#emc | Logs for 2011-08-01

[00:00:34] -!- syyl__ has quit [Quit: Leaving]
[00:01:53] -!- grandprix [grandprix!~What@] has parted #emc
[00:04:05] -!- theorb [theorb!~theorb@] has joined #emc
[00:04:17] -!- theorbtwo has quit [Ping timeout: 246 seconds]
[00:04:33] theorb is now known as theorbtwo
[00:28:13] <danimal_garage> JT-Shop: i am now
[00:28:19] <danimal_garage> whats up?
[00:58:15] -!- robh__ has quit [Ping timeout: 258 seconds]
[01:01:06] -!- FinboySlick has quit [Quit: Leaving.]
[01:06:10] -!- sumpfralle [sumpfralle!~lars@] has joined #emc
[01:08:01] -!- sumpfralle1 has quit [Ping timeout: 260 seconds]
[01:12:11] -!- Tom_itx has quit [Ping timeout: 246 seconds]
[01:12:52] -!- Tom_itx [Tom_itx!~Tl@unaffiliated/toml/x-013812] has joined #emc
[01:26:04] -!- mozmck has quit [Remote host closed the connection]
[01:26:08] -!- mozmck [mozmck!~moses@client-] has joined #emc
[02:09:06] -!- ries has quit [Quit: ries]
[02:12:06] -!- Tom_itx has quit []
[02:13:22] -!- Tom_itx [Tom_itx!~Tl@unaffiliated/toml/x-013812] has joined #emc
[02:22:04] -!- mendelbuild [mendelbuild!~MendelBui@77-95-78-56.bb.cust.hknett.no] has joined #emc
[02:22:43] -!- sumpfralle1 [sumpfralle1!~lars@e183155179.adsl.alicedsl.de] has joined #emc
[02:24:27] -!- sumpfralle has quit [Ping timeout: 240 seconds]
[03:05:36] -!- sumpfralle1 has quit [Ping timeout: 276 seconds]
[03:10:26] -!- factor [factor!~factor@] has joined #emc
[03:15:37] -!- danimal_garage has quit [Ping timeout: 246 seconds]
[03:16:10] -!- danimal_garage [danimal_garage!~kvirc@ip68-111-138-74.sd.sd.cox.net] has joined #emc
[03:18:36] -!- Jymmm has quit [Ping timeout: 276 seconds]
[03:23:03] -!- Jymmm [Jymmm!~jymmm@unaffiliated/jymmm] has joined #emc
[03:34:41] -!- jbunch has quit [Ping timeout: 240 seconds]
[03:35:35] -!- jbunch [jbunch!~jim@] has joined #emc
[04:02:14] -!- theos has quit [Ping timeout: 257 seconds]
[04:23:47] -!- theos [theos!~theos@unaffiliated/theos] has joined #emc
[04:26:28] -!- Guest67124 has quit [Ping timeout: 250 seconds]
[04:42:33] -!- ve7it has quit [Remote host closed the connection]
[05:04:48] -!- mhaberler [mhaberler!~mhaberler@extern-181.stiwoll.mah.priv.at] has joined #emc
[05:17:43] -!- awallin [awallin!~quassel@cs27057046.pp.htv.fi] has joined #emc
[05:39:38] -!- mhaberler has quit [Quit: mhaberler]
[05:41:55] -!- mhaberler [mhaberler!~mhaberler@macbook.stiwoll.mah.priv.at] has joined #emc
[05:43:12] -!- vladimirek [vladimirek!~vladimire@bband-dyn106.178-40-170.t-com.sk] has joined #emc
[05:45:07] -!- dvdm [dvdm!~dvdm@dsl-240-198-112.telkomadsl.co.za] has joined #emc
[05:49:40] <Loetmichel> mornin'
[06:05:57] <anonimasu> morning
[06:06:23] -!- IG-garage [IG-garage!~ig@unaffiliated/mazafaka] has joined #emc
[06:06:47] <IG-garage> anonimasu: here?
[06:07:31] <IG-garage> my sample workflow with lathe is on Youtube now, three videos - 10 min, 3 min, then 18 min.
[06:08:56] <anonimasu> hehe, nice your getting faster
[06:09:50] <anonimasu> im reading trajectory planner files
[06:13:57] <awallin> anonimasu: do write down in the wiki if you get it all sorted out in your head!
[06:16:25] <anonimasu> awallin: just from initial thought what keeps you from reading a few segments ahead and adding a extra decel segment if the calculated accel for the next move will exceed machine limits?
[06:18:06] <awallin> there is a desire to have feed override respond very quickly to the ongoing motion. that restricts how much you can plan ahead
[06:18:26] <awallin> but you could probably plan ahead and then throw it all away if the user changes feed override
[06:19:07] <nicko> anyone here a Pro-E user ?
[06:19:33] <toastydeath> i wish
[06:19:39] <toastydeath> i liked pro/e a lot better than solidworks
[06:20:01] <nicko> we're learning it at uni ... getting a student version soon
[06:20:02] <anonimasu> well, you can always make it re-calculate the move if it's flagged as a move inserted to decel the machine
[06:20:15] <anonimasu> awallin: does that make sense?
[06:20:38] <toastydeath> pro/e isn't really appropriate for learning
[06:20:50] <toastydeath> it's monsterous and nonintuitive
[06:20:52] <toastydeath> but a lot more powerful
[06:21:08] <anonimasu> ie instead of blending segment the next move before the direction change happens so you can keep within machine constraints
[06:21:19] <anonimasu> feed override really dosent matter your still limited in max accel.
[06:24:06] <anonimasu> what you do is simply add a move so that the next move can be made within constraints of the physical machine
[06:26:06] <IG-garage> youtube removed this long video
[06:27:10] * anonimasu yawns
[06:27:13] <anonimasu> http://www.imakenews.com/gefanuc/e_article000735539.cfm?x=b11,0,w
[06:27:18] <anonimasu> something interesting.. also
[06:27:52] <awallin> anonimasu: not sure.. this isn't exactly an easy problem to solve. cradek suggested the segmentqueue paper as a starting point for something new (sonya mcfarlane wrote that?)
[06:28:59] <awallin> the problem gets significantly easier if you restrict to 3-axis trivkins and no (or 0-100%) feed override.
[06:29:40] <anonimasu> yeah, it'd be a better solution but as I understood it's heavy to implement in emc because it's of the way things are implemented
[06:31:25] <awallin> it probably needs to be done by someone who really has a fast fast machine and needs it...
[06:31:59] <anonimasu> how fast is that? 15m/min
[06:32:17] <anonimasu> :]
[06:32:32] <awallin> not sure, if you can do >1g accelerations then its fast
[06:32:54] <anonimasu> probably im able to do so with the setup im having for the lathe, and for the mill
[06:35:10] <anonimasu> got a 30kg table moved by 1.9kw continous servos
[06:35:28] <anonimasu> with 9.5Nm cont and 13Nm peak
[06:36:34] <anonimasu> probably the best would be to re-implement the planner, but i guess improving it a bit would also work..
[06:37:38] -!- capricorn_one has quit [Remote host closed the connection]
[06:39:00] toastydeath is now known as princess-celeste
[06:39:36] princess-celeste is now known as princess-celesti
[06:39:44] princess-celesti is now known as princess-celeste
[06:41:25] princess-celeste is now known as rainbow-dash
[06:54:35] -!- nullie [nullie!~nullie@dhcp249-232.yandex.net] has joined #emc
[07:10:00] * anonimasu yawns
[07:10:13] <anonimasu> and back to flame cutting stuff at work -_-
[07:13:51] -!- factor has quit [Ping timeout: 255 seconds]
[07:16:43] <nicko> 'flame' is a high end linux based video editing system
[07:16:52] <nicko> that you 'cut' video with ...
[07:18:31] -!- vladimirek has quit [Remote host closed the connection]
[07:18:36] -!- awallin_ [awallin_!~quassel@2001:708:110:1020:224:7eff:feda:7c7d] has joined #emc
[07:24:32] -!- robh__ [robh__!~robert@5ace7001.bb.sky.com] has joined #emc
[07:29:09] <anonimasu> no it's a oxyfuel torch...
[07:30:32] * IG-garage uses L.I.V.E.S. to split the video. So far so good.
[07:32:47] -!- mendelbuild has quit [Ping timeout: 240 seconds]
[07:35:08] -!- IG-garage [IG-garage!~ig@unaffiliated/mazafaka] has parted #emc
[07:52:00] <Valen> acceleration is more important than top speed normally
[07:59:24] <nicko> on a technical drawing - what takes precendence >> a centre line or a hidden line ?
[07:59:35] <nicko> ...if they both occupied the same line that is
[08:00:31] <nicko> jeez
[08:00:37] <nicko> sorry - google answers it for me
[08:00:42] <nicko> hidden lines do
[08:01:39] -!- uwe_ has quit [Remote host closed the connection]
[08:16:43] -!- jbunch has quit [Read error: Connection reset by peer]
[08:16:52] -!- jbunch [jbunch!~jim@] has joined #emc
[08:32:35] -!- jbunch_ [jbunch_!~jim@] has joined #emc
[08:32:36] -!- jbunch has quit [Read error: Connection reset by peer]
[08:40:15] -!- jbunch [jbunch!~jim@] has joined #emc
[08:40:15] -!- jbunch_ has quit [Read error: Connection reset by peer]
[08:50:48] -!- syyl [syyl!~syyl@p4FD142CA.dip.t-dialin.net] has joined #emc
[08:54:52] -!- mhaberler has quit [Quit: mhaberler]
[09:17:58] -!- awallin has quit [Read error: Operation timed out]
[09:20:24] -!- awallin [awallin!~quassel@cs27057046.pp.htv.fi] has joined #emc
[09:27:35] <awallin_> did IG-garage have any new nice videos online? links?
[09:54:09] -!- mhaberler [mhaberler!~mhaberler@macbook.stiwoll.mah.priv.at] has joined #emc
[10:02:18] <anonimasu> I have no idea :S
[10:06:48] * anonimasu reads trajectory planning while having coffee
[10:07:27] * anonimasu wishes that cradek would show up
[10:10:52] -!- jbunch_ [jbunch_!~jim@] has joined #emc
[10:10:52] -!- jbunch has quit [Read error: Connection reset by peer]
[10:14:04] -!- jbunch [jbunch!~jim@] has joined #emc
[10:14:05] -!- jbunch_ has quit [Read error: Connection reset by peer]
[10:15:22] -!- maximilian_h has quit [Quit: Leaving.]
[10:17:29] -!- nicko has quit [Ping timeout: 250 seconds]
[10:20:01] -!- jbunch_ [jbunch_!~jim@] has joined #emc
[10:20:01] -!- jbunch has quit [Read error: Connection reset by peer]
[10:21:50] <awallin_> anonimasu: did you get anywhere with understanding blending as it is now in emc2, or blending with tolenrace?
[10:32:19] <anonimasu> a tiny bit
[10:34:30] <anonimasu> awallin_: from what I gather it begins decelerating like in a normal segment and when you are within accel limits(close enough to follow the path) you start blending the next segment in
[10:35:43] <anonimasu> and then it adds a blend velocity to the move
[10:37:58] <anonimasu> awallin_: I "think" that it would be possible to look a few moves ahead and calculate blending for thoose moves, even before they happen
[10:38:06] <anonimasu> the same way it's donw now
[10:38:07] <anonimasu> done
[10:38:34] <anonimasu> but probably there's a reason why it's not done atm.
[10:41:30] -!- factor [factor!~factor@] has joined #emc
[10:45:03] -!- toastyde1th [toastyde1th!~toast@c-69-140-223-139.hsd1.md.comcast.net] has joined #emc
[10:48:11] -!- rainbow-dash has quit [Ping timeout: 258 seconds]
[11:00:03] <anonimasu> toastyde1th: awake?
[11:00:39] <anonimasu> toastyde1th: http://see.stanford.edu/materials/aiircs223a/handout6_Trajectory.pdf
[11:04:10] -!- Valen has quit [Quit: Leaving.]
[11:45:48] -!- Loetmichel has quit [Ping timeout: 276 seconds]
[11:46:53] -!- mhaberler has quit [Quit: mhaberler]
[11:50:21] -!- Loetmichel [Loetmichel!Cylly@p54B13359.dip.t-dialin.net] has joined #emc
[12:04:00] -!- theos has quit [Ping timeout: 276 seconds]
[12:06:10] -!- skunkworks has quit [Ping timeout: 264 seconds]
[12:07:32] -!- theos [theos!~theos@unaffiliated/theos] has joined #emc
[12:18:08] -!- elmo40 [elmo40!~Elmo40@CPE0017ee09ce6d-CM001bd7a89c28.cpe.net.cable.rogers.com] has joined #emc
[12:32:02] -!- BSaboia [BSaboia!bd5aa03c@gateway/web/freenode/ip.] has joined #emc
[12:32:18] -!- BSaboia [BSaboia!bd5aa03c@gateway/web/freenode/ip.] has parted #emc
[12:44:19] -!- theos has quit [Disconnected by services]
[12:44:33] -!- theos [theos!~theos@unaffiliated/theos] has joined #emc
[12:44:58] -!- El_Matarife has quit [Quit: Nettalk6 - www.ntalk.de]
[12:49:19] -!- theos has quit [Disconnected by services]
[12:49:36] -!- theos [theos!~theos@unaffiliated/theos] has joined #emc
[12:56:30] -!- emcrules_mill has quit [Quit: Ex-Chat]
[12:56:50] -!- theos has quit [Ping timeout: 252 seconds]
[13:02:26] -!- JT-Shop_ [JT-Shop_!~chatzilla@216-41-156-49.semo.net] has joined #emc
[13:03:53] -!- JT-Shop has quit [Ping timeout: 250 seconds]
[13:04:07] JT-Shop_ is now known as JT-Shop
[13:09:34] -!- WalterN [WalterN!~walter@tiwake.com] has joined #emc
[13:11:46] -!- FinboySlick [FinboySlick!~shark@] has joined #emc
[13:16:09] <JT-Shop> hi ho it's back to work I go
[13:16:31] * JT-Shop goes back to changing caps out on the power supply
[13:17:14] <Jymmm> CODE WORD
[13:17:41] <JT-Shop> LOL
[13:27:55] -!- ries [ries!~ries@] has joined #emc
[13:32:30] -!- boot|necklad [boot|necklad!~tmgcleave@host-78-151-63-214.as13285.net] has joined #emc
[13:32:46] boot|necklad is now known as bootnecklad
[13:44:56] -!- skunkworks [skunkworks!447329d2@gateway/web/freenode/ip.] has joined #emc
[13:51:08] <JT-Shop> dang it on of the solder pads came off on the next to last cap and the trace is on the cap side :/
[14:03:23] -!- capricorn_one [capricorn_one!~capricorn@zima.linwin.com] has joined #emc
[14:07:01] -!- theos [theos!~theos@unaffiliated/theos] has joined #emc
[14:08:30] -!- awallin_ has quit [Remote host closed the connection]
[14:19:24] <pcw_home> Well if you have not torn the trace off the PCB yet sometimes you can gently tease it off the component lead
[14:19:26] <pcw_home> so you lost a un-attached pad on the solder side and the cap connects to a top side trace?
[14:19:27] -!- skunkworks_ [skunkworks_!~chatzilla@68-115-41-210.static.eucl.wi.charter.com] has joined #emc
[14:24:52] <skunkworks_> logger[psha]:
[14:25:20] <L84Supper> if the through hole is plated the solder will wick itself up the hole, if it's not plated you can solder a blob onto the cap lead, let it cool, then reheat the lead after stuffing the cap
[14:36:32] <JT-Shop> pcw_home: yes the small pad on the opposite side of the trace came off
[14:37:10] -!- acemi [acemi!~acemi@unaffiliated/acemi] has joined #emc
[14:37:18] <JT-Shop> there are two caps side by side that share the same trace so I put an 18 awg jumper between the two...
[14:37:59] <JT-Shop> L84Supper: I almost tried that but didn't know if it would work or not
[14:38:53] <pcw_home> If you did not tear out the plated through hole it usually works
[14:40:49] <pcw_home> if you still have the pad you can often solder it back on. or make a little copper wire ring of you worry about the cap pulling put
[14:41:00] <pcw_home> if you
[14:41:39] -!- vladimirek [vladimirek!~vladimire@bband-dyn106.178-40-170.t-com.sk] has joined #emc
[14:43:01] <JT-Shop> yes, the through hole looks like the plating is gone
[14:44:49] <pcw_home> Yeah so maybe your wire is the best thing
[14:46:54] <JT-Shop> ok, thanks
[14:50:46] <bootnecklad> oh look a PDP-11/23 :D http://marc.cleave.me.uk/pdp11/pdp11/001.jpg
[14:50:51] <bootnecklad> sat behind me :D
[14:52:09] -!- elmo40 has quit [Read error: Connection reset by peer]
[14:52:48] <jdhNC> bootnecklad: I still have two 11/73's in production use.
[14:53:25] <bootnecklad> oh nice!
[14:53:53] <bootnecklad> what do they do?
[14:54:04] <jdhNC> 'nice' is one way to put it, not what I would pick though.
[14:54:26] <jdhNC> ultrasonic data collection
[14:54:57] <bootnecklad> can I ask what for?
[14:55:10] <bootnecklad> I guess the term if it ain't broke don't fix it applys here.
[14:57:39] <pcw_home> I have DECstation doing DNS at work but thats _Modern_
[14:57:48] <jdhNC> it's not broken, but should still be replaced. The computer is fine but the analog electronics are not replaceable.
[14:58:18] <jdhNC> pcw: a mips box?
[14:58:23] <bootnecklad> Ooo
[14:58:25] <jdhNC> *bsd?
[14:58:50] <pcw_home> Yeah 5000/33 Personal DECStation NetBSD
[14:58:53] -!- Dave911 has quit [Ping timeout: 258 seconds]
[14:59:28] <jdhNC> at some point, I would have thought doing that was cool, now it just makes you a sorry old fart like me :)
[14:59:46] <pcw_home> Ha
[15:00:01] <bootnecklad> rofl
[15:00:27] <jdhNC> I used to have a vaxstation2000 with netbsd
[15:00:33] <bootnecklad> i've got an RX02 emulator up and running
[15:01:11] <jdhNC> you could just write the 1's and 0's down by hand,then erase as needed
[15:01:22] <pcw_home> Its got an old unpatched version of Netsbd but when script kiddies run their x86 specific exploits they usually just crash the app
[15:02:28] <bootnecklad> so this '11 is fitted with 256k memory
[15:03:06] <jdhNC> toss the HW and run it in a emu instead
[15:03:17] <jdhNC> mine run RT11XM+
[15:03:23] <pcw_home> I have VS2000 but its buried in the junk pile and I have not looked at if for a few years
[15:03:38] <bootnecklad> noooooo jdhNC
[15:03:58] <jdhNC> I netbooted mine with a mop server on linux
[15:04:12] <bootnecklad> err *not* really supposed to mention this I guess but mine runes RT11 V4
[15:04:13] <bootnecklad> i thin
[15:04:18] <bootnecklad> but i didn't say that :p
[15:11:17] -!- sumpfralle [sumpfralle!~lars@g226085221.adsl.alicedsl.de] has joined #emc
[15:12:28] -!- sumpfralle1 [sumpfralle1!~lars@] has joined #emc
[15:15:46] -!- sumpfralle has quit [Ping timeout: 264 seconds]
[15:19:42] <cncbasher> has anyone fitted a digispeed GX , with 7i43 .only getting a slow rpm any pointers ? (using pwm gen )
[15:21:16] -!- WalterN has quit [Read error: Connection reset by peer]
[15:25:57] -!- acemi has quit [Quit: WeeChat 0.3.2]
[15:29:03] -!- n2diy has quit [Read error: Operation timed out]
[15:29:29] -!- n2diy [n2diy!~darryl@] has joined #emc
[15:33:11] <skunkworks_> your using the software pwmgen though the 7i43?
[15:33:12] <anonimasu> heh.. I used to have a vaxstation too
[15:33:14] <anonimasu> :)
[15:33:45] * anonimasu just machined a 550mm long knife out of hardox 400
[15:34:24] <cncbasher> skunkworks > correct via the 7i43
[15:34:52] <skunkworks_> I think that will cause you issues... The 7i43 is only gettting updated at 1khz
[15:34:59] <skunkworks_> by default.
[15:35:16] <skunkworks_> why can't you used one of the pwm's on the 7i43?
[15:36:39] <skunkworks_> what frequency are you trying to run the software pwmgen?
[15:37:18] <cncbasher> the spindle already had digispeed controlling the vfd
[15:37:35] <cncbasher> i set pwm-gen to 50
[15:38:55] -!- Dave911 [Dave911!~chatzilla@] has joined #emc
[15:39:19] <skunkworks_> 50hz?
[15:40:00] <skunkworks_> that should go through the 7i43 just fine I would think
[15:40:22] -!- elmo40 [elmo40!~Elmo40@CPE0017ee09ce6d-CM001bd7a89c28.cpe.net.cable.rogers.com] has joined #emc
[15:41:19] <cncbasher> i believe the digispeed is 50Hz
[15:41:56] <cncbasher> so i set pwm_frequency50
[15:42:08] <skunkworks_> I would look at it in halscope and see what the siganl looks like
[15:42:42] <cncbasher> iv'e just try'd it seems to spin then stall back to a crawl after a few seconds
[15:43:48] <cncbasher> would it be better to build a pwm to analog input and ditch the digispeed ?
[15:44:15] <cncbasher> i dont think the 7i43 has analog out , but i may be wrong
[15:44:23] <skunkworks_> I would think the digispeed should work..
[15:44:44] <skunkworks_> you would have to buy a daughter board like the 7i33 for analog out
[15:45:18] -!- sumpfralle1 has quit [Quit: Leaving.]
[15:45:59] <cncbasher> yea thought as much , hence why i left the digispeed in
[15:46:53] <cncbasher> i'll see if i can find info on the vfd it may accept pwm directly
[15:46:53] <pcw_home> What frequency PWM does the digispeed want?
[15:46:54] <pcw_home> And are you usng the software PWMgen?
[15:47:42] <cncbasher> hi , digispeed is 1hz to 50hz
[15:47:57] <cncbasher> yes using software pwmgen
[15:48:41] <pcw_home> 1Hz! must have a really slow filter
[15:49:37] <cncbasher> well they are used under windows with mach
[15:49:39] <skunkworks_> is the pwmgen running at 1khz thread
[15:49:56] <pcw_home> The software PWMGEN should work at those frequencies. Did you enable the output pin
[15:50:21] <pcw_home> And how is the input wired is it and OPTO?
[15:50:31] <pcw_home> an OPTO
[15:50:51] <cncbasher> i'll paste bin my hal
[15:51:04] <cncbasher> output via 7i42ta
[15:51:05] <skunkworks_> if the pwmgen is running at 1khz thread and set to 50hz - that should give you 20 levels.
[15:51:24] <pcw_home> just need to make sure you have an IS_OUTPUT on the PWM pin
[15:51:38] <cncbasher> ok checking
[15:54:46] <anonimasu> what mini itx computers are good except for the d510mo ?
[15:55:16] -!- syyl_ [syyl_!~sg@p4FD142CA.dip.t-dialin.net] has joined #emc
[15:55:59] -!- JT-Shop has quit [Read error: Connection reset by peer]
[15:57:03] -!- JT-Shop [JT-Shop!~chatzilla@216-41-156-49.semo.net] has joined #emc
[16:03:45] -!- mhaberler [mhaberler!~mhaberler@macbook.stiwoll.mah.priv.at] has joined #emc
[16:05:54] <cncbasher> pcw_home > got not found so looking further hm2_7i43.0.gpio.039.is_output true
[16:09:07] <JT-Shop> anonimasu: I've got a D525 running 10.04 with 2.5, very slightly worse latency ... I'm not running anything atm with it but plan on putting one in the plasma when I move it over here
[16:10:26] <JT-Shop> well changing the capacitors in the power supply section of the Simodrive 611 didn't fix anything with the spindle or break anything which is good
[16:25:11] <skunkworks> that is goodish
[16:29:30] -!- acemi [acemi!~acemi@unaffiliated/acemi] has joined #emc
[16:30:14] packrat is now known as skinniestjeans
[16:33:01] -!- JT-Shop has quit [Read error: Connection reset by peer]
[16:33:33] -!- JT-Shop [JT-Shop!~chatzilla@216-41-156-49.semo.net] has joined #emc
[16:47:43] -!- hatch789 [hatch789!~hatch789@] has joined #emc
[16:50:03] -!- JT-Shop has quit [Ping timeout: 252 seconds]
[16:50:27] -!- Dave911 has quit [Ping timeout: 276 seconds]
[16:50:49] -!- JT-Shop_ [JT-Shop_!~chatzilla@216-41-156-49.semo.net] has joined #emc
[16:50:51] JT-Shop_ is now known as JT-Shop
[16:50:53] -!- mhaberler has quit [Quit: mhaberler]
[16:51:17] <pcw_home> JT-Shop: what is the actual drive fault?
[16:52:09] skinniestjeans is now known as packrat
[16:53:55] <anonimasu> the motors go up to 2800rpm
[16:54:52] <anonimasu> at 310v
[16:54:56] <anonimasu> maximum of my supply
[16:58:21] <anonimasu> they are running cold now with a bit more proper tuning :)
[16:58:26] <anonimasu> pcw_home: thanks for the help!
[16:58:27] <JT-Shop> pcw_home: it's pretty vague, "axis drive fault"
[16:58:28] <hatch789> hi guys I'm wondering if anyone here could help me please. I finally have the cash to purchase the required boards for my EMC retrofit of my Tree Milling Machine. I just want to be sure I am getting the right parts that I'll need for it.
[16:58:57] <hatch789> Here's the link to my Tree that I'm retrofitting for EMC: http://oupower.com/index.php?dir=_My_Projects/_Shop_Projects/Tree%20Journeyman%20200R%20Milling%20Machine
[16:59:08] <JT-Shop> I get the same message if I put S3000 M3 or sometimes when it gets warm during a rapid Z up
[16:59:31] <Spida> /last traj
[16:59:34] <Spida> /last traj
[16:59:41] <Spida> oergs.
[17:00:46] <anonimasu> hatch789: what kind of motors?
[17:01:10] <anonimasu> and drives?
[17:03:48] <anonimasu> http://ah.io23.net/ul/files/hardox_knife.jpg
[17:03:49] <anonimasu> :]
[17:04:15] -!- nullie has quit [Quit: Ex-Chat]
[17:04:56] <cradek> hatch789: I agree that looks like a resolver. but I can't tell anything about the motors/amps from your photos.
[17:04:59] <cradek> bbl
[17:06:43] <hatch789> they appear to be servo motors but we're not sure of the make. I actually have a serial number off of one of them hold on
[17:07:26] <danimal_garage> hi
[17:07:36] <alex4nder> danimal_garage: sup
[17:07:51] -!- IchGuckLive [IchGuckLive!~chatzilla@95-89-104-26-dynip.superkabel.de] has joined #emc
[17:08:27] <hatch789> cradek/ you still here?
[17:08:48] <danimal_garage> nada, you?
[17:09:00] <danimal_garage> hi JT-Shop
[17:09:11] <JT-Shop> hi danimal_garage
[17:09:36] <alex4nder> danimal_garage: eating a bagel, contemplating world domination
[17:09:54] <JT-Shop> Dan, have you been inside of any quad rear axles? do they have a differential like a car rear axle
[17:10:18] <Loetmichel> re @ home
[17:11:59] <danimal_garage> i have not John
[17:12:19] <danimal_garage> i suspect probably a trail quad might
[17:14:14] <danimal_garage> JT-Shop: :http://forums.atvconnection.com/can-am-brp/338142-open-rear-differential-limited-slip-locker.html
[17:14:50] -!- sumpfralle [sumpfralle!~lars@] has joined #emc
[17:14:51] <JT-Shop> I assume the quad would be a better match for the power (lack of) than the car axle
[17:14:59] <hatch789> I think this is all good on this side. My thoughts are that I need to rip everything out of the "computer" side of my machine but I think my "power" side is still just fine. Here's the page where I show the power side. http://oupower.com/index.php?dir=_My_Projects/_Shop_Projects/Tree%20Journeyman%20200R%20Milling%20Machine&PageNum=5
[17:15:18] <danimal_garage> polaris with turf mode
[17:15:29] <danimal_garage> ha maybe a lawn tractor
[17:15:32] <hatch789> I think PIC #2 is the driver boards for my servo motors. You can click the picture for a larger image.
[17:15:54] <danimal_garage> at least to scavange the rear diff parts
[17:16:00] <hatch789> If anyone can help me make a final determination on the parts I need, that would be great
[17:17:11] -!- Tom_itx has quit [Ping timeout: 240 seconds]
[17:17:52] <JT-Shop> Dan looking at it
[17:18:16] <danimal_garage> JT-Shop: if you want to make your own, there's also the differential in front wheel drive transmissions for cars
[17:18:18] -!- rooks [rooks!~rooks@102-bem-18.acn.waw.pl] has joined #emc
[17:18:30] <danimal_garage> fairly compact gearing
[17:18:49] <JT-Shop> hmmm, more to ponder
[17:18:52] <danimal_garage> the one from a 4l60-e would be good
[17:20:50] <danimal_garage> get a burned up tranny for cheap
[17:21:02] <hatch789> anyone know the guys user name on here who is in charge of Mesa Electronics?
[17:21:09] <danimal_garage> i wish i still had my spare kicking around
[17:21:23] <danimal_garage> pcw
[17:21:41] <hatch789> danimal_garage: ok Thank you
[17:22:02] <hatch789> danimal_garage: is he usually on most evenings?
[17:22:07] <JT-Shop> danimal_garage: there's a differential in the transmission?
[17:22:24] <danimal_garage> JT-Shop: yes
[17:22:35] <danimal_garage> it's usually called a final drive
[17:22:42] <JT-Shop> ok
[17:22:44] <JT-Shop> thanks
[17:22:49] <danimal_garage> the one i mentioned is probably 5" in diameter
[17:22:59] <danimal_garage> about 5-6" long
[17:23:47] <JT-Shop> that is small enough
[17:24:03] <danimal_garage> it has areas for thrust bearings on both sides
[17:24:16] <JT-Shop> hatch789: have you determined what your drives require for input?
[17:24:52] -!- Connor has quit [Ping timeout: 252 seconds]
[17:24:57] <danimal_garage> you'd have to cut your own axles
[17:25:08] <danimal_garage> but i know a place that can do it
[17:25:19] <JT-Shop> cut a spline?
[17:25:27] -!- theorb [theorb!~theorb@] has joined #emc
[17:25:30] -!- fatpandas has quit [Ping timeout: 260 seconds]
[17:25:31] <danimal_garage> possibly
[17:26:00] -!- Connor [Connor!~Connor@] has joined #emc
[17:26:07] <danimal_garage> i'd get some axles from a car rear end with the bolt patrtern you want, and cut them shorter and respline
[17:26:26] <danimal_garage> use the wheel bearings for those axles
[17:26:43] <danimal_garage> all you'd have to make is your own housing
[17:26:55] -!- krushia has quit [Ping timeout: 250 seconds]
[17:27:00] <danimal_garage> and a way to mount it
[17:27:04] <JT-Shop> all the trike conversions I've seen so far use sealed bearings everywhere
[17:27:17] -!- fatpandas [fatpandas!~fatpandas@c-98-219-42-32.hsd1.ga.comcast.net] has joined #emc
[17:27:29] <danimal_garage> wheel bearings are cheaper and proven
[17:27:33] <JT-Shop> and they pack the diff with grease so no oil is used
[17:27:38] <danimal_garage> ah
[17:27:50] <JT-Shop> with the opening for the sprocket you have to go that way
[17:29:03] <danimal_garage> ah
[17:29:57] -!- theorbtwo has quit [Ping timeout: 250 seconds]
[17:30:08] theorb is now known as theorbtwo
[17:31:21] <JT-Shop> scroll down a bit here http://www.triketalk.com/forum/trinity-trike/5295-trinity-silver-wing-trike.html
[17:34:34] <hatch789> JT-Shop: no, is there an easy way to do that? We are suspecting they are analog input. Look at page 5 of my project and page 8. I give some of the needed information there
[17:35:20] <hatch789> Page 5: http://oupower.com/index.php?dir=_My_Projects/_Shop_Projects/Tree%20Journeyman%20200R%20Milling%20Machine&PageNum=5
[17:35:25] <hatch789> Page 8: http://oupower.com/index.php?dir=_My_Projects/_Shop_Projects/Tree%20Journeyman%20200R%20Milling%20Machine&PageNum=8
[17:36:06] <cncbasher> ok i'm tearing my hair out anyone see a problem http://pastebin.com/1TC9Ayba
[17:36:38] <cncbasher> why no pwm output
[17:37:24] -!- Dave911 [Dave911!~chatzilla@] has joined #emc
[17:37:36] <cncbasher> as previous chat with pcw_home
[17:37:46] <JT-Shop> hatch789: I can't tell by looking at the pictures what you might have, have you determined what boards are connected to the servos (trace wires)?
[17:38:50] <hatch789> I think I have. It appears to be the 2 boards on page 5 -Pic #2
[17:39:31] <danimal_garage> differential braking? sounds sketchy, i'd rather have 2 disks
[17:39:46] <JT-Shop> Dan, yes me too
[17:39:48] <danimal_garage> not that you do a ton of braking in the rear i guess
[17:40:03] <hatch789> JT-Shop: it's hard to trace things (i'm sure you know) because of all the tight connections. I will try harder to trace that back tonight. It's an important questions I'm sure.
[17:40:03] <JT-Shop> even if only 25% comes from the rear wheels
[17:40:08] <danimal_garage> it wouldnt be hard to do twin brakes at the wheels
[17:40:12] -!- krushia [krushia!~krushia@h69-131-92-179.cntcnh.dsl.dynamic.tds.net] has joined #emc
[17:40:31] <danimal_garage> it would be a lot more withcar wheels
[17:40:45] <danimal_garage> probably closer to 40-50%
[17:40:51] -!- IchGuckLive [IchGuckLive!~chatzilla@95-89-104-26-dynip.superkabel.de] has parted #emc
[17:40:54] <hatch789> pcw_home: are you also user pcw?
[17:41:30] * JT-Shop needs to go make a part or take a nap...
[17:41:38] -!- ve7it [ve7it!~LawrenceG@S0106009027972e37.pk.shawcable.net] has joined #emc
[17:41:59] <danimal_garage> fry's had a mega sale on blu-ray players yesterday
[17:42:05] <pcw_home> cncbasher can you scope the pin?
[17:42:45] <danimal_garage> they're prettyy sweet, you can stream netflix and pandora through them
[17:44:01] <cncbasher> pcw_home>can do with halscope or a real one
[17:44:44] <pcw_home> real better (or just set it to 5Hz or so and use a LED)
[17:45:16] <cncbasher> pcw_home can you clarify making the pin an output , i may have the syntax incorrect
[17:45:41] <cncbasher> as it shows in dmesg as correct pin ,
[17:46:14] <cncbasher> yea can do with a led
[17:46:17] <JT-Shop> cncbasher: http://www.linuxcnc.org/docview/html/drivers_hostmot2.html#r1_12
[17:46:38] <JT-Shop> 1.12.2 Parameters
[17:46:53] <pcw_home> You were not clear before, you are not using a software PWMGEN, but rather the hardware PWMGen. These outputs will be enable automatically if the module is enabled
[17:47:04] <pcw_home> enabled
[17:47:39] <cncbasher> pwmgen within the 7i43 is what i'm using
[17:47:45] <anonimasu> what is the most sane way to wire limit switches?
[17:47:47] <anonimasu> (hard)
[17:48:01] <anonimasu> to the enable pin on the drives?
[17:48:07] <JT-Shop> anonimasu: N/C
[17:48:07] <anonimasu> in a nc setup
[17:48:24] <skunkworks> anonimasu: should cut power to the drives.
[17:48:35] <anonimasu> skunkworks: the enable pin does just that
[17:48:47] <skunkworks> I don't trust the enable pins...
[17:49:05] <skunkworks> and I am sure that is not 'code' for some definition of code...
[17:49:15] <pcw_home> And active high so a short to ground looks like and open switch
[17:49:24] <pcw_home> an open
[17:49:44] <anonimasu> yes
[17:49:57] <pcw_home> (and blows your 24V field power fuse)
[17:50:57] <anonimasu> complex..
[17:51:09] <anonimasu> looks like a safety relay is in place..
[17:52:08] <pcw_home> For safe drive shutdown I would have a relay + brake resistor across the DC power so you brake all motors in event of a fault
[17:52:56] <anonimasu> how big resistor do you think i would need?
[17:54:42] <pcw_home> Depends on how many axis and what the tolerable brake current could be per axis but say 50 Ohm or so maybe 50w (bigger if you like to cycle the power often)
[17:55:01] <anonimasu> well not under normal conditions
[17:55:15] <pcw_home> peak power is KW but short duration
[17:55:35] -!- nullie [nullie!~nullie@nullie.telenet.ru] has joined #emc
[17:55:45] <anonimasu> i have a feeling my wiring box will be too small.
[17:56:49] <pcw_home> the way the works is that the HBridge diodes make a 3 phase rectifier of the generated motor voltage regardless of any thing else being turned on
[17:56:58] <pcw_home> way this
[17:57:06] <anonimasu> hm what about 2x 50w 50ohm..
[17:57:40] <anonimasu> or 2x100ohm paralell to make 50ohm and disspasiate 25w each..
[17:58:10] <pcw_home> sure put them is series if you are not so brave (So ~1KW peak at 320V)
[17:58:23] <anonimasu> yep
[17:58:41] <anonimasu> question, can you run the drives on 400v?
[17:58:49] <anonimasu> if I figure I need more speed.
[17:58:56] <pcw_home> Just make sure that the relay id NC and contacts are open when powered up
[17:59:28] <anonimasu> err safely.
[18:00:01] <pcw_home> 400V should be OK but leaves little margin Absolute max is 450
[18:00:52] <anonimasu> the drives are not terribly expensive so I can survive with _less_ margin
[18:01:25] <anonimasu> how can that 450v happen if powered by 400v?
[18:01:54] <pcw_home> It can happen every time you deccelerate
[18:02:33] <danimal_garage> is there an inhibit pin on the drives?
[18:02:45] <hatch789> pcw_home: you're user pcw normally ...right?
[18:02:54] <anonimasu> there's a enable input
[18:03:04] <hatch789> pcw_home: if so I need to ask you a few questions about some boards I need to buy from you
[18:03:11] <pcw_home> in that case the drive acts as a boost converter, dumping energy back into your PS capacitor
[18:03:38] <danimal_garage> if the limit switch is normally closed, i would assume putting it in line with the ennable pin would make sense
[18:03:50] <pcw_home> hatch789 perhaps best to email sales@mesanet.com
[18:04:05] <danimal_garage> but that's according to my very limited kowledge
[18:04:07] <anonimasu> I used to have a smart cap for taking care of that in one of the systems i built
[18:04:46] -!- The_Ball has quit [Ping timeout: 260 seconds]
[18:05:11] <pcw_home> The brake resistor is designed for this
[18:05:48] <anonimasu> do i just buy one and wire up on the drive inputs?
[18:06:04] <pcw_home> but with Multi Axis a common brake could be used
[18:06:48] <anonimasu> well, they are on a common power supply
[18:07:03] <pcw_home> Yes its another ~50 Ohm but 250W or more resistor
[18:07:32] <anonimasu> im thinking im going to buy a dc motor brake resistor off my supplier off that kind of stuff
[18:08:26] <anonimasu> but how does it work? do I just have a resistor sitting on my power lines?
[18:08:35] <anonimasu> or do I need diodes and stuff for handling that?
[18:08:51] <anonimasu> to keep them from being loaded normally
[18:08:52] <cncbasher> pcw_home> ok got a pulse , but at steady rate not variable ,
[18:09:25] <cncbasher> so i'm presuming it's got to be scaling
[18:10:25] <anonimasu> pcw_home: (im talking about handling regenerative energy not the one from hitting estop)
[18:10:31] <anonimasu> or I dont have to worry if my cap is big enough
[18:11:30] -!- The_Ball [The_Ball!~The_Ball@123-2-12-83.static.dsl.dodo.com.au] has joined #emc
[18:12:16] -!- isssy [isssy!~isssy@78-83-51-185.spectrumnet.bg] has joined #emc
[18:13:26] -!- mendelbuild [mendelbuild!~MendelBui@77-95-78-56.bb.cust.hknett.no] has joined #emc
[18:14:03] <pcw_home> Your cap is unlikely to be big enough for a fast deccel with a heavy load
[18:15:00] <pcw_home> cncbasher: The rate is supposed to be constant, the duty cycle changes
[18:15:33] <anonimasu> so probably it's sane keeping the meter measuring peak on the power cables..
[18:15:49] <cncbasher> pcw_home>yea not thinking straight
[18:16:03] <cncbasher> it's buggin me
[18:16:41] <cncbasher> think i'll see if i can run this vfd directly with pwm
[18:17:44] <pcw_home> well you had the PWM rate set to 10 KHz that will probably fail with with your 1-50 Hz analog widget
[18:17:47] <cncbasher> pcw_home> at the moment using the digispeed to control the pot
[18:26:01] -!- newbynobi [newbynobi!~norbert@p4FC9B2F6.dip.t-dialin.net] has joined #emc
[18:27:40] <pcw_home> Yeah didn't you say that the digispeed expected 1-50 Hz PWM?
[18:27:44] -!- Tom_itx [Tom_itx!~Tl@unaffiliated/toml/x-013812] has joined #emc
[18:27:52] <Tom_itx> logger[psha]
[18:29:19] <Tom_itx> JT-Shop, did you fix your solder trace?
[18:29:53] <Tom_itx> nm. catching up here..
[18:35:38] <JT-Shop> Tom_itx: I jumped it out with some 18awg
[18:35:57] -!- mendelbuild has quit [Ping timeout: 276 seconds]
[18:36:16] <JT-Shop> seems to work as well as it did before the cap change so I didn't seem to cause any more damage
[18:36:54] <Tom_itx> is the trace on both sides of the board or just the bottom?
[18:37:01] -!- n2diy has quit [Quit: Ex-Chat]
[18:37:04] <Tom_itx> make sure the top is connected if it is on both
[18:38:24] -!- n2diy [n2diy!~darryl@] has joined #emc
[18:39:08] <cncbasher> pcw_home> 1 -50hz is correct , just downloaded the manual to check
[18:39:56] <JT-Shop> Tom_itx: the one I pulled off was on the trace is on the cap side and I pulled off the ring and hole plating on the bottom solder side
[18:46:06] <L84Supper> JT-Shop: hardware engineers have to fix mistakes on pcb's all the time using http://www.google.com/products/catalog?hl=en&safe=off&client=firefox-a&hs=Jmd&rls=org.mozilla:en-US:official&q=blue+wire+wire+wrap&bav=on.2,or.r_gc.r_pw.&biw=973&bih=855&um=1&ie=UTF-8&tbm=shop&cid=14225822260353982474&sa=X&ei=QvQ2TpWoKcuAsgKu55SICw&ved=0CEMQ8gIwBQ
[18:46:26] <L84Supper> should have shortened that url :)
[18:48:03] -!- mboehm [mboehm!~mike@95-90-230-242-dynip.superkabel.de] has joined #emc
[18:48:12] -!- syyl_ has quit [Quit: Verlassend]
[18:50:27] <danimal_garage> ugh, i should never have messed with this machine, i should have just left the 7i33 in there...
[18:50:44] <danimal_garage> now i have a runaway axis for some reason... and it's not even enabled
[18:51:40] <Tom_itx> lasso it
[18:52:40] <anonimasu> treat it like a kid, ignore it.
[18:53:14] * anonimasu is not serious.
[18:53:31] * Tom_itx O_o
[18:53:36] <danimal_garage> fixed it, forgot to change the firmware back
[18:53:44] <Tom_itx> user error
[18:54:02] <danimal_garage> already did a number on a home switch though
[18:54:04] <JT-Shop> PEBKAC
[18:54:17] <anonimasu> danimal_garage: is it premiere for that?
[18:54:32] <Tom_itx> JT-Shop, what do you use for a tool setter?
[18:54:52] <Tom_itx> i was looking at that renishaw thing and thought about making one of those
[18:55:01] <Tom_itx> it's rather similar to the touch probes
[18:55:22] <Tom_itx> only horizontal with a tool pad on it
[18:55:30] <skunkworks> danimal_garage: you need overtravel switches...
[18:55:38] <danimal_garage> seriously, how hard is it to touch off a tool?
[18:55:51] <Tom_itx> not
[18:56:03] <Tom_itx> it's more of a project than a need
[18:56:06] <anonimasu> Tom_itx: want to collaborate on designing one?
[18:56:10] <danimal_garage> skunkworks: what good would that have done with the wrong firmware? emc wasn't even on
[18:56:11] <Tom_itx> sure
[18:56:22] <anonimasu> 4 way triggerable?
[18:56:31] <Tom_itx> did you see the renishaw one?
[18:56:36] <danimal_garage> it would have just smashed the limit switches
[18:56:41] <anonimasu> no
[18:56:44] <Tom_itx> it can touch the top or any side
[18:56:45] <skunkworks> danimal_garage: the overtravel limits would be hooked though the estop loop killing power to the drives
[18:56:47] <Tom_itx> it's a round disk
[18:57:00] <anonimasu> ah that kind
[18:57:01] <JT-Shop> Tom_itx: on my mill I use one made by edge it's a dial indicator
[18:57:05] <danimal_garage> skunkworks: again, emc was off
[18:57:08] <anonimasu> http://img.directindustry.com/images_di/photo-g/tool-setter-for-cnc-machining-centers-354486.jpg
[18:57:08] <JT-Shop> works well for me
[18:57:42] <anonimasu> Tom_itx: it cant be hard to make that kind of thing
[18:57:53] <skunkworks> danimal_garage: doesn't matter - the limit switches should be hardwired into whatever is powering the amps..
[18:58:28] -!- awallin has quit [Ping timeout: 258 seconds]
[18:58:46] <anonimasu> Tom_itx: I have a idea,
[18:58:48] <skunkworks> I would have turned our machine inside out if I didn't have hard limit switches by now ;)
[18:59:22] <anonimasu> a contact plate on the inside and a slide with a spring return
[18:59:31] <Tom_itx> http://www.youtube.com/watch?v=8xWrT12NY_g
[18:59:37] <Tom_itx> there's the renishaw one
[18:59:39] <JT-Shop> Tom_itx: I use this one on the mills http://www.edgetechnologyproducts.com/pro-touch-off-gage.html
[19:00:27] <Tom_itx> i considered getting one of those 2" height things with an led and modding it
[19:00:38] <Tom_itx> they're like $175 though
[19:00:45] <anonimasu> mhm
[19:01:15] <anonimasu> i think a contact plate a a uniform pivot would work great.
[19:02:30] <Tom_itx> JT-Shop, did you see my probe since i added the adjuster screws?
[19:03:55] -!- vladimirek has quit [Remote host closed the connection]
[19:05:20] <anonimasu> but that setup wouldnt trigger on twise.
[19:05:22] <anonimasu> twist.
[19:07:43] <anonimasu> Tom_itx: whatever goes on inside it'd look like your probe does
[19:11:49] -!- hatch789 has quit [Quit: Leaving]
[19:13:27] -!- syyl_ [syyl_!~sg@p4FD142CA.dip.t-dialin.net] has joined #emc
[19:16:18] <JT-Shop> Tom_itx: I think so, but I spent the weekend in Memphis so my mind is not right yet...
[19:17:41] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/emc/probe_index.php
[19:17:44] <Tom_itx> last couple pics
[19:18:34] -!- newbynobi has quit [Quit: Verlassend]
[19:19:00] <Tom_itx> i checked around the plastic tip (no probe tip yet) and got the last word to stop moving almost completly
[19:20:01] <JT-Shop> cool, I see your moving the delrin around with your screws... mine will be on top just because it worked out that way
[19:20:12] -!- tom3p [tom3p!~tomp@74-93-88-241-Illinois.hfc.comcastbusiness.net] has joined #emc
[19:20:33] <JT-Shop> Tom_itx: looking good
[19:21:04] <Tom_itx> now to make one with a strain gage
[19:21:51] <Tom_itx> ~1/4 the size
[19:21:54] -!- chester88 has quit [Ping timeout: 276 seconds]
[19:22:56] -!- roberth_ [roberth_!~robert@5ace7001.bb.sky.com] has joined #emc
[19:25:25] -!- robh__ has quit [Ping timeout: 250 seconds]
[19:27:37] -!- nullie has quit [Quit: Ex-Chat]
[19:27:44] <JT-Shop> neat!
[19:28:25] <danimal_garage> yay mill is back in running shape again
[19:28:36] <danimal_garage> now no more messing with it
[19:29:26] <danimal_garage> or else
[19:30:47] -!- PCW [PCW!~chatzilla@] has joined #emc
[19:45:30] -!- isssy has quit [Ping timeout: 276 seconds]
[19:58:31] -!- awallin [awallin!~quassel@cs27057046.pp.htv.fi] has joined #emc
[20:01:10] * JT-Shop smacks danimal_garage with a frozen mackerel so he won't forget
[20:01:39] <JT-Shop> I wonder if I should put the drill press or the file cabinet next to my desk???
[20:01:41] <Loetmichel> o, JT-Shop: micromanipulator at hand?
[20:02:08] <Loetmichel> (thinking of the size of a mackerel) ;-)
[20:02:24] <danimal_garage> lol
[20:02:32] <danimal_garage> lunch time, bbl
[20:02:39] <Loetmichel> JT-Shop: definetly the drill press.
[20:02:58] -!- syyl_ has quit [Quit: Verlassend]
[20:02:59] <Loetmichel> the file cabinet can stand anywhere
[20:03:11] <Loetmichel> (if its a digital one ;-)
[20:03:20] <archivist> but then the swarf gets in the sandwiches on the desk
[20:04:09] <DaViruz> if you don't like swarf in your sandwich you're not hardened enough
[20:04:40] <archivist> I just have the fingermarks
[20:05:55] <Loetmichel> DaViruz: right said ;-)
[20:07:05] <Loetmichel> on the other side: MY sandeiches dont survive from kitchen to my desk. they get eaten on the way ;-)
[20:07:14] <Loetmichel> sandwiches
[20:07:53] <archivist> not enough breakfast error at line one, redo from start
[20:11:37] <JT-Shop> Yep, what he said works for me
[20:20:59] <Connor> Whoho, Got my Tooling today.. Vice, Parallel Bars, V-Blocks, Coring Head with Bars, Endmills, a few Bits
[20:29:32] -!- mendelbuild [mendelbuild!~MendelBui@77-95-78-56.bb.cust.hknett.no] has joined #emc
[20:30:24] -!- micges has quit [Ping timeout: 252 seconds]
[20:32:58] <anonimasu> where are all the m2 taps -_-
[20:33:27] <anonimasu> I have limits switches.
[20:33:28] <anonimasu> :)
[20:33:41] <anonimasu> they just need a bit of cleaning
[20:42:32] <anonimasu> tomorrow is wiring time for them
[20:43:03] <JT-Shop> cool
[20:43:40] <anonimasu> and maybe i'll machine one motor mount
[20:43:50] <anonimasu> and get one of the servos on the z axis :)
[20:44:12] -!- boot|necklad [boot|necklad!~tmgcleave@host-78-151-63-214.as13285.net] has joined #emc
[20:44:40] <anonimasu> I think, i will wire up the home switches, so I can home the absolute encoders with them at first startup.
[20:45:09] <anonimasu> is that nescessary or unnescessary work?
[20:45:29] <anonimasu> tho, it might be nice in case of swapping belts.
[20:46:24] <anonimasu> i wonder if it's possible to hack the homing stuff so that if there's no absolute offset set up that based on a home switch.
[20:47:54] -!- bootnecklad has quit [Ping timeout: 276 seconds]
[20:47:56] <danimal_garage> yay, got my 7i37 today, so at least i can start wiring the home and limit switches
[20:48:24] -!- bootnecklad [bootnecklad!~tmgcleave@host-78-151-63-214.as13285.net] has joined #emc
[20:48:31] <anonimasu> haha
[20:48:34] <anonimasu> why are they so important :D
[20:49:02] * anonimasu is joking
[20:49:29] <anonimasu> then there's the vfd.. via modbus..
[20:49:32] <anonimasu> or 0-10v.
[20:50:47] -!- boot|necklad has quit [Ping timeout: 240 seconds]
[20:52:37] <anonimasu> I have no idea what's better
[20:52:40] <anonimasu> or easier
[20:53:01] -!- mendelbuild has quit [Ping timeout: 240 seconds]
[20:53:08] <cpresser> i guess 0-10Volt is easier
[20:53:33] <cpresser> but via modbus, you are able to read back data of the VFD (for example the current load)
[20:53:44] <anonimasu> http://www.wallacecompany.com/machine_shop/EMC2/modbus/
[20:53:46] <cpresser> (at least, if your VFD supports such features)
[20:53:50] <anonimasu> it does
[20:53:50] -!- skunkworks has quit [Ping timeout: 252 seconds]
[20:53:57] <anonimasu> also, jogging of the spindle
[20:54:04] <anonimasu> and pid
[20:54:50] <anonimasu> I guess I could always spend time writing a realtime comp to handle serial communications.
[20:54:59] <anonimasu> even if it havent been done.
[20:59:11] <cpresser> i wonder if the serial port is fast enough for a closed loop PID
[20:59:47] <cpresser> modbus in userspace is already implemented more than one time
[21:00:20] <anonimasu> probably not.
[21:01:23] <anonimasu> might be better to use a rc filer and a opamp to generate the voltage for the vfd.
[21:02:34] <anonimasu> it's got a output port for error.
[21:02:40] <anonimasu> and a output one for at speed
[21:03:07] -!- FinboySlick has quit [Quit: Leaving.]
[21:03:26] <cpresser> since the complete chain has some delay (there is a controller in the VFD, mechanical momentum...) it might not matter how fast the serial port really is; the real speed is anyways measured by encoders
[21:03:52] <anonimasu> true.
[21:04:33] <cpresser> also: the 'at speed' output only tells you that the desired frequency is set; not how fast the motor rotates
[21:04:58] -!- chester88 has quit [Ping timeout: 258 seconds]
[21:04:59] <cpresser> an asyn AC-motor is slower than the frequency most of the time :)
[21:05:23] -!- tom3p [tom3p!~tomp@74-93-88-241-Illinois.hfc.comcastbusiness.net] has parted #emc
[21:07:55] <anonimasu> true.
[21:08:04] <JT-Shop> I'm pretty sure my "at-speed" comes from the spindle encoder...
[21:09:00] <anonimasu> that's true, but the vfd acknowledges that the setpoint has been reached
[21:09:09] <anonimasu> (from what it can tell)
[21:09:12] <cpresser> i was more refering to a vfd, not other hardware
[21:09:32] <anonimasu> I have a 0-5v input, can I pwm a opamp somehwo
[21:10:16] <anonimasu> or use rc filter from the mesa board to get that working
[21:11:05] <cpresser> my setup uses a RC-filter with a single op-amp stage (voltage follower) to generate a 0-10Volts out of a single PWM-pin
[21:11:38] <toastyde1th> anonimasu, re: trajectory planning
[21:11:48] <toastyde1th> i have no doubt that is a valid way of planning certain types of machines in certain operations
[21:11:51] <anonimasu> can I do the same, with 0-5v and just a rc filter?
[21:12:08] <ds3> wow, ball bearing chucks are big
[21:12:12] <toastyde1th> but that's not how fanuc, okuma, mitsubishi, mazak, ikeagi, nakamura-tome, etc do it
[21:12:17] <anonimasu> toastyde1th: im not looking at options to re-write the whole planner just tweak the current one a bit if it can be made to make it better...
[21:12:21] <toastyde1th> ah
[21:12:27] <toastyde1th> godspeed
[21:12:27] <anonimasu> I dont have that kind of skill
[21:12:29] <anonimasu> :)
[21:12:32] <cpresser> anonimasu: since the 0-5Volt input is most likely high impedance, that might work
[21:12:48] <anonimasu> 50k ohm
[21:12:58] <cpresser> but you cant be sure, how much current that input draws and how it affects your RC-filter
[21:13:11] <anonimasu> > 50k ohm
[21:13:17] <anonimasu> the manual states
[21:13:39] <anonimasu> toastyde1th: I wish I did, but im realistic
[21:14:02] <cpresser> if you source enough current from your pwm-pin that might work. but it also makes the RC-Filter slow
[21:14:41] <anonimasu> cpresser: well, I have a 5vI have a 7i47 as driver
[21:15:01] <anonimasu> the inputs will source 20mA
[21:15:12] <cpresser> you want your RC-filter to respond fast, so a smaller cap might work better. to be hones, i did my design by trial and error ^^
[21:15:28] <cpresser> i was much to lazy to to the calculation of the filter
[21:15:30] <toastyde1th> anonimasu, even a small improvement will drastically help a lot of people
[21:15:38] <anonimasu> toastyde1th: that's what I were figuring too :)
[21:16:04] <anonimasu> re-implementing the planner is probably a good thesis project for someone
[21:16:15] <toastyde1th> hahahaha
[21:16:33] <anonimasu> I dont have the math skills that's required
[21:17:19] <anonimasu> but talking from the mcfairlane paper it looks like they are calculcaing polynomials for blending to control the error.
[21:18:47] <JT-Shop> anonimasu: I though you were talking about "at-speed" for EMC
[21:19:21] <anonimasu> JT-Shop: yeah from the vfd, but since there will be a encoder as somoene mentinoed that's kind of pointless
[21:19:45] <anonimasu> too much coffee.
[21:30:01] <JT-Shop> yea, my drive as an at speed bit but like I said I just use the encoder velocity I think
[21:30:15] <JT-Shop> #spindle sync
[21:30:16] <JT-Shop> net spindle-index motion.spindle-index-enable <= hm2_5i20.0.encoder.02.index-enable
[21:30:18] <JT-Shop> net spindle-velocity motion.spindle-speed-in <= hm2_5i20.0.encoder.02.velocity
[21:30:20] <JT-Shop> net spindle-revs motion.spindle-revs <= hm2_5i20.0.encoder.02.position
[21:30:47] <anonimasu> i didnt wire it yet so that's why im wondering :)
[21:30:54] <JT-Shop> nope wrong section
[21:31:14] <JT-Shop> # spindle at speed
[21:31:16] <JT-Shop> setp near.0.difference 0.1
[21:31:17] <JT-Shop> net spindle-velocity near.0.in1
[21:31:19] <JT-Shop> net com-spindle-vel near.0.in2 <= motion.spindle-speed-out-rps
[21:31:20] <JT-Shop> net spindle-at-speed near.0.out <= motion.spindle-at-speed
[21:31:48] <anonimasu> http://sim.okawa-denshi.jp/en/PWMtool.php
[21:35:15] <anonimasu> at 15khz, I need a 200k ohm resistor and a 2.5f cap...
[21:35:21] <anonimasu> :)
[21:36:15] <anonimasu> and then the question is what the 100k load will do to that.
[21:36:25] <anonimasu> err 50k ohm.
[21:37:16] <anonimasu> tho feeding a opamp with that would be a good move.
[21:43:09] <cpresser> you may want to use a rail2rail opamp, so you dont need additional supply-voltages for the opamp
[21:44:27] <anonimasu> im doing calcs now, it looks like i dont have to
[21:48:23] <anonimasu> no, i can just use it as is
[21:49:39] <anonimasu> 3f and a resistor around 50k +/- a bit will work :)
[21:50:16] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust1037.basl.cable.virginmedia.com] has joined #emc
[21:50:16] -!- Fox_Muldr has quit [Read error: Connection reset by peer]
[21:50:39] -!- mendelbuild [mendelbuild!~MendelBui@77-95-78-56.bb.cust.hknett.no] has joined #emc
[21:53:25] -!- MOGLI [MOGLI!~MOGLI@] has joined #emc
[21:54:53] -!- Fox_Muldr [Fox_Muldr!quakeman@frnk-5f742614.pool.mediaWays.net] has joined #emc
[21:59:15] -!- ries has quit [Ping timeout: 250 seconds]
[22:03:27] -!- ries [ries!~ries@] has joined #emc
[22:08:03] -!- mboehm [mboehm!~mike@95-90-230-242-dynip.superkabel.de] has parted #emc
[22:08:53] -!- factor has quit [Read error: Connection reset by peer]
[22:11:14] -!- ries has quit [Ping timeout: 252 seconds]
[22:20:44] -!- ries [ries!~ries@] has joined #emc
[22:22:11] -!- roberth_ has quit [Ping timeout: 240 seconds]
[22:23:08] -!- factor [factor!~factor@] has joined #emc
[22:31:40] <L84Supper> anyone have a good source for granite tables/beds?
[22:31:55] <Tom_itx> a quarry
[22:31:58] <Tom_itx> sorry no
[22:32:11] <L84Supper> self serve?
[22:32:15] <Tom_itx> travers.com probably has them
[22:32:57] <Tom_itx> what size?
[22:33:11] -!- Gensor has quit [Ping timeout: 240 seconds]
[22:33:13] <L84Supper> 48" x 48"
[22:33:22] <L84Supper> x8" thick or so
[22:33:33] <Tom_itx> maybe not that thick
[22:33:42] <Tom_itx> i think the one we had was about 6"
[22:33:53] <L84Supper> tarvers go's to 24 x 36 x 4
[22:34:01] <L84Supper> Travers even
[22:34:02] <Tom_itx> http://www.msdiscount.com/columnar.aspx?cat_id=1524&category_site=STARTOOL
[22:34:05] <PCW> anonimasu: around 1K series resistance and 1 uF would get you about 98% of 5V with a 50K load
[22:34:37] <PCW> that at 100KHz would have less than .2% ripple
[22:34:47] <L84Supper> there we go x 8 x10 even x12
[22:35:03] <anonimasu> PCW: very neat!
[22:35:05] <anonimasu> :)
[22:35:18] <anonimasu> looks like my machine is getting closer and closer
[22:35:56] <PCW> An op amp is better (as is a reference supply for the PWM)
[22:36:47] <anonimasu> i dont get why that is better..
[22:37:36] <L84Supper> PCW: have you looked at any of the Actel smart fusion devices that have ARM M3 hard cpu's + 60K-500K gates?
[22:37:41] <PCW> thats how our PWM analog doohickies work (FPGA --> reference powered gate --> 2 to 5 pole filter)
[22:38:36] <PCW> I looked at Actel a long time ago, Now we would rather do our own CPUs
[22:39:16] <L84Supper> PCW: the difference with these is the new hard 100 MHz ARM cores vs soft cores
[22:40:16] -!- i_tarzan [i_tarzan!~i_tarzan@] has joined #emc
[22:40:41] <PCW> 100 MHz is dismal a softcore in a Spartan 6 can beat that easily
[22:41:13] <L84Supper> sure, but these are for lower cost <$10 in oem volumes devices
[22:41:55] <PCW> same for SP6
[22:42:29] -!- frallzor [frallzor!~frallzor@81-231-233-185-no56.tbcn.telia.com] has joined #emc
[22:42:54] <andypugh> Would this circuit work? (I have never got round to building it). http://www.linuxcnc.org/images/fbfiles/images/PWM_to_Voltage.png
[22:42:59] <L84Supper> have SP6's dropping price that much?
[22:43:10] <PCW> and whatever the SP7 is called sporenix or some such
[22:43:30] <L84Supper> PCW: how many gates does their softcore ARM cortex take up?
[22:44:10] <PCW> Dont know ARM is not FPGA optimized
[22:45:14] <L84Supper> do you use microBlaze or?
[22:45:27] <PCW> Andy, should work, I'd have to think about the funny slow off fast on characteristics of OPTOs there
[22:46:37] <PCW> WE normally use our own for DSP or comm work (tiny maybe 400 -700 LUTs depending on features)
[22:46:37] <anonimasu> PCW: we'll see if there's no real downside to running the vfd straight like that i might do so
[22:47:07] <L84Supper> PCW: ok, I see
[22:47:22] <andypugh> The main problem with that circuit is what it does with no connected input signal.
[22:47:28] -!- JT-Hardinge [JT-Hardinge!~chatzilla@216-41-156-49.semo.net] has joined #emc
[22:48:53] <PCW> anonimasu: Advantage of op amp is low output impedance so output does not vary with load,
[22:48:54] <PCW> advantage of reference is you do not depend on noisy varying 5V bu clean reference voltage
[22:49:28] <anonimasu> PCW: but the vfd shouldnt vary the input resistance at all
[22:50:11] <andypugh> If you run the spindle closed-loop you can get away with a dreadfully non-linear conversion, as long as it is monotonic.
[22:51:04] <PCW> Andy I think our simple PWM output circuit on the 7I47S uses a fast OPTO with a
[22:51:05] <PCW> pullup resistor driving a 4000 series gate (gated by an enable OPTO)
[22:51:17] <anonimasu> it's also always possible to write a comp that takes nonlinear curve and makes it linear
[22:51:46] <anonimasu> andypugh: how many weeks/years before the sserial stuff ends up in main?
[22:51:59] <anonimasu> and bldc..
[22:52:04] <andypugh> You think it's up to me?
[22:52:09] <PCW> up to the powers that be I guess
[22:52:47] <anonimasu> andypugh: you seem to be the person working on that :)
[22:53:08] <andypugh> There is actually an update/bugfix to sserial in the works at the moment. It needs a bit more sigining off by folk with hardware.
[22:53:46] <anonimasu> andypugh: where can I get that to test?
[22:54:06] <andypugh> I am working on a bunch of stuff. (SPI, UARTS, an ISA card) but if and when it will get accepted is a mystery to me.
[22:54:09] <anonimasu> im testing on emc 2.6.0-pre and official is 2.4.6 :)
[22:54:28] <anonimasu> is there a list of bugs for the bldc driver or something?
[22:54:47] <PCW> Seems to be working for Matt last time I heard (and we've run it here for 12 hours or so at 5KHz with no timeouts/crc errors)
[22:55:39] -!- theos has quit [Ping timeout: 258 seconds]
[22:55:49] <PCW> Andy, anonimasu has your latest sserial patches (and a v24 bitfile)
[22:55:51] <andypugh> Yes, I think it is probably ready to be submitted.
[22:56:01] <anonimasu> im not so extremely _into_ coding and stuff but i have a working motor now so i can test things if you need
[22:56:52] <andypugh> I would like to see it in 2.5 which is semi-imminent
[22:56:58] <anonimasu> me too
[22:57:22] <PCW> Only bug Ive noticed (and anonimasu also noticed) is the jump at the end of Q init which causes a FERROR
[22:57:24] -!- boot|necklad [boot|necklad!~tmgcleave@host-78-151-63-214.as13285.net] has joined #emc
[22:57:26] <andypugh> I am rather hoping that Matt will push it, when he is happy that it is an improvement.
[22:57:35] <anonimasu> it dosent happen now with the proper pid tuning
[22:57:45] <anonimasu> but my ferror is 100mm :D
[22:57:55] <PCW> Ha!
[22:58:02] <anonimasu> however there's still a jump.
[22:58:32] <andypugh> It might not f-error with the PID index-enable pin connected.
[22:58:32] <PCW> This is a BLDC thang not SSERIAL
[22:58:48] <anonimasu> im aware
[22:59:01] <anonimasu> I can try to squeeze it down tomorrow evening and see how big it is if that helps at all
[22:59:56] -!- bootnecklad has quit [Ping timeout: 260 seconds]
[23:00:47] <andypugh> I will try to have a look at it. But the intention is that PID and motion are off during initialisation. The init-done pin should be wired such that motion is not enabled during the init.
[23:01:12] <andypugh> (If there is a jump _after_ init-done is set, then that counts as a bug)
[23:01:33] <anonimasu> im using the sample config (5i20green
[23:01:41] <anonimasu> so it might not be wired anywhere
[23:02:03] <PCW> Well I think you can see it with halmeter, maybe it s just a matter of clearing the encoder so it matches the commanded position
[23:02:47] <PCW> before you turn control over to the PID loop
[23:02:57] <anonimasu> how do I do that?
[23:03:07] <anonimasu> stick a extra pid comp that watches the bldc init-done pin?
[23:03:11] <anonimasu> err extra comp..
[23:03:13] <anonimasu> sorry.
[23:03:19] <PCW> probably in the BLDC comp
[23:03:39] <jbunch_> has anybody played with feeback on your encoders? it appears that my new drives output a voltage in proportion to current.
[23:03:41] <anonimasu> or to my abs comp
[23:04:09] <anonimasu> tho, as that deprecates magnetic homing it's probably a non issue
[23:04:58] <PCW> Have to find the magic way to clear the encoder without messing with the raw counts
[23:05:33] <PCW> Dont know if the driver does that, would have to look at the source
[23:05:59] <anonimasu> anyway, bedtime, im gonna look tomorrow and tell you if i see what's happening in the scope
[23:06:38] <anonimasu> btw, when i do magnetic homing my emc position changes..
[23:07:01] <anonimasu> instead fo the motor homing and the axis position starting from 0
[23:07:10] <anonimasu> (I dont know if that's relevant)
[23:07:18] <anonimasu> anyway, bedtime
[23:08:15] <PCW> Yeah, however its done the commanded/actual need to match before control is handed over to the PID loop
[23:08:16] <PCW> 'nite
[23:08:30] <andypugh> I think rawcounts is sacrosanct in all the encoder interfaces. (ie it doesn't have any jumps in it, whereas counts oes)
[23:10:38] <PCW> Yes you never actually want to physically clear the counter (it breaks velocity estimation as well)
[23:12:06] <PCW> so is commanded position at 0 at EMC startup?
[23:12:59] <andypugh> I assume that it will be current position, as reported by the encoders.
[23:13:22] <andypugh> In fact, I am pretty sure i read that bit of code last week looking for something else.
[23:14:28] <PCW> you mean EMC moves its commanded position around following the encoders when not enabled?
[23:15:39] <danimal_garage> PCW: i just got the boards out in the mail
[23:15:48] <danimal_garage> you should have them in 2 days
[23:16:05] <andypugh> I think that is the case, yes.
[23:17:08] <PCW> danimal_garage is it possible you didn't re-wire your 7I33 connectors for the 7I48 (they have very different pinouts)
[23:17:08] <danimal_garage> i think i want to skip the 7i48 and just go with the 7i33, i can't afford to have the machine down again
[23:17:16] -!- Valen [Valen!~Valen@ppp121-44-39-100.lns20.syd6.internode.on.net] has joined #emc
[23:17:47] <danimal_garage> no, it is not possible.
[23:18:18] <PCW> We tried to break one today, 6V PS in and you have to short out 3 analog outputs to get into trouble
[23:18:21] <danimal_garage> everything was moving fine before the smoke, aside from the pulsation
[23:18:53] <ds3> pushing genies back into the bottle?
[23:18:56] <PCW> Pulsation does indicate somthing was wrong
[23:19:11] boot|necklad is now known as bootnecklad
[23:19:42] <PCW> The 7I48 has cleaner outputs than the 7I33 (5 pole vs 2 pole filter)
[23:19:43] <danimal_garage> i tripple checked the wiring since the board wasnt marked
[23:20:33] <danimal_garage> the encoders run under their own power (5v), and it seemed to keep count ok, for as long as i could test it before the pop
[23:21:29] <danimal_garage> it was all 3 axii
[23:21:33] <PCW> Well we will check it out, Thats the first return we've had on the 7I48
[23:23:06] <PCW> andy The only reason I think its not the case is then I dont think there would be a jump when PID was enabled
[23:23:15] -!- syyl has quit [Read error: Connection reset by peer]
[23:23:52] <PCW> andy: interested in looking at the parameter discovery stuff?
[23:24:04] <Valen> danimal_garage: break something?
[23:24:22] <PCW> we would really like this is 2.5 if theres any chance
[23:24:29] <PCW> in 2.5
[23:25:28] <andypugh> You might need to find someone with push access if you are in a hurry.
[23:25:37] <danimal_garage> I'll buy a 7i33 for now just so i can get the thing running... if it was deemed my fault for the damage, then just send it back broken, i'll scavange it for terminals or something
[23:26:44] <Valen> danimal_garage: thats the pport one yes?
[23:27:04] <danimal_garage> no
[23:28:09] <Valen> ahh duh
[23:28:13] <Valen> my bad
[23:31:14] <PCW> No its under warranty I just want to know as much as possible in case there is a card problem
[23:33:01] <PCW> You have to do something pretty egregious to violate warranty (or be a regular offender)
[23:34:48] -!- acemi has quit [Quit: WeeChat 0.3.2]
[23:35:53] <PCW> Rev25 and > SSLBP firmware supports parameter discover on those remotes that support it (everything but the 8I20 and 7I64 and eventually they will as well)
[23:36:08] <Tom_itx> you should send him a plaque back as being the firstr
[23:36:32] <andypugh> Send me the docs, I can look at it.
[23:37:02] <andypugh> Have you had the chance to do any testing on the 7i65 driver?
[23:37:26] <andypugh> (Also, I am puzzled how 8i20 can ever document itself...)
[23:37:54] <PCW> OK will do tommorow
[23:37:56] <PCW> I need to set that up, now that I have a working 2.6 dev setup ill try sometime this week
[23:40:15] <danimal_garage> appreciate it, but if i screwed up somewhere, i don't expect a freebie.
[23:40:53] <PCW> parameter discovery works like this
[23:40:55] <PCW> in user 2 after start on remotes that support discovery you will have 3 new parameters
[23:40:56] <PCW> 1. byte # parms ,
[23:40:58] <PCW> 2. word: offset to 1st Parm,
[23:40:59] <PCW> 3. byte parm record stride
[23:41:28] <danimal_garage> however i am confident in the wiring. The only thing i noticed is the polarity for the analog out was reversed from the 7i33 i believe
[23:43:15] <PCW> each parm has an associated record with
[23:43:16] <PCW> 1 size in bits
[23:43:18] <PCW> 2 type (bits,signed unsigned) with input/output flag
[23:43:20] <PCW> 3 Name
[23:44:31] <andypugh> OK, so you can read out lots of info from the card firmware?
[23:44:58] <andypugh> I suppose there is space in there for a decent novel, now I think about it.
[23:45:26] <PCW> with this information its possible for the driver to create and name all the HAL pins without specific knowledge of the remote
[23:45:27] <PCW> all new cards will be like this so the driver need not change to accomodate new remotes
[23:45:36] <andypugh> Sounds good.
[23:46:34] <andypugh> Is there info on scaling, max, min, that sort of thing?
[23:46:39] <PCW> Yeah there are certain rules like data is packed in the mailbox (users) registers in the parameter order
[23:47:29] <PCW> We have room for that open to (polite) suggestions...
[23:48:14] <PCW> also funny bit sizes will be rounded to bytes in the mailbox reg packing
[23:49:22] <Tom_itx> this 'discovery' will be on the anything io boards?
[23:49:38] <PCW> SSERIAL remotes
[23:49:44] <Tom_itx> ok
[23:50:10] <PCW> I wish we had done this on daughter cards...
[23:50:24] <Tom_itx> make your life easier
[23:51:47] <andypugh> Depending on how complicated you want to make it, you could copy the ASAP2 conventions.
[23:52:00] <PCW> Have a plan for next gen DBs with a serial EEPROM on each but haven't got to it yet
[23:52:55] -!- tlab0 [tlab0!~tlab@c-98-223-20-100.hsd1.in.comcast.net] has joined #emc
[23:53:35] <andypugh> That defines byte length and signing, max and min in engineering units, then a choice of encoding (none, slope/offset, inverse slope/offset, polynomial) and the parameters of that scaling.
[23:53:52] <PCW> Thst looks a little too big for this
[23:54:49] <andypugh> Almost everything ends up being scale/offset. And there is always the option of scale=1, offset=0
[23:58:57] <PCW> And its either a lot of text or floating point numbers neither of which are easy to accomodate