#emc | Logs for 2007-04-03

[00:55:42] <ejholmgren> zomg ... netslpit
[00:59:20] <ejholmgren> http://www.homestead.com/tool20895/CABIN1.html
[01:01:02] <toastydeath> wow that is impressive hootnannery
[01:01:37] <toastydeath> also I have determined I hate CNC lathes without tailstocks
[01:02:23] <jmkasunich> I like the 2nd pic down in the left column
[01:02:56] <jmkasunich> I bet it works ;-)
[01:10:01] <ejholmgren> hrrmmnn...
[01:10:06] <ejholmgren> could be dangerous ;)
[01:10:16] <eric_u> you go first
[01:10:30] <ejholmgren> as operator ... or target
[01:10:46] <ejholmgren> sure as hell doesn't look like nerf =)
[01:10:58] <jmkasunich> eric_u: we're talking about http://www.homestead.com/tool20895/files/22cal_gattling_gun.jpg
[01:13:02] <ejholmgren> I still have yet to figure out what exactly that first one is
[01:13:37] <jmkasunich> which first one? the top left pic?
[01:13:39] <skunkworks> looks like the start of the second picture - or similar
[01:13:48] <jmkasunich> its the crankcase for the top right
[01:14:10] <ejholmgren> ah
[01:14:55] <ejholmgren> that configuration is usually used in aircraft, right?
[01:14:59] <jmkasunich> yeah
[01:15:05] <jmkasunich> was rather than is
[01:15:06] <ejholmgren> what's the technical term?
[01:15:15] <jmkasunich> radial engine
[01:15:21] <toastydeath> tru.
[01:15:33] <jmkasunich> think WW2 bombers and DC-3s
[01:15:42] <ejholmgren> what's the benefit
[01:15:49] <jmkasunich> compared to a V-12 or something?
[01:15:52] <toastydeath> torque
[01:15:57] <toastydeath> low rpm torque
[01:16:02] <jmkasunich> good cooling by the airstream, light weight
[01:16:16] <toastydeath> and even then a lot of figher aircraft used V12s
[01:16:25] <ejholmgren> why is the sky blue?
[01:16:30] <toastydeath> rayleigh scattering
[01:16:32] <ejholmgren> I kid >:)
[01:16:35] <toastydeath> oh
[01:16:39] <jmkasunich> for fighters the narrow profile of a V helped with drag
[01:16:44] <jmkasunich> for bombers, HP was more important
[01:17:40] <jmkasunich> they made a 4 row one (28 total cylinders) that made 3000 HP from 4360 cu in
[01:17:52] <skunkworks> I want to buld one of these - http://www.homestead.com/tool20895/files/dif_engine.jpg
[01:17:57] <skunkworks> then a 4 cylinder bigger one.
[01:18:01] <jmkasunich> oh, actually later versions made 4300 HP
[01:19:32] <skunkworks> a free piston one would be really cool - but I want to start easy ;)
[01:27:06] <jmkasunich> http://en.wikipedia.org/wiki/Pratt_%26_Whitney_R-4360
[01:27:17] <jmkasunich> amazing technology for 60 years ago
[01:27:32] <jmkasunich> 1.11 HP/lb
[01:30:16] <ejholmgren> close to 1hp/cu inch as well
[02:00:02] <tomp> jmkasunich: i dont know where i got 'em but there are prints for the 'edwards radial 5' available (I'm looking at a hard copy now). the 'model air museum says they have them but i cant find them yet http://engine-museum.com/. and here's a radial bike http://www.rotecradialengines.com/Osh/Osh06/3RotecTeam0.htm
[02:04:12] <tomp> ah, i might have got them here http://groups.yahoo.com/group/R_and_R_engines/ ( same guy drew this set ) very detailed, 36 drawings.
[02:44:02] <eric_u> my kinematics prof used to joke that he took a job at Va. Tech because they had a cutaway radial aircraft engine
[02:44:25] <eric_u> that thing was pretty neat, I hope they still have it.
[02:47:12] <eric_u> omg, the book from that course now costs $128
[02:50:08] <CIA-19> 03cradek 07TRUNK * 10emc2/configs/sim/axis.ini: support for starting a user-configurable editor on the loaded program or tool table
[02:50:08] <CIA-19> 03cradek 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: support for starting a user-configurable editor on the loaded program or tool table
[02:50:11] <CIA-19> 03cradek 07TRUNK * 10emc2/share/axis/tcl/axis.tcl: support for starting a user-configurable editor on the loaded program or tool table
[02:59:48] <CIA-19> 03cradek 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: removed an entry above, so this index has to change
[03:10:21] <CIA-19> 03cradek 07TRUNK * 10emc2/lib/python/rs274/author.py: use tolerance mode
[03:10:19] <CIA-19> 03cradek 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/image-to-gcode.py: use tolerance mode
[03:15:43] <ejholmgren> http://www.grizzly.com/products/category.aspx?key=78
[03:15:49] <ejholmgren> what pray tell do those do?
[03:16:37] <eric_u> cut things
[03:16:47] <eric_u> those are planer heads, right?
[03:17:05] <eric_u> they are just using machining inserts to do their job instead of blades
[03:17:58] <ejholmgren> that's a lot of inserts
[03:18:52] <eric_u> probably just enough to do a decent job
[03:55:36] <steve_stallings> steve_stallings is now known as steves_logging
[04:03:27] <bytecolor> any other tkinter gurus besides jepler about?
[04:26:25] <Jymmm> Jymmm is now known as Jymmmmmmmmm
[04:27:06] <Jymmmmmmmmm> Jymmmmmmmmm is now known as MrAsshole
[04:27:33] <Skullworks_> Sir Mr A
[04:27:55] <MrAsshole> MrAsshole is now known as Jymmm
[04:29:51] <Jymmm> Skullworks_ yes sir?
[04:30:17] <Skullworks_> were you looking for me earlier?
[04:30:44] <Skullworks_> Skullworks_ is now known as Skullworks_PGAB
[04:31:29] <Skullworks_PGAB> NIC changed while power company yanked my meter...
[04:32:27] <Skullworks_PGAB> lazy SOB's put in wireless transmitters so they only have to drive down the alley to "collect" meter readings.
[04:32:52] <Skullworks_PGAB> short range like bluetooth.
[04:33:44] <Jymmm> Yeah, should make less errors on the meter reaidng.
[04:37:30] <Jymmm> I say put transmitters all over and make everyones nill go negative =
[04:39:33] <Skullworks_PGAB> To bad there wasn't a way to pull a Tesla... Figure out how to absorb all those emissions and use it as a remote power source.
[04:40:08] <Skullworks_PGAB> Old Nic was a crazy coot
[04:41:56] <Skullworks_PGAB> Burned down the city power plant with one of his experiments, then build the city a replacement generator that could make the amount of power he reuired.
[07:07:52] <Jymmm> Jymmm is now known as Red70sShow
[07:07:52] <Red70sShow> Red70sShow is now known as Jymmm
[07:30:20] <ds3> Hmmm
[07:30:43] <alex_joni> wwwwH
[07:30:48] <Jymmm> ?
[07:31:06] <Jymmm> mmmyJ
[07:31:18] <ds3> no one at ESC?
[07:31:27] <Jymmm> just SWPadnos
[07:31:30] <ds3> in SJ for ESC I mean
[07:31:31] <ds3> oh
[07:31:43] <Jymmm> I have an invitation, but no time
[07:33:04] <ds3> oh
[07:40:34] <Jymmm> i think if you register online, you can get in for free.
[07:40:43] <Jymmm> "A $95 value"
[07:41:46] <Jymmm> SWPadnos ears ringing?
[07:53:00] <ds3> yeah, but I got the post card
[08:03:32] <Jymmm> ds3 the big tri-fold one?
[08:09:46] <Jymmm> anonimasu
[08:12:29] <alex_joni> Jymmm:
[08:12:38] <Jymmm> Mom?
[08:13:06] <alex_joni> not here
[08:13:18] <Jymmm> Son!
[08:13:28] <Jymmm> (daughter?)
[08:13:38] <Dallur> You guys need help
[08:13:51] <Jymmm> Dallur you're here, what more do we need
[08:14:41] <Dallur> coffee ?
[08:14:57] <Jymmm> Got coco
[08:17:06] <Dallur> hmmm. a complete plasma welding system on ebay for $1000,,,, tempting
[08:18:02] <Dallur> http://cgi.ebay.com/THERMAL-DYNAMICS-PLASMA-WELDING-SYTEM-NEW-PRICE_W0QQitemZ260102413273QQcategoryZ25295QQrdZ1QQcmdZViewItem
[08:22:47] <alex_joni> Dallur: indeed :D
[08:23:16] <alex_joni> but it's in the US
[08:25:25] <ds3> Jymmm: yep
[08:25:48] <Jymmm> ds3 You should go harass Steve
[08:25:56] <ds3> Steve? who's steve?
[08:26:07] <Jymmm> SWPadnos_ = Steve, he's at the ESC
[08:26:13] <ds3> Oh. didn't know his name
[08:26:19] <ds3> yeah, I am going there
[08:26:45] <alex_joni> Stephen Wille Padnos :)
[08:27:20] <ds3> Ohhhhhh... thought it was something like Software Padnos..
[08:29:44] <alex_joni> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Stephen_Wille_Padnos
[09:04:32] <anonimasu> what?
[09:04:36] <anonimasu> hi
[09:04:37] <anonimasu> lunch
[09:06:36] <alex_joni> good idea :P
[09:07:25] <lerneaen_hydra> 'lo all
[09:09:43] <alex_joni> hi
[10:07:33] <alex_joni> Bananasplit sterling.freenode.net <-> irc.freenode.net quits: skunkworks
[10:23:52] <anonimasu> bzzt*click*
[11:45:23] <alex_joni> bzzt*dang*
[11:56:42] <a-l-p-h-a> bzzt*dead*
[11:56:48] <a-l-p-h-a> electrocuted.
[12:05:40] <Skullworks_PGAB> must go to work :(
[12:37:06] <a-l-p-h-a> same...
[12:37:13] <a-l-p-h-a> somehow I joined the working world.
[12:52:21] <anonimasu> nice
[12:56:27] <alex_joni> a-l-p-h-a: welcome
[12:57:13] <tomp> who had the h bridge-on-a-cap design?
[12:57:22] <alex_joni> skunkworks did
[12:57:27] <tomp> thanks
[12:57:49] <alex_joni> maybe he should patent the name :D
[12:57:58] <alex_joni> h-bridge-on-a-cap
[12:58:03] <tomp> skunkworks: how many watts was your h bridge? :)
[12:58:27] <alex_joni> * alex_joni thinks 4-500
[12:59:59] <tomp> could i drive them in pwm right from the mesa 5i20? ( or opto drive them )
[13:00:13] <alex_joni> skunkworks drove them directly from the parport
[13:00:22] <alex_joni> so I see no problem in doing that from the mesa
[13:00:28] <tomp> wicked, thank you very much
[13:00:33] <alex_joni> I'd recommend optos though
[13:00:41] <tomp> :)
[13:00:44] <tomp> will do
[13:00:50] <alex_joni> 4-500W meets 2-300$
[13:00:52] <alex_joni> :P
[13:00:59] <anonimasu> LOL
[13:01:22] <alex_joni> anonimasu: ^_^
[13:02:35] <skunkworks> mesa is pwm+dir - my bridge was pwm+pwm. hoping someone would make that an option for the mesa.
[13:03:28] <alex_joni> yeah, but you can generate the second pwm from the first.. right_
[13:03:29] <alex_joni> ?
[13:03:36] <alex_joni> one inverter should be enough
[13:04:06] <skunkworks> right - that was the plan
[13:04:29] <skunkworks> (I am going to play with the pluto first which does pwm+pwm)
[13:05:24] <tomp> is there a visual on this? ( a scope output showing what is desired on the drive signal )
[13:05:31] <skunkworks> btw - the cap on bridge design was jmkasunich idea.
[13:05:57] <tomp> I've seen HBOAC before :)
[13:06:33] <alex_joni> tomp: you're driving the 2 H-halves directly
[13:06:47] <alex_joni> can't have them on at the same time
[13:06:57] <tomp> so 2 out of phase signals
[13:07:02] <alex_joni> right
[13:07:11] <tomp> gotcha, thanks
[13:07:16] <skunkworks> emc lingo is up/down :)
[13:07:16] <alex_joni> if dutycycle1 < dutycycle2 it runs CW
[13:07:24] <alex_joni> if dutycycle1 > dutycycle2 it runs CCW
[13:07:40] <alex_joni> if dutycycle1 = dutycycle2 it doesnt turn
[13:08:28] <tomp> ok, i got the scope picture in my head now :)
[13:08:53] <alex_joni> * alex_joni launches IRCScope v0.1
[13:09:08] <tomp> this is called up/down ? ( picture got fuzzy suddenly )
[13:09:22] <alex_joni> yes and no
[13:09:42] <alex_joni> up/down is very similar but doing digital stuff (steps on steppers)
[13:10:32] <tomp> ok, would a hardware inverter cause significant delay ( if only 40nS or so ) ?
[13:11:43] <tomp> significant = anythingtoworryabout
[13:15:02] <skunkworks> don't you need 2 'and' gates and an inverter?
[13:15:11] <skunkworks> just thinking out loud
[13:16:17] <alex_joni> tomp: if you use a HC or HCT chip, I suspect it's ok
[13:16:32] <tomp> pencil sketch looks like its just inverted, thanks
[13:19:39] <tomp> did not know 'skunkworks' was a regular slang term, translates to something like the slang 'government job' ?
[13:19:43] <skunkworks> I am talking conveting pwm+dir to pwm+pwm. 2
[13:19:47] <skunkworks> oops
[13:20:12] <alex_joni> tomp: you can always run the other leg through a simple and gate (for the same delay)
[13:20:33] <skunkworks> 2 and gates - one leg of each hooked to the pwm out - one of the other legs hooked to direction - and the other leg hooked thru an inveter to direction.
[13:20:45] <alex_joni> pwm-> and-> h-bridge-in0
[13:20:54] <alex_joni> pwm-> not-> h-bridge-in1
[13:21:01] <tomp> again, thanks, (doh!)
[13:24:32] <skunkworks> tomp: no clue - I used only because I have a dodge stealth - stealth plane was made by 'skunkworks'
[13:24:49] <skunkworks> usually it means a top secret project of sorts.
[13:25:18] <alex_joni> it always reminds me of the related animal
[13:25:47] <tomp> ah, a REAL government job ( the shop joke when a toolmaker was working on his own stuff was, he said it was a 'government job' )
[13:26:41] <skunkworks> we use that term also - 'goverment job'
[13:26:53] <tomp> :)
[13:29:53] <tomp> skunkworks: can you point me to your files? i know i found your schematic and jpegs last night, but cant even find my downloads today
[13:33:49] <anonimasu> lol
[13:41:46] <tomp> old logs had it http://www.electronicsam.com/images/KandT/servostart/
[13:42:48] <alex_joni> * alex_joni always smiles at the endoder
[13:44:30] <skunkworks> sorry - was away.
[13:44:48] <skunkworks> I didn't even notice it until cradek mentioned it to me ;)
[13:46:14] <skunkworks> where was that site that you can sketch things like pastebin?
[13:46:40] <tomp> http://www.drawblog.com/
[13:48:50] <rayh> http://imagebin.org/ will let you post images.
[13:49:00] <tomp> i got 3 heavy duty copier slides, cast alum frame, nsk rails, 1mm pitch national ball screws, 1um tll linear scale, pittman dc motors with spiral couplers. will try your h bridge and the mesa on it.
[13:50:16] <skunkworks> this is sort of waht I was thinking 'http://www.drawblog.com/images/20070403064924794.jpg' usemap=#20070403064924794
[13:50:47] <skunkworks> http://www.drawblog.com/images/20070403064924794.jpg
[13:51:09] <alex_joni> nice handwriting :P
[13:51:43] <tomp> looks good, thanks, the xtra drive is good idea, with alex's even delays , and optos... save and strong enuf to overcome shop noise
[13:54:19] <alex_joni> http://www.drawblog.com/images/20070403065356764.jpg
[13:54:24] <skunkworks> tomp: http://www.cnczone.com/forums/showpost.php?p=274587&postcount=9
[13:54:41] <skunkworks> that is the actual part list I am using. I would try to find some lower rds mosfets.
[13:54:49] <alex_joni> I call it .. 'boredom'
[13:55:03] <skunkworks> alex_joni: nice.
[13:56:46] <tomp> ah, great, thanks!
[13:58:40] <skunkworks> tomp: what are you planning to do with it?
[13:59:16] <tomp> if i can find my niMh batteries will post pix of the slide i will drive
[13:59:55] <skunkworks> cool. alex_joni: there are opto's on my bridge.
[14:03:16] <alex_joni> skunkworks: then it's great
[14:30:34] <tomp> i was planning on testing an idea with it.
[14:30:47] <tomp> if emc can position a simple xyz machine tool, then get hal to move an axis without emc control ( but emc knowing the postiion is 'ok' like Dallur does with his torch height control). then moving to another position
[14:31:05] <tomp> ( emc would be whats called 'point to point control; ).
[14:31:11] <tomp> The Hal motion would be EDM with feedback from the process to the amplifier, and positions reported to EMC.
[14:31:17] <tomp> Hal motion would return the tooltip to the original position before returning to 'emc' mode.
[14:31:24] <tomp> Here's the test slide http://imagebin.org/7964. heres one of the machine tools to retrofit http://imagebin.org/7965
[14:31:41] <anonimasu> tomp: yeah that would probably work
[14:31:48] <anonimasu> tomp: but that's still a ugly hack ;)
[14:32:43] <tomp> the alternative is to rewrite the trakjectory planner so it wont try to enforce 'ontime & in position' which ruins edm
[14:33:08] <tomp> i cant untangle that code so i work around it
[14:33:28] <anonimasu> tomp: well, it's this stupid idea with where motion comes from :/
[14:34:28] <tomp> its good for mills and lathes, just not for dynamics which want to be 'on path wherever the process says'
[14:34:41] <anonimasu> well, or toolchanging..
[14:34:55] <anonimasu> unless you write toolchanges as macros..
[14:35:19] <rayh> I'd have thought that adaptive feedrate would allow on time and in position to accomodate edm.
[14:35:26] <tomp> i think toolchangin has some new possibilities, if this hack works, then a rack tool changer can work too
[14:35:41] <anonimasu> well, you can script that in a macro...
[14:36:06] <anonimasu> tomp: g0 (T*10)
[14:36:30] <anonimasu> for 10 times the number of the tool to the next raco position
[14:36:37] <anonimasu> rack..
[14:36:45] <anonimasu> ;)or something..
[14:37:51] <anonimasu> rayh: been thinking about another kind of UI?
[14:38:04] <rayh> what kind?
[14:38:23] <anonimasu> the more machine-ish kind we spoke about
[14:38:45] <rayh> okay. what did you come up with?
[14:39:01] <anonimasu> nothing yet
[14:39:06] <anonimasu> my mind's all at workstuff right now
[14:39:18] <tomp> rayh: as mydynac (?) demonstarted, edm that always progresses can work with that idea. but real edm doesnt lag behind the process. it follows the process and needs to 'jiggle ( not retract ) edm is not feedrate, edm is process a jiggly signal. we been here before.
[14:39:20] <anonimasu> I've been thinking about a sane toolkit..
[14:39:44] <anonimasu> tomp: how about introducing jitter on the output position?
[14:39:53] <tomp> anonimasu: if this hack works, then rack tool changers and t(#1001 * #1002) is ok
[14:39:57] <anonimasu> in hal, before the signal gets output?
[14:40:56] <anonimasu> maybe im all on the wrong track..
[14:41:27] <tomp> that jitter would have to stay inside emc's inposition term. we'd just be 'loosening up the control to allow it. i want tight position control based on the process. thats when edm runs well.
[14:41:40] <anonimasu> * anonimasu nods
[14:42:40] <tomp> better i do it with a hack and evaluate the result :)
[14:43:21] <anonimasu> yep
[14:48:20] <tomp> thanks all: off to work:)
[14:49:05] <anonimasu> laters tomp
[14:52:40] <alex_joni> bbl
[17:12:11] <maddash> should i disable apic support in my kernel?
[17:12:26] <maddash> i remember reading that apic's were bad for realtime
[17:16:06] <alex_joni> that was ACPI
[17:16:30] <maddash> no, acpi != A. programmable interrupt controller, I think
[17:23:42] <Jymmm> alex_joni: In the background, is there snow in the mountains? http://dsplabs.cs.utt.ro/~juve/blog/index.cgi-files/photography/01168244631/IMG_9508.JPG
[17:23:49] <Jymmm> s/in/on/
[17:27:20] <alex_joni> hmm.. the white stuff in that pic are clouds
[17:27:47] <alex_joni> but there was snow on that mountains, can
[17:27:52] <alex_joni> but there was snow on that mountains, can't see it because of the trees
[17:28:51] <SWPadnos__> SWPadnos__ is now known as SWPadnos
[17:33:41] <maddash> oh crap, the processor i bought has netburst
[17:36:36] <lerneaen_hydra> maddash: how's RT on netburst?
[17:38:31] <maddash> lerneaen_hydra: makes me wanna burst
[17:39:00] <lerneaen_hydra> maddash: netburst isn't really all that fun
[17:39:32] <maddash> lerneaen_hydra: i'll post up some latency results after I finish recompiling
[17:42:48] <maddash> what x86 hardware do you guys use? i had quite a difficult time locating a motherboard w/o onboard video
[17:43:10] <cradek> I've been happiest with fast PIII machines
[17:43:21] <cradek> onboard video doesn't matter as long as you can ignore it and put in a video card
[17:43:52] <Jymmm> server mobo w/ onboard video seem to work
[17:44:08] <lerneaen_hydra> PIII here too
[17:45:02] <alex_joni> the best one I had is a Athlon 1600+
[17:47:01] <maddash> hm, so my first board (Pent D, intel graphics, sound , onboard lan) wasn't so bad after all?
[17:47:52] <cradek> I had one machine with onboard i8xx graphics that was good for realtime
[17:48:43] <cradek> it was also PIII
[17:48:58] <maddash> then what's the deal with http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Hardware_Requirements?
[17:50:02] <cradek> I don't understand the question
[17:50:44] <maddash> the wiki cries wolf on onboard video
[17:51:13] <cradek> yes it's the most common problem people have
[17:51:21] <cradek> sometimes you get lucky
[17:51:30] <cradek> If you have realtime problems with a system using onboard video, the first thing to do is disable it and plug in a video card.
[17:51:34] <cradek> ^^ very good advice
[18:07:54] <maddash> wow, in < 15 min.
[18:08:14] <maddash> brbb
[18:42:36] <maddash> anyone want to see the netburst latencies? http://pastebin.ca/422816
[18:43:08] <cradek> looks great
[18:43:20] <cradek> did you exercise it? move windows, load programs, etc?
[18:43:30] <cradek> be sure to run opengl stuff too (like a few glxgears)
[18:43:53] <maddash> okey...
[18:44:19] <maddash> did i mention that a) i don't have X11 installed and b) the board has no video capabilties whatsoever?
[18:45:07] <alex_joni> maddash: ook
[18:45:20] <cradek> ha that'll help
[18:45:55] <maddash> haha.
[18:46:08] <maddash> also was able to disable onboard sound from the bios
[18:46:50] <maddash> netburst's kernel latencies: http://pastebin.ca/422822 vs. Pentium D's kernel latencies (from a month ago): http://pastebin.ca/379779
[18:47:55] <alex_joni> maddash: maybe watch a movie using libascii ?
[18:48:04] <Jymmm> lol
[18:49:04] <maddash> alex_joni: fullscreen youtube in w3m
[18:49:34] <alex_joni> or lynx
[18:50:06] <maddash> hmm, all i have to do now is figure out what those numbers mean
[18:54:38] <Jymmm> OVL MAX
[18:54:49] <Jymmm> Overall Max Latency
[18:55:39] <Jymmm> 20000nS is about as high as you ever want to go.
[18:57:20] <maddash> Jymmm: 20000ns under what load?
[18:57:36] <Jymmm> Yes
[18:57:44] <Jymmm> any/all loads
[18:59:11] <Jymmm> Also, be sure to disab;e any serial, usb, IDE controllers, etc in the BIOS that oyu're not using.
[19:00:58] <maddash> while compiling a kernel, i'm awfully close to Jymmm's 20ms limit: http://pastebin.ca/422844 (kernel) and http://pastebin.ca/422846 (user)
[19:02:27] <Jymmm> and it's NANOseconds, not Miliseconds btw =)
[19:02:55] <maddash> Jymmm: isn't 20 * 10^3 nanosec = 20 microsec?
[19:03:10] <alex_joni> maddash: yeah
[19:03:15] <Jymmm> yeah, but no need to convert
[19:03:16] <alex_joni> nano, micro, milli,
[19:03:41] <maddash> Jymmm: did i mention that my zero key is almost broken in half?
[19:03:52] <alex_joni> femto, pico, nano, micro, milli..
[19:03:52] <Jymmm> you have two of them
[19:04:10] <maddash> Jymmm: laptop
[19:04:15] <alex_joni> maddash: even there
[19:04:22] <Jymmm> you STILL have two of them =)
[19:04:32] <alex_joni> usually on 'M'
[19:04:55] <Jymmm> 18KnS
[19:05:07] <Jymmm> * Jymmm snickers *
[19:05:11] <maddash> my boss is going to be warpath when he sees this data
[19:05:37] <Jymmm> maddash your still at 18000, not THAT bad
[19:06:11] <maddash> well, because of my insistence to find a board w/o video, the ovl max has nearly doubled..
[19:06:25] <maddash> s/doubled/tripled/
[19:06:44] <Jymmm> maddash I found a Dell P3 500 w/o onboard video
[19:07:09] <Jymmm> my laptop does 100,000 nS, but under DOS using TurboCNC it does fine.
[19:07:28] <maddash> Jymmm: hm, so rm -rf emc*?
[19:07:34] <maddash> Jymmm: :)
[19:07:52] <Jymmm> maddash No, it has BDI on it, just not too usefully for RT
[19:08:10] <robin_sz> meep?
[19:08:13] <maddash> how's the avg latency ccalculated?
[19:08:26] <Jymmm> fiik
[19:09:14] <maddash> hmm, maybe i can focus my boss's attention on lat avg, which has nearly fourthed
[19:09:22] <Jymmm> Spam Subject: Be like Ron Jeremy
[19:09:23] <Jymmm> lol
[19:09:33] <alex_joni> no thanks :)
[19:10:08] <maddash> robin_sz: how much did that plasma cutter cost?
[19:10:19] <Jymmm> He seems like a very nice guy... just happens to be hung like a horse is all
[19:10:26] <Jymmm> literally
[19:10:40] <alex_joni> hi robin
[19:10:40] <robin_sz> maddash, which plasma cutter?
[19:11:06] <robin_sz> hi alex
[19:11:23] <maddash> robin_sz: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Robin_Szemeti's_Plasma
[19:12:00] <robin_sz> blank page
[19:12:13] <maddash> huh?
[19:12:24] <maddash> http://www.redpoint.org.uk/plasma.jpg
[19:12:31] <alex_joni> For some reason, I built a 2.5 x 1.25m plasma cutter running EMC as the controller.
[19:12:43] <maddash> alex_joni: that's yours?
[19:12:46] <alex_joni> robin_sz: never got anything back from your mate
[19:12:51] <alex_joni> maddash: no, robin built that
[19:13:07] <robin_sz> alex_joni, ? he said he mailed you asking for more info onthe spec
[19:13:19] <alex_joni> robin_sz: really?
[19:13:23] <robin_sz> yeah
[19:13:23] <alex_joni> I never got anything..
[19:13:33] <robin_sz> hmmm .. I cc'd you on the mail I sent, right?
[19:13:39] <alex_joni> right
[19:13:42] <alex_joni> I got that one
[19:13:46] <robin_sz> so he had the address OK,
[19:13:56] <alex_joni> and I think I replied to him a couple of days later
[19:13:59] <alex_joni> * alex_joni checks
[19:14:10] <robin_sz> oh, right ... I'll kick him for you
[19:15:12] <robin_sz> maddash, that was a VERY early prototype
[19:15:22] <robin_sz> they ended up a lot neater than that
[19:15:38] <maddash> er, you built that?
[19:15:47] <maddash> from scratch?
[19:17:07] <alex_joni> no, from good parts
[19:17:37] <robin_sz> maddash, yeah
[19:18:48] <robin_sz> http://www.redpoint.org.uk/photos/machine2/DCP_1200.JPG
[19:18:51] <robin_sz> later variant
[19:19:27] <maddash> pretty
[19:20:13] <maddash> erm, how is v2_1_branch diff from RELEASE_2_1_*?
[19:20:56] <cradek> the RELEASE_2_1_* tags are all along v2_1_branch
[19:21:07] <cradek> v2_1_branch is always the latest thing waiting to be released
[19:21:42] <maddash> so current_tag = RELEASE_2_1_4 == current_tag = v2_1_branch?
[19:22:42] <cradek> I don't understand - v2_1_branch is the latest thing on that branch, which might be the same as the last release, or it might have bugfixes that are waiting for the next release
[19:22:55] <cradek> today there are a couple bugfixes on it that will be in 2.1.5
[19:23:15] <alex_joni> http://cvs.linuxcnc.org/cvs/emc2/VERSION?graph=1
[19:23:33] <maddash> so it'll be v2.1branch for me, then...
[19:28:31] <robin_sz> maddash, did you like that?
[19:28:55] <maddash> robin_sz: v2.1branch? sure
[19:29:01] <robin_sz> no the machine silly
[19:29:17] <alex_joni> lol
[19:29:29] <robin_sz> wanna see the current version?
[19:30:20] <maddash> it's quite...red..
[19:30:34] <robin_sz> well, yes
[19:30:54] <robin_sz> current version: http://www.rapidcut.co.uk/products/PL2515.html
[19:31:12] <maddash> did you quit redpoint?
[19:31:14] <alex_joni> 404 Document Not Found
[19:31:21] <maddash> ditto
[19:31:31] <alex_joni> that means robin doesn
[19:31:37] <alex_joni> that means robin doesn't do them anymore
[19:32:21] <maddash> or it could mean rapidcut fired him :)
[19:32:30] <maddash> err...s/:)/:P/
[19:32:37] <robin_sz> maddash, no, redpoint is my computer consultancy company, rapidcut is my fabriaction and sheet metal business
[19:32:47] <robin_sz> we just do fabrications, not plasmas anymore
[19:32:51] <robin_sz> too much hassle
[19:33:34] <maddash> why?
[19:33:44] <alex_joni> tech support
[19:34:15] <robin_sz> yeah, driving around the country answering questions asked by idiots
[19:34:19] <maddash> does that mean i should avoid plasma cutters? because i've always dreamed of getting one...
[19:34:29] <robin_sz> well, no not idiots, thats unfair
[19:34:50] <robin_sz> but at least soem of them didnt have any clue how to work one
[19:34:59] <robin_sz> no, you should build one
[19:35:10] <robin_sz> with a clue about CNC its easy to use one
[19:35:23] <robin_sz> with no clue, ... well, its not going to be fun/
[20:03:07] <xemet> hi all
[20:03:54] <xemet> first question: halui spindle-override pins are available in the 2.1.4 release or only in the cvs-head?
[20:04:30] <alex_joni> it is in 2.1.4
[20:04:34] <alex_joni> most of halui is in 2.1.x
[20:04:43] <alex_joni> only the new analog jogging pins aren
[20:04:47] <alex_joni> 't
[20:05:10] <xemet> ok
[20:05:29] <xemet> second question what exactly is the pin halui.spindle-override.value
[20:05:54] <xemet> thinking about the feed override in AXIS I can modify it from 0 to a max%
[20:06:46] <xemet> if I want for example the override be 120% what should be the value for that pin?
[20:07:18] <skunkworks> 1.2? my guess ;)
[20:07:47] <crepincdotcom`> whee :-) new Z-axis motor
[20:07:47] <crepincdotcom`> a REAL one
[20:07:53] <crepincdotcom`> its uber fast and accurate
[20:07:55] <crepincdotcom`> * crepincdotcom` likes
[20:07:59] <alex_joni> (FLOAT) halui.spindle-override.value - current SO value
[20:08:02] <skunkworks> sweet
[20:08:17] <alex_joni> xemet: you can only read that pin, not write to it
[20:08:28] <xemet> uhm...S0 is?
[20:08:36] <xemet> spindle speed?
[20:08:37] <alex_joni> spindle override
[20:08:47] <alex_joni> spindle override = SO
[20:08:53] <xemet> so if I want 120% I've to?
[20:08:55] <alex_joni> feed override = FO
[20:09:03] <alex_joni> you need to set it to 120%
[20:09:17] <xemet> ok, not 1.2
[20:09:16] <alex_joni> set scale to 0.1, and hit increase 2 times
[20:09:45] <xemet> in the wiki there is an error, halui.spindle-override.value float //current FO value
[20:09:45] <alex_joni> xemet: yes, 1.2
[20:09:52] <xemet> it says F0, not S0
[20:09:54] <alex_joni> xemet: fix it :)
[20:09:56] <xemet> ok
[20:10:06] <alex_joni> xemet: maybe I'm not beeing clear
[20:10:14] <alex_joni> the idea is the following
[20:10:34] <alex_joni> you can set SO (spindle override) from more than one place
[20:10:44] <alex_joni> GUI (AXIS, tkemc, etc), halui, ..
[20:10:46] <xemet> ok
[20:11:01] <alex_joni> for that to work you cannot use fixed settings
[20:11:11] <alex_joni> like halui.pin-whatever = 1.2
[20:11:30] <alex_joni> that is not permitted / doesn't work
[20:11:46] <xemet> yes, but if I want to change it...?
[20:12:03] <xemet> I want a slider in AXIS made with pyVCP to change it
[20:12:25] <alex_joni> you already have a slider in AXIS which changes it directly
[20:12:31] <xemet> no...
[20:12:43] <xemet> there is one for the feed override
[20:12:49] <xemet> not one for spindle
[20:12:54] <alex_joni> there is one for spindle speed override too
[20:13:01] <alex_joni> but you need to enable it in the ini
[20:13:03] <xemet> uhm...I don't ghave it
[20:13:12] <xemet> Ah, how to enable it?
[20:13:31] <alex_joni> http://wiki.linuxcnc.org/uploads/lathe_sim_pyvcp.png
[20:13:37] <skunkworks> woo hoo - server is fault tolerant again..
[20:14:23] <alex_joni> there is a MIN_SPINDLE_OVERRIDE and MAX_SPINDLE_OVERRIDE
[20:14:23] <xemet> well, the picture is beautiful, but how to enable it?
[20:14:30] <alex_joni> xemet: rtfm :P
[20:14:36] <alex_joni> it's a fine manual :D
[20:15:12] <xemet> well, there is the chat :)
[20:15:41] <alex_joni> http://www.linuxcnc.org/docs/devel/html/config/ini_config/index.html
[20:15:58] <xemet> great
[20:16:51] <xemet> a question, WHY not all the wiki pages are in the wiki table of contents
[20:16:54] <xemet> ?
[20:17:06] <alex_joni> xemet: because there are 240 pages
[20:17:24] <alex_joni> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl/emcinfo.pl?action=index
[20:17:43] <xemet> well, exactly this the reason because they should be in the table of contents...
[20:17:58] <xemet> so one can find things...
[20:18:23] <alex_joni> you cannot have a 240 pages table of contents
[20:18:36] <xemet> not 240 pages...but sections
[20:18:40] <xemet> with other tables...
[20:18:48] <xemet> like abook
[20:18:50] <xemet> a book
[20:19:30] <jtr_> maybe put the link to the index in the table of contents?
[20:19:31] <alex_joni> xemet: the wiki is community work :)
[20:19:43] <skunkworks> I put it under my pillow in hopes of osmosis..
[20:19:45] <xemet> so I could do that
[20:19:47] <alex_joni> jtr_: the link to the index is at the beginning of every page
[20:20:07] <alex_joni> jtr_: it's called PageIndex
[20:20:31] <alex_joni> xemet: you can start a page where you put the structure you propose
[20:20:39] <xemet> what's jtr_?
[20:20:42] <alex_joni> if it's ok, we'll move it to the front page
[20:20:49] <alex_joni> what's xemet?
[20:20:54] <jtr_> No wonder I missed it...
[20:20:58] <xemet> ah
[20:21:13] <xemet> sorry, not seen the ,message before
[20:22:39] <jtr_> jtr and jtr_ are me - I'm at an airport and initially hooked up to a (bad?) access point.
[20:23:11] <jtr_> So i guess there's a zombie jtr out there for a while.
[20:23:48] <alex_joni> jtr_: if you have a registered nick you can type "/msg NickServ GHOST nick"
[20:24:31] <alex_joni> or GHOST pass? something like that
[20:27:59] <Jymmm> [04/03 13:27:52] -NickServ- Syntax: GHOST <nickname> [password]
[20:27:57] <Jymmm> [04/03 13:27:52] -NickServ-
[20:27:57] <Jymmm> [04/03 13:27:52] -NickServ- If you suddenly get disconnected from your isp or
[20:27:57] <Jymmm> [04/03 13:27:52] -NickServ- the irc server, sometimes a ghosted client will
[20:27:57] <Jymmm> [04/03 13:27:52] -NickServ- be left behind. This command can be used to
[20:27:59] <Jymmm> [04/03 13:27:52] -NickServ- kill the ghosted client so you can get your
[20:28:01] <Jymmm> [04/03 13:27:52] -NickServ- nickname back. If <nickname> has SET SECURE
[20:28:03] <Jymmm> [04/03 13:27:52] -NickServ- OFF, and you match a hostmask on <nickname>'s
[20:28:04] <Jymmm> [04/03 13:27:52] -NickServ- ACCESS LIST, or if you are identified to a linked
[20:28:07] <Jymmm> [04/03 13:27:52] -NickServ- nickname, you do not need to supply a password.
[20:28:08] <Jymmm> [04/03 13:27:52] -NickServ- Otherwise, you have to supply the correct
[20:28:11] <Jymmm> [04/03 13:27:52] -NickServ- password.
[20:34:34] <xemet> ok, thank you alex, I go, good night
[20:59:17] <Jymmm> On an automated SMT soldering machine, how are the compents kept in place without moving before being soldered?
[20:59:24] <Jymmm> pick n place
[20:59:26] <alex_joni> glue
[20:59:33] <Jymmm> seriously?
[20:59:36] <alex_joni> yes
[20:59:59] <alex_joni> but there are different technologies
[21:00:09] <alex_joni> the most common is glue under the capsule
[21:00:16] <Jymmm> I've seen videos, I never saw any glue being used when a component is grabbed from it's contianer
[21:00:16] <alex_joni> then solder
[21:00:30] <alex_joni> I think it's on the board
[21:00:33] <Jymmm> or is it on the PCB before placing?
[21:00:36] <Jymmm> ah
[21:01:46] <alex_joni> the other way I've seen is they use some solder paste on the contacts
[21:02:00] <alex_joni> place teh component on it (sticky) then input some heat
[21:44:37] <maddash_> maddash_ is now known as sudo_maddash
[21:47:12] <tomp> jmkasunich: could you post a component list for the h-bridge design? 'info' in eagle does not yield any values. i can add the values to the schema and post it back.
[21:55:50] <Jymmm> Ya know, soldering a wire to a SMT LED AIN'T THAT EASY!!!
[22:03:54] <maddash> where is `man usrmot`?
[22:04:26] <maddash> i can't find the documentation in linuxcnc.org/docs
[22:07:05] <Jymmm> wiki.linuxcnc.org
[22:08:53] <maddash> wiki's got nothing...
[22:09:07] <alex_joni> there is no man usrmot
[22:09:16] <tomp> if you have built from source, then ~/emc2-head/src/emc/usr_intf/usrmot.c can be read for some info ( like accepted commands )
[22:10:23] <alex_joni> http://cvs.linuxcnc.org/cvs/emc2/src/emc/usr_intf/usrmot.c?rev=1.21;content-type=text%2Fplain
[22:10:38] <Jymmm> * Jymmm needs a vise that'll hold a 1.1mm x 1.6mm part w/o crushing it. What would MacGyver do???
[22:11:03] <tomp> chewing gum and ice box
[22:12:08] <cradek> Jymmm: clothespin
[22:12:16] <Jymmm> http://en.wikipedia.org/wiki/List_of_problems_solved_by_MacGyver
[22:12:29] <cradek> actually sounds like you need tweezers, not a vice
[22:12:31] <maddash> GUT
[22:12:59] <Jymmm> cradek that's not a bad idea at all. I could even toss in a bolt/nut to adh the tension
[22:13:00] <cradek> watchmakers will often shellac a part that size onto a much larger carrier, for finishing etc
[22:13:16] <Jymmm> adjust
[22:13:57] <cradek> then soften and throw it in alcohol to get the shellac off without abrasion when done
[22:14:42] <Jymmm> cradek most tweezers won't give enough tension. to hold it secure (I think), at least not from what I've tried so far.
[22:15:03] <cradek> all depends what you want to do to it
[22:15:09] <Jymmm> solder
[22:15:20] <cradek> solder what to what?
[22:15:29] <Jymmm> solder a wire to a SMT LED
[22:15:57] <alex_joni> Jymmm: I use fingers for that
[22:16:04] <cradek> yeah that's easy :-)
[22:16:04] <Jymmm> correction, solder lots of wires to lots of SMT LEDs
[22:16:11] <cradek> I thought you wanted to machine it
[22:16:17] <alex_joni> lots of fingers?
[22:16:18] <Jymmm> machien what?
[22:16:24] <cradek> the 1mm part
[22:16:41] <tomp> overarm spring tension, like a 'cats whisker', leave solder area exposed
[22:16:46] <Jymmm> ah, no. just hold long enough to solder two wires to SMT led
[22:17:01] <Jymmm> tomp I was thinking something like that
[22:17:27] <alex_joni> Jymmm: when I need to do stuff like that, I secure the wire
[22:17:32] <Jymmm> Hey, I could use the clothes pin as the tension, then drill a couple tiny holes to hook
[22:17:45] <alex_joni> then grab the LED with my left hand, and soldering gun with my right one
[22:17:52] <Jymmm> alex_joni and the wire?
[22:17:57] <Jymmm> with the 3rd hand
[22:17:59] <alex_joni> Jymmm: secure the wire
[22:18:09] <Jymmm> alex_joni ha, yeah right....
[22:18:23] <alex_joni> usually on the desk, some sticking out
[22:18:26] <Jymmm> were talking 30ga, never is straight
[22:18:38] <cradek> I think that's what I'd do too
[22:18:47] <alex_joni> put a weight on it
[22:18:54] <cradek> yeah sit the spool of solder on the wire
[22:19:01] <alex_joni> cradek: exactly
[22:19:12] <alex_joni> I do the exact same thing
[22:19:13] <Jymmm> I have forcepts, but the wire still flows to it's natural bended state
[22:19:25] <Jymmm> Remember, I have to position the wire within 1.6mm
[22:19:29] <Jymmm> err half of that
[22:19:53] <cradek> quit planning and just go solder it, sheesh :-)
[22:19:53] <alex_joni> Jymmm: you can get http://www.eng.iastate.edu/twt/Courses/Undergrad/packet/images/sensor/helping%20hands%20with%20sensor.jpg
[22:20:07] <Jymmm> cradek I did, took me forever
[22:20:15] <maddash> errrr
[22:20:28] <maddash> sample output from `ls -l /dev/rt*`:
[22:20:36] <maddash> crw-rw-rw- 1 root root 10, 254 2007-04-04 11:12 /dev/rtai_shm
[22:20:38] <alex_joni> Jymmm: it shouldn't take more than 1 second to solder it
[22:20:50] <Jymmm> alex_joni Yeah, helping hands are nice, but the alligator clips usually have larger gabs in the teeth
[22:20:53] <maddash> shouldn't rtai_shm be a symlink?
[22:21:02] <maddash> at least, according to http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TroubleShooting
[22:21:07] <Jymmm> alex_joni soler is instant, getting everything aligned is the problem
[22:21:16] <tomp> lots of them? make a jig to hold the leds (several) make a jig to hold the wires (several) then both hands fre to hold pen and solder
[22:21:20] <cradek> maddash: where those come from depends a lot on your rtai version
[22:21:40] <cradek> worse, some are uppercase, some lowercase
[22:21:43] <cradek> I forget which one emc wants
[22:21:53] <Jymmm> tomp That's the idea, just need to figure out a way to actually clamp the SMT LED and the wires
[22:21:55] <maddash> cradek: does emc expect rtai_shm to be a symlink?
[22:21:56] <alex_joni> lowercase I think
[22:22:02] <cradek> maddash: that doesn't matter
[22:22:18] <alex_joni> * alex_joni goes to bed
[22:22:23] <alex_joni> good night all
[22:22:24] <cradek> goodnight alex
[22:22:26] <tomp> nite alex
[22:22:26] <Jymmm> alex_joni see you in a few
[22:23:10] <alex_joni> Jymmm: one of my first soldering jobs (way back) was to make a model of a couple of street corners with street lights.. soldered about 150 LEDs back then
[22:23:21] <Jymmm> alex_joni SMT?
[22:23:34] <alex_joni> no.. regular
[22:23:40] <alex_joni> with regular resistors
[22:23:52] <Jymmm> alex_joni Yeah, if I use regular, I'd just wirewrap and be done with it
[22:23:59] <alex_joni> yuck.. wirewrap
[22:24:15] <Jymmm> fits PERFECTLY on a led leads
[22:24:21] <Jymmm> s/a//
[22:24:40] <alex_joni> ok.. night
[22:25:17] <tomp> the spring clamps used to hold microscope slides http://www.standa.lt/products/catalog/optical_mounts?item=82
[22:25:52] <tomp> (just a piece of thin metal)
[22:26:11] <Jymmm> I was almost thinking the clips used to hold model rocket igniter leads
[22:27:12] <Jymmm> BU-34/BU-34C http://www.muellerelectric.com/alligator_clips.html
[22:27:19] <tomp> and 'fahenstock clips' to hold the wires http://www.crystalrad.com/parts.htm
[22:29:15] <tomp> i'd relieve the lower jaw of a regular allig8r clip and mount the shank end rearward of where the led die went, so only the top jaw pressed down over the die. then you'd have a nice thumb button to release it after soldering
[22:31:23] <tomp> http://www.drawblog.com/images/20070403033041664.jpg
[22:32:22] <Jymmm> did you see those clips I linked to... flat, no teeth
[22:33:13] <tomp> yep. BU-34/BU-34C
[22:33:30] <Jymmm> http://www.muellerelectric.com/images/BU34.jpg
[22:34:11] <Jymmm> could use those to hold the led and the wires too
[22:35:09] <tomp> i think you might want a little 'well' under the top jaw to place the led into ( a locator divot ), so the lower jaw might be removed
[22:36:27] <Jymmm> yeah, will have to think about this, I don't have any bits that small, would have to use a razor blade.
[22:37:52] <robin_sz> removing your lower jaw would be a bit extreme, makes eating difficult
[23:28:28] <tomp> skunkworks: could you point me to a full parts list for the h-bridge?
[23:28:32] <tomp> i'd like to order the parts and cant determine the values for any caps of resistors, or the chip used for q4 ( suspect opto-isol8r), also the 15v-reg is not on the pcb as far as i can tell.
[23:52:02] <jmkasunich> hi tomp
[23:52:07] <tomp> skunkworks: jmkasunich: wait. isee the parts already total 120$ for 1 amp while mesa offers a dual 400 watt hbridge fpr 161.
[23:52:13] <tomp> hi john
[23:52:17] <jmkasunich> you will need to get that info from skunkworks
[23:52:24] <jmkasunich> I just laid it out ;-)
[23:52:42] <jmkasunich> where are you getting $120 from?
[23:52:49] <jmkasunich> and where are you getting 1A from?
[23:52:57] <jmkasunich> skunk's bridge is good for about 20A IIRC
[23:53:57] <tomp> ok, spreadsheet said 120, digikey & mouser...and 1 amp(lifier) not 1 amp(ere)
[23:54:19] <jmkasunich> IIRC he's using a pair of IR2110 or similar gate drive chips, I would hope they're under $5 each (maybe $2 each?)
[23:54:34] <jmkasunich> the FETs should be $5-10
[23:55:36] <jmkasunich> skunks bridge is designed for up to about 200VDC input
[23:55:46] <tomp> fet 16.52 each, rurg 7.41 @ , driver 7.56 each....
[23:55:52] <jmkasunich> at 20A thats 4000 watts (peak)
[23:55:54] <jmkasunich> ouch
[23:56:00] <jmkasunich> digikey prices?
[23:56:07] <tomp> i only need 24V at 9amp
[23:56:15] <tomp> digikey & mouser
[23:56:24] <jmkasunich> only 24V?
[23:56:47] <jmkasunich> that seems like high amps and low volts - not good for efficiency
[23:56:48] <tomp> pittman 30.3V motor rated 19A stall
[23:56:54] <jmkasunich> but if its what your motor wants...
[23:57:09] <tomp> but it's an assy that i have already, just to test an idea
[23:57:16] <jmkasunich> right
[23:57:31] <jmkasunich> what is the 9A based on?
[23:58:05] <tomp> the power curve suggested they'd run at a reasonable speed/torq with those values
[23:58:24] <tomp> not maxing anything, middle of the road stuff
[23:58:32] <jmkasunich> with such low voltage, I'd give some serious thought to paralleling '298s, especially if you don't really need that much current
[23:59:08] <jmkasunich> I'm not real familiar with 298s, (cradek is) but I think they work up to 40V, and are good for a couple amps
[23:59:26] <tomp> seemed 298 was near 80-100watt, i'll look to see how to parallel them tho
[23:59:37] <tomp> thanks
[23:59:40] <jmkasunich> watts is deceptive
[23:59:41] <tomp> ?
[23:59:45] <jmkasunich> use amps and volts
[23:59:55] <tomp> watt = va