Back
[00:04:05] <AR_> stop fighting
[00:33:25] <r00t4rd3d> nah, drink beer and kill each other
[00:35:03] <tjb1> Quiet r00t4rd3d
[00:35:16] <r00t4rd3d> ive watched a terminal window for like 3 days
[00:36:16] <tjb1> Shhh
[00:51:11] <ReadError> watched a terminal window huh'
[01:36:00] <tjb1> seen r00t4rd3d
[01:57:09] <Valen> tjb1: about 4 hours ago
[01:57:20] <tjb1> lol
[01:57:30] <tjb1> That works in all the other fancy IRC rooms
[01:57:42] <Valen> it worked here what are you bitching about
[01:57:53] <Valen> (typically one also prefaces seen with a ! as i recall)
[02:01:22] <tjb1> !seen r00t4rd3d
[02:01:22] <the_wench> last seen in #linuxcnc 2013-01-10 06:22:40GMT 01:26:06 ago, saying ive watched a terminal window for like 3 days
[02:01:33] <Valen> yeah like that
[02:01:44] <tjb1> I need not perform that slave labor in other rooms
[02:03:43] <tjb1> You got a 3d printer Valen ?
[02:03:56] <tjb1> ReadError fails to help with my decision
[02:03:56] <Valen> pshaw,
[02:04:02] <Valen> 3d printing is for chumps
[02:04:07] <tjb1> Too busy playing with wd-40 and aluminum cans
[02:04:15] <Valen> turning huge hunks of Al into wafers is where its at
[02:07:13] <Valen> i've seen a few though tjb1 and i'm thinking of mounting a print head on our mill ;->
[02:07:54] <tjb1> I'd strap one on the plasma but I can't afford a 4'x8' heated bed ;)
[02:08:01] <Valen> heh
[02:08:10] <Valen> we are making a new light mill/router
[02:08:12] <Valen> 1200x1200mm
[02:10:45] <tjb1> JT-Shop and PCW_ are going to have a bunch of questions soon
[02:10:55] <tjb1> G540 is on its way back from Cali
[02:19:57] <DJ9DJ> moin
[02:20:56] <tjb1> Mornin!
[02:21:22] <DJ9DJ> hi there :)
[02:40:43] <tjb1> I have a serious urge to buy a gameboy color and pokemon...
[02:48:01] <tjb1> Is this what they call nostalgia
[02:48:25] <DJ9DJ> hrhr, yep :)
[02:49:06] * DJ9DJ played sid meiers civilization a few days ago. the original one from 1995 or so (free now), using dosbox emulator :)
[02:51:59] <Loetmichel> mornin'
[02:58:18] <tjb1> I would use an emulator but its not the same as the gameboy…probably wont play it anyway
[02:59:08] <DJ9DJ> just hang it at the wall ;)
[03:01:39] <Loetmichel> tjb1: the last use i have for my GBA:
http://www.cyrom.org/palbum/main.php?g2_itemId=4183
[03:01:57] <Loetmichel> ... but i still have the tetris cartridge in it sometimes ;-)
[03:02:08] <tjb1> scope?
[03:02:13] <Loetmichel> yes
[03:02:16] <DJ9DJ> hehe
[03:02:20] <tjb1> how hard is that to get working
[03:02:24] <Loetmichel> max 50khz or so, though
[03:03:43] <Loetmichel> would not be reasonable today, you can get better scopes for less money than the scope cardridge has costed. it was about 300 eur
[03:04:08] <Loetmichel> it was a kit from "Elektor"
[03:04:17] <Loetmichel> (google gbdso)
[03:04:49] <tjb1> I have a rigol
[03:05:27] <tjb1> DS1052e hacked to 100Mhz
[03:14:36] <tjb1> I dunno how to use it really
[03:37:28] <Loetmichel> tjb1: hrhr, know that feeling;-)
[06:45:36] <jthornton> http://www.kickstarter.com/projects/crystalfontz/cfa-10036-open-hackable-linux-arm-embedded-gpio-mo
[06:51:54] <L84Supper> arm9 :(
[08:17:28] <jthornton> Yippie I'm moving tabs now
http://gnipsel.com/glade/glade08b.html
[09:52:07] <ReadError> hsmworks seems pretty nice
[10:15:08] <ReadError> just need to disable this g28
[10:33:08] <JT-Shop> well the motor gearbox seems ok... on to the control board
[11:17:25] <JT-Shop> how would you check this to see if it is borked
https://www.relayspec.com/specs/006401/m_output_modules.pdf
[11:18:03] <JT-Shop> the ones I have are gordos M-OAC5AH
[11:19:31] <archivist> I would expect a short if dead across the triac
[11:20:04] <gene__> Got a strange error just now, I added a "net probe parport.0.pin-10.in-not motion.probe.input"
[11:20:30] <gene__> only to be told motion.probe.input does not exist???
[11:20:51] <archivist> JT-Shop, pin 1 to 2 is it open or short
[11:21:48] <JT-Shop> 1 to 2 are 25-30 M ohms except for one that is 20 ohms
[11:22:18] <pcw_home> sounds dead
[11:23:30] <archivist> not healthy
[11:23:51] <JT-Shop> and that one goes to JP10 pin 3B the one that runs the motor!
[11:24:31] <skunkworks> gene motion.probe-input
[11:24:34] <JT-Shop> well that is part of the puzzle solved
[11:24:40] <skunkworks> gene__, motion.probe-input
[11:25:20] <gene__> Ok, but in halconf, its, ahh duh. I'll get me coat
[11:26:59] <ReadError> who ever recommended hsmworks, thanks
[11:27:02] <ReadError> its pretty nice stuff
[11:27:13] <archivist> JT-Shop, noe tell me where I got my worm calculations from 4 years ago....must make notes in the resulting gcode
[11:32:29] <kwallace> What G76 options are commonly used? Compound angle, Spring Passes?
[11:32:43] <JT-Shop> I use all of them
[11:33:38] <JT-Shop> have you seen my threading subroutine?
[11:34:04] <archivist> does it do worms :)
[11:34:40] <JT-Shop> just night crawlers
[11:35:04] <archivist> I have to do a .5mod lh plastic worm
[11:35:35] <archivist> not sure whether to mill or turn at the moment
[11:35:45] <kwallace> I should look, but I don't want to include an option if it does more harm than good. I'd like to keep to the 90% applications
[11:35:52] <JT-Shop> I'm even less sure that you
[11:36:40] <gene__> Possibly silly Q, is motion sensitive to probe inputs while doing a backlash comp move?
[11:37:25] <jdh> doesn't all motion stop if it sees the probe during a non-probe-move?
[11:38:14] <JT-Shop> yep
[11:38:19] <kwallace> JT-Shop: May I get a hint to where your subs are?
[11:38:33] <skunkworks> well - I think if you hook it up.. The K&T just gives me a message.
[11:38:35] <JT-Shop> gnipsel.com
[11:38:55] <kwallace> Okay. brb
[11:39:06] <jdh> does gnipsel mean somethign?
[11:39:13] <skunkworks> my probe trips ever once in a while during rapids.
[11:39:25] <JT-Shop> yep
[11:39:30] <gene__> it appears so. My Z has a lot of backlash, and once contact is made, it takes about 50 or more taps on the key to take up the backlash and actually break contact
[11:41:49] <gene__> skunkworks, that would almost have to be noise pickup I'd think.
[11:42:25] <skunkworks> gene__, really doubt it.. The input is a pretty stout opto.
[11:42:43] <skunkworks> and the probe is pretty sensitive
[11:42:58] <gene__> I just checked here, with z set for 1"/min, it takes 13 taps on the right arrow key to reopen the contact
[11:43:55] <skunkworks> http://www.youtube.com/watch?v=U61ub6mtpH4
[11:44:05] <gene__> x, with the ball screw in it, has no problem, even at .1"/min motion
[11:45:09] <gene__> this probe is the tool touching a piece of dbl-sided pcb super-glued to the face of the gage I just made
[11:45:58] <gene__> at 5 volts from the pullup on the C1G, s/b perhaps .0001 accurqate?
[11:46:55] <kwallace> JT-Shop , I can't find it, sorry.
[11:47:11] <gene__> Also, I can't use it to measure backlash as the axis cannot bew zeroed when the contact is in effect
[11:47:19] <JT-Shop> http://gnipsel.com/shop/hardinge/hardinge.xhtml
[11:47:49] <gene__> Thats a bummer for us guys with sow's ear hardware.
[11:48:03] <kwallace> Got it, thanks.
[11:48:07] <JT-Shop> np
[11:48:20] <jdh> gene: call it a digital input instead of a probe
[11:48:33] <jdh> display the input value on a panel
[11:49:13] <JT-Shop> jdh it is what Will Stockdale said to the sargent in the mess hall after the eye test
[11:49:28] <gene__> Bok, but will that work with G38.2 then?
[11:49:54] <gene__> ok, but will that work with G38.2 then?
[11:49:59] <JT-Shop> no
[11:50:20] <gene__> there goes my auto-calibrate idea...
[11:50:52] <JT-Shop> hmmm
[11:51:52] <JT-Shop> didn't someone block the probe input in HAL for some reason?
[11:52:51] <gene__> If you don't "net" it, I'd assume it is ignored.
[11:54:24] <gene__> I need to check the ini to see what I have set for backlash, whatever it is, axis shows it as around .013" in addition to whatever its set for. Grrrrr.
[11:54:47] <JT-Shop> yikes!
[11:55:44] <skunkworks> backlash is evil
[11:56:31] * JT-Shop heads to the cocina to whoop up a sammich
[11:57:50] <jdh> I see.
[11:58:15] <gene__> TEll me about it, z on a 7x12
[11:58:28] <gene__> pure junque!
[11:59:04] <Connor> 7x12 mini lathe ?
[11:59:38] <jdh> if you are using the stock screw, you can tighten it
[12:01:39] <gene__> It is under a slight compression lengthwise, and the nut is just loose enough to allow it to turn. Part of the problem is the very poor and tapered fit of the nut in the apron casting.
[12:03:14] <gene__> This thing is over a decade old now, would I fair any better if I ordered a new nut assembly and the apron casting
[12:04:07] <Connor> gene__: You talking about a 7x12 mini lathe ?
[12:04:11] <gene__> yes
[12:04:23] <Connor> You CNC it ?
[12:04:27] <gene__> yes
[12:04:35] <jdh> put a ballscrew on it.
[12:05:14] <Connor> for the X, that's no problem.. but, how do you do that for the Z? Do they make them small enough ?
[12:05:19] <gene__> A screw that long is most of the price of this thing when I bought it.
[12:05:23] <t12> connor: are you the same connor as on cnczone?
[12:05:33] <Connor> t12: Yes
[12:05:39] <Connor> Well. Connor9220
[12:05:49] <t12> from the uh
[12:05:55] <t12> g0704 thread?
[12:06:00] <Connor> yes.
[12:06:26] <t12> can i bombard you with ?'s here sometime?
[12:06:40] <t12> pondering getting one for cnc conversion
[12:06:43] <Connor> Yea, or jdh, he's got one too.
[12:07:31] <jdh> I think people use 8mm ballscrews for z's ?
[12:08:49] <Connor> gene__: You tighten up the gibs or redo them ?
[12:09:01] <Connor> those can cause lots of issues..
[12:09:24] <IchGuckLive> hi all B)
[12:09:44] <gene__> I have an 8mm on the x, courtesy one of the guys here, but its not long enough for Z by about a foot. And yes, the gib, such as they are, have been scraped.
[12:10:18] <Connor> Wait.. What are you calling the Z and what are you calling the X ?
[12:10:37] <jdh> the other ones :)
[12:10:45] <jdh> unless he's got a helluva cross slide
[12:10:54] <gene__> It woul;d take about 24.5" for Z, which is carraige drive, x is crossfreed.
[12:11:05] <Connor> You have that backwards.
[12:11:20] * archivist giggles at the lathe mill confusion
[12:11:29] <Connor> X is the carraige.
[12:11:38] <Connor> Z is the crossfeed.
[12:11:42] <skunkworks> no - that is correct - X is the crossfeed - Z is the carrage... Z is in-out from spindle.
[12:11:46] <gene__> on your machone maybe
[12:11:51] <skunkworks> heh
[12:11:52] <jthornton> lol
[12:12:02] <kwallace> gene__ , It seems a lot of the slop is from parts and joints from bits that could be removed on a CNC lathe. Can you mount the nut directly to the carriage or slide?
[12:12:03] <Connor> Right Z is in-out.
[12:12:05] <skunkworks> z is always too and fro the spindle.
[12:12:10] <jdh> heh
[12:12:21] <jdh> perhaps this is why mine is manual.
[12:12:29] <gene__> and x is across the spindle axis
[12:12:38] <skunkworks> just like a mill
[12:12:54] <jthornton> gene__, like this one?
http://www.mini-lathe.com/Mini_lathe/Tuning/tuning.htm
[12:13:16] <gene__> My x nut is directly mounted to the cross slide
[12:14:13] <Connor> grubmle.. I got all confused when I tried running my mill as a Lathe.. All my axises were strange... :)
[12:16:13] <jdh> you could get a chinese 16mm screw for it pretty cheap
[12:16:15] <Connor> WHY do they use Z instead of Y anyway ?
[12:16:36] <IchGuckLive> gene__: where are you from USA Europ ...
[12:16:37] <gene__> I did folllow quite a bit of that link, but I had to start by making the way groove in the bottom of the saddle match the angles of the ways.
[12:17:28] <gene__> when the v was pulled down into good contact, the back edge of the saddle was 1/8" up in the air! North central WV, USA
[12:17:39] <skunkworks> because Y isn't what goes towards and away from the spindle - Z is.
[12:18:01] <gene__> No Y on this setup.
[12:18:20] <gene__> Its a lathe, not a mill
[12:18:34] <IchGuckLive> lathe has only ZX in the interpreter to call
[12:18:39] * jthornton thinks a nap is in order now
[12:19:00] <IchGuckLive> jthornton: get a toast also
[12:22:12] <kwallace> Has anyone done any left hand threading?
[12:23:00] <archivist> I am about to if I make the job I have in hand in the lathe
[12:23:56] <kwallace> I'd like to take that button off my screen and maybe move it to an extended screen.
[12:24:27] <gene__> I have not using G76, but I believe its invertable, check the man page. I know its usable for internal as well as external threading
[12:24:30] <IchGuckLive> kwallace: what tread
[12:24:44] <IchGuckLive> size
[12:25:31] <kwallace> http://www.wallacecompany.com/tui/Screenshot-6.png
[12:25:33] <gene__> G76 is universal, you have to tell it what thread pitch in the units yor machine is setup for
[12:26:37] <archivist> kwallace, do I see something missing :)
[12:27:44] <kwallace> The problem is that the graphic needs to accommodate the feature. Plus the software needs to check input sanity, so I only want to support what really gets used.
[12:28:25] <gene__> I've cut, so far, 6-32, 1/4-28, and 5/8-27 on my toy, worked well as long as the tool is sharp.
[12:29:04] <gene__> kwallace, looks great, could be handier than bottle beer and sliced bread combined
[12:30:40] <gene__> This is weird, I can see the actual air breakdown in the LED for pin 10 on the C1G as I creep from good contact back to open. :)
[12:31:25] <IchGuckLive> get a debounce in
[12:32:28] <gene__> The ledcreeps up to about 10% britness before its fully open again.
[12:33:06] <archivist> kwallace, you will want depth progression on coarser work or you will be breaking inserts/tooling or you will have to take small bites per pass
[12:34:52] <Connor> kwallace: That looks like a mach wizard...
[12:35:36] <IchGuckLive> Connor: tkuinter can do this also on linuxcnc
[12:35:45] <Connor> I know.
[12:35:46] <archivist> kwallace, you wont need a safety groove with linuxcnc :)
[12:35:50] <IchGuckLive> glade is also available on this
[12:36:55] <kwallace> It's supposed to look familiar.
[12:37:30] <Connor> kwallace: I noticed some Tormach screen shots in that directory too..
[12:37:44] <kwallace> Oops.
[12:40:16] <skunkworks> crap - already fixed ;)
[12:40:16] <archivist> kwallace, why miss out the taper in out as that is very useful
[12:41:05] <Connor> kwallace: Why did you remove them?
[12:41:14] <kwallace> That's why I'm asking for input.
[12:41:40] <archivist> there is no way I could have done
http://www.collection.archivist.info/archive/DJCPD/PD/2012/2012_05_06_Leawood/IMG_1240.JPG
[12:42:00] <archivist> without a taper out, no safety groove
[12:42:26] <kwallace> It's not up to me to say what gets presented.
[12:43:08] <Connor> huh?
[12:43:11] <archivist> I thought this was your creation
[12:43:18] <IchGuckLive> archivist: reprap
[12:43:25] <kwallace> The taper feature doesn't have a safety zone?
[12:43:28] <skunkworks> kwallace is helping out tormach...
[12:43:28] <FinboySlick> archivist: You do that on your lathe?
[12:43:38] <archivist> FinboySlick, yes
[12:43:46] <Connor> OH.
[12:44:40] <FinboySlick> I really need to get my mill fixed.
[12:44:40] <archivist> kwallace, its coming out in the last threads
[12:45:46] <kwallace> I'll have to try the taper threading to check. Thanks for the warning.
[12:46:02] <archivist> its not taper threading
[12:46:12] <archivist> its the in out move
[12:47:06] <kwallace> Oh, the normal lead in and out?
[12:47:17] <archivist> in or out I should say
[12:47:24] <archivist> yes
[12:47:54] <archivist> but a taper on the out removes the need for a groove
[12:49:41] <kwallace> I've always put a groove in, so I haven't considered otherwise. I may need to.
[12:50:12] <kwallace> It seems safer to always have one.
[12:51:33] <archivist> but this is cnc not manual :)
[12:52:55] <gene__> How about this screw for my 7x12?
http://www.ebay.com/itm/SFU1605-Ball-Screw-L650mm-Ballscrew-With-SFU1605-Single-Ballnut-For-CNC-/290776364520?pt=LH_DefaultDomain_0&hash=item43b39f35e8
[12:54:32] <Connor> 16mm, with 5mm pitch.. should be fine.
[12:54:47] <Connor> LMB2008 probably has the same thing..
[12:54:52] <Connor> he's on ebay too
[12:56:56] <kwallace> You might want to check if this has a double nut or adjustable preload.
[12:57:04] <jdh> how long is the 8mm screw?
[13:06:03] <gene__> That one is 16mm I believe, if I read the engrish right, single row nut, rolled, would still be a 10x improvement in accuracy
[13:06:52] <gene__> I didn't find him on ebay under ball screw listings
[13:07:31] <gene__> 16mmx5mm pitch, might take a re-think in my gear ratios.
[13:08:00] <gene__> ATM its only 2/1 from motor to screw
[13:10:11] <gene__> lets me run 40" rapids
[13:11:19] <gene__> with a 16tpi screw & 38 volts on the motors
[13:14:46] <gene__> That screw above, how much more than a 1" square tube with a slot in the bottom to allow the nut mount to slide in, would I need to keep the crap off the threads?
[13:15:24] <mazafaka> http://mazafaka.posterous.com/tony-hawks-american-wasteland-russian-winter had been riding in the night, now fear to awake late, will sit 3 hours and awaiting for the morning...
[13:19:55] <kwallace> Covering the screw should be doable.
[13:20:29] <JesusAlos> hi people
[13:21:24] <IchGuckLive> Hi how is the weather in spain
[13:21:35] <kwallace> http://www.kabelschlepp.com/conicalspring.htm
[13:25:09] <elmo40> IchGuckLive: how is the oil leak in the Gulf? :-P I bet it is still happening, just a lot less then before
[13:26:07] <IchGuckLive> B)
[13:26:09] <Connor> gene__:
http://stores.ebay.com/linearmotionbearings?_trksid=p2047675.l2563
[13:32:36] <JesusAlos> a bit cold, but less than germany
[13:36:51] <archivist> gene__, balls screw cover imagine a clock spring pulled to a tube
http://www.collection.archivist.info/archive/DJCPD/PD/2009/2009_09_10_Starturn_cnc_lathe/P1010039.JPG
[13:40:16] <elmo40> it os a common system
[13:40:33] <elmo40> Toshiba uses it on their large boring mills
[13:41:44] <archivist> http://www.dqr.co.uk/ballscrew_covers.html
[13:47:34] <kwallace> I made up a belows for my HNC from a manila folder to just get by. It has been working well for six years now. Maybe a paper screw cover would work too.
[13:50:18] <gene__> Conner, that guy has one that is maybe 3/4" too long, but I can machine that. His prices for some items are very good.
[13:50:40] <gene__> What backlash could one expect from a C7 rated screw?
[13:51:37] <gene__> For screw covers, hiw about the spring out of an old, long tape measure?
[13:52:19] <Connor> gene__: Those C7's are the one people are using on the G0704 conversions.. people report .001 to .002 backlash..
[13:52:39] <Connor> For a 7x12, that should be fine.
[13:53:12] <gene__> It would be a 20 thou improvement :)
[13:53:31] <Connor> gene__: You can have him do the machining.. for a little extra $$ I forogt to ask on a set and they came cut flat.. they are surface hardened..
[13:54:14] <Connor> It's a pain to machine too.. because, you have to remove the nut.. and it can be tricky getting it back on if both ends are machined..
[13:56:12] <kwallace> The screw may make the backlash better, but my guess is that once you get used to it, the remaining backlash may become a bother.
[13:56:18] <AR_> that moment when you add up what you have spent on your machine........
[13:56:34] <AR_> i should have just bought a mill
[13:57:05] <AR_> i've definitely spent at least $500
[13:57:10] <AR_> and it isnt even together yet
[13:57:51] <Connor> kwallace: Can't he then use backlash compensation ?
[14:00:11] <gene__> Damn fleabay, they will not let me check out as a guest because my address is alreay known, and there is no way in hell I can recall the throwaway passwords I use with these turkeys.
[14:00:30] <kwallace> In my opinion, no because cutting can pull the axes into the cut. Backlash comp is a way to try and make backlash not feel so bad.
[14:00:42] <Connor> gene__: which one are you looking to get?
[14:01:06] <gene__> leave it sit I guess, while I go scan the wall in the house in case I scribbled it down.
[14:01:12] <Connor> kwallace: Even manual mills/lathes have backlash..
[14:02:18] <gene__> link is <
http://www.ebay.com/itm/1antibacklash-ballscrew-RM1605-675mm-C7-BK-BF12-coupler-/250773562835?pt=LH_DefaultDomain_0&hash=item3a6344e5d3>
[14:02:37] <kwallace> But with manual mills, the operator is trained to lock axes and only feed in a direction that doesn't allow the axis to be drawn in.
[14:04:08] <Connor> He has that one machined.. so your good.. It's made to fit those blocks.
[14:05:13] <Connor> If you use those, you'll want to make sure the pockets on the one for the AC bearings aren't too deep.. or you can end up with some lash from that.. I had to put a shim on mine. and make the bearing spacer a bit longer..
[14:09:28] <kwallace> Angular bearings come with different angles, so one needs to balance axial stiffness with max. RPM.
[14:10:39] <Connor> kwallace: Problem was in the bearing block.. they were machined a tad too deep and had some slop in it.. couldn't pre-load them.. had to add a shim between them..
[14:11:36] <kwallace> Bummer.
[14:11:53] <Connor> then the bearing nut spacer was too short.. caused the lock nut to rub against the rubber bearing shield on the bearing block.. had to have some a bit longer..
[14:11:58] <Connor> got it all worked out in the end..
[14:12:11] <Connor> these was a few years ago now.. I think they've improved their quality since..
[14:15:14] <kwallace> Another thing I learned the hard way, I very carefully machined the bore for a bearing in an aluminum block for a light press fit. While holding it in my hand, the heat made it grow to a slide fit.
[14:16:18] <Connor> heat expansion can be a pain in the a$$
[14:17:39] <kwallace> I should have known and made the fit for mounting in a preheated mount.
[14:18:39] <gene__> Aware of that. I just tried to call ebay and resolve a bad bid that is in the way of my buying this screw, and the bitch claimed she couldn't hear me for the static and hung up, half a damned hour wasted.
[14:45:00] <skunkworks> http://www.youtube.com/watch?v=ix7zAX1byaA
[14:47:17] <ReadError> strange
[14:47:21] <ReadError> canadians must like the G0704
[14:47:29] <ReadError> all the youtube videos are by canadians it seems
[14:48:39] <kwallace> That's pretty cool, but now what?
[14:49:40] <skunkworks> THe world!
[14:50:11] <skunkworks> I think maybe he has metric/english issues... The tool seems very large for the size of the movement
[14:51:24] <archivist> is that some form of chair, human, for the excitement of
[14:52:30] <alex4nder> haha
[14:52:54] <MercuryRising> is there a gcode to home the axis at the current position? say I wanted the final x point in my program to be a 'new home'
[14:54:36] <MercuryRising> i have a 3d printer that uses the A axis, and when I g28 at the end, it a) goes really slowly, and b) unwinds my filament, if I could say "a axis, you're home now!" at the end of my gcode, it would be nice
[14:55:24] <kwallace> I would tend to use a windshield wiper worm drive instead of the direct stepper drive.
[14:59:41] <kwallace> Your using the A axis to feed the filament?
[14:59:55] <MercuryRising> yeah
[15:01:32] <kwallace> Off hand, I would not use motion but maybe a custom component, but I've only given it 30 seconds of thought.
[15:02:28] <skunkworks> can you do G10L20P0A0?
[15:03:02] <skunkworks> (assuming home is also 0)
[15:04:02] <MercuryRising> yeah that worked great!
[15:04:11] <MercuryRising> thanks kwallace and skunkworks!
[15:05:50] <skunkworks> kwallace, I would think you would want motion - I would think the blending of all axis's would be a plus
[15:06:42] <kwallace> Oh I missed that part, one needs the coordinated motion with XYZ. Oops.
[15:13:59] <MercuryRising> hmmmm that actually didn't work
[15:14:46] <MercuryRising> it 'zerod' my axis, but it did not rehome it, so when i load up a new gcode file, it runs from the negative the final starting point to 0
[15:17:33] <skunkworks> change so - what is at the begining of the gcode? it must somehow set the a to machine coordinates...
[15:19:04] <MercuryRising> g90
[15:19:22] <MercuryRising> here's the first lines before it starts moving around with g0's
[15:19:24] <MercuryRising> G28 ; home all axes
[15:19:26] <MercuryRising> G90 ; use absolute coordinates
[15:19:28] <MercuryRising> G21 ; set units to millimeters
[15:20:05] <skunkworks> g28 isn't home all axis - it is - go to a predefined location
[15:21:04] <MercuryRising> the comments are from slic3r (although I thought it was home) and it's mainly developed for the reprap based gcode, it might be a little different
[15:21:28] <MercuryRising> they have an 'output for emc/mach3' option that i'm using
[15:21:36] <skunkworks> right - I have seen that on other controllers - G28 is home - but on linuxcnc it is not.
[15:21:49] <skunkworks> http://linuxcnc.org/docs/2.5/html/gcode/gcode.html#_g28_g28_1_go_to_predefined_position_a_id_sec_g28_g28_1_a
[15:22:17] <MercuryRising> you can have 9 axes of movement in linuxcnc?
[15:23:20] <skunkworks> yes
[15:23:28] <skunkworks> xyzabcuvw
[15:23:35] <MercuryRising> what kind of machine would use all 9?
[15:23:41] <skunkworks> robot arm..
[15:24:01] <skunkworks> uvw are usually based on tool tip direction
[15:24:38] <Connor> G28 and G30 are alike... ?
[15:24:46] <skunkworks> yes
[15:24:54] <Connor> No G29 ? :)
[15:28:51] <skunkworks> you should be fine without the g28 if you home the machine when you first start it then G10L20P0A0 at the begining of the gcode will set A to 0
[15:29:19] <skunkworks> each run
[15:32:00] <MercuryRising> alright, i'll try that out, thanks!
[15:50:20] <andypugh> I was trying to think of a way to make a hexagonal socket in the end of a shaft.
[15:50:42] <andypugh> I thouhgt of pressing in a turned-down socket or the head of an allen bolt.
[15:50:53] <andypugh> Then I saw some rotary broaches on eBay
[15:51:16] <andypugh> Only then did I remember that my lathe can bore hxagonal holes...
[15:55:47] <kwallace> But that's one of the things you are famous for.
[15:57:59] <andypugh> Yes, strange thing to forget really
[15:58:34] <andypugh> I have been adding faceting to my lathe config more permanently this evening.
[15:59:02] <andypugh> (A pyvcp box with "number of facets" in it :-)
[15:59:41] <DJ9DJ> gn8
[16:03:15] <ReadError> you can bore hex holes?
[16:03:21] <skunkworks> yah - your going to have to do another video...
[16:04:30] <ReadError> oh theres a video ?
[16:04:31] <skunkworks> ReadError,
http://www.youtube.com/watch?v=T4q8gCpeY1A
[16:05:53] <ReadError> oh thats cool
[16:06:03] <ReadError> how does it detect the position/rate ?
[16:09:55] <kwallace> The thing is that it goes only on spindle and X encoder ticks. Andy is turning the spindle by hand.
[16:12:53] <r00t4rd3d> http://www.youtube.com/watch?v=i0HE7dAw2ns&feature=youtu.be&hd=1
[16:17:00] <kwallace> Turning a hex on a lathe reminds me of a video of a synchronized live tool and spindle cutting a hex. ... here's one at 1:30:
http://www.cnczone.com/forums/general_metalwork_discussion/54494-cutting_hex_cnc_lathe_live_tooling.html
[16:22:48] <andypugh> This machine is specifically for that job, and turns out some interesting shapes. (not just polygons)
http://youtu.be/1uZ241GRreg
[16:30:13] <JT-Shop> what is the meaning of pose in this? " g5x_offset
[16:30:13] <JT-Shop> offset of the currently active coordinate system, a pose. Tuple of floats.
[16:30:13] <JT-Shop> "
[16:34:44] <kwallace> A pose is a set of joint positions? So Home would describe a particular pose.
[16:35:10] <JT-Shop> ah ok
[16:35:21] <kwallace> Just a guess.
[16:35:31] <JT-Shop> better than my guess
[16:42:09] <andypugh> A pose is a set of _axis_ positions.
[16:42:33] <andypugh> (As I understand it)
[16:42:47] <andypugh> XYZABC
[16:48:59] <JT-Shop> thanks
[16:50:34] <JT-Shop> just doesn't seem a logical description of a tuple of floats to me
[16:51:14] <andypugh> PCW: If I comment out the flash_start call, does that mean I can dry-run the rest of the code without worrying about breaking stuff?
[16:51:16] <cradek> I don't think it's helpful in a documentation sense to call it a pose
[16:51:40] <andypugh> JT-Shop: A "tuple of floats" is a description of a Pose, not the other way round.
[16:51:41] <JT-Shop> kinda what I was thinking
[16:52:08] <andypugh> Some tuples of floats are not poses.
[16:52:10] <cradek> it's not at all clear to me that adding "poses" makes sense. adding tuples makes sense, and adding offsets to a pose makes sense (you get a different pose)
[16:52:24] <cradek> yeah I think we're all saying the same thing
[16:52:25] <JT-Shop> ok, using pose like one would pose for a photograph
[16:52:37] <cradek> yeah
[16:52:57] <andypugh> I personally had no idea what a "tuple" was until last year, so I would also suggest removing that from the docs. It isn't a normal word for normal people.
[16:52:58] <cradek> "everyone move left a bit!" is not also a pose
[16:53:23] <cradek> this is a bit esoteric
[16:53:58] <cradek> 'tuple' has specific meaning in the python language, so it might make sense to leave that, if you're describing the types as well as what the values mean
[16:54:26] <cradek> but if you're only describing what the values mean, yeah it's unnecessarily technical
[16:55:09] <andypugh> Other languages use "vector" for "tuple". Which doesn't make any more sense.
[16:55:11] <archivist> I hate the spread of the word tuple
[16:55:17] <JT-Shop> to use the stat you need to know it returns a tuple
[16:55:28] <cradek> what do you mean spread?
[16:55:33] * cpresser uses both. vector and tuple.
[16:56:13] <cpresser> in fact, they are both used together as 'vector tuple' (at least in the german language)
[16:56:14] <archivist> a set of values that are related and belong to each other (database users over use it too)
[16:56:47] <cradek> archivist: it's the name of a particular type in python, and aside from that I have no opinion :-)
[16:56:55] <cpresser> but since i am no native english speaker, my input to this conversation is not that valuable :)
[16:57:06] <cradek> I haven't heard it in everyday speech otherwise. maybe it'll get here eventually.
[16:57:22] <archivist> I have never seen it as a vector till now
[16:57:28] <cradek> cpresser: it's not an English question so much as a python question, I think
[16:59:31] <cpresser> in my experience touple is not an everyday word. my coworker here also doesnt know its meaning. i did learn about tuples at university. so it might not be a common word
[16:59:47] <cradek> I agree it's not
[16:59:52] <cpresser> but still, i like the word :)
[17:00:00] <andypugh> We should use tuples of tensors :-)
[17:00:32] <archivist> you are about to be beaten with a wet fish for that
[17:00:35] <cpresser> tensors are just vectors of higher dimension :D
[17:00:41] <cradek> pair, triple, and (to a lesser extent) quad are everyday words, but the generic form tuple (or n-tuple) is not
[17:01:49] <kwallace> I think tuples are in music too.
[17:02:11] <archivist> a quintuplet
[17:02:21] <JT-Shop> http://linuxcnc.org/index.php/english/forum/41-guis/17806-gscreen-a-gtk--glade--python-based-screen?start=294#28561
[17:02:23] <cradek> kwallace: I only know of triplets and duples in music
[17:02:43] <JT-Shop> is a duple a double tuple?
[17:02:48] <JT-Shop> :)
[17:02:58] <kwallace> Duples, that's it. Sorry
[17:36:03] <kwallace> The eBay mill down the road from me went cheap:
http://www.ebay.com/itm/160949107892 . It would look so much better in my shop.
[17:38:35] <Valen> you giant bastard
[17:38:49] <Valen> that would have sold for like $15k in australia :-<
[17:43:57] <kwallace> Valen: It was hard, but I didn't even bid on it. I'm bummed out too.
[18:05:10] <kwallace> Yay, it's snowing:
http://www.wallacecompany.com/E45/cam2_archive/2013-01-10T15:45-0800_E45c2f.jpg
[18:14:16] <DaViruz> no, it isn't
[18:14:24] <Valen> its 42C here
[18:14:47] <Valen> our weather department has just added 2 colours to its temperature display map to account for over 55C
[18:18:41] <t12> https://www.dropbox.com/s/usu4rywt851lewn/2013-01-10%2015.15.13.jpg
[18:18:44] <t12> look at all them disks
[18:21:48] <Valen> theres the odd one there
[18:21:52] <Valen> they doing anything interesting?
[18:22:15] <t12> bulk em storage
[18:22:27] <t12> we have some insane microscope camera that does 8kx8kx25fps effectively
[18:22:35] <t12> so even transient storage of the datasets takes alot
[18:22:40] <Valen> thats a pretty decent camera
[18:22:41] <t12> like 10tb/24hours or so sometimes
[18:22:49] <Valen> compressed at all or raw?
[18:22:52] <t12> the raw camera is 400fps 4kx4k
[18:22:54] <Valen> 10tb isn't that much
[18:23:08] <Valen> 3.5Gb drives are only $150 now
[18:23:10] <t12> than its mathed up to 8k8k
[18:23:20] <Valen> i remember when I got a 2Gb drive
[18:23:22] <t12> its not alot, but the logistics of dealing with it is hard
[18:23:31] <t12> like how do you actually move the dataset around effectively
[18:23:38] <t12> 10gbe works, but costs alot and
[18:23:43] <Valen> now you can get a 2TB drive for half what i paid for it, a quater in real terms
[18:23:51] <t12> like 10gbe sfp+ fiber is about 3.5k/link
[18:23:59] <Valen> yeah screw that lol
[18:24:10] <t12> thats not including the nics
[18:24:20] <t12> then theres backing it all up..
[18:24:29] <Valen> you can get cheap infiniband cards these days
[18:24:36] <t12> ya
[18:24:41] <t12> theres alot of convergance 10gbe/IB cards
[18:24:44] <t12> where you can use them for either
[18:25:06] <Valen> thats new
[18:25:12] <t12> i think mellanox is into that
[18:25:19] <Valen> I wanted the infiniband for latency reasons with DRBD
[18:25:34] <t12> we use it for big simulations sometimes
[18:25:48] <t12> IB latency is nice
[18:27:14] <t12> did you get to try it for drbd
[18:27:22] <t12> i'd expect actual disk latency to be the issue before 1gbe
[18:27:26] <t12> unless its for ssd or something
[18:27:55] <Valen> its just the way drbd is done
[18:28:15] <Valen> the write is only confirmed when its flushed on the disk attached over the network
[18:28:38] <t12> ah ok
[18:28:50] <Valen> so if i can shave any time off it at all for a few hundred $ why not
[18:29:13] <t12> were you just gonna do ib host to host
[18:29:34] <Valen> yeah
[18:29:42] <t12> ah yeah thats def worthwhile
[18:29:52] <t12> ib switch cost is heavy until you're saturating the switch upfront
[18:30:26] <Valen> its not totally needed, but its really cool lol
[18:30:47] <Valen> I do wish I could mangle the thing thats running DRBD for me so i could use raw IB, not IPoIB
[18:31:23] <t12> hum
[18:31:24] <Valen> I run bonded-rr gbit at the moment
[18:31:48] <t12> i guess it depends on how modular the drbd code is
[18:31:59] <t12> if the network layer can be abstracted transport over ib wouldnt be so bad
[18:32:10] <t12> i dunno if ib has native reliable transport streams or not
[18:32:35] <PCW> "PCW: If I comment out the flash_start call, does that mean I can dry-run the rest of the code without worrying about breaking stuff?"
[18:32:36] <t12> otherwise you kinda have to reimplment tcp on top of ib
[18:33:00] <ReadError> man im lovin hsmworks
[18:33:05] <PCW> Andy no that wont work, you need to skip the "commits"
[18:33:57] <Valen> oh drbd can do it just fine
[18:34:21] <Valen> i'm using ganeti to manage my VM's and it takes care of setting up DRBD for me
[18:34:27] <Valen> but ganeti doesn't know about IB
[18:35:36] <t12> i'm just assuming drbd depends on tcp for stream integrity from a gif on their page
[18:36:41] <Valen> I know they have added IB support recently
[18:36:56] <Valen> well not that recent
[18:37:13] <Valen> http://www.drbd.org/users-guide/s-replication-transports.html
[18:37:31] <t12> oh yeagh
[18:37:39] <t12> cool it directly supports it then
[18:38:02] <Valen> so what are you doing taking high speed microscope photos for 24 hours anyway?
[18:39:49] <t12> EM tomography and EM single particle reconstruction
[18:40:19] <t12> em tomo is essentially a CT scan of a single cell or object by electron microscope
[18:40:37] <t12> em single particle reconstruction is prepare a sample of say, some protein
[18:40:43] <t12> take a zillion pictures of the single proteins
[18:40:58] <t12> align the pictures to eachother and you can gradually back out the outside surface shape
[18:41:12] <t12> i think the record is down to around 4A for some virus coats now
[18:43:20] <t12> half of it is the science of the output of all that
[18:43:37] <t12> the other half is the development of that camera and various things about the microscope to make everything work
[18:50:15] <Valen> I spose your going to be doing a fair bit of seeking style access not so much streaming
[18:50:51] <Valen> has anybody managed to get round-robin style bonding working through a switch yet?
[18:52:34] <t12> like just non LACP aggregates?
[18:52:39] <t12> its not so seeky
[18:52:48] <t12> alot of it just jamming huge datasets into memory and working on them there
[18:53:07] <t12> the micrographs are large, but you end up cutting out little 200x200ish pixel images and thats where all the math happens
[18:53:10] <t12> so you can put alot in ram
[18:53:20] <Valen> i mean if you lacp it you can get 2x data rate to 2 hosts but not to 2 hosts
[18:53:27] <Valen> bah
[18:53:30] <Valen> not to 1 host
[18:53:39] <Valen> so 2x gbe ports != 2gbe
[18:53:44] <t12> yeah with non-lacp ittl just RR though
[18:53:47] <t12> (i think)
[18:53:59] <Valen> yeah but that won't go through a switch
[18:54:08] <Valen> switches tend to out-of order the packets
[18:54:10] <t12> ittl go through a real switch!
[18:54:21] <Valen> doesn't go through my cisco 2960s
[18:55:09] <t12> hum i dunno what the keywords would be for ios
[18:55:15] <t12> i know on my hp and dell switches
[18:55:20] <t12> i just specify trunks as being non-lacp
[18:55:41] <Valen> I'll have another google at it
[18:55:56] <Valen> it doesn't matter too much now I have the rr crossover connection
[18:56:27] <t12> or its more like unless you say it IS lacp, it isnt
[18:56:32] <andypugh_> PCW: So I am safe without the commits? That's fairly easy.
[19:11:08] <Mr_Wolfsl> hello
[19:11:28] <Turkishviking> hi
[19:23:41] <kwallace> I got a new toy from Spark fun in the mail. 433mHz transmitter and receiver to fix a rain gauge, I hope.
http://www.wallacecompany.com/tmp/Image.jpg
[19:39:08] <skunkworks> I have to stay away from sparkfun
[19:39:16] <skunkworks> I have way too many other projects...
[19:49:07] <frysteev> how does one get anolog inputs in?
[19:53:15] <r00t4rd3d> my names billy and i hack stuff
[19:54:35] <tjb1> r00t4rd3d: shh
[19:56:55] <r00t4rd3d> i was cutting some purple heart earlier and a splinter got jammed up in my bearings and my shit stalled out
[19:57:05] <tjb1> dwp611?
[19:57:08] <tjb1> or the machine
[19:57:13] <r00t4rd3d> x axis
[19:57:22] <tjb1> single or dual?
[19:57:25] <r00t4rd3d> single
[19:57:42] <tjb1> oh yeah you use ballscrew dont you
[19:57:43] <tjb1> or acme
[19:57:54] <r00t4rd3d> hardware store for my x
[19:57:59] <r00t4rd3d> what ever that stuff is
[19:58:00] <tjb1> all thread?
[19:58:18] <tjb1> My rack just skips if it gets stuck :)
[19:58:44] <r00t4rd3d> my nut will allow for it also
[19:58:50] <tjb1> lol
[19:59:48] <r00t4rd3d> one bad thing about cncrp linear carriages
[20:00:35] <r00t4rd3d> i was vacuuming too
[20:03:48] <tjb1> you arent running the brushes though
[20:04:10] <tjb1> Why dont you get some of that from mcmaster and throw it on
[20:04:32] <r00t4rd3d> i have one
[20:04:37] <r00t4rd3d> wasnt using it
[20:04:45] <tjb1> the brush that cleans the rail?
[20:04:47] <r00t4rd3d> i had my air deflector on
[20:05:05] <tjb1> put brushes in front of your carriages
[20:05:12] <r00t4rd3d> oh
[20:05:37] <r00t4rd3d> the piece that jammed me up landed in the middle
[20:05:44] <tjb1> let me find a pic
[20:05:54] <r00t4rd3d> i know what i am going to do
[20:06:11] <r00t4rd3d> this usmmer
[20:06:14] <tjb1> http://www.cncrouterparts.com/images/ProximityLimitSwitch.JPG
[20:06:14] <r00t4rd3d> summer*
[20:06:46] <r00t4rd3d> yeah thats nice
[20:07:09] <tjb1> I need to get better proximity sensors...
[20:07:14] <Valen> http://www.youtube.com/watch?v=fhiniF4DCUo
[20:07:24] <Valen> Adam talks about Jamie at Wootstock
[20:07:26] <r00t4rd3d> i have to constantly clean my bearings off
[20:07:43] <tjb1> But at $130 each I dont think I am buying those ones either
[20:08:00] <Valen> nsfw btw
[20:11:56] <r00t4rd3d> lol
[20:12:01] <HowardZ> Anyone used
http://rankone.us before?
[20:13:02] <r00t4rd3d> what a scumbag
[20:32:04] <skunkworks> frysteev: I use an arduino...
[20:32:36] <skunkworks> If it doen't have to be real realtime.. (I use it for spindle temp to compensate for the spindle growth)
[20:42:04] <frysteev> hmm
[20:42:57] <frysteev> i only have 2 to worry about right now, torch neight air pressure
[20:45:42] <Tom_itx> skunkworks in what way does spindle growth affect a cut?
[20:45:46] <skunkworks> I don't know how fast torch hight has to be... Like I say - usb is not realtime.
[20:45:51] <Tom_itx> length? diameter?
[20:45:51] <skunkworks> Tom_itx: yes
[20:45:55] <skunkworks> length
[20:46:02] <Tom_itx> how much?
[20:46:21] <Tom_itx> have you figured a relationship between temp and length?
[20:47:03] <jdh> definite relationship between cold and length
[20:49:11] <skunkworks> http://www.youtube.com/watch?v=h-CdFd2Zakc&feature=share&list=UUHk52YjGT8HryRYmJKSl-lg
[20:49:27] <skunkworks> yes - graphed it - ended up being linear
[20:50:44] <RyanS> I read some quote "Yet another TB6560 driver board" .. What's wrong with the TB6560?
[20:52:00] <Tom_itx> nice
[20:52:43] <Tom_itx> how much overall change do you allow?
[20:53:11] <skunkworks> I limit it to +/-.005
[20:53:35] <Tom_itx> and do you reset it to zero say you change a tool in the middle of a run for whatever reason
[20:54:05] <skunkworks> no - the spindle sticks out for the rails about a foot - that is all I am compinsating for
[20:54:38] <Tom_itx> i noticed our lathe would vary throughout a day's run
[20:54:53] <ReadError> dang skunkworks
[20:54:55] <ReadError> hardcore stuff
[20:55:13] <Tom_itx> not alot but enough when you're running 'close' parts
[20:55:14] <skunkworks> that is why linuxcnc is so cool :) what ever your heart desires..
[20:55:19] <skunkworks> yes
[20:56:47] <Tom_itx> where is the temp sensor mounted?
[20:57:31] <skunkworks> in the spindle ;) it is the original sensor.
[20:57:40] <Tom_itx> oh
[20:58:30] <ReadError> going to try the hero3 and get some video tonight
[20:59:04] <jdh> I need a black hero3
[20:59:11] <ReadError> ya thats what i got ;)
[20:59:34] <jdh> d00d said the wifi-to-phone stuff worked underwater
[21:00:05] <ReadError> i got it monday, just havent filmed anything with it
[21:06:28] <r00t4rd3d> put it on your trumpet
[21:06:45] <ReadError> this doesnt feel right
[21:06:51] <ReadError> its jan, almost 10pm
[21:06:55] <ReadError> and 63F outside
[21:07:02] <r00t4rd3d> https://www.youtube.com/watch?v=LxX7xrazxLc
[21:07:27] <tjtr33> skunkworks, does the temp eventually feed an offset comp?
[21:07:36] <ReadError> i see you read the reddits r00t4rd3d
[21:07:38] <jdh> I just turned the a/c on
[21:07:58] <ReadError> jdh: feels wrong dont it?
[21:08:06] <jdh> not here
[21:08:27] <jdh> gulfstream keeps it fairly warm in the winter anyway
[21:09:00] <MercuryRising> it's jan 10 and it's raining outside!
[21:09:06] <jdh> ocean temp is 74F
[21:09:37] <ReadError> winter is pretty non existant so far this year
[21:11:08] <jdh> upper 70's this w/e, I'm going diving sunday :)
[21:13:02] <skunkworks> raining here too
[21:13:10] <skunkworks> should be -10f
[21:13:29] <MercuryRising> where you from skunkworks?
[21:13:33] <skunkworks> WI
[21:13:52] <MercuryRising> i now dislike you, I'm from MN :O
[21:14:05] <frysteev> silly muricans
[21:14:31] <MercuryRising> they beat us in football last weekend
[21:14:40] <skunkworks> I am in trempealeau
[21:14:49] <ReadError> 'MURICA
[21:18:18] <jdh> leave them alone, it helps them to feel superior.
[21:18:39] <skunkworks> eh - I don't watch sports...
[21:19:05] <frysteev> cnc football?
[21:19:14] <skunkworks> oooooh
[21:19:15] <skunkworks> maybe
[21:47:39] <RyanS> Not sure what to search for, I want to research what signals etc DB25 breakout boards communicate to the stepper drivers? Are breakout boards nothing more than an opto-isolated pass through between the PC and drivers?
[21:48:35] <jdh> mostly
[21:51:57] <jdh> http://www.nbglin.com/cnc25.htm
[22:03:25] <tjtr33> a Linuxcnc parport can be cfg'd to be in or out or 'x' mode, that changes what is presented to your breakout board right away. see Hal manual Ch.10
[22:04:00] <tjtr33> then, look at your stepper driver, they have similar functions, but wildly differnt pin locations
[22:04:14] <jdh> unless you have a g540 :)
[22:05:25] <tjtr33> becuz its a 4in1?
[22:36:35] <ReadError> im pretty happy with my g540
[22:36:44] <ReadError> small, clean, gives me no issues
[22:37:41] <skunkworks> did the charge pump work fr you?
[22:38:29] <ReadError> not sure
[22:38:36] <ReadError> never noticed anything not working i can tell
[22:38:54] <ReadError> i have it off currently
[22:38:59] <ReadError> per the switch
[22:50:59] <jdh> where does one find those?
[22:51:57] <ReadError> find what
[22:52:38] <jdh> wrong channel... but 1972 nagashima(?) dark side of the moon live recording
[23:04:07] <tjb1> charge pump works for me
[23:08:27] <r00t4rd3d> charge your face
