#emc | Logs for 2009-04-09

[00:33:30] <cradek> would feeler gauges rust in gasoline? they are attracted to a magnet so I think that means they are not stainless.
[00:34:20] <cradek> I know I could get stainless, but I don't seem to have any.
[00:34:53] <BigJohnT> if the gasoline had any water in it they would
[00:35:14] <cradek> yeah, I'm pretty sure gasoline normally has a bit of water
[00:36:49] <ds3> 400 series SS is attracted to magnets
[00:37:28] <BigJohnT> are you testing the rusting of feeler gauges in gasoline?
[00:38:10] <cradek> ds3: hmmm
[00:38:13] <Skullworks> and currently with the Ethanol content it can absorb even more water without drops forming at the bottom
[00:38:37] <cradek> I think I've never seen rusty feeler gauges, so I'm suspicious that they are stainless
[00:38:46] <cradek> a little rust would ruin them pretty fast
[00:38:59] <cradek> BigJohnT: considering using them to make flapper valves for my fuel pump
[00:39:12] <cradek> turns out there's not enough height available to make a ball+spring valve.
[00:39:29] <BigJohnT> get some stainless shim stock...
[00:39:40] <Skullworks> 2nd
[00:39:57] <BigJohnT> what kind of fuel pump?
[00:40:10] <cradek> old fashioned carburetor pump
[00:40:54] <BigJohnT> for check valves
[00:41:34] <cradek> hmmmmm I have .005 brass shim stock
[00:42:08] <BigJohnT> wabro uses some kind of fiber inpregenated material on their check valves
[00:42:11] <Skullworks> some .02" g10?
[00:42:28] <SWPadnos> I think there's a SS shim stock kit from MSC that's got 0.0005 - 0.012, somthing like 15 or 20 sheets
[00:42:30] <SWPadnos> $30 or so
[00:42:46] <cradek> that's tempting
[00:43:15] <SWPadnos> oops, 0.001, 0.0015 ... to 0.015
[00:43:16] <BigJohnT> don't some floppy drives have a flat ss band that is used to move the head
[00:43:19] <SWPadnos> http://www1.mscdirect.com/CGI/NNSRIT?PMPXNO=1647761&PMT4NO=61705617
[00:43:34] <eric_unterhause1> is using a mill DRO on a lathe a horrible thing to do?
[00:44:03] <cradek> eric_unterhause1: it'll read radius
[00:44:18] <BigJohnT> or was that an old hard drive I smashed up...
[00:44:47] <SWPadnos> unless you can adjust the scale on the two axes independently
[00:45:01] <BigJohnT> you can get an 8 x 12 sheet of 0.001 ss shim from mcmaster for $7
[00:45:01] <eric_unterhause1> I think I can live with radius
[00:45:14] <cradek> now we play "try to cut some off the roll without bleeding all over it"
[00:45:24] <SWPadnos> this is 12 6x12 sheets for $27.18 total
[00:45:52] <BigJohnT> cradek: are you restoring something?
[00:46:10] <SWPadnos> I'd send you some (brass or stainless), but I'd proably send the wrong thickness or size or something
[00:46:10] <cradek> just fixing the fuel pump in my 52 chevy POS
[00:46:37] <BigJohnT> you can't get a rebuild kit for it?
[00:46:40] <cradek> it can still pump fuel, but not air, and it loses its prime
[00:46:52] <cradek> BigJohnT: it would be as bad as the original - I've already replaced it once
[00:47:08] <BigJohnT> ok
[00:47:57] <BigJohnT> sounds like my 81 chevy I could take it apart and put it back in my sleep I did it so many times
[00:47:57] <cradek> replacements pumps are only $110 or so, but experience shows they only last a few years :-)
[00:49:25] <BigJohnT> I could snip you off a piece next time I go to briggs and put it in an envelope for a few cents postage if you want to try some
[00:49:38] <cradek> hm, with brass I can solder the flapper to the flat
[00:49:44] <cradek> I'm going to try this stuff
[00:49:50] <BigJohnT> they have every size of brass shim stock from 0.001 to 0.031
[00:49:53] <cradek> might be super simple if it works
[00:50:03] <cradek> thanks for the offer - this .005 I have seems fine
[00:50:15] <cradek> I also have .001 but that's really flimsy
[00:50:31] <Skullworks> When I found out where the fuel pump was on my Ford Ranger (blocked by steering gear box) I just used a low pressure electronic inline pump.
[00:50:54] <Skullworks> $29
[00:51:09] <cradek> Skullworks: that's definitely plan B :-)
[00:51:41] <BigJohnT> cradek: do you have a picture of the pump opened up?
[00:51:50] <eric_unterhause1> people with the model of taurus that I had cut a hole in the floorboard to get to the pump
[00:52:13] <cradek> BigJohnT: I'll take one...
[00:52:17] <BigJohnT> cool
[00:56:02] <cradek> man this camera sucks
[00:57:58] <cradek> http://timeguy.com/cradek-files/emc/pict6438.jpg
[00:58:34] <cradek> the flapper valve is a rubber washer with a spring on top (pressing down from that top cup)
[00:58:44] <cradek> two of these were pressed in those holes
[00:59:02] <cradek> I destroyed one because I couldn't get to both sides to press it out
[01:00:04] <cradek> you can see how little height there is to work with on the inlet valve (the near one)
[01:00:14] <BigJohnT> * BigJohnT hums the final jepordy song while the picture downloads
[01:01:16] <BigJohnT> I see what you mean
[01:01:54] <cradek> you can sort of see how the edges of the rubber washer are curled up - that's the failure
[01:02:03] <cradek> the spring can't press on the edges
[01:02:42] <BigJohnT> ok, I see that
[01:03:19] <BigJohnT> ever thought about using check valves screwed into the fitting holes?
[01:03:42] <cradek> interesting, that could work
[01:04:37] <BigJohnT> you can get brass check valves in 1/8 npt up
[01:05:03] <BigJohnT> viton seat 1/8 ntp male x female is only $10
[01:05:10] <BigJohnT> from mcmaster
[01:05:26] <BigJohnT> 7768k55
[01:08:23] <BigJohnT> * BigJohnT heads up to veg out in the easy chair for an hour before bed
[01:08:30] <BigJohnT> talk to you later
[01:13:04] <cradek> heh, holes are 0.997 - american car
[01:28:00] <eric_unterhause1> how do you write a web page such that text is blocked by a column?
[01:32:35] <eric_unterhause1> http://travisgoodspeed.blogspot.com/2007/11/ti-ez430-in-linux-with-iar-kickstart.html
[01:33:21] <eric_unterhause1> somebody check that page to see if it's cut off so I don't have to go downstairs and check on another comp
[01:35:23] <SWPadnos> tie right side of the monospace text is cut off on firefox 3 (on Windows)
[01:35:40] <eric_unterhause1> thanks
[01:35:42] <SWPadnos> s/tie/the/
[01:35:45] <SWPadnos> sure
[01:36:25] <SWPadnos> but I can see it all if I select "No Style" from the view -> Page Style menu
[01:37:19] <eric_unterhause1> didn't know you could do that, thanks
[01:37:39] <SWPadnos> oh, sure
[01:38:52] <eric_unterhause1> i want to make a bicycle light controlled with the msp430
[01:39:41] <eric_unterhause1> I figure I can spare 6 milliamps
[01:39:42] <SWPadnos> seems like overkill for a light
[01:39:57] <eric_unterhause1> the chinese guys are using AVRs and PICs
[01:40:07] <SWPadnos> yep, that's what I'd do ;)
[01:40:25] <Valen> PICs are nice and easy ;->
[01:40:26] <SWPadnos> or maybe one od the Z8-style chips, I think they have a low power series
[01:40:34] <eric_unterhause1> not that low
[01:40:44] <SWPadnos> the PicoPower AVRs may have the best power consumption though
[01:40:56] <Valen> AVR's probably have greater community support but a higher learning curve to get into
[01:40:59] <eric_unterhause1> plus you can turn the msp on and off very fast
[01:41:04] <eric_unterhause1> without losing anything
[01:41:19] <Valen> a nanowatt PIC can "run" off a coin cell until the cell goes flat from age
[01:41:28] <SWPadnos> presumably you still have to maintain 0.7V or 1.5V or something to keep the internal RAM alive ...
[01:41:49] <SWPadnos> Valen, the picopower AVRs are in a similar class, I think
[01:42:08] <Valen> yes you keep it at full voltage, it just doesn't draw much current
[01:42:10] <Valen> pico amps
[01:42:31] <eric_unterhause1> last I checked, msp was the best at sleeping, but time marches on
[01:42:40] <SWPadnos> picoamps would surprise me
[01:42:51] <SWPadnos> nanoamps I can believe easily though
[01:43:34] <Valen> I could be wrong, but in their maximum sleep mode i believe you can get current draw that low
[01:44:11] <SWPadnos> ok. that's not running in my book :)
[01:44:19] <Valen> you have to do crazy things with regards pin output states and the like I believe to ensure your truly in the lowest power mode
[01:45:01] <Valen> If you want it to do something then you need to spend some power ;-P
[01:45:08] <SWPadnos> yep, 50 nA@2V typical in standby
[01:45:22] <ds3> but the ez430 dev kit is just cool ;)
[01:45:26] <Valen> If your driving a PWM out your going to be sucking some power
[01:45:28] <SWPadnos> (for the PIC16F631/677/685/687/689/690
[01:45:31] <SWPadnos> )
[01:45:44] <Valen> are they nanowatt parts?
[01:45:50] <SWPadnos> yes
[01:45:56] <SWPadnos> http://ww1.microchip.com/downloads/en/DeviceDoc/41262E.pdf
[01:46:12] <SWPadnos> there may be others, I just chose the 16F series because they've been the lowest end in my experience
[01:46:13] <eric_unterhause1> the nice thing about ez430 is you can get a working computer for $3
[01:46:15] <ds3> wonder how well would they run off of 2 dissimiliar metal rods in coolant
[01:46:17] <SWPadnos> heh
[01:46:23] <Valen> lol
[01:46:28] <SWPadnos> do you get 2V from that?
[01:46:36] <SWPadnos> isn't it around 0.9-1.1V?
[01:46:44] <Valen> there are pics and AVRs that will run off .7
[01:46:44] <ds3> donno...the TI booth was using grapes
[01:46:48] <SWPadnos> heh
[01:46:52] <Valen> they have a built in boost converter
[01:46:59] <SWPadnos> grapes are almost as good as ptotatoes :)
[01:47:05] <SWPadnos> lots of acid in there
[01:47:18] <Valen> they might be better? pretty watery less internal resistance ;->
[01:47:28] <SWPadnos> yeah, they may be
[01:47:36] <eric_unterhause1> I should go see if my vfd will spin my lathe fast enough to cut the light housing
[01:47:38] <SWPadnos> I've never experimented in that realm :)
[01:47:46] <eric_unterhause1> or else the controller will have to be naked
[01:49:45] <Valen> w000 naked pics
[01:50:08] <eric_unterhause1> the first time I fired up the leds for the light, I was seeing spots for about an hour
[01:50:24] <SWPadnos> 130000 mCd?
[01:50:31] <eric_unterhause1> 140 lumen
[01:50:36] <Valen> got pics of this light?
[01:50:37] <eric_unterhause1> I don't do no mCds
[01:50:40] <SWPadnos> heh
[01:50:43] <SWPadnos> what size LED?
[01:50:52] <eric_unterhause1> it's on a 20mm board
[01:51:01] <Valen> whats 140 lumen equivalent in incandescent watts?
[01:51:03] <SWPadnos> ok, similar to the Luxeon?
[01:51:05] <eric_unterhause1> Cree XPE R2 binning
[01:51:13] <SWPadnos> I think it's about like 10W
[01:51:23] <SWPadnos> maybe 15
[01:51:37] <eric_unterhause1> I think it's more like 100w
[01:51:46] <eric_unterhause1> incandescent
[01:51:56] <eric_unterhause1> car lights are 600 lumen
[01:52:00] <Valen> yeah that sounds plausible
[01:52:09] <Valen> car lights are (halogens) only 50W
[01:52:20] <Valen> so thats ~100W incandecant
[01:52:22] <SWPadnos> Ive got a box of 100W light bulbs in front of me, and it says 1530 lumens
[01:52:25] <eric_unterhause1> W not a good measure, obviously
[01:52:31] <Valen> true that
[01:52:39] <Valen> but it has a "feel" to it
[01:52:45] <eric_unterhause1> lux are probably better
[01:52:51] <eric_unterhause1> nobody uses them though
[01:53:02] <SWPadnos> the intensity from the LED will be much higher due to the smaller size
[01:53:05] <Valen> candles works ;->
[01:53:18] <eric_unterhause1> those aren't used very often either
[01:53:18] <SWPadnos> but the total illuminance is ~1/10 of the 100W bulbs
[01:53:36] <Valen> I just want to know when i can put LED headlights in my car
[01:53:41] <eric_unterhause1> soon
[01:53:44] <Valen> I know somebody who had a sample of one
[01:53:55] <Valen> had to be real careful hooking it up
[01:53:57] <eric_unterhause1> they have multiple emitters on a single substrate now
[01:54:07] <eric_unterhause1> 4 and 6
[01:54:21] <SWPadnos> yeah, the difference between headlights and all the others is that the others are used as indicators, whereas the headlights are illuminators
[01:54:29] <eric_unterhause1> one of the 6 emitters are supposed to put out 1000 lumen
[01:54:39] <SWPadnos> indicators in that others need to see them, they don't have to light up the surroundings
[01:54:56] <SWPadnos> you need big heatsinks on those
[01:55:03] <SWPadnos> the ones I've seen anyway
[01:55:10] <eric_unterhause1> one guy built a helmet light with 2
[01:55:15] <eric_unterhause1> he used a fan
[01:55:34] <eric_unterhause1> that will light up the road
[01:55:45] <eric_unterhause1> actually, it's too bright, cops will stop you for that
[01:56:26] <eric_unterhause1> car headlights cut off the top of the light
[01:56:54] <SWPadnos> car headlights don't flash all over the place when youturn your head ;)
[01:57:17] <eric_unterhause1> yeah, the high powered ones are for off-road use
[01:57:41] <eric_unterhause1> I wear a 45lumen light, it's good for getting people to dim their lights
[01:58:27] <eric_unterhause1> the people that don't get to stare at it for a good long time, which can't be pleasant
[02:00:02] <eric_unterhause1> I'm a little worried about the light I'm building, haven't figured out how to dim it yet
[02:00:12] <eric_unterhause1> uses 3 leds
[02:03:45] <SWPadnos> PWM?
[02:04:06] <eric_unterhause1> right, but current controlled
[02:04:38] <eric_unterhause1> generator hub is a current source too, so that part is not easy to get my head around
[02:04:53] <SWPadnos> ok, carry on :)
[02:11:03] <eric_unterhause1> the real trick will be carrying my oscilloscope along for the ride with a very long extension cord
[02:18:18] <SWPadnos> battery + car inverter
[02:18:24] <SWPadnos> + lots of padding
[02:18:41] <eric_unterhause1> driving the car and the bike at the same time as reading the scope is a no-no
[02:18:53] <SWPadnos> you need an assistant
[02:20:05] <jdhNC> I have a 1500 lumen flashlight
[02:20:18] <jdhNC> but, it's HID, not LED.
[02:23:07] <eric_unterhause1> since I'm powering the light myself, it has to be led
[02:23:46] <jdhNC> 2amps, 12vdc, just pedal harder.
[02:24:31] <jdhNC> they make an 1100 lumen led version, but it takes almost as much
[02:24:39] <eric_unterhause1> reminds me of the old saying "nothings a problem for the guy that doesn't have to do it"
[02:25:30] <eric_unterhause1> you can get about 15w out of a good generator
[02:25:55] <jdhNC> or in this case, pay for it... the led one is us$1300.00
[02:43:35] <Valen> you need to be PWMing it based on current anyway to handle variances in supply current so just change the "target current"
[02:44:01] <eric_unterhause1> true
[02:44:28] <eric_unterhause1> I want to make some kind of precision controlled rectifier
[02:44:52] <eric_unterhause1> because even the low Vf rectifier diodes take up a lot of voltage
[02:59:02] <John_F> how about a buck regulator with synchronus rectifiers
[02:59:25] <eric_unterhause1> thanks, I was trying to remember what it was I wanted to do
[02:59:43] <eric_unterhause1> thought of synchronous regulators the other day but forgot to look them up
[02:59:48] <John_F> use current mode control. all you need is the inner current loop
[04:11:26] <JustinXJS2_> JustinXJS2_ is now known as JustinXJS2
[04:26:00] <Valen> Try to get one that will do cycle by cycle current limiting if your going to be getting sourced from a variable supply
[04:26:50] <fenn> just use a voltage regulator and resistor.. sheesh
[04:27:41] <tomp> any examples of sending nml msgs direct to motion? (are they text? binary? )
[04:27:44] <fenn> going from 300mA to 1000mA gets you like 5% gain in brightness
[04:28:05] <fenn> tomp: look at halui?
[04:28:13] <tomp> thx
[04:38:31] <eric_unterhausen> nice, somewhere along the way ubuntu upgraded the fonts in firefox
[04:40:05] <Skullworks> I just wish there was a way to steal a driver from Xandros and use it in Ubuntu.
[04:49:09] <eric_unterhausen> how hard can stealing a driver be exactly?
[04:50:06] <Skullworks> I'm thinking it was compiled into the specific kernal
[04:50:26] <Skullworks> and Xandros isn't open source
[04:51:00] <eric_unterhausen> drivers are almost always available from somewhere other than the distribution
[04:51:14] <eric_unterhausen> there is a module building package for ubuntu
[04:51:25] <eric_unterhausen> I assume
[04:53:50] <tomp> Skullworks, whats the function youd like?
[04:54:00] <Skullworks> I wish - they have cobbled up an open source version - but on a dual boot system you can see that the driver in Xandros is far superior to the Ubuntu open source offering. - This is for a Wireless "G" card
[04:55:29] <tomp> try lsmod to see your drivers, then ask the author for that good difference, sometimes they're very responsive
[04:57:04] <tomp> and theres a lot to wireless running well beside the hdwr driver, lotsa infrastructure. any diff with wired on the 2 os's?
[04:58:04] <SWPadnos> Xandros may have different firmware from Ubuntu, and that's not covered by the GPL (so they don't have to share)
[04:59:00] <Skullworks> I have the Asus EeePC 900A with the 1.6GHz Intel Atom N270 CPU and it has an Altera 802.11G - the hardwired eth0 is 1000tx and works great in both O/S
[05:04:59] <fenn> Skullworks: xandros isn't open source/
[05:05:17] <fenn> that's ridiculous
[05:05:34] <tomp> ah i was using a 701 this morning, dropping skype like a hot potato (pickitup, pickitup)
[05:05:45] <renesis> i have a 701
[05:06:14] <tomp> it eoesnt have a '' key (puzzle)
[05:06:29] <renesis> ?
[05:06:35] <fenn> a what key?
[05:07:04] <tomp> the 'd' btn top is missing, i have to lean on the little brass bits
[05:07:52] <Skullworks> yeah Xandros isn't open source and Asus isn't trying to help.
[05:07:58] <fenn> sue!
[05:08:30] <fenn> seriously, if you got a kernel with a binary blob compiled into it, they have to give you the source code
[05:08:49] <Skullworks> you calling a Pig or a lawyer?
[05:09:26] <SWPadnos> it's unlikely that the binary blob is compiled in
[05:09:34] <SWPadnos> it's most likely loaded via the firmware interface
[05:09:55] <SWPadnos> so it's bundled along with the open source driver, it's not open source itself
[05:10:06] <SWPadnos> at least that would be my assumption
[05:10:17] <tomp> btw is by the way and back to work
[05:10:59] <fenn> boiling tea water
[05:11:14] <SWPadnos> bing-time wanker
[05:11:16] <Skullworks> I think Xandros came down like MAC OSX, Maybe BSD or other roots
[05:11:18] <SWPadnos> err, big
[05:11:28] <SWPadnos> no, Xandros is Linux
[05:11:34] <SWPadnos> at least it was at the start
[05:13:01] <SWPadnos> Skullworks, have you looked here? ftp://ftp2.xandros.com/src/
[05:16:21] <Skullworks> I had the system running dual boot with eeebuntu, and I think they used what info they could get when they built the existing open driver.
[05:16:36] <Skullworks> Ibex
[05:18:07] <Skullworks> had EMC sim running real nice with mini/tkemc/xemc - Axis was a known no go.
[05:18:53] <SWPadnos> http://support.asus.com/download/download.aspx?modelname=Eee%20PC%20900/XP&SLanguage=en-us
[05:19:08] <Skullworks> but the eeebuntu wireless was weak, the hardware was just not being managed properly.
[05:20:06] <SWPadnos> wifi cards need firmware. that firmware may be different for the Xandros distro, due to NDA/licensing agreements. you might be able to copy the Xandros firmware over to the Ubuntu partition and use it (after suitable renaming, possibly)
[05:20:59] <SWPadnos> there are even issues with disconnects on the supported Intel wifi chipsets, but they're working on it (and it's the same on all distros, since that driver is part of the kernel)
[05:21:31] <Skullworks> that would be the ticket - if I was that good a hacker.
[05:22:13] <SWPadnos> cp /path/to/xandros/partition/whatever_file_it_is /lib/firmware/wherever_it_goes_on_ubuntu
[05:22:41] <Skullworks> I real tweaked out the eeebuntu install - it was really nice - but people kept killing it.
[05:22:51] <SWPadnos> there probably aren't a lot of files in the Xandros firmware tree, since they know exactly what hardware will be
[05:23:41] <fenn> how exactly do you kill ubuntu?
[05:24:37] <Skullworks> I had it running from a 16GB class 6 SDHC in the built in card reader slot - it really gets scrambled when someone removes the SDHC card while the PC was running.
[05:25:27] <fenn> ouch
[05:25:57] <Skullworks> funny thing is the Class 6 SDHC card was faster than the cheapo OED 4GB internal SSD it came with.
[05:26:07] <fenn> so i hear..
[05:26:08] <Skullworks> OEM
[05:26:23] <fenn> * fenn impatiently waits for 16GB card to arrive
[05:26:57] <Skullworks> got mine for $24.95 shipped from Newegg.
[05:28:29] <fenn> me too, probably about the same time you ordered it
[05:29:24] <fenn> april 3
[05:29:50] <Skullworks> I got mine 2wk of JAN
[05:31:36] <Skullworks> What was the real pain was I was down in Guaymas when the unit went FUBAR. No external Opticals with me to effect repairs
[05:34:58] <Skullworks> I want to get one of these, and rig it to dual boot with XP - I still need XP for Solidworks. http://www.mydigitaldiscount.com/ProductDetail.jsp?LISTID=8000092A-1231822185#productphotos
[06:16:06] <Valen> if your using the 901 take a squiz at moblin
[06:16:17] <Valen> If its got SSE3 in it you can run it
[06:16:22] <Valen> and get a 5 second boot
[06:20:51] <Skullworks> I have the 900A
[06:20:59] <Valen> whats the cpu?
[06:21:13] <Skullworks> Intel Atom 1.6GHz
[06:21:55] <Valen> yeah it should work, give it a try
[06:22:01] <Valen> ~300mb and boots off a usb key
[06:22:40] <Skullworks> is it a read only or full R/W?
[06:22:48] <Valen> dunno
[06:23:02] <Valen> i only have a 701 so it didn't boot propperly
[06:23:11] <Valen> i *think* the usb image might be RW
[07:23:24] <pjm_> good morning
[07:45:18] <Skullworks> That sounds like an assumption.
[07:46:22] <pjm_> hahh
[07:46:22] <archivist> being at work baad
[07:46:32] <pjm_> well its good for me, i have a week off work, yay!
[07:47:27] <archivist> boss was talking yesterday about next week being a "holiday"
[07:47:45] <pjm_> not a permenant one i hope
[07:48:20] <pjm_> i cant believe how many small engineering businesses round here are going pop, its bloody sad, there wont be any left soon!
[07:48:26] <archivist> preparing for
[08:53:35] <micges1> good morning
[12:53:17] <skunkworks186> It's a beautiful day in the neighborhood....
[12:53:43] <skunkworks186> skunkworks186 is now known as skunkworks_
[12:53:57] <archivist> you could be indoors breathing the smoke from your drive :)
[12:54:27] <skunkworks_> Smoke? what is this smoke you are talking about?
[12:55:11] <skunkworks_> I wonder if anyone sells small rack mounts locally.
[12:56:22] <skunkworks_> walmart? ;)
[13:00:31] <Valen> you have a CnC machine go get some angle iron ;-P
[13:00:58] <skunkworks_> heh
[13:01:14] <skunkworks_> I wonder if I could talk the machine shop into making a simple frame..
[13:27:16] <Valen> I'm sure money could convince them ;->
[13:54:38] <eric_unterhausen> skunkworks_ you want the rack or the box?
[13:54:49] <eric_unterhausen> music stores often have racks
[13:55:05] <Optic> smoop
[13:58:49] <skunkworks_> ordered one from tiger direct
[13:59:42] <skunkworks_> suprisingly newegg doesn't have small ones.
[13:59:50] <skunkworks_> That I could find
[15:28:19] <fenn_> fenn_ is now known as fenn
[15:51:30] <alSMT_> just playing around with halui_pyvcp using emc2.2.8 went and got the xml hal files from the CVS and I get emc/task/emctask.cc 312: interp_error: Unable to open file
[15:51:30] <alSMT_> Unable to open file
[15:51:30] <alSMT_> emc/task/emctaskmain.cc 2629: can't initialize interpreter
[15:51:30] <alSMT_> emc/usr_intf/emcsh.cc 3813: can't connect to emc
[15:51:30] <alSMT_> Shutting down and cleaning up EMC2...
[15:52:24] <alSMT_> do I need to wait for 3.2
[15:53:52] <SWPadnos> I'm not sure what you did to get that error
[15:53:54] <alex_joni> alSMT_: if you want to use halui_pyvcp with 2.2.8 there's a minor change you need to make
[15:54:01] <alex_joni> SWPadnos here will tell you what to do ;)
[15:54:12] <SWPadnos> I'm also unsure why you had to go to CVS to get an XML file
[15:54:13] <alex_joni> it's the missing SHMEM KEY
[15:54:16] <SWPadnos> thanks a lot Alex :)
[15:54:34] <alex_joni> SWPadnos: gotta run, let me grab a reverse diff
[15:54:38] <SWPadnos> ok, SHMEM_KEYs have nothing to do with XML :)
[15:54:42] <SWPadnos> ok, do as you like :)
[15:54:48] <alex_joni> no, but they do with the whole config
[15:54:53] <alex_joni> which didn't exist for 2.2.8
[15:54:58] <SWPadnos> oh, and hal, which is right after xml :)
[15:55:43] <alex_joni> http://cvs.linuxcnc.org/cvs/emc2/configs/motenc/motenc.ini.diff?r1=1.14;r2=1.15;f=h
[15:55:47] <alex_joni> http://cvs.linuxcnc.org/cvs/emc2/configs/common/core_servo.hal.diff?r1=1.11;r2=1.12;f=h
[15:56:06] <SWPadnos> ok
[15:56:11] <alex_joni> alSMT_: basicly add the things that have gotten removed on the left
[15:56:11] <alSMT_> ok
[15:56:17] <SWPadnos> was the pyvcp demo not included in 2.2.8?
[15:56:21] <alex_joni> no
[15:56:25] <alex_joni> only halvcp
[15:56:28] <SWPadnos> ah, ok
[15:56:35] <alex_joni> bbl for real now ;)
[15:56:41] <alSMT_> thanks
[15:56:42] <SWPadnos> alSMT_, even better would be to install 2.3beta2 :)
[15:57:01] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UpdatingTo2.3
[15:57:40] <alSMT_> I am using vcp and that gets removed so I'm looking at the options
[15:57:46] <SWPadnos> ah, right
[16:26:48] <alex_joni> alSMT_: let us know what options you are missing
[16:26:52] <alex_joni> I doubt there are nay..
[16:26:53] <alex_joni> any
[16:27:58] <alSMT_> core_stepper.hal:7: exit value: 1
[16:27:59] <alSMT_> core_stepper.hal:7: insmod failed, returned -1
[16:27:59] <alSMT_> Shutting down and cleaning up EMC2...
[16:28:30] <alex_joni> check dmesg
[16:30:16] <alSMT_> i,m using core_stepper.hal
[16:30:36] <alSMT_> emc/task/emctask.cc 312: interp_error: Unable to open file
[16:30:36] <alSMT_> Unable to open file
[16:30:36] <alSMT_> emc/task/emctaskmain.cc 2629: can't initialize interpreter
[16:30:36] <alSMT_> emc/usr_intf/emcsh.cc 3813: can't connect to emc
[16:41:45] <alex_joni> did you modify core_stepper.hal ?
[16:42:06] <alSMT_> yes
[16:42:15] <alex_joni> the Unable to open file sounds like you're missing a .var file
[16:42:18] <alex_joni> or a .tbl file
[16:42:30] <alSMT_> looking
[16:42:32] <alex_joni> can you put the config somewhere online?
[16:42:40] <alex_joni> (all files in that folder)
[16:44:58] <alSMT_> it w3as the var file thanks
[16:45:09] <alSMT_> *was
[16:59:44] <maddash> what's a reasonable update rate for a keypad?
[16:59:56] <maddash> should I scan for input every 100ms?
[16:59:59] <maddash> 100 us?
[17:00:35] <maddash> and the keypad is designed for normal folk, not the ones afflicted with parkinson's
[17:01:54] <archivist> debounce is the mantra for kb scanning
[17:02:21] <archivist> and that depends on key technology
[17:03:10] <alex_joni> maddash: start with 50-100ms, then see how it feels
[17:03:21] <alex_joni> it depends on how many keys you have, how they are connected, etc
[17:04:16] <archivist> some of the click type can oscillate on contact
[17:05:04] <archivist> seen some evil scope traces
[17:07:09] <fenn> less than 100ms
[17:07:31] <fenn> i guess 10ms would be ok
[17:08:39] <fenn> here's a fun project that will answer your question: scan at a super high speed and log all the data, then graph keypresses and see what the typical delay between keystrokes is
[17:10:53] <maddash> rofl
[17:13:09] <maddash> how am i supposed to debounce with a microcontroller?
[17:13:53] <archivist> Im abusing my cnc pc, using it as a build slave while emc is running to see if it handles it
[17:14:33] <archivist> catch closure wait, then poll for release
[17:15:17] <SWPadnos> maddash, it depends on why you're debouncing
[17:15:30] <maddash> SWPadnos: shitty switch?
[17:15:31] <SWPadnos> also on the speed you expect keys to be pressed
[17:15:40] <SWPadnos> ok, that's different from "noisy wire"
[17:16:29] <SWPadnos> if you expect not much induced noise - ie no false trips - but you may have noisy contacts, then you can respond immediately to a switch closure
[17:16:33] <archivist> seen switches that make a noisy wire look quiet
[17:16:51] <SWPadnos> or you can do something simple like AND the previous reading and the current reading and using that
[17:16:56] <archivist> hence the wait for release
[17:17:04] <SWPadnos> (that gives a 2-sample requirement0
[17:17:17] <SWPadnos> right, then you have a timer on the release
[17:17:25] <archivist> bounce can be 16ish khz
[17:17:27] <SWPadnos> or a delay before another keypress will be recognized
[17:17:54] <SWPadnos> I usually have a 1ms to 4ms scan loop in microcontroller projects
[17:18:05] <SWPadnos> and I use between 2 and 10 samples
[17:18:46] <SWPadnos> if you don't care about responding absolutely immediately, you can do the following (but it takes a little more time and memory than the simple AND):
[17:18:52] <archivist> those dimple click switches are probably the worst
[17:18:58] <SWPadnos> keep a buffer of N samples
[17:19:19] <SWPadnos> each time a new sample is read, compute the AND of all samples and the OR of all samples
[17:19:29] <SWPadnos> if the two values are the same, you have a stable reading
[17:19:31] <maddash> bah, disregard that, i found my solution
[17:19:37] <maddash> SWPadnos: thanks
[17:19:41] <SWPadnos> sure
[17:19:46] <SWPadnos> a cap often works better anyway :)
[17:19:57] <maddash> SWPadnos: hey how'd you know?
[17:20:13] <maddash> actually, i'm trying to avoid a software-based method because I'm running on a microcontroller
[17:20:16] <SWPadnos> well, I do this kind of thing from time to time
[17:20:33] <SWPadnos> an AVR can do the simpler calculations in about 8 clock cycles
[17:20:34] <maddash> resistor + capacitor (in series) + schmitt trigger
[17:20:42] <SWPadnos> that's assuming you need to access SRAM
[17:21:04] <Valen> The way i have done it in the past is to trigger on the "release" of the button
[17:21:20] <SWPadnos> that works but can be counterintuitive for the user
[17:21:25] <Valen> or to trigger on the instant, but then lockout for a set period
[17:21:30] <SWPadnos> "I pressed the button and nothing happened"
[17:21:48] <Valen> yeah, just gotta edumacate people
[17:21:57] <SWPadnos> it also prevents repeat or operations that depend on hold time
[17:21:58] <Valen> in the application it works better
[17:21:59] <Valen> a timer
[17:22:01] <SWPadnos> impossible!
[17:22:29] <Valen> i meant its a timer application
[17:22:35] <Valen> it seems to work ok
[17:22:41] <maddash> oh crrap, repeats
[17:22:44] <SWPadnos> yeah, it depends on the application
[17:22:53] <maddash> my arrow just flew off my lcd
[17:23:10] <archivist> hehe delay!
[17:23:15] <SWPadnos> in some applications, you want repeat for some keys (or in some modes), but not for others
[17:23:21] <Valen> Like i said, easy way is to trigger on the press then just let it sit for a while
[17:23:37] <Valen> in my timer app i had an event loop that ran at about 4Hz or so
[17:23:51] <SWPadnos> slooooooooow :)
[17:24:24] <Valen> If there was a change in state on the pin it started acting and flagged the event loop for 2 cycles to ignore it untill it cleared
[17:24:50] <Valen> actually i think you are right, that was the LED update loop
[17:25:03] <Valen> It actually ran at 200Hz or so
[17:27:45] <alex_joni> http://divajutta.com/doctormo/ubunchu/ubunchu-episode-01.pdf
[17:27:54] <maddash> 'UBUNCHU'?
[17:34:04] <Guest163> hello, Im new to the whole emc2 software and I'm tring to do some tutorals found in the integrators hand book, I try to use the "halscope" and get a "halscope gtk warning **:locale not supported by C library. Using the fallback 'c' library"....any suggestions on how to fix this issue??
[17:34:25] <cradek> what is your locale?
[17:35:05] <Guest163> don't know, im a newbi to this whole system
[17:35:10] <SWPadnos> it looks like you're in Miami. do you have your PC set up for a non-english language?
[17:35:14] <alex_joni> what language do you use?
[17:35:21] <Guest163> english
[17:35:38] <alex_joni> where do you see that warning?
[17:36:08] <Guest163> when i try to run hal scope in the terminal window
[17:36:27] <cradek> in your shell type `echo $LANG'
[17:36:36] <Guest163> kk one sec
[17:37:22] <Guest163> en_US.UFT-8
[17:38:00] <alex_joni> try echo $LANGUAGE too
[17:38:22] <Guest163> en
[17:38:35] <alex_joni> that seems ok..
[17:38:48] <alex_joni> is there also an error? or just that warning?
[17:39:02] <Guest163> just the error
[17:39:08] <cradek> en_us.UFT-8 is invalid
[17:39:11] <cradek> are you sure that's what it says?
[17:39:25] <cradek> err en_US.UFT-8 I mean (what you typed)
[17:39:27] <Guest163> one sec ill try again
[17:39:54] <cradek> that would both explain the error and be extremely weird
[17:40:24] <Guest163> en_US.UTF-8 sorry mistyped
[17:41:22] <cradek> ok, that's a very normal locale... I don't understand your error.
[17:41:55] <Guest163> maybe a bad install of emc2
[17:42:13] <SWPadnos> how did you install?
[17:42:28] <SWPadnos> (emc2 liveCD, Ubuntu liveCD then EMC2 install, other method ...)
[17:42:50] <Guest163> Ubuntu liveCD
[17:43:03] <cradek> cradek has kicked IRSeekBot3 from #emc
[17:43:15] <alex_joni> be gone foul foe
[17:43:28] <cradek> yay, I remembered how to do that
[17:43:36] <archivist> hehe it just had a kickban in #mysql
[17:43:37] <alex_joni> Guest163: is there another error message?
[17:43:55] <alex_joni> does halscope come up?
[17:44:10] <Guest163> no, just that one about the bad locale C
[17:44:58] <Guest163> yes it halscope comes up but I cant connect it to anyhing i set up
[17:47:47] <Guest163> well ill try to re install eveything and start from scrach
[17:47:56] <alex_joni> hang on a bit
[17:47:59] <Guest163> kk
[17:48:07] <alex_joni> maybe there's something usefull/needed from you :)
[17:48:17] <alex_joni> (if you're not in a hurry ..)
[17:48:51] <Guest163> no, just sitting here at work, so im getting paid to chill
[17:51:04] <Guest163> im rebuilding an old cnc and hoping to retrofit it with a pc with emc2 to it
[17:52:52] <alex_joni> Guest163: what you could try is doing it from the LiveCD
[17:53:18] <alex_joni> that way you don't bork the install (which might be ok), and you prove that it works
[17:54:46] <Guest163> ok
[17:55:52] <Guest163> well if i have anymore issues ill be back as <jojo> insted of <Guest###>
[17:56:02] <alex_joni> you're most welcome
[17:56:09] <Guest163> thanks everyone
[17:58:54] <maddash> ubunchu is the japanese word for 'Debian is too hard for me'
[17:59:16] <alex_joni> swell word
[18:01:08] <maddash> is it just me or is that pdf full of words/thought bubbles randomly scattered on each page?
[18:01:42] <maddash> are there supposed to be pictures to tie down the dialogue bubbles?
[18:05:07] <alex_joni> maddash: it's just you
[18:07:31] <skunkworks_> http://www.youtube.com/watch?v=MV6ogckf3n8
[18:08:28] <ds3> n
[18:29:21] <maddash> :(
[18:31:25] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[18:33:50] <L84Supper_> L84Supper_ is now known as L84Supper
[19:39:29] <alex_joni> haha, great engineering: http://failblog.org/2009/04/09/camaro-fail/
[19:42:47] <SWPadnos> huh. now why would you put huge back tired on a front wheel drive car?
[19:42:51] <SWPadnos> tires
[19:43:09] <SWPadnos> (I'm assuming it's FWD due to the failure mode)
[19:43:29] <alex_joni> I don't think so
[19:43:37] <alex_joni> I'd say it's RWD
[19:43:50] <alex_joni> but too much torque on the wheels
[19:43:55] <SWPadnos> I would expect RWD, but then the car should have flipped, rather than leaving the back tires behind
[19:43:57] <SWPadnos> I suppose
[19:44:17] <alex_joni> if the fixing would have been ok, the nose should have come up
[19:44:30] <alex_joni> like in a regular dragster
[19:47:51] <archivist_attic> there was an anti rotation bar to the axle that failed :)
[20:02:56] <skunkworks_> camaros are all rwd
[20:05:26] <SWPadnos> I thought that, but didn't have a mental picture of how that failure could occur
[20:06:02] <cradek> I think the oversize tires hit the body of the car hard enough cause the axle to rip away
[20:06:08] <cradek> to cause
[20:23:16] <alSMT_> is this a problem? Starting EMC2...
[20:23:16] <alSMT_> ULAPI: WARNING: module 'HAL_classicladder' failed to delete shmem 03
[21:22:00] <alex_joni> good night all
[21:22:04] <pjm__> gn
[22:22:55] <BigJohnT> Sweet, I got a freq in
[22:29:41] <BigJohnT> if I'm sending a pulse into encoder.0.phase-A should the velocity be 0?
[22:30:14] <BigJohnT> or do I need the encoder to be updated on the base thread?
[22:31:59] <BigJohnT> looks like the pulses are on for about 600uSec on the hal scope
[22:34:51] <BigJohnT> or do I need to have both functions addf to the servo-thread
[22:36:18] <BigJohnT> well the last guess was correct :)
[22:38:08] <Optic> Hihi
[22:55:29] <BigJohnT> should the single pulse go into A or B and should X4 be off?
[22:58:46] <BJT-Hardy> http://imagebin.ca/view/C9gpRvo.html
[22:58:54] <wildrice> Been looking at the HAL Handbook and specifically, the NET command. I think I understand it, but are the following NET commands equivalent?
[22:58:59] <wildrice> net signalname component.source-pin => component.reader-pin
[22:58:59] <wildrice> net signalname component.reader-pin <= component.source-pin
[22:59:17] <BigJohnT> yep
[23:00:08] <BigJohnT> which manual are you reading?
[23:00:27] <wildrice> So since you can only have one source, it would be best to always make it first?
[23:00:45] <BigJohnT> what ever is easier for you to read
[23:01:38] <wildrice> The Feb 2009 Hand book, and the March 2009 Integrators Manual.
[23:02:59] <wildrice> The Hand book did not say (that I could find) that there could only be one source pin, but the IM did. Makes sense, but I searched the Handbook long an hard before I looked in the IM.
[23:03:58] <BigJohnT> do you mean the users manual? and the integrators manual
[23:04:14] <BigJohnT> if it says handbook I think it is old
[23:04:51] <wildrice> Is net signalname component.reader1-pin <= component.source-pin signalname component.reader2-pin valid, but weird?
[23:05:36] <wildrice> The first page says:
[23:05:38] <wildrice> The HAL (Hardware Abstraction Layer) Handbook
[23:05:38] <wildrice> February 4, 2009
[23:05:46] <BigJohnT> ok
[23:09:48] <BigJohnT> is this the section you were reading in the integrators manual http://www.linuxcnc.org/docview/devel/html//hal_basic_hal.html
[23:12:33] <wildrice> Yes 1.1.4 on your web page, but I was looking at 9.1.4 net in the PDF version. The PDF version is easier to search.
[23:13:36] <wildrice> But the information looks to be the same, even though the section numbers do not match.
[23:14:04] <BigJohnT> yea, they get hosed up when I cut them up for the html's
[23:15:17] <BigJohnT> did you find the 9.1.4 section easy to understand?
[23:17:31] <wildrice> Yes it was good. I had seen it before but could not remember it, so I was searching the Handbook to find it again. Then I realized it must have been in the IM. All the different manual's content sometimes run together when things are new.
[23:19:24] <BigJohnT> Thanks, I've spent a lot of time trying to make each one unique but there is some overlap
[23:19:26] <wildrice> I was adding some comments to a new component that I cloned from jepler's Arduino to help me understand it.
[23:19:37] <BigJohnT> cool
[23:20:18] <wildrice> So far, I think that they are pretty good! There is just a lot of information to absorb.
[23:20:36] <BigJohnT> that's for sure
[23:22:34] <BigJohnT> now if I could just figure out why I'm not getting what I expeced I'll understand the encoder component more than I do now :)
[23:22:36] <wildrice> It would be nice if you could package all the PDF books up in a zip file so that I can download and refresh my doc versions. I sometimes wonder if I have the latest.
[23:23:32] <BigJohnT> I'm constantly updating them so the latest ones are on the web site