] <cnc> they are rounded off already but the internal representation is whats causing the problem
[08:43:52] <cnc> archivist, even a double would have the same issue, just smaller lol
[08:44:20] <cnc> decimal is the only type that wont have it because it has implicit rounding
[08:44:29] <archivist> yes Im wondering where the error is becoming so large
[08:45:18] <cnc> I think its a cordinate system offset
[08:45:22] <archivist> I have though for a time EMC needs exactness and not float types
[08:45:34] <cnc> we jogged to place then touched off
[08:45:58] <cnc> when we touched off it did so off the comanded position
[08:46:07] <cnc> which was .000something
[08:46:37] <cnc> meaning there is a .000something offset meaning the encoder counts never line up
[08:47:16] <archivist> there will be some error due to the resolution of pwm on the drive
[08:48:16] <cnc> that wont cause the ferror to be in more signifigant digits than the scale
[08:48:21] <cnc> thats thecore of the problem
[08:48:55] <cnc> working in machine cordinates it works
[08:49:16] <cnc> but the I term continiously adds up the error in the offset
[08:49:25] <cnc> making the axis hunt
[08:50:28] <cnc> right I have G53 ed to a .000 then touched off to 3 digits and now the cordinate systems are aligned
[08:51:05] <cnc> the ferror is in .001s
[08:51:49] <cnc> its Valen here btw
[08:52:00] cnc is now known as Valen_cnc
[09:35:17] -!- e-jones [e-jones!~jkastner@nat/redhat/x-jeinoweybnttvrro] has joined #emc
[09:36:23] -!- Dallur [Dallur!~root@85-220-120-40.dsl.dynamic.simnet.is] has joined #emc
[09:39:46] -!- L84Supper has quit [Ping timeout: 255 seconds]
[09:43:19] -!- robh__ [robh__!~robert@5ace70af.bb.sky.com] has joined #emc
[09:50:22] -!- Valen_cnc has quit [Remote host closed the connection]
[09:52:34] -!- L84Supper [L84Supper!~ly@unaffiliated/l84supper] has joined #emc
[11:15:39] -!- awallin_ [awallin_!~quassel@2001:708:110:1020:224:7eff:feda:7c7d] has joined #emc
[11:23:36] -!- mk0 [mk0!~tabor@fiztech.basnet.by] has joined #emc
[11:36:40] -!- Dallur has quit [Quit: Leaving.]
[11:39:02] -!- adb [adb!~Moldovean@178-211-232-101.dhcp.voenergies.net] has joined #emc
[11:47:43] -!- Techrat has quit [Quit: Leaving]
[12:03:07] <mrsunshine_> fenn, what kind of rubber glue do you use ?
[12:04:15] -!- Valen [Valen!~Valen@ppp121-44-129-24.lns20.syd7.internode.on.net] has joined #emc
[12:07:15] <fenn> something i got from the office supply store
[12:09:17] <mrsunshine_> know what the agent in it is ?
[12:09:22] <mrsunshine_> or wahtever i should call it
[12:10:07] <alex_joni> yo fenn
[12:10:56] <mk0> how to convert dxf to gcode for foam cutting? (milling, lathe does not suite). after dxf2gcode (it's the best tool) still a heck lot of work.
[12:12:17] <alex_joni> do you cut 2d, 2.5d or 3d?
[12:12:39] <alex_joni> I used sheetcam for 2d and 2.5d stuff, and it worked great
[12:20:23] <mk0> 2D
[12:21:50] -!- servos4ever [servos4ever!~chatzilla@173-87-48-6.dr01.hnvr.mi.frontiernet.net] has joined #emc
[12:30:21] <mk0> alex_joni, what lib does it need? i get "Could not find 'Support libraries for SheetCam TNG'. Try using the native package manager for Ubuntu 10.04.1 LTS (apt-get) to install a package with similar name to 'sheetcamlibs'."
[12:42:15] <mk0> sorry
[12:45:28] -!- psha has quit [Quit: leaving]
[12:46:10] -!- Valen has quit [Quit: Leaving.]
[12:49:49] <alex_joni> mk0: http://www.sheetcam.com/downloads.shtml
[12:49:54] <alex_joni> The first time you install you will also need these libraries:
[12:49:54] <alex_joni> Download sheetcamlibs-
[12:52:16] <mk0> yes this is why i said sorry :) but it does not start after installing, crashes and asks to send debugging info :(
[12:58:13] -!- skunkworks has quit [Ping timeout: 255 seconds]
[13:07:40] -!- mk0 has quit [Ping timeout: 255 seconds]
[13:28:41] -!- psha [psha!~psha@] has joined #emc
[13:33:31] -!- skunkworks [skunkworks!447329d2@gateway/web/freenode/ip.] has joined #emc
[13:37:54] -!- e-jones has quit [Quit: Ex-Chat]
[13:39:13] -!- e-jones [e-jones!~jkastner@nat/redhat/x-cplzaxkgzqbnsnkv] has joined #emc
[13:46:06] -!- e-jones has quit [Quit: Ex-Chat]
[13:59:53] -!- Dallur [Dallur!~root@85-220-120-40.dsl.dynamic.simnet.is] has joined #emc
[14:01:47] -!- toastyde1th has quit [Quit: Leaving]
[14:06:29] <mrsunshine_> fenn, no answer tot hat? :)
[14:06:30] <mrsunshine_> that
[14:10:52] -!- Dallur has quit [Quit: Leaving.]
[14:21:56] <skunkworks> http://cnczone.com/forums/product_announcements_manufacturer_news/121274-geckodrive_changes_new_products_2011_a.html
[14:27:41] -!- adb has quit [Ping timeout: 255 seconds]
[14:28:21] -!- JT-Work [JT-Work!~chatzilla@216-41-154-151.semo.net] has joined #emc
[14:36:02] -!- Tom_itx has quit [Ping timeout: 265 seconds]
[14:48:23] -!- micges_cnc has quit [Quit: Leaving]
[14:56:02] -!- JT-Work has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.13/20101203075014]]
[15:06:23] -!- adb [adb!~Moldovean@178-211-232-101.dhcp.voenergies.net] has joined #emc
[15:07:26] -!- pingufan [pingufan!~xerox@goliath.hantsch.co.at] has joined #emc
[15:39:01] -!- psha has quit [Quit: Lost terminal]
[16:08:28] -!- Dannyboy [Dannyboy!~Dan@] has joined #emc
[16:20:03] <skunkworks> come on now! why isn't anyone talking smart?
[16:20:29] <kb8wmc> good day skunkworks
[16:20:34] <skunkworks> Good day!
[16:20:38] <skunkworks> friday!
[16:20:47] <kb8wmc> yes, very good day
[16:20:49] <Jymmm> skunkworks: You dumb us down by pasting cnczone links all the time. What do you expect?
[16:20:52] * skunkworks may have had too much coffee
[16:21:06] <skunkworks> time for some tea!
[16:23:26] -!- odiug [odiug!~guido@p57BD1253.dip0.t-ipconnect.de] has joined #emc
[16:28:13] -!- micges [micges!~micges@acuc219.neoplus.adsl.tpnet.pl] has joined #emc
[16:28:42] -!- micges [micges!~micges@acuc219.neoplus.adsl.tpnet.pl] has parted #emc
[16:29:51] <Jymmm> skunkworks: Ok, I take that back JUST THIS ONCE. Will be interesting to see the GM504 thugh.
[16:31:20] <Jymmm> Guess that mothballs EMC2.
[16:31:40] * Jymmm puts up the FOR SALE sign on linuxcnc.org
[16:31:57] <archivist> not enough axes dont be silly
[16:32:14] <archivist> and other sillyness
[16:32:14] -!- Tom_itx [Tom_itx!~Tl@unaffiliated/toml/x-013812] has joined #emc
[16:33:15] <Jymmm> I'm sure that newer firmware would allow for chaining 2 or 3 GM540's for 8 or 12 axis
[16:33:37] <archivist> I am less sure
[16:36:06] <Jymmm> Seriously though, not having to find the right mobo for latency would be nice
[16:36:36] <skunkworks> Jymmm: I really have not had too much issue with latency on motherboards.
[16:36:58] <Jymmm> skunkworks: you never had to use an ext video card?
[16:37:09] <skunkworks> sure - and that fixes it
[16:37:12] <skunkworks> usually
[16:37:26] <Jymmm> Exactly
[16:37:58] <pcw_home> And the Intel atoms work fine with their on board video
[16:38:03] <skunkworks> yes
[16:38:13] <Jymmm> Oh, anyone what laser engraved chocolate bar for valentines?
[16:38:27] <Jymmm> s/what/want/
[16:43:05] -!- pingufan has quit [Remote host closed the connection]
[16:44:33] -!- isssy [isssy!~isssy@78-83-51-185.spectrumnet.bg] has joined #emc
[16:44:50] <skunkworks> pcw_home: nice info on pid
[16:45:13] -!- wobblybootie [wobblybootie!5c11dd75@gateway/web/freenode/ip.] has joined #emc
[16:48:15] -!- micges has quit [Quit: Ex-Chat]
[16:48:36] <mrsunshine_> aparently heptan does not disolve the pink building foam atleast =)
[16:50:04] <Jymmm> mrsunshine_: But acetone does =)
[16:50:27] <mrsunshine_> Jymmm, but this time i was looking for something that doesnt =)
[16:50:45] <Jymmm> wheres the fun in that?
[16:52:12] <mrsunshine_> that i do not have to do my foam pattern over again ? :P
[16:52:34] <Jymmm> pattern of what?
[16:54:30] <mrsunshine_> for spindle motor for the mill
[16:54:32] <mrsunshine_> for casting
[16:54:51] <Jymmm> your casting a motor ?!
[16:54:54] <Jymmm> cool!
[16:55:00] <mrsunshine_> no
[16:55:01] <mrsunshine_> mount
[16:55:02] <mrsunshine_> :P
[16:55:07] <mrsunshine_> not the actual motor :P
[16:55:16] <Jymmm> does it including the windings too? ;)
[16:55:28] <mrsunshine_> sigh ... you are not listening to me! :P
[16:55:30] <Jymmm> oh, bummer
[16:55:42] <mrsunshine_> "oo motor lets stop reading and start rambling" :P
[16:55:59] <mrsunshine_> Jymmm, just simple shapes to get the hang of stuff =)
[16:56:34] <Jymmm> gotcha, its the molten shape that scares me =)
[16:56:56] <mrsunshine_> why ?
[16:57:04] <mrsunshine_> its no more dangerous then water ;)
[16:57:48] <Jymmm> ok, I'll pour water on your left arm, and molten aluminum on your right, deal?
[16:57:50] <mrsunshine_> never had molten alu blow up in my face, molten salt on the other hand
[16:57:53] <mrsunshine_> kablooom!
[16:58:06] -!- Tom_itx has quit [Ping timeout: 250 seconds]
[16:58:16] <mrsunshine_> i guess metal will do the same thing at the right circumstances :P
[16:58:23] <Jymmm> mrsunshine_: Better you than me
[16:58:24] <mrsunshine_> but, so far nothing has happened to me :P
[16:58:36] <Jymmm> lets keep it that way too!
[16:58:53] <mrsunshine_> lucky i had my glasses on that day, it blew a 1 inch fire brick clean in two and threw it like 5 meters in each direction :P
[16:59:12] <mrsunshine_> just one little drop of stupid water :P
[16:59:32] * Jymmm points to his FULL face shield - fsck the glasses crap!
[17:00:00] <mrsunshine_> aye will get myself a face shield ... as i feel a bit exposed with glasses =)
[17:00:38] <Jymmm> Even the one from HF is ANSI approved, and the shield itself is replceable
[17:01:52] <mrsunshine_> so is my eyes!
[17:01:52] <mrsunshine_> and face
[17:01:53] <mrsunshine_> dont need em anyways
[17:01:53] <mrsunshine_> got some spare parts int he cellar
[17:02:29] <Jymmm> TMI
[17:02:51] <mrsunshine_> tmi ?
[17:02:59] <mrsunshine_> haha
[17:03:02] <mrsunshine_> to much information ? :P
[17:03:04] <Jymmm> TMI == Too Much Information
[17:05:26] <Jymmm> http://geckodrive.net/ark-4/support.html?pid=32&id=111
[17:07:02] <PCW> 10 uSteps is cheezy (and noisy)
[17:07:53] <skunkworks> Jymmm: any kind of load on the stepper will make it not positoin between the full steps acuritley
[17:07:58] <skunkworks> wow
[17:08:11] <skunkworks> I suck at typing fast and spelling correclty
[17:08:13] <skunkworks> heh
[17:08:33] <cradek> 'acuritely' is particularly creative
[17:08:43] <Jymmm> lol
[17:09:07] <skunkworks> :)
[17:09:50] <PCW> When we were messing with it 32 uSteps seemed about the minimum for a quiet drive
[17:11:35] -!- odiug has quit [Ping timeout: 276 seconds]
[17:12:56] <skunkworks> Jymmm: those motion devices like what gecko is developing is neat... but they are very inflexable compared to emc2
[17:13:26] <skunkworks> if all you need is 3 axis and a few i/o that you may or may not be able to do other things with - then great.
[17:20:29] * skunkworks hugs emc again.
[17:22:31] <Jymmm> but you have a freak of a machine.
[17:23:05] <cradek> we prefer to call it "differently-abled"
[17:23:15] <Jymmm> hahahahahahahahaha
[17:23:42] <cradek> (I still think of horizontal machines as odd)
[17:24:05] <Jymmm> Speaking of tape loving freaks... You got it setup and running yet?
[17:24:28] <cradek> nope - I exercised it a bit, but it will be a few days (or more) before it's hooked to a computer.
[17:24:56] <Jymmm> cradek: you are using the web interface though, right?
[17:25:02] <Jymmm> now that is
[17:25:19] <skunkworks> heh - Chips just fall away on a horizontal....
[17:25:29] <Jymmm> I'm sure it could use firmware upgrades
[17:25:46] <Jymmm> multiple firmwre upgrades
[17:26:48] <DaViruz> skunkworks: on an upside-down vertical too.
[17:27:53] <skunkworks> DaViruz: sure - I have not seen one. do you have pictures?
[17:29:31] <mshaver> Jymmm: web interface?
[17:29:35] <DaViruz> i haven't seen one either
[17:29:40] <DaViruz> but i'd like to :)
[17:30:46] <Jymmm> mshaver: this thing called a browser to view and control ancient technology from the dark ages
[17:31:53] <mshaver> Right :) I understand what the "web" is, just wondered what you meant. Is there an EMC GUI that runs in a browser?
[17:32:07] <Jymmm> mshaver: in your dreams =)
[17:32:42] -!- adb has quit [Ping timeout: 272 seconds]
[17:34:14] -!- isssy has quit [Quit: Visitor from www.linuxcnc.org]
[17:34:32] <mshaver> Well, yes: http://www.mattshaver.com/emcgui/index.htm (not really functional)
[17:35:21] <Jymmm> No thanks, I dont do java
[17:35:36] <mshaver> I don't anymore either...
[17:35:58] <Jymmm> I mean I have java disabled in my browsers
[17:36:09] <mshaver> But an HTML5 GUI would be neato!
[17:37:10] <Jymmm> not really
[17:39:00] -!- mshaver [mshaver!~mshaver@c-68-48-14-94.hsd1.dc.comcast.net] has parted #emc
[17:40:56] -!- Tom_itx [Tom_itx!~Tl@unaffiliated/toml/x-013812] has joined #emc
[17:59:56] -!- Dallur [Dallur!~root@85-220-120-40.dsl.dynamic.simnet.is] has joined #emc
[18:00:30] -!- pingufan [pingufan!~xerox@goliath.hantsch.co.at] has joined #emc
[18:04:01] -!- vezza [vezza!~andrea__@host253-44-dynamic.17-87-r.retail.telecomitalia.it] has joined #emc
[18:10:33] -!- vezza has quit [Quit: Sto andando via]
[18:11:53] -!- pingufan has quit [Remote host closed the connection]
[18:26:28] -!- adb [adb!~Moldovean@178-211-232-101.dhcp.voenergies.net] has joined #emc
[18:38:41] * bill2or3 thinks his fireballv90 got lost in snowpocalypse.
[18:39:41] <skunkworks> bill2or3: you ordered a fireball?
[18:39:46] <skunkworks> neat
[18:39:50] <bill2or3> y.
[18:40:12] <bill2or3> but the UPS tracking says it left illnoise wednesday evening, and no furthur update.
[18:40:22] <bill2or3> it was supposed to be here by end-of-day, yesterday.
[18:41:43] <skunkworks> yeck
[18:41:46] <skunkworks> I hate that
[18:43:04] <skunkworks> bill2or3: how are the kid(s)?
[18:43:08] <atmega> I have a similar layout router... anyone know how you line up parts for 2nd operations, or recuts?
[18:43:29] <bill2or3> neglected. :-)
[18:43:32] <skunkworks> heh
[18:43:34] <bill2or3> I just have one.
[18:43:45] <skunkworks> I thought so - but couldn't quite remember
[18:43:45] <bill2or3> but the robots are doing great.
[18:45:20] <skunkworks> atmega: mill a pocket?
[18:45:41] <atmega> sure... or flip a part over and pocket the other side
[18:45:55] <atmega> how do you make sure things are really lined up?
[18:46:08] <atmega> or do you have to make fixtures for everything?
[18:46:22] <skunkworks> I meant mill a pocket to set the part into.
[18:46:34] <atmega> oh, that would be the fixture part :)
[18:46:42] <skunkworks> yes
[18:46:51] <atmega> yeah, but then I have to align the fixture
[18:47:32] <atmega> is that usually a mechanical operation, or by rotating axes
[18:47:57] <skunkworks> usually mechanical - but if you had a probe - now you could do coordinate rotation...
[18:48:36] <atmega> I have some doughnut sort of shaped parts I want to cut out, and then radius the edges
[18:49:26] <skunkworks> if you had home switches - putting reference pins on everything helps..
[18:49:39] <skunkworks> pin it to the table some how
[18:49:51] <atmega> I have home switches, but dunno how repeatble they really are
[18:50:17] <atmega> I think I could do the radiusing with a pair of inside and outside pockets
[18:50:33] <atmega> but, then I'm stuck for squaring the fixture
[18:51:44] <skunkworks> we have done a lot of jogging the axis back and fort and indicating it
[18:51:53] <atmega> I was hoping CNC stuff would obviate the need for me to have any real machinists skills
[18:52:12] <skunkworks> I think a lot of people think that ;)
[18:52:15] <bill2or3> sadly, no.
[18:52:24] <bill2or3> but now you can screw stuff really fast!
[18:52:25] <bill2or3> :-)
[18:53:46] <skunkworks> crunch! 8 tons of force really messes things up quick
[18:53:54] <atmega> yeah... I made some beautiful cannister lids out of 1" cast acrylic with a pocketed recess for a switch, but I need to flip them over and make a pocket on the other side
[18:54:33] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust1037.basl.cable.virginmedia.com] has joined #emc
[18:58:11] -!- danimal_garage has quit [Ping timeout: 276 seconds]
[18:58:58] -!- danimal_garage [danimal_garage!danimal_ga@ip98-176-169-90.sd.sd.cox.net] has joined #emc
[19:01:59] <atmega> anyone used meshcam?
[19:02:52] <andypugh> I assume _somebody_ has.
[19:02:57] <andypugh> :-)
[19:03:10] -!- danimal_garage has quit [Ping timeout: 240 seconds]
[19:03:38] -!- danimal_garage [danimal_garage!danimal_ga@ip98-176-169-90.sd.sd.cox.net] has joined #emc
[19:05:19] -!- odiug [odiug!~guido@p57BD1253.dip0.t-ipconnect.de] has joined #emc
[19:05:43] <atmega> since I'm not a machinist, could I have a 4 axis mill (4th being a vertical mounted rotary), could I model a 3d object and have some software generate all the code to cut whatever, from whatever side it needed to be cut from? and if so, what SW?
[19:06:28] <andypugh> Software will do that, but generally rather expesive software
[19:06:33] <bill2or3> yes, but as to the the "What software" part, I have no idea.
[19:06:58] <bill2or3> but I think the answer has at least one comma in it.
[19:07:17] <atmega> comma is fine, if it is euro notation
[19:08:02] -!- danimal_garage has quit [Ping timeout: 255 seconds]
[19:08:05] <bill2or3> nope, sorry.
[19:08:07] <PCW> Hey Andy, got a linear motor running with the 8I20 and your BLDC comp about 3 ft long, 1 u resolution strip type encoder
[19:08:43] <atmega> http://www.golemgear.com/images/mav_ox2.jpg
[19:09:06] <atmega> that has two threaded holes in teh top, one on the side (with a step down inside) and one on the bottom
[19:09:26] -!- acemi [acemi!~acemi@unaffiliated/acemi] has joined #emc
[19:10:51] <andypugh> PCW: Great! Nice to know that the system works with oddball stuff too.
[19:12:35] <andypugh> atmega: Something like that would be easy enough to do in 2.5D CAM as 4 sides. and put in the workpiece rotations by hand.
[19:12:49] -!- danimal_garage [danimal_garage!danimal_ga@ip98-176-169-90.sd.sd.cox.net] has joined #emc
[19:14:31] <andypugh> Where you need clever and expensive software is stuff like http://www.youtube.com/watch?v=Fjzhygzafr0&feature=related
[19:18:15] <danimal_garage> i use mastercam x3
[19:18:58] <danimal_garage> me likey
[19:26:40] <atmega> I wanted the coordinates to be aligned automagically also
[19:30:22] <Tom_itx> does anyone know who actually makes those 'blackbox' stepper drivers you see all over? i've seen them on at least 3 sites and they all look identical except the branding
[19:31:09] -!- Tom_L [Tom_L!~Tl@unaffiliated/toml/x-013812] has joined #emc
[19:31:14] <Tom_L> http://www.kelinginc.net/KLDriver.html
[19:31:17] <Tom_L> one such example
[19:31:47] <Tom_L> and are they worth a crap?
[19:33:36] <atmega> if you want a really good stepper driver, don't you really want a servo instead?
[19:34:05] <Tom_itx> probably
[19:34:13] <Tom_itx> cost is somewhat of a concern on this one
[19:34:23] <andypugh> I use them, they are absolutely fine.
[19:35:04] <Tom_L> what about the gecko G540 4 axis board?
[19:35:05] <atmega> there is a chinese place on ebay that sells similar looking ones paired with motors pretty cheap
[19:35:17] <Tom_L> servo?
[19:35:27] <atmega> no, those blakcbox stepper drivers
[19:35:36] <andypugh> No experience of the G540, but it has a lot going for it.
[19:35:55] <andypugh> The PWM to 10V is very handy for soindle control.
[19:36:16] <Tom_itx> what's the rated supply v on those?
[19:36:31] <Tom_itx> i'm lookin at a 50v supply right now
[19:37:18] -!- micges [micges!~micges@acuc219.neoplus.adsl.tpnet.pl] has joined #emc
[19:39:39] <Tom_L> i don't know enough about sizing servos but i'm looking at a 282in oz stepper for this project
[19:39:54] <Tom_L> any suggestions for a comparable torque servo?
[19:39:55] <andypugh> G540 is 50V, I think.
[19:40:19] <Tom_L> http://www.kelinginc.net/KL23H276-30-8B.pdf
[19:40:22] <Tom_L> those steppers
[19:40:46] <Tom_L> unless i get a bunch of flack here on them
[19:42:28] <danimal_garage> you wont regret servos
[19:42:39] <Tom_L> well i know that
[19:42:54] <danimal_garage> what are you putting them on?
[19:42:57] <Tom_L> i just need to size some comparable to those steppers that i can afford
[19:43:17] <Tom_L> right now, i have a sherline but i'm possibly building something a bit beefier
[19:43:35] <Tom_L> the sherline is just a toy
[19:43:48] <Tom_L> but i don't need or have room for a mori seki etc
[19:44:00] <Tom_L> so i'm looking for something inbetween those two :D
[19:44:43] <danimal_garage> i got approximately 650 in/oz servos and they move my very heavy shizuoka knee mill around
[19:45:19] <atmega> geared down?
[19:45:34] <danimal_garage> probably 400-500in/oz would be fine for a bridgeport
[19:45:38] <danimal_garage> no, direct drive
[19:45:45] <danimal_garage> .200"/rev
[19:46:21] <Tom_L> do you use double lead screws?
[19:46:39] <danimal_garage> no
[19:46:48] <Tom_L> what pitch?
[19:46:50] -!- seb_kuzminsky has quit [Ping timeout: 240 seconds]
[19:47:15] <atmega> you won't regret ball screws either
[19:47:20] <danimal_garage> i don't know, whatever ball screws came on the mill. .200"/rev.
[19:47:50] <danimal_garage> lead screws would be a nightmare on a bigger machine
[19:47:55] <Tom_L> i picked up a couple catalogs today from a local bearing supplier
[19:53:34] -!- ve7it [ve7it!~LawrenceG@S0106009027972e37.pk.shawcable.net] has joined #emc
[19:54:41] -!- Tom_L has quit []
[19:57:04] <cradek> http://i451.photobucket.com/albums/qq232/fasto_tt/dsc01288.jpg
[19:57:10] <cradek> someone was asking about one of the hurcos the other day
[19:57:21] <cradek> it has backplot preview!
[19:57:25] <mrsun> cradek, looks like a doomsday device =)
[19:57:36] <Jymmm> MONOCHROME!!!!!!!!!!!! EWWWWWWWWWWWWWWWWWWW
[19:58:43] <Jymmm> how in the hell did they do the graphic though.... Hercules doens't have graphic capability, maybe vesa.
[19:58:52] <bill2or3> 1-bit ftw
[19:59:18] <mrsun> Jymmm, mode 13h ? :P
[19:59:24] -!- awallin_ has quit [Remote host closed the connection]
[19:59:25] <atmega> hercules had fine graphics,
[19:59:27] <cradek> hercules certainly had mono graphics
[19:59:38] <cradek> the right side *might* be vector?
[19:59:40] <Jymmm> mrsun: It's been a LONG time, I dont recall
[19:59:51] <Jymmm> I'd still say VESA
[19:59:55] <atmega> but, were they using PC stuff for that?
[19:59:56] <mrsun> mode 13h is a 8bi graphics mode
[19:59:58] <mrsun> bit
[20:00:08] <mrsun> VGA mode
[20:00:41] <mrsun> isnt VESA banked and supports higher resolutions ?
[20:00:53] <Jymmm> maybe CGA
[20:00:54] <mrsun> mode 13h is 320x240 or something
[20:01:02] <Jymmm> VESA is awesome imo
[20:01:16] <Jymmm> mrsun: that would be CGA
[20:01:20] <mrsun> oki =)
[20:01:35] <mrsun> its actualy quite nice to plot in 13h =)
[20:02:01] <mrsun> mov ah, 0h; mov al, 13h; int 10h
[20:02:06] <mrsun> kablaaam! =)
[20:02:20] <mrsun> now thats easier then trying to set up a modern graphics card :P
[20:02:31] <cradek> http://www.abltechnology.com/equipment/images/thumbs/50308a.jpg
[20:02:32] <Jymmm> I like VESA; would make a killer iface to many proggys including emc =)
[20:03:03] <cradek> I like the calculator glued to it
[20:03:08] <cradek> and the huge video game trackball
[20:03:24] <mrsun> i like the direct memory access of atleast 13h, vesa as far as i know is banked and stuff gets a bit more hairy, but still linear memory for each bank =)
[20:03:57] <Jymmm> cradek: and the adding machine on the left of the calculator. Now, which has more power? =)
[20:04:06] <mrsun> lol cradek looks like they used the start cycle button some :P
[20:04:14] <Jymmm> Oh, and love the burn-in on the right crt
[20:04:16] <Tom_itx> any recomended sites for hunting down small servos?
[20:04:18] <mrsun> "wtf it doesnt work, get a stick!"
[20:08:09] <danimal_garage> Tom_itx, ebay
[20:08:22] <Tom_itx> yeah
[20:08:52] <danimal_garage> there's some drives on there right now that are pretty slick... they're the same ones as mine, except for use with encoders
[20:09:04] <danimal_garage> they don't need a power supply, and they'll run ac and dc
[20:09:25] <danimal_garage> 6a rms, 12a max
[20:09:47] <danimal_garage> they're run off of single phase 115v or 230v or 3 phase
[20:09:59] <danimal_garage> they're $70 each
[20:10:23] <danimal_garage> he's got like 7
[20:10:30] <Tom_itx> beam me up scotty. got a link?
[20:11:43] -!- Tom_L [Tom_L!~Tl@unaffiliated/toml/x-013812] has joined #emc
[20:12:24] <Tom_itx> i nearly sold enough junk this week to cover em
[20:12:24] <danimal_garage> http://rover.ebay.com/rover/0/e11051.m43.l1123/7?euid=676337d64af942cb8b6ab1b3e1e05107&loc=http%3A%2F%2Fcgi.ebay.com%2Fws%2FeBayISAPI.dll%3FViewItem%26item%3D190400345571%26ssPageName%3DADME%3AX%3ARTQ%3AUS%3A1123
[20:12:52] <danimal_garage> make him an offer for like $60 each if you buy a few. He'll combine shipping too
[20:13:51] <danimal_garage> the manual is online on the kollmorgen website
[20:13:54] <Tom_L> those may be bigger than what i need
[20:14:07] <danimal_garage> they'll still run smaller motors, right?
[20:14:22] <Tom_L> We lack the means to test these further as we do not have large enough servo motors available
[20:14:24] <danimal_garage> they're pretty small in size
[20:14:25] <Tom_L> i dunno
[20:14:53] <danimal_garage> and you dont need a power supply, just plug them into the wall
[20:15:26] <Tom_L> where's he at?
[20:15:50] <danimal_garage> they may be overkill, but for the price, it's cheaper than stepper drives and you can always use them on a bigger machine down the road
[20:15:56] <danimal_garage> i dunno
[20:16:00] <danimal_garage> usa somewhere
[20:16:09] <Tom_L> but would they work on small servos?
[20:16:40] <Tom_L> says 230v ac
[20:16:40] <danimal_garage> i dont see why they wouldnt, i think the limitation would only be if the servos were bigger
[20:16:45] <Tom_L> not 120
[20:16:48] <danimal_garage> they run on 115v too
[20:16:57] <danimal_garage> read the manual
[20:17:31] <danimal_garage> mine are all running on 115v right now
[20:18:35] <PCW> Hercules entire point was that they could do graphics on the IBM mono display (720 x something)
[20:18:46] <danimal_garage> anyone wanna buy a gecko 203x stepper drive?
[20:18:58] <danimal_garage> used for about a week
[20:19:05] <Tom_L> tell me about it?
[20:19:11] <Tom_L> single?
[20:19:15] <Tom_L> or multi
[20:19:19] <danimal_garage> single
[20:19:23] <Tom_L> pfft
[20:19:25] <Tom_L> no
[20:19:28] <Tom_L> :)
[20:20:02] <danimal_garage> i bought it to replace one of my leadshine drives, but it turns out the leadshine just needed more heatsink grease
[20:20:19] <danimal_garage> and i bought servos the next week anyways
[20:20:56] -!- vezza [vezza!~andrea__@host253-44-dynamic.17-87-r.retail.telecomitalia.it] has joined #emc
[20:20:58] <danimal_garage> i'll sell it for half price, which i think is like $70
[20:21:32] -!- vezza has quit [Remote host closed the connection]
[20:21:42] <Tom_L> ps-33xx servo drive pdf?
[20:23:22] <danimal_garage> http://www.kollmorgen.com/website/com/eng/products/drives/ac_servo_drives/ps33_series_manuals.php
[20:24:07] <danimal_garage> mine is labeled as a pacific scientific pc3406ai, but they're pretty much exactly the same drives
[20:24:25] <danimal_garage> i read both manuals because i was going to buy the api ones as well
[20:25:30] <skunkworks> hmm - I think we have some noise issues.
[20:25:41] <skunkworks> We have another harddrive going bad.
[20:26:32] <Tom_L> 90-253vac
[20:26:51] <danimal_garage> see?
[20:27:40] <danimal_garage> that sucks skunkworks
[20:28:03] <Jymmm> skunkworks: dd that bitch!
[20:28:06] <skunkworks> or it could be the really really cheap computer power supply running it
[20:28:15] <skunkworks> dd?
[20:28:22] <Jymmm> skunkworks: man dd
[20:29:08] <Jymmm> skunkworks: dd if=/dev/sda1 of=/dev/sdaX
[20:30:08] <Tom_L> is skewed rotor design better on a dc servo?
[20:30:21] <Tom_L> does that give it smoother torque?
[20:32:46] <skunkworks> The smart errors (relocated sector counts) is rising
[20:32:55] <Tom_L> do servo motors have builtin feedback? ie hall effect or encoder?
[20:33:05] <cradek> skunkworks: can you use a CF card? the 8GB are super cheap now, and plenty big enough
[20:33:07] <Jymmm> skunkworks: DD it NOW!!
[20:33:12] <cradek> (they're not fast, though)
[20:33:20] <skunkworks> cradek: yes - we could
[20:33:31] -!- Dannyboy has quit [Ping timeout: 240 seconds]
[20:33:44] <andypugh> Why am I racking my brains trying to make the RTAI usb drivers compile?
[20:34:01] <andypugh> (Apart from the fact I can't let things go)
[20:34:07] <cradek> tell us about your mother
[20:34:11] <skunkworks> Jymmm: we have all the pertanenat files backed up - we can do a fresh install in less than an hour
[20:34:28] <Jymmm> skunkworks: Oh, then smash the sucker!
[20:34:36] <Jymmm> for the magnets
[20:35:08] <skunkworks> cradek: heh
[20:38:48] -!- Dannyboy [Dannyboy!~Dan@] has joined #emc
[20:40:02] <skunkworks> Maybe the vfd is getting into things. It is funny that it is just the hd. (there is a lot of crap going on in there ;)
[20:40:36] <skunkworks> I think they where both used... iirc
[20:46:32] <skunkworks> cradek: you can get a 8gb ssd for $44
[20:48:33] -!- psha [psha!~Pavel@] has joined #emc
[20:49:07] -!- psha has quit [Client Quit]
[20:59:40] <Tom_L> would you consider these servos? http://tom-itx.dyndns.org:81/~webpage/motors/motor3.jpg
[20:59:55] <Tom_L> i stole the encoders off the back but they were 1024 quadrature
[21:00:08] <danimal_garage> does anyone run a memory card instead of a hdd with one of those little atom mobos?
[21:00:20] <Tom_L> that board probably wouldn't drive those
[21:01:24] <danimal_garage> what board, the one i linked you?
[21:01:43] <Tom_L> yes
[21:01:51] <danimal_garage> i'm sure it would, it needs a tach though.
[21:02:23] <danimal_garage> either need brushless with hall sensors or brushed with a tach
[21:02:25] <Tom_L> well i can put the encoders back on em
[21:02:35] <danimal_garage> still needs a tach
[21:02:43] <Tom_L> it has an index channel too
[21:02:43] <skunkworks> huh?
[21:03:13] <danimal_garage> still needs a tachometer
[21:03:19] <Tom_L> bak in a while
[21:03:21] <skunkworks> you don't need a tach. depends on thea mp
[21:03:22] <skunkworks> amp
[21:03:24] <Tom_L> gotta run
[21:03:28] -!- Tom_L has quit []
[21:03:31] <danimal_garage> that drive does
[21:03:38] <skunkworks> was there a link?
[21:03:42] <Tom_itx> sure was
[21:03:47] <danimal_garage> http://www.kollmorgen.com/website/com/eng/products/drives/ac_servo_drives/ps33_series_manuals.php
[21:04:16] <Tom_itx> well youall are trying to talk me into servos but i know little about em
[21:04:25] <danimal_garage> i have pretty much the same drives, it says you need a tach if you're using brushed motors
[21:04:59] <danimal_garage> you gotta start somewhere Tom_itx
[21:05:17] <Tom_itx> i did, now i'm steppin it up possibly
[21:05:18] <danimal_garage> i dont know much about them either. i learned when i converted my lathe
[21:05:36] <skunkworks> danimal_garage: a few people have actually taken the high quality velocity signal from the mesa encoder and sent it out of a +/-10v output to the servo drive
[21:05:59] <danimal_garage> skunkworks, that's awesome!
[21:15:24] -!- e3m [e3m!~IceChat7@] has joined #emc
[21:23:32] -!- Dannyboy has quit [Remote host closed the connection]
[21:32:28] -!- Guest790 [Guest790!~Guest790@78-83-51-185.spectrumnet.bg] has joined #emc
[21:44:09] -!- acemi has quit [Quit: WeeChat 0.3.2]
[21:45:32] -!- Guest790 has quit [Quit: Visitor from www.linuxcnc.org]
[22:04:39] -!- Dannyboy [Dannyboy!~Dan@] has joined #emc
[22:18:30] -!- skunkworks has quit [Ping timeout: 245 seconds]
[22:22:13] <andypugh> I wonder how nasty Mini-DIN would be for resolver connection? I want shielded 6-wire cable to PCB.
[22:22:40] -!- skunkworks [skunkworks!447329d2@gateway/web/freenode/ip.] has joined #emc
[22:23:16] <atmega> cat6/rj45?
[22:23:26] <andypugh> It looks like a straight choice between £25 Lemo and 40p Mini-DIN
[22:23:50] <andypugh> I don't think RJ suits the wire size/insulation/strand
[22:23:53] <skunkworks> logger_emc: bookmark
[22:23:53] <danimal_garage> resolver cable is usually individually sheilded i think
[22:23:54] <logger_emc> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2011-02-04.txt
[22:24:44] <danimal_garage> i use cat 6 for encoder feedback
[22:25:30] -!- tlab [tlab!~tlab@c-69-243-187-136.hsd1.in.comcast.net] has joined #emc
[22:27:42] <andypugh> crikey! http://uk.rs-online.com/web/search/searchBrowseAction.html?method=getProduct&R=7133475
[22:27:58] <andypugh> That's more expensive than normal
[22:32:10] -!- voxadam has quit [Quit: voxadam]
[22:33:15] <andypugh> on reflection, there is not a lot wrong with ye olde D-sub
[22:36:24] <PCW> DE-9 is a decent encoder/resolver connector
[22:37:21] <andypugh> I have some servos that use one for encoder, and the other for power and hall.
[22:37:42] <andypugh> Current/voltage rating of DB-9 is surprisingly high
[22:39:21] <andypugh> (300V, 7.5A)
[22:48:53] <Tom_itx> unless it would be in a place someone would mistake it for rs232
[22:50:09] <cradek> or ega
[22:50:38] <skunkworks> heh
[22:52:29] <Tom_itx> at 7.5A something would melt
[22:52:41] <andypugh> Hmm, 13W3 would work very nicely, wouldn't it? (though 0W3 would be better)
[22:54:46] <Tom_itx> newegg is cheaper on that atom than ebay
[22:54:47] -!- Fox_Muldr has quit [Ping timeout: 276 seconds]
[22:55:58] -!- Fox_Muldr [Fox_Muldr!quakeman@frnk-4d01d3c1.pool.mediaWays.net] has joined #emc
[23:03:30] -!- skunkworks has quit [Ping timeout: 245 seconds]
[23:04:53] -!- Valen [Valen!~Valen@ppp121-44-129-24.lns20.syd7.internode.on.net] has joined #emc
[23:05:00] -!- Dannyboy has quit [Remote host closed the connection]
[23:06:13] <danimal_garage> which atom?
[23:06:32] <danimal_garage> i wanna get an atom for my mill
[23:06:39] <danimal_garage> for the new enclosure
[23:07:14] <danimal_garage> does anyone boot from a memory card with an atom? i'd like to ditch the hdd
[23:08:01] <Valen> I dont but theres no reason you couldnt
[23:08:39] <Valen> main thing you need to watch for is write lifetime with those things
[23:08:44] <danimal_garage> ahh
[23:08:45] <andypugh> D501MO has onboard support for USB boot
[23:08:48] <danimal_garage> good point
[23:09:08] <andypugh> I can't find many actual devices though.
[23:09:45] <Valen> so anybody have any tips on my problem from before
[23:10:12] <Valen> when touching off it does it to a commanded position which is fine, but it'll do it to an offset from my encoder
[23:10:17] <andypugh> (It is an 8-pin header for a "flash drive")
[23:10:30] <Valen> IE i have .001 encoders but it'll touch off to .0003455
[23:10:49] <Valen> meaning i can never get .000 ferror making the machine always oscilate by one count
[23:11:13] <andypugh> Mine boots happily from a SATA DOM thing. (SSD that plugs directly into an SATA port)
[23:11:13] <danimal_garage> can't you adjust that by changing the deadband?
[23:11:35] <Valen> I'd rather have deadband at 0
[23:11:38] <danimal_garage> andypugh, thats good to know
[23:11:48] <Valen> as it will happily hit .000 when its not cutting
[23:12:10] -!- JT-Shop [JT-Shop!~chatzilla@216-41-143-50.semo.net] has joined #emc
[23:14:34] <Valen> it seems also that when playing with deadband it just leaves the motor settings as they were when it hits the deadband, and it just oscilates through it
[23:15:35] <Valen> (ie if its moving as it enters the deadband, it just keeps moving until it comes out the other side)
[23:15:37] <andypugh> danimal_garage: http://www.lambda-tek.com/componentshop/index.pl?origin=gbase4.5&prodID=B564243
[23:16:16] <andypugh> That's the supported form factor on the MB, but there is no reason you couldn't connect any old device with wires.
[23:16:28] <andypugh> I booted mine from a back panel connector with no problems.
[23:17:00] <andypugh> Or any SATA interfaced SSD will work too.
[23:18:01] <danimal_garage> which atom baords do you guys go with?
[23:18:39] <andypugh> D510MO
[23:18:51] <andypugh> But that might be old-hat now
[23:18:58] <danimal_garage> http://www.newegg.com/Product/Product.aspx?Item=N82E16813121398&cm_re=atom-_-13-121-398-_-Product
[23:19:15] <Valen> D945GCLF2
[23:20:48] <andypugh> I have http://www.newegg.com/Product/Product.aspx?Item=N82E16813121399&Tpk=D510MO
[23:21:18] <andypugh> There is a version with onboard LVDS for flat panel screens, which is potentially useful.
[23:22:34] <danimal_garage> cool, i think i was looking at that one before
[23:23:06] <danimal_garage> and you run emc on that mem card with no problems?
[23:23:20] <Valen> to use those though you need to be dealing with the bare panel though yes? not like a DVI interface?
[23:23:58] <Valen> its only 4Gb but that should be enough
[23:24:02] -!- theorbtwo has quit [Ping timeout: 276 seconds]
[23:24:08] <Valen> they don't give details about write life there though
[23:25:28] <andypugh> I am accidentally stress-testing my SSD, the machine has become a development machine. It must have compiled EMC2 5000 times in the last few months.
[23:25:42] <Valen> heh
[23:25:56] <Valen> doing better than I would have expected then lol
[23:27:07] <Valen> so any idea how to get touch off's to be based on the same resolution as the machine?
[23:27:30] <Valen> feels like it could be a bug
[23:27:34] <danimal_garage> always thought it was
[23:27:43] <Valen> doesn't seem to be
[23:27:45] -!- EDocToor has quit [Read error: Connection reset by peer]
[23:27:56] <andypugh> Valen: It does seem odd
[23:28:01] <Valen> do a touch off with more precision than your machine and look at your ferror
[23:28:05] <Valen> it can never be 0
[23:28:16] <cradek> touch off has nothing to do with ferror
[23:28:35] <andypugh> Can f-error ever be zero except by accident?
[23:28:47] <Valen> if you touch off based on a jogged position the "commanded position" can be in between counts of the encoder
[23:29:22] <cradek> yes and that's a feature
[23:29:35] <Valen> andypugh: yeah, if we manually touch off after doing a G53 to a .000 position ferror goes to 0
[23:29:40] <cradek> you can command positions between encoder counts
[23:30:02] <cradek> (but again it has nothing to do with touch off)
[23:30:08] <Valen> yes but it makes the mill dither around an encoder position as it can never hit 0 ferror
[23:30:33] <cradek> you fix that in your pid tuning, with a deadband setting
[23:30:50] -!- tlab has quit [Ping timeout: 240 seconds]
[23:31:06] <Valen> with deadband it just pushes through the deadband, if its moving into the deadband it just keeps moving until it comes out the other side then corrects it
[23:31:57] <Valen> so your saying you must always have a deadband of at least one encoder count to fix touch off being more precise than encoder counts
[23:31:57] <cradek> deadband will cause a zero pid output when the command and feedback are close enough
[23:32:08] <cradek> if your machine moves during that, then yes you'll see dither
[23:32:11] -!- odiug has quit [Ping timeout: 255 seconds]
[23:32:18] <cradek> fourth time: it has nothing to do with touch off
[23:32:19] -!- e3m has quit [Quit: Light travels faster then sound, which is why some people appear bright, until you hear them speak]
[23:32:30] <Valen> ok look here is how it does
[23:32:37] <cradek> it's a matter of commanding position that doesn't correspond exactly to an encoder count
[23:32:53] <cradek> what offsets are in effect are only very tangentially related to that
[23:33:05] <cradek> to say it's a bug in touch off is a gross misunderstanding of the issue
[23:33:21] <Valen> right which you can do during jog and touch off, meaning that after touch off all your G1 z23.000 are now aiming inbetween encoder counts
[23:33:25] <andypugh> cradek: I think his point is that you can touch-off such that position can _never_ exactly match counts.
[23:33:41] <Valen> it isnt "caused" by touch off
[23:33:52] <Valen> but that is what exposes the problem
[23:34:20] <cradek> no it isn't - if my encoders are .001 resolution and I write G0 G53 X.0005, it has nothing to do with touch off
[23:34:22] <andypugh> Though I would think you would get the same issues with, for example, a metric leadscrew and imperial machine.
[23:34:34] <Valen> cradek: thats correct
[23:35:00] <cradek> it's by no means generally true that all commonly programmed gcode positions have corresponding exact encoder counts
[23:35:23] <cradek> on my mill, the encoders are .001 mm, I program in .0001 inches
[23:35:33] <Valen> however if I jog, touch off, then go a G0 z100.000 the mill wont ever be able to reach 100.000 it will be 100.000345 or 99.000564
[23:35:59] <cradek> yep
[23:36:24] <cradek> because you've commanded a position between those two
[23:36:24] <Valen> so it will sit there hunting between those positions
[23:36:36] <cradek> then you should fix your pid tuning to solve that
[23:36:57] <Valen> I should fix my PID tuning to correct for setting a commanded value in between counts
[23:37:12] <andypugh> deadband = "that'll do , pig"
[23:37:28] <cradek> yes - your machine should come as close to the commanded position as possible
[23:37:29] <danimal_garage> lol
[23:37:30] <Valen> I have set deadbands and it doesn't actually seem to help
[23:37:40] -!- psha [psha!~Pavel@] has joined #emc
[23:37:43] <archivist> someone seems to be expecting accuracy to be equal to resolution
[23:37:54] <danimal_garage> perhaps your bias is off so it wanters when it's in the deadband?
[23:37:59] <danimal_garage> wanders*
[23:38:04] <Valen> its got nothing to do with accuracty archivist
[23:38:22] <Valen> there is a bias term in it danimal_garage
[23:38:29] <cradek> yes if a zero pid output causes the machine to move, it'll dither. danimal_garage is right.
[23:38:39] <danimal_garage> wow thats a first.
[23:38:40] <cradek> deadband won't work if zero pid output causes motion
[23:38:48] <Valen> it only moves if its moving when it enters the deadband
[23:38:52] <Valen> stiction
[23:39:19] <cradek> Valen: try getting it to dither, then fiddle your offset pot until it stops (or slows way down)
[23:39:39] <Valen> offset pot?
[23:39:50] <cradek> with deadband = 1/2 count
[23:40:06] <danimal_garage> he might not have a pot if his drives are like mine
[23:40:12] <Valen> deadband to 1/2 a count I hadn't thought of that
[23:40:18] <danimal_garage> i had to use the bias in EMC
[23:40:26] <cradek> Valen: see the pid manpage re deadband
[23:40:35] <Valen> I use mesa, so no pots
[23:40:42] <cradek> on the amp, I meant
[23:40:50] <danimal_garage> bias in the ini worked for me
[23:40:56] <Valen> no pots on amps, its all digital
[23:41:06] <danimal_garage> same here
[23:41:26] -!- Techrat [Techrat!~TRat@c-71-232-246-244.hsd1.ct.comcast.net] has joined #emc
[23:41:26] -!- Techrat has quit [Changing host]
[23:41:26] -!- Techrat [Techrat!~TRat@unaffiliated/techrat] has joined #emc
[23:41:32] <cradek> a curse!
[23:41:37] <danimal_garage> lol
[23:41:41] <danimal_garage> bias worked fine
[23:41:43] <Valen> it should make things better
[23:41:44] <cradek> yeah, use pid's bias then :-)
[23:41:47] <danimal_garage> i got mine dialed in
[23:42:00] <danimal_garage> it wandered before
[23:42:02] <Valen> I am using the bias in EMC to correct for some stupid axis
[23:42:15] <Valen> we have an air system to counterweight the Z axis
[23:42:21] <danimal_garage> ah
[23:42:43] <Valen> thing is if we only counterweight it we get massive oscilation Z up and glass smooth Z down
[23:42:49] <andypugh> air counterweight seems the ideal solution to me
[23:43:03] <Valen> so we wind the air up to half again whats needed for pure counterweight and it is smooth up and down
[23:43:04] -!- Guest584 [Guest584!~Guest584@78-83-51-185.spectrumnet.bg] has joined #emc
[23:43:16] <Valen> but needs a bias term back down to get the PID close
[23:43:25] <andypugh> (I intend fitting one, pushing on the bottom of the leadscrew, because I hate to do conventional)
[23:43:28] <danimal_garage> do you use an air piston with a flow control valve looped to both chambers of the psiton?
[23:43:33] <cradek> andypugh: mine's got a ton of lead - not so ideal, but sure is foolproof
[23:44:09] <andypugh> A weight-based counterweight doubles your inertia though.
[23:44:15] <Valen> we use a cylinder hooked up to a bigass fire extingusiher tank
[23:44:36] -!- theorbtwo [theorbtwo!~theorb@] has joined #emc
[23:44:46] <Valen> as a buffer
[23:44:53] <andypugh> If you use a relieving regulator, then you have a true constant-force counterbalance.
[23:45:23] <andypugh> (At the cost of some air consumtion)
[23:45:38] <Valen> variation due to position is about 3-5%
[23:45:56] <Valen> the buffer tank is large in relation to the cylinder
[23:45:58] -!- Guest584 has quit [Client Quit]
[23:46:13] <andypugh> That's another way, and probably quieter
[23:47:37] <andypugh> I want to find a high pressure, silent, compressor. My current one is a bit bulky, but nice and quiet. However my drawbar turned out to need 150psi, so I have the pressure switch adjusted to max, and that concerns me.
[23:47:54] <cradek> silent!?
[23:47:59] <Valen> 150PSI is a little high for ours, we use an air conditioner compressor
[23:48:19] <Valen> might do 150, never tried
[23:48:28] <andypugh> Well, quiet at least.
[23:48:31] <cradek> you can get to 175 with a common two stage piston compressor - but silent it is not
[23:48:32] <Valen> give it a little squirt of WD40 at the start of the day
[23:48:59] <cradek> quietness is proportional to radius^2
[23:49:08] <andypugh> I have hurt myself too many times recoiling in shock when my dad's direct drive compressor fires up.
[23:49:13] <Valen> lol
[23:49:22] <cradek> like bagpipes
[23:49:24] <Valen> thats why I like that one
[23:49:42] <Valen> we have a full sized one for when you want allot of air
[23:50:06] <Jymmm> Valen: 25+ CFM ?
[23:50:26] <Valen> nfi what the cfm on either of them is lol
[23:50:36] <Valen> the little one isn't bad though
[23:50:50] <Valen> it'll fill a small compressor tank in a few minutes
[23:51:00] <Valen> say its half of a regular (small) compressor
[23:51:07] <danimal_garage> 80 gallos plus or bust!
[23:51:11] <danimal_garage> gallons*
[23:51:31] -!- micges has quit [Quit: Ex-Chat]
[23:51:42] <andypugh> OK, this is just silly: http://cgi.ebay.co.uk/Super-silent-compressor-Jun-Air-Type-36-150-/150557013903
[23:52:10] <andypugh> Reminds me of a tractor-pulling machine, bolt on more V12s
[23:52:24] <Jymmm> I have this dinky 1gal compressor for nail/brad gun, tires, etc. barely enough volume and loud. I think I might just get rid of it and use the nitrogen tank instead.
[23:52:33] <Valen> V8's tractor pulling is an american thing, they wouldn't be refined enough for a V12
[23:53:33] <danimal_garage> getting a real compressor was the best thing i ever did for the shop
[23:53:44] <danimal_garage> MUCH quieter and runs less
[23:54:06] <andypugh> Valen: http://experiencewyre.files.wordpress.com/2010/09/tractor-pull.jpg
[23:54:12] <cradek> my lathe pushed me over the edge - it uses air like crazy
[23:54:29] <cradek> but yeah, it's nice for compressed air to be in the "solved problem" category
[23:54:59] <Jymmm> I barely even need a compressor at all. Blowing things off, I have a foot pump for tires, really no need. And now that I have the nitrogen tank I can fill tires that most compressors can't.
[23:55:19] <danimal_garage> mine developed a leak in the toolchanger, but the leak somehow dissapeared
[23:55:26] <danimal_garage> so it's quiet again
[23:55:54] <Jymmm> like leaks jsut magically disappear never to be thought of again
[23:56:02] -!- morfic has quit [Changing host]
[23:56:02] -!- morfic [morfic!~morfic@unaffiliated/morfic] has joined #emc
[23:56:03] <andypugh> I used to watch tractor pulling on Eurosport, my favourite quote was "He has taken off one of his Allison V12s to compete in the lighter class, it will be back on later for the unlimited class"
[23:56:21] <danimal_garage> i disconnected the collet chuck a long time ago... i never use collets in that lathe
[23:56:24] <Valen> how hard would it be to add an exponential to ff1?
[23:56:34] <Jymmm> danimal_garage: ah
[23:56:42] <andypugh> Valen: Easy. Why?
[23:56:57] <danimal_garage> jymmm, i think it wasa chip in there or something
[23:57:09] <Jymmm> danimal_garage: gotcha
[23:57:09] <danimal_garage> i dont know... but it went away and hasnt been back
[23:57:21] <Valen> i can tune to ~0 ferror at a given speed with FF1
[23:57:35] <Valen> but when i change the speed the ferror comes back
[23:57:42] <Jymmm> danimal_garage: I still like my 2000PSI at my disposal =)
[23:58:02] <cradek> Valen: amps are in velocity mode?
[23:58:09] <Valen> (i use I to take it out at the moment, but I'd rather "fix the problem" than take care of it afterwards)
[23:58:25] <Valen> cradek: again using mesa hardware, PID out == PWM to the motors
[23:58:42] <cradek> ok, more like torque mode then
[23:58:52] <Valen> I spose that may explain it
[23:59:04] <Valen> as velocity increases amps decrease for a given PWM out
[23:59:07] <cradek> that's harder to tune well
[23:59:11] <Valen> back EMF of the motor
[23:59:40] <Valen> so as velocity increases torque for a given PWM out decreases