#robotics Logs
Sep 15 2012
#robotics Calendar
00:07 rue_shop3 I was briefly worried about it, but the manual says that strnlen is C99
00:23 rue_house TELunus,
00:23 rue_house char * strndup (const char *s, size_t n) {
00:23 rue_house size_t len ; char *out ;
00:23 rue_house len = strlen (s, n); if (n < len) len = n);
00:23 rue_house if((out = malloc (len + 1)) != NULL){ strncpy (out, s, len); out[len] = '\0'; }
00:23 rue_house return out;
00:23 rue_house }
00:25 rue_house can windows compile THAT!?
00:27 TELunus Well, you mismached your parenthases on your third line, but correcting for that...
00:28 rue_house len = strlen (s, n); if ((n < len) len = n);
00:28 rue_house :) sorry
00:28 rue_house er
00:28 rue_house haha
00:28 rue_house len = strlen (s, n); if(n < len) len = n;
00:28 TELunus ya.
00:28 rue_house 8th time is the charm, really
00:30 TELunus I bet you it's going to complain about strncpy. It doesn't seem to like strn.
00:31 rue_house if so, we use memcopy
00:31 rue_house which you HAVE to supply how many bytes to transfer
00:32 TELunus it's telling me I gave strlen too many args. Drop the n?
00:32 rue_house er
00:32 rue_house yes
00:33 rue_house http://www.ebay.ca/itm/370527794989?ssPageName=STRK:MEWNX:IT&_trksid=p3984.m1497.l2649
00:33 rue_house dont buy those
00:33 rue_house the little pcb isn't mounted properly in them
00:33 TELunus compiled. 3 warnings, no errors.
00:33 rue_house so the cards dont lign up
00:33 rue_house warnings are ok
00:33 rue_house see if it runs
00:34 TELunus runs fine.
00:34 rue_house cool
00:34 TELunus output is:
00:34 rue_house so did you see how easy it is to set up variables and apply an equation?
00:34 rue_house in main.c
00:35 rue_house I have not seen a library yet that a) can read/write variables like that, b) dosn't require waaay more code to pre-parse the equation and stuff
00:35 TELunus 0: (RW) d = 5.000000
00:35 rue_house read/writable variable, index 0 named d value 5
00:36 TELunus (d==6)|(d==7) -->> 0
00:36 TELunus and then the first line again.
00:36 rue_house the result of the equation (d==6)|(d==7) was 0
00:37 rue_house and the value of d was left at 5 when it was done
00:37 rue_house throw some new equations at it, tell me how it breaks
00:37 TELunus Maybe later, I think I should head to bed.
00:38 rue_house ok
00:38 rue_house its half the code you need to write a computer language
00:42 rue_house my computer case is obsolete
00:42 katsmeow-afk Plains to Buy BP Oil and Gas Assets for $5.55 Billion; Unveils $560 Million Deal with Shell
00:42 rue_house how I know this:
00:42 rue_house my case is white, all my drives are black
00:43 rue_house chumpchange
00:45 katsmeow-afk Today’s transaction brings BP’s total asset sales since the beginning of 2010 to more than $32 billion
00:49 rue_house !time
00:49 tobbor My watch says its 10:37PM Fri Sep 14 2012
00:49 rue_house and I dont think I can stay awake
00:49 rue_house I finished fixing and modifying the flash drive
00:49 rue_house the pcb stays where it should and the cord is as long as I need it
00:50 rue_house and now I ahve a front panel usb jack
02:12 rue_house ok I want to recap an amp too
02:13 rue_house sleeep would be nice too
02:24 Kiiwinki rue_bed: Recap as in capacitor replacement?
02:28 rue_bed yes
02:29 rue_bed I have an amplifer which is apparently antique
02:29 rue_bed it uses a single amp chip,
02:29 rue_bed but the amp chip is ok
02:29 rue_bed the problem is just the capacitors
02:29 rue_bed its a nice looking amp
02:29 Kiiwinki Which chip?
02:29 rue_bed so I'll fix it
02:29 rue_bed dont recall
02:30 rue_bed ask me in 8 hours
02:31 Kiiwinki I've been meaning to build a tube preamp for my bass amp. Nice little thing; even has an indepedently-boarded power amp
02:32 Kiiwinki Unless, maybe, something really trivial is wrong with it
02:33 Kiiwinki Switched some Sennheiser headphone elements to a pair of torn-speakered Pioneer SE-305 headphones I've had for years
02:33 Kiiwinki Pretty pleased
03:30 theBear speaking of which, saw a commercial product yesterday that runs a 12ax7a off +/-20vdc !!!! the foolish fools
03:30 theBear and now, to the bottleshop
04:05 Kiiwinki Har
04:05 Kiiwinki Hm
04:05 theBear wow ! a real live tellie tubbie !
04:05 Kiiwinki theBear: Is there some kind of a back-to-back 12AX7 arrangement?
04:05 theBear errr, how you mean back to back ?
04:06 Kiiwinki How'd they arrange the positive/negative power supply?
04:06 theBear didn't look too hard, was working, but i suspect they had the filiament/cathode at -20v
04:06 Kiiwinki Or, rather, how were the tubes arranged for the double-sided voltage
04:07 Kiiwinki Man - what's the word for that?
04:07 theBear oh, it wasn't anything like you're thinking... opamps do it by basically being a full push-pull (a/b) amplifier setup between the rails
04:07 Kiiwinki for +/- voltage arrangemet
04:07 theBear split/dual rail maybe
04:07 Kiiwinki The latter seems more logical
04:08 Kiiwinki Hm
04:08 Kiiwinki I just remembered that I have some subminiature pentodezs
04:08 Kiiwinki pentodes
04:08 Kiiwinki Might do something with those
04:08 theBear mmm, they probably WOULD be ok with 40v across them, just barely
04:09 Kiiwinki I probably have voltage/current charts for these somewhere
04:15 Kiiwinki Man
04:15 Kiiwinki Realised something the other day
04:16 Kiiwinki I should've started archiving electronics-related material off the internet a long time ago
04:16 Kiiwinki For example, there's this excellent page on HV probe design
04:17 theBear mmm maybe, as the years go on i find it's quicker to find most stuff with a quick search than even in my own well sorted library
04:17 Kiiwinki hosted on geocities or somesuch. I dare say I was rather boggled by the realisation that I had not even considered them as volatile information before
04:18 sherlock there is a giant geocities archive
04:18 sherlock with like 80% of geocities
04:18 sherlock eta
04:18 theBear yeah, but there's nothing usefull on geocities and never was :)
04:18 sherlock http://archive.org/web/geocities.php
04:18 Kiiwinki Thanks, sherlock. This, however, does not solve the problem itself
04:18 theBear hehe
04:19 Kiiwinki You are incorrect, theBear. I have found much useful information on geocities
04:19 Kiiwinki This is the case because usefulness is subjective
04:19 theBear hehe, glad you qualified that :)
04:19 Kiiwinki Unless, perhaps, you are aware of some objective definition for usefulness
04:20 Kiiwinki Btw
04:20 Kiiwinki Any ideas on what I should do with these 40uF/2500V caps?
04:20 theBear ummm, ummmmm
04:20 Kiiwinki Hefty bastards - I think 1.5kg or so in weight
04:20 theBear if you got a LOT of them you could make one of those sparkey cockroft-walton like things
04:21 sherlock use them to build a stargate
04:21 Kiiwinki theBear: Four of them I have
04:22 theBear that's not a lot... maybe use them for an intermediate stage in a 2 stage traditional (non-ss) tesla coil ?
04:22 theBear you know, to bounce off the coil and the sparkgap
04:23 theBear oh wait, they're polarised tho aren't they ?
04:23 Kiiwinki Yup
04:23 Kiiwinki Paper-oil caps
04:23 theBear mmm... i dunno, but i do need a nap :) bbl
04:23 Kiiwinki http://www.ebay.com/itm/40UF-2-5KV-DC-BOSCH-OIL-PAPER-CAPACITOR-/200454374490
04:24 Kiiwinki 1.8kg
11:22 rue_house it seems I nolonger know how to convert 40uF * 5000V to joules
11:22 rue_house or 12V * 40Ah
11:26 m_itx wonders what to bring to class t
11:27 theBear banana
11:27 katsmeow-afk a 40uf cap charged to 5kv
11:27 katsmeow-afk and toilet paper
11:30 theBear hehe
11:32 Tom_itx fresh change of tidy whities
11:38 rue_house timy26 programmed to act as dual stepper sequencers
11:38 rue_house clock in on an interrupt line, dir on a generic io
11:38 theBear c'mon, class can't be that scary
11:38 rue_house 4 phase outputs, half stepping sequence
11:39 theBear rue_house, i can string together random technical words too
11:39 rue_house nothing is random
11:39 rue_house some things are just a mix of a lot of simple stuff
11:39 theBear and some things are random :)
11:41 rue_house this morning is right on my heels
11:41 katsmeow-afk how about a way to make that olde 120vac line communication scheme work with synthetic pwm'd "sinewaves"?
11:41 rue_house time to go clean up the shop
11:41 tsmeow-afk ponders cleaning it down ins
11:41 rue_house katsmeow-afk, the new stuff uses the low voltage on the sine wave just before and after zero crossing
11:42 katsmeow-afk and the zero-crossing on a pwm signal is where?
11:42 rue_house may I suggest a triagle wave is friendlier on transformers?
11:42 katsmeow-afk yes, you may
11:42 theBear you may not !
11:42 katsmeow-afk we have a tie!
11:42 theBear you like thai ? i like shirt
11:43 katsmeow-afk nice shirt!
11:43 theBear thanks
11:44 katsmeow-afk i did not know today was going to be Saturday
11:44 rue_house start laundry, robotics, lunch, do bank deposits, go to landfill for wood and golf clubs...
11:45 katsmeow-afk i was going to do Thursday stuff
11:45 rue_house I have ~15 mins
11:45 rue_house to brush my teeth and get dressed
11:45 tsmeow-afk wants to see rue make it to the landfill and back in 15 min
11:45 theBear i never could quite get the hang of thursdays
11:46 katsmeow-afk that's why i planned for today being thursday, i don't plan for Saturdays
11:49 tsmeow-afk tosses the plans out the window and wonders what to do now ,,,, the window was closed, the plans are now a mess all over the window and f
11:50 theBear hehe, time to get a dog
11:51 katsmeow-afk :-/
11:52 rue_house !time
11:52 tobbor My watch says its 09:40AM Sat Sep 15 2012
11:55 tsmeow-afk ponders singing "songs", instead of singing the actual songs, placing all the words in Michael Jackson's "Beat It" with "so
11:55 theBear i like sounds like noo nee noo personally
11:56 katsmeow-afk i don't know that one
11:56 theBear you can't know nonsense sounds
11:56 katsmeow-afk i know a few!
11:57 theBear but too few to mention ?
11:57 katsmeow-afk like doe ramey falso latte doe!
11:57 theBear hehe
11:57 katsmeow-afk doh
12:01 tsmeow-afk spent a lot of last nite wondering when a cow wasn't a cow any
12:01 theBear when it jumps over the moon of cours
12:01 roboman2444 if i cut my brushless motor wires about 1 inch shorter, will that screw it up?
12:01 katsmeow-afk it orbits forever?
12:02 roboman2444 i have this large bl motor, and it coggs and overall doesnt feel like its at its full potential
12:02 katsmeow-afk are the wires threaded?
12:02 roboman2444 threaded?
12:02 roboman2444 yea, i think
12:02 roboman2444 they are an extension of the windings
12:02 theBear err, cogging and wires are pretty unrelated
12:03 katsmeow-afk you need special tools to cut threaded wires to eliminate the cogs
12:03 theBear what am i missing here ?
12:03 katsmeow-afk insanity
12:03 katsmeow-afk cogging has ZERO to do with an inch of wire
12:04 roboman2444 hm
12:04 theBear ahh, good
12:04 roboman2444 so what is my issue? speec control timing?
12:04 katsmeow-afk afaik, all "bl" motors cog
12:04 theBear also, if these wires are not windings, wtf are you talking about ?
12:04 theBear roboman2444, you know cogging happens when the motor is unpowered right ?
12:04 katsmeow-afk unless they are ac motors with no pm in them
12:05 roboman2444 no i mean if i accelrate to fast, the motor kinda squeals and wont accelerate
12:05 theBear and "doesn't feel like its at its full potential", aside from missing some punctuation, isn't enough to answer
12:05 katsmeow-afk then accelerate only to medium fast
12:06 theBear what you just described sounds like you losing sync with the motor... that could be somewhere between imperfect drive and not enough power to accelerate that fast (either input power, or power->rotation efficiency kinda power)
12:06 roboman2444 hm
12:07 katsmeow-afk yeas, if the load is too much, the stepper will just sit there and osscilate
12:07 roboman2444 odd
12:08 katsmeow-afk it's almost knewn
12:08 roboman2444 what about rc car bl motors?
12:08 theBear what about them ?
12:08 roboman2444 they can accelerate fast enough to do backflips from a standstill
12:08 katsmeow-afk nice
12:09 katsmeow-afk glad that's settled then
12:09 theBear back to my previous answer inverted, they either have more supply power/efficiency of power-> motion conversion, or are driven better
12:09 roboman2444 so controller timing?
12:09 theBear also they use feedback, so either way, if they lose sync they get it back within probably 2 phases worth of drive
12:09 theBear zero crossing/backemf style feedback specifically, at least on a 3 wire one
12:09 katsmeow-afk we knew it was controller "timing", but see if it's the feq or the pulse width or the duty cycle or not enough current
12:10 roboman2444 well it is a cheap controller
12:10 boman2444 is building an electric sco
12:10 katsmeow-afk possibly a ID1-0T issue
12:10 roboman2444 hm?
12:11 katsmeow-afk bbbl
12:11 theBear ok then, so does it happen more/less with/without a load ?
12:11 roboman2444 much more
12:11 theBear that's not an answer !
12:11 roboman2444 when im riding it, i have to carefully accelerate
12:11 roboman2444 or it gets out of sync and i have to throttle down
12:11 theBear orright, your driver is stupid and shit, that much is for sure
12:11 anannie theBear: How do you measure the back-emf?
12:11 theBear and by stupid, i mean the opposite of intelligent
12:12 theBear and by shit, i mean 'of poor quality/design'
12:12 theBear anannie, general approach is something like an opamp connected to the winding
12:12 theBear in some schemes you might do the same across a current sense resistor
12:12 anannie theBear: What if the spike is too variable and too great?
12:13 theBear anannie, err, scale it with resistors ?
12:13 theBear tho from what i gather in traditional schemes you only care about levels around zero crossing, so you just need brutal clamping and a series resistor
12:14 anannie theBear: I have a thing where a human being pushes down against the current, producing a large backemf in the system... The problem is that I want to accurately measure how much did the person push down, the distance X travelled and the force involved in real time.
12:16 theBear anannie, erm, i think that's forward emf, it's too late to workout the maths and if they are simple or mid-complex, but i'd say measure the current change vs the drive voltage
12:17 anannie theBear: Thank you! I have been stuck on this for a while now and I didn't know what the answer was. How can I learn to model this and build this myself?
12:17 theBear umm, lookup shunt resistor, and err, workout how to do the maths with either/and opamp/micro i guess
12:17 izua hello
12:18 izua is there a standard language for describing pulse generating/ramping for CNC control?
12:18 izua say I use a computer to calculate toolpaths and G-Code, and from that G-Code, I'll run through my acceleration profile and corner tolerances
12:18 izua the computer can't actually generate the pulses, but it can describe them
12:19 theBear izua, not that i know of, but generally in cases like that, tool speed is more desirable than complex pulse control
12:19 izua for example, 10000 pulses, starting at 100Hz, ramp up to 500Hz by incrementing the speed by 5Hz every 10 pulses (a linear and rather unrealistic example I guess, but still)
12:19 izua theBear: I'm looking if there is a standard
12:20 izua for example, if I want to make a universal mach3/emc^2 usb interface, do i need to create my own protocol, or is there some standard
12:20 theBear i'd be VERY surprised if there was
12:20 theBear it just doesn't make sense in cnc-world
12:20 izua it doesn't?
12:20 izua asuming cheap hobby machines, not expensive ones with servo control
12:22 theBear any ones, why would you need to do more than choose tool speed in X number of axis ?
12:24 izua oh.
12:24 izua i'm not thinking of tool speed
12:25 izua i'm thinking of - say, on a single axis - a standard method of describing the acceleration ramp
12:26 izua so a computer could generate the toolpaths and acceleration/decceleration profiles, and simply send them to an fpga/controller, instead of implementing a nck and bitbanging on the parallel port
12:26 theBear by tool speed in X axis, i meant in any axis(s) ... but why would you want to describe the acceleration ramp ? that should be handled by the machine itself, which knows itself infinitely better than the code fed to it
12:26 izua hmm, I suppose one could go the upper layer and dump G Code directly in the machine
12:29 izua i was thinking this would be the middle ground because a) modern computers don't really have a suitable I/O interface b) old computers with a parallel port/PCI paraport cards aren't that powerful and c) one could do a lot of experimenting and tweaking by sending commands do an intermediate layer in the numerical kernel, instead of having the whole kernel in a box
12:32 theBear usb is a suitable interface for simple text commands
12:36 izua yep. like ones describing how to generate a train of pulses
12:38 izua one more question
12:39 theBear :)
12:40 izua what would be the effect on a stepper motor if the step signals sent to the driver are generated in a bresenham line algorithm fashion? they have a constant pulses/second rate, but the time spent between succesive pulses varies
12:41 izua i've ran a simulation and some fourier analysis, then flashed a chip and sent the actual pulses to a driver+motor. sure enough, the motor was buzzing on several frequencies
12:41 theBear oh, like funny-pwm (==bresenham line algo?) i would think umm, it wouldn't be good
12:41 izua took out a mike, recorded it a bit, seems the first three harmonics as in the simulated fourier analysis were there
12:41 izua funny-pwm? heh
12:42 theBear you mean effective harmonics, or 3 harmonics from the simulation (un)excited the motor as expected ?
12:42 izua yeah, i guess it is. the speed varies by minute amounts between pulses, but it's constant over ~100 or more
12:43 theBear oh, is his algorithm more like 'random' pwm ? i was thinking more pcm vs pwm
12:44 izua the first three harmonics from the sim
12:44 izua one moment, i'll upload some pics
12:44 izua http://imgur.com/a/tTWho
12:44 izua first pic shows the pulses
12:45 izua for 1337Hz (yay for magic numbers), it does something like in the first picture. counting the pulses over 1 second, you'd get 1337, and every 124 seconds or so, 1336
12:45 izua which turns out to be correct in terms of the math
12:46 izua next, i checked what extra harmonics would the "funny pwm" induce, 1.86k, 4.53k, 5.04k
12:46 theBear mmm... any reason for attempting this outside of curiosity ?
12:47 izua put the program in an actual chip, generated the signal, and put a mike near the motor, then ran a fft on the recorded signal; sure enough, the harmonics were there
12:47 izua oh
12:47 izua i'm just curious if one can do a decent 3-4-5 axis pulse generator (and maybe the interpreter) without funky 16 bits counter/16 bits prescaler pwm generators
12:48 theBear what about single funky 16 bits counter and 'soft' 'comparators' ?
12:48 izua well
12:49 izua if you do software pwm, driven by a hardware timer, you'd need a lot of speed to get a decent resolution, namely the product of your maximum speed and desired resolution
12:50 izua i did that some while ago with a 90pwm3b, and earlier, with a stm32/f103. but both have 16 bit high speed counters :P
12:51 theBear i dunno how much other processing you doing, but it's pretty achievable inside a regular speeded avr with spare time to think
12:52 izua well.
12:52 izua here are what i stand by
12:53 izua if you have one timer, generating an ISR at 3200KHz, so you can get at the very least, 1600KHz out, you'd have a compare value of 1
12:53 izua load that in a counter, decrement every ISR, if it hits 0, trigger the pin high, otherwise, make it low
12:54 izua now, this is the resolution problem.
12:54 izua at 3200Hz (whoops, i mean hz, not khz)
12:54 izua you'd be getting a 1600 Hz signal
12:54 izua if your compare value is 2, you'd be getting a 1066Hz out
12:57 izua but if you increase the speed of your ISR, say to 160000, you'd need your software compare value set to 100 to generate 1600Hz out. the following value, 101, will generate about 1584hz out - better speed "resolution" at the cost of only 100 cycles between ISRs
12:58 theBear what what ? there should be no difference in your frequency
12:59 izua i'm not considering doing a lot of processing, but you know, with context saving and reloading, and the actual signal generation logic, plus the fact that you're working with 16 bits on an 8 bit micro, i succeded in getting this down to around 60 cycles. not really useful
12:59 izua i mean, except if you read data and load it :P
12:59 izua i'm talking about the range of available speeds
13:00 izua (as a "resolution" of sorts)
13:00 izua if you go down from 1600Hz to 1066Hz, you can't really say your top speed is 1600Hz, because your next available speed is really close to 50% of 1600
13:02 theBear what what ? pwm should NEVER change freq
13:03 izua i think the academic term for this issue is "long tail", since your output frequency follows a Fisr/n curve
13:03 izua ah.
13:03 theBear it what ? you doing pwm wrong
13:03 izua sorry if i'm not clear, english isn't my first language
13:03 izua well, it's more like PFM.
13:04 theBear indeed, do pwm instead, it all works out nicely with a single counter and minimal error
13:04 izua but i'm not sure there are better methods of generating this in software, except by comparing your timer, every clock, with some values
13:04 izua well, software pwm is easier.
13:05 theBear what about setting a non-timer-resetting interrupt each time you pass a channel pwm value ?
13:05 izua a neat trick i've found is to sort the values, and have a mask for each change. thus, if your current timer value matches a value, apply the mask. all the comparisons are done "before" the actual generating
13:05 izua loop
13:06 izua that works, but you'll only get a single channel?
13:07 izua ah, i know the right way is to simply have timer/compare+match hardware at a decent speed
13:08 theBear no no, i'm just saying do a sort each cycle and set the channel to interrupt sequentially
13:08 izua ah; now i figure out what you said. i wasn't thinking of changing the PWM speed mid-session; i was just thinking of using a higher speed by default.
13:08 izua esentially, the tradeoff between speed and resolution
13:09 izua yeah, that works great for pwm
13:12 izua this is an exercise in futility i guess.
13:14 izua farnell lists a mega2560 @ 20MHz, with 6 channels @ 17.48GBP pcs1, while a 32f103 @ 72MHz, with 8 channels, all 16 bit and 16 bit prescaler @ 6.76GBP pcs 1.
13:14 theBear i dunno 2560, but i'd guess it has HUGE ram/rom compared to a 32f
13:18 izua 2560: 256K flash, 8k ram, 4k eeprom; f103: flash 128k, ram 20k, no eeprom but ~64 bytes of battery backup registers
13:20 izua nxp had some announcments of a LPC1114 in DIP, can't remember details, but it ran around 50MHz
13:29 theBear erg gnihgt
14:10 rue_house by the time you compare al the processors their all about the same, just pick a brand and stick with it
14:11 rue_house avrs are a good fit for smart io control or control of amall things
14:11 rue_house dont try to put an os on them and use them for image recognition
14:11 rue_house thats what arms are for
14:13 rue_house I switched to atmel cause microchip was screwing me over
14:14 rue_house I use x86 for anything big thinking, and avr for smart io control
14:15 rue_house the pwm from the dac of a cd player changes duty and amplitude
14:15 rue_house its evil
14:19 izua yeah, it's a delta sigma thing
14:20 izua come to think of it, even bresenham acts like a delta-sigma modulator of sorts, since it keeps track of the "rounding" error
14:20 izua heh, i know, but there are so many to choose from
14:21 izua pics are great for simple stuff, and they're incredibly cheap; at the cost of having to work in that horrible assembly for anything remotely useful and lack of decent free compilers
14:23 izua avrs are great, and you can do some 'big thinking' on them too, if you don't go overboard with floats, printfs and such
14:23 izua i did find myself in a similar situation to yours, simply using controllers for some smart io
14:35 joink I've fallen in love with the cortex m#s :P
17:40 Tom_itx rue_shop3, you saw trif's calibration page?
17:51 katsmeow-afk "The lack of linguistic resources adequate is recognized as one of the main obstacles to the success of research and development in the linguistic resources." yeas, and the lack of having Jupiter aliens is an obstacle to research into them, and a lack of me having copious amounts of gold is an obstacle to me spending it.
18:07 rue_shop3 I swear I'm gonna post that javascript as a page, I cant find it ANYWHERE online
18:08 rue_shop3 its stupid to post the jdavascript and not a page with it
18:12 Tom_itx rue_shop3, what are you working on this fine evening?
18:22 rue_shop3 reprap, amongst other things
18:22 rue_shop3 have to clear out some firewood
18:22 rue_shop3 and I want to recap an amp
18:32 katsmeow-afk place pcb into reprap, press "recap" button, enter pcb serial number, press "go" button
18:39 Tom_itx rue_shop3, i think reprap is alot of guessing
18:39 Tom_itx no real organized data seems to be kept
18:43 rue_shop3 yea
18:43 rue_shop3 I dont know if I should get him to understand the sillyness or not
18:43 Tom_itx you'll just get ridiculed
18:43 rue_shop3 ok now I stop
18:43 rue_shop3 hahaha
18:43 rue_shop3 that says it all
18:44 Tom_itx i mean, they run their axis loose on purpose and think it's good
18:45 rue_shop3 I suppose this is what happens when you release this stuff into the masses
18:45 Tom_itx ignorant masses
18:47 Tom_itx the 4pi board looks pretty nice really
18:47 rue_shop3 I was seeing how far I could push his understanding
18:47 Tom_itx it's got the good a4982 drivers on it
18:47 rue_shop3 his misunderstanding
18:48 Tom_itx keep pushing, i'm bored
18:48 Tom_itx http://koti.kapsi.fi/~kliment/photos/4pi/IMG_7421.JPG
18:48 rue_shop3 na, I have things to accomplish
18:48 Tom_itx 4pi
18:49 Tom_itx i got the schematic, i was gonna rob the driver section and make a board layout for it
18:49 Tom_itx i may get a 2560 one of these days
18:51 rue_shop3 I think the tables are interpolated
18:52 rue_shop3 so if I make a table of like 6 values that I measure, I think it'd work
18:52 Tom_itx you'd be ahead to ask about pid in linuxcnc
18:52 rue_shop3 I just need to know how to come up with a new table for the marlin firmware
18:52 Tom_itx but they're kinda locked into the linuxcnc mindset too
18:53 Tom_itx what file is it in?
18:53 Tom_itx these screwballs sure don't know how to draw schematics
18:54 rue_shop3 there is a file called thermistortables.h
18:54 rue_shop3 there is a program provided to regenerate it
18:54 Tom_itx i don't have original marlin
18:54 Tom_itx mine is for the delat
18:54 Tom_itx delta
18:55 rue_shop3 huh, I think I'm on the wrong computer
18:56 Tom_itx there's about 10 tables in it
18:56 Tom_itx none fit yours?
19:22 rue_shop3 damnit, the marlin code wont compile again
19:23 rue_shop3 no, I'm using a thermistor from a photocopier, its way more sensitive than theirrs
19:35 rue_shop3 FFFFFFFFFFFFFFFFFFFFFFFFFFFFF arduino java FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFf
19:35 Tom_itx gotta love it
19:37 rue_shop3 so EVERY? time I want to compiel the marlin software I need to re-unpack it, remodify all the files ?
19:37 Tom_itx no clue
19:37 Tom_itx are you using their environment?
19:38 rue_shop3 0.23
19:38 Tom_itx i've got 1.0.1
19:38 rue_shop3 marlin will not compile under 1.01
19:38 Tom_itx why are you using such an old one?
19:38 rue_shop3 marlin will not compile under 1.01
19:38 Tom_itx umm
19:38 rue_shop3 you HAVE to use 0.23
19:38 Tom_itx i just did it
19:39 rue_shop3 then they modified yours
19:39 Tom_itx gimme the link to the marlin you have
19:40 Tom_itx and i'll try it
19:40 rue_shop3 I'll just do that I did lat tiem
19:40 Tom_itx i'd like to try it once
19:40 Tom_L i think i had it once but got the delta one this time
19:40 rue_shop3 0ErikZalm-Marlin-Marlin_RC3-0-g538189c.tar.gz
19:40 rue_shop3 look for that
19:41 Tom_L i may have it in a dir here
19:41 Tom_L ErikZalm-Marlin-Marlin_RC3-42-g0e58ef6.zip
19:41 Tom_L what's the difference?
19:42 rue_shop3 after its compiled or SOMETHING, SOMETHING does cjurrupt in a round() call and it wont compile anymore
19:42 Tom_L what's the difference?
19:42 Tom_L what's the difference?
19:42 rue_shop3 I dont kljlkmkwo if I'm see if thaty comielse or now
19:43 rue_shop3 no even an new unpack is corrupt or soemthing I'll try redownlaod or soemthing and see if tshtde the envornomane or not
19:43 rue_shop3 java CRAP
19:49 Tom_itx ok i grabbed the zip
19:58 Tom_itx rue_shop3,
19:58 Tom_itx how do i load it again?
20:01 Tom_itx nm
20:01 Tom_itx load configuration.h first
20:03 Tom_L `__vector_default':
20:03 Tom_L (.vectors+0xb0): relocation truncated to fit: R_AVR_13_PCREL against symbol `__vector_44' defined in .text section in c:/avr stuff/arduino/arduino-1.0.1/hardware/tools/avr/bin/../lib/gcc/avr/4.3.2/../../../../avr/lib/avr6/crtm2560.o
20:03 Tom_L i get that error with 1.0.1
20:22 Tom_itx ok rue, i got .23 and it compiles fine
20:22 Tom_itx 47576 bytes
21:23 rue_shop3 I cant compile it
21:23 rue_shop3 gona try antoher machine
21:32 rue_shop2 what I have to go thru EVERY time i want to compile that firmware is really pissing me off
21:33 rue_shop2 !time
21:33 tobbor My watch says its 07:21PM Sat Sep 15 2012
21:40 katsmeow-afk Spanner is a scalable, globally-distributed database designed, built, and deployed at Google. Spanner is designed to scale up to millions of machines across hundreds of datacenters and trillions of database rows.
21:54 katsmeow-afk http://gigaom2.files.wordpress.com/2012/09/screen-shot-2012-09-12-at-8-09-53-am.png
21:55 katsmeow-afk walmart has 65 MEGAwatts of solar installed
21:56 katsmeow-afk Apple is building a 20 MW system in North Carolina for its data center.
22:27 rue_shop2 I'm giving up on the reprap till I can regain my temper with the crap its based on
22:27 rue_shop2 work on fixing this amp
22:35 katsmeow-afk i agree
23:06 ace4016 http://i.imgur.com/WUbbU.jpg
23:08 katsmeow-afk quite literally, smaller bits
23:11 katsmeow-afk "But GM will need to sell 120,000 Volts ........" i didn't know we could sell volts
23:12 rue_house aparently the amp chip is half blown
23:12 rue_house its a 20W chip
23:12 rue_house I cant find a reasonable replacement for cheap
23:12 rue_house but I can find its 30W counterpart
23:12 rue_house so I got that, $6
23:14 katsmeow-afk http://arstechnica.com/tech-policy/2012/09/dutch-court-rules-linking-to-photos-is-copyright-infringement/
23:14 katsmeow-afk $6 ? shame you couldn't get it for cheap
23:14 Tom_itx rue_house, are you compiling under linux or windows?
23:15 katsmeow-afk rue_house, are you compiling ON linux or windows?
23:15 Tom_itx under
23:15 Tom_itx over
23:16 rue_house in
23:16 rue_house linux
23:17 Tom_itx hmm
23:17 rue_house I thought that sticking with the program and using the arduino ide was the right thing to do
23:17 rue_house but this is ****
23:17 Tom_itx i could try that i suppose but i'm not sure i have gcc etc installed
23:17 Tom_itx i installed .23 or whatever it was and it compiled fine in windows
23:19 katsmeow-afk Pi supercomputer: The whole system cost under £2,500 (excluding switches) and has a total of 64 processors and 1Tb of memory (16Gb SD cards for each Raspberry Pi). http://www.southampton.ac.uk/mediacentre/features/raspberry_pi_supercomputer.shtml
23:19 katsmeow-afk http://www.southampton.ac.uk/~sjc/raspberrypi/pi_pictures.htm
23:20 SavageSimian any solidworks gurus here? i'm a noob trying to cut a hole in a nonplaner surface and i'm a little stumped.
23:23 Tom_itx SavageSimian maybe in #seattlerobotics or #linuxcnc however the time isn't the best for it
23:23 katsmeow-afk those are really rare, google hits for nonplaner surface is only 13,100
23:24 Tom_itx i've used it a bit but not enough to help
23:25 SavageSimian omg! google??? is that something on the intranet what help people like find stuff? i never though of that!!!
23:26 SavageSimian Tom_itx: ty. i'll give it a try.
23:26 Tom_itx jt_shop in #linuxcnc could help you for sure however he's asleep
23:26 katsmeow-afk yeas, but only 13,100 hits
23:27 Tom_itx alot to sort thru for one answer
23:27 tsmeow-afk s
23:28 katsmeow-afk he'd prolly find better data if he knw how to spell it
23:28 katsmeow-afk i swear
23:29 Tom_itx i've learned spelling varies in countries or language barriers get in the way of proper english sometimes
23:30 SavageSimian yea. i've looked at about 2 dozen sites. nothing specific to my problem. and, being a noob at solidworks i'm looking for something that can go straight to the heart of the problem without having to go thru half a chapter to get to it... was hoping there was a simple answer somebody with some smarts could answer.
23:30 Tom_itx check there in the morning and you will get one
23:30 SavageSimian anyway, thanks for your help Tom_itx.
23:33 Tom_itx http://www.3dcadforums.com/solidworks-forum/5047-extruded-cut-non-planar-solid.html?s=03f7b0ad447ad2c99612e7c851ccbe9f
23:35 SavageSimian i saw that link. it didn't help. i didn't understand if they were talking about starting a new sketch unrelated to the current part or what. afaik, you can't just make a sketch and link it to existing geometry.
23:36 SavageSimian of course being a noob, i'm probably wrong.
23:36 Tom_itx jt should be able to help
23:36 SavageSimian t. when is a good time to check?
23:37 Tom_itx in the morning CST
23:37 Tom_itx 6am ish
23:37 SavageSimian cool
23:44 GuShH bah i can't find "xml unleashed by michael morrison" anywhere online :( bookstores here dnt carry it
23:56 rue_house unleashed, it ran away
23:57 GuShH lol
23:57 GuShH its not for me