#robotics Logs

Jan 08 2018

06:15 AM rue_bed: tho not quite the same
06:18 AM rue_bed: if (i >= nn) i -= (nn-1); is WAY FASTER than i %= (nn-1);
06:40 AM anonnumberanon: neat
06:42 AM anonnumberanon: so yeah, everyone in the US is about to get their tax refunds, they will buy a lot of crypto, buy now, wait till they do, 3 or 4 weeks, then sell
06:43 AM anonnumberanon: right now also, you should buy some of the cryptos that are currently coming up and are still a dollar or less
06:43 AM anonnumberanon: the ones in the top 20 ranking
06:43 AM anonnumberanon: like ripple, stellar, cardano, and a few others
06:43 AM anonnumberanon: i already said too much
06:43 AM anonnumberanon: but for those, hold them for a year or more
06:54 AM rue_bed: OMG, the origional fortran fn was called four1 , the numerical recipies book disn't change it
06:56 AM rue_bed: omg, and NRIC (numerical recipies in c) got the sign of the transform backwards
06:59 AM rue_bed: omg and there it is
07:00 AM rue_bed: https://www.google.ca/url?sa=t&rct=j&q=&esrc=s&source=web&cd=12&ved=0ahUKEwjslZWrs8jYAhXp7IMKHS_fCaU4ChAWCDUwAQ&url=http%3A%2F%2Fciteseerx.ist.psu.edu%2Fviewdoc%2Fdownload%3Fdoi%3D10.1.1.942.7100%26rep%3Drep1%26type%3Dpdf&usg=AOvVaw1dC-KtlWZ9XZaSJ-uxzgM9
07:18 AM squirrel1 is now known as veverak
08:52 AM rue_: ok I prettymuch been up all night
09:24 AM rue_: sot eh origional code multiplexed the values cause thats the only way you could do it in the version of fortran it was written in
09:25 AM rue_: says in the origional notes
01:26 PM Tom_itx is now known as Tom_L
08:48 PM Tom_L: not much to report today.
08:48 PM Tom_L: i did get the right screws for the limit switches
09:05 PM rue_: 4 packages from china today
09:06 PM rue_: lm393
09:06 PM rue_: inductor kit
09:07 PM rue_: bicolour leds
09:07 PM rue_: er, no? just green?
09:08 PM rue_: odd...
09:08 PM rue_: and LM358
10:02 PM Tom_L: pondering what to work on next
10:03 PM rue_: cable guides
10:03 PM Tom_L: i'm gonna have to make the same type of switch arrangement on Y as i did on X and probably Z too
10:04 PM Tom_L: too early for cable guides
10:04 PM rue_: sit dow with it under some subdued lighting and talk wire
10:04 PM Tom_L: getting the hard limits etc figured out first then the electrical
10:05 PM Tom_L: the X limits turned out good though
10:06 PM Tom_L: i do need to change the plug on the sherline though cause i don't have another of that type to use i don't thiin
10:06 PM Tom_L: k
10:06 PM Tom_L: i should look i guess but i really don't like that plug so much
10:07 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/rue/Control2.jpg
10:07 PM Tom_L: it's the odd plastic one between the din and db25
10:08 PM Tom_L: it works i guess
10:08 PM Tom_L: i think i'll run the wires thru some of that blue flex conduit to the control though where it comes out the drag chain
10:09 PM Tom_L: i think i'll mount the control on the wall somewhere since it's sitting on top of the sherline box right now
10:13 PM Tom_L: rue_, yours is running of the parallel port right?
10:13 PM Tom_L: where's your control for it located?
10:17 PM Tom_L: http://sherline.com/product/5090-mill-accordion-way-cover-set/
10:17 PM Tom_L: haha cute
10:18 PM rue_: the ooooooooorigional machine runs from two parallel ports, but it also runs dos software I wrote, the other two machines run grbl
10:18 PM rue_: over usb
10:19 PM Tom_L: where'd you put the stepper drivers etc?
10:19 PM Tom_L: http://www.aliexpress.com/store/group/Flexible-accordion-type-protective-shield/1202171_256871181.html
10:22 PM rue_: the OOOOOOOOOOOrigional machine has an external floppy drive case that sits on an ATX power supply, all the drivers are in there. The sherline and the other machine both have external boxes with drivers/controller in them
10:22 PM rue_: funny enough, ont hat last machine, its another external floppy drive box, ... might hav been a tape drive actually.... its rather big
10:23 PM rue_: the NNNNNEW machine uses a case from an emergency light
10:23 PM Tom_L: that reminds me i still have the original serial control box for the sherline
10:23 PM rue_: serial!
10:23 PM Tom_L: yeah cereal
10:23 PM Tom_L: does that tell you how long i've had it?
10:24 PM Tom_L: https://www.aliexpress.com/store/product/U-type-custom-made-cnc-machine-protection-bellows-way-cover-W156mm-x-H67mm-x-Lmax750mm/1202171_32800993721.html?spm=2114.12010612.0.0.24b7de772t4SLK
10:24 PM Tom_L: custom way covers
10:29 PM Tom_L: garden underlayment plastic sheet it pretty heavy too
10:30 PM Tom_L: might work for a while
10:30 PM Tom_L: i think the neoprene sheet i can get may be too heavy to fold out of the way
10:30 PM Tom_L: iirc it's ~1/16" thick or so
10:34 PM rue_: hey, in the hunt for this fft stuff, I found out the code I'm working on was origioanlly written in 1967
10:35 PM Tom_L: fortran?
10:35 PM rue_: wtf were they doing processing fft's on computers in 1967!?
10:35 PM rue_: YEA
10:35 PM Tom_L: a step ahead of their time
10:35 PM Tom_L: what exactly is fft anyway?
10:35 PM rue_: a step and about 20 years ahead
10:36 PM rue_: fft is spectrum analizer
10:36 PM Tom_L: i think they did more with slide rules back in the 50's and 60's than we'll ever do on a computer
10:36 PM rue_: adc data to freq array
10:36 PM rue_: fortran wasn't for slide rules!
10:36 PM Tom_L: no, i'm not saying it was
10:37 PM Tom_L: i don't know anything about fortran or cobol
10:37 PM rue_: what I knew is that its apparently just above assembler
10:37 PM Tom_L: efficient?
10:37 PM rue_: get this, the big ibm systems in 1967 were 8 bit, used hexdecimal and could do floating point
10:38 PM rue_: it reads like basic, but with strange twists
10:38 PM Tom_L: is this a funventure for you?
10:38 PM rue_: IF(I-J)1,2,2
10:38 PM rue_: yea
10:39 PM rue_: and there are a million spinoffs
10:39 PM rue_: did you know about rotating greycode for position sensors?
10:39 PM Tom_L: so ^^ if the result is positive use the first positon var?
10:39 PM rue_: I'm not really sure
10:39 PM Tom_L: i do immediate if's in clipper
10:39 PM rue_: so far I'm not trying to translate the fortran
10:40 PM Tom_L: and they can be pretty complicated
10:40 PM rue_: tho, for the history of it, I might type this out
10:40 PM Tom_L: and run them thru a code block to iterate an array or file
10:40 PM Tom_L: i still like clipper
10:41 PM rue_: http://ruemohr.org/~ircjunk/datasheets/four1-fortran.pdf
10:41 PM rue_: about page 23? function called FOUR1
10:41 PM Tom_L: forboden
10:41 PM rue_: its still called that today
10:41 PM rue_: what?
10:42 PM Tom_L: You don't have permission to access /~ircjunk/datasheets/four1-fortran.pdf on this server.
10:42 PM rue_: hold up
10:43 PM rue_: fixed
10:43 PM Tom_L: did MIT write it?
10:43 PM rue_: side note of an amazing peice of hisotry
10:44 PM rue_: its from a IEEE paper, yea, mit I think
10:49 PM rue_: one of the docuemnts I was following today had credits to someone who helped retrieve them back off punchcard
10:49 PM Tom_L: i wonder what the used for debuggers back then
10:50 PM rue_: fortran on punchcard?
10:50 PM rue_: Tom_L, fly swatters :)
10:50 PM Tom_L: i could have taken a punchcard class in HS
10:50 PM rue_: did they make high density punchards?
10:50 PM Tom_L: i didn't get into it so i don't know
10:50 PM Tom_L: i know the first HP calculator came out then
10:51 PM rue_: its right in the era of the 4040
10:51 PM Tom_L: the math teacher had one but we had to use slide rules
10:51 PM Tom_L: i can't find mine anymore. it kinda makes me sad
10:52 PM Tom_L: this was done before that though
10:52 PM Tom_L: this meaning your document
10:52 PM rue_: this is crazy old
10:53 PM rue_: which is part of what astounds me, its such an advanced application for computers of that era
10:53 PM Tom_L: they had to use their minds
10:53 PM Tom_L: not so much nowdays
10:53 PM rue_: its actaully really messy code
10:54 PM rue_: there are a number of things they could have done way better
10:54 PM rue_: its almost like its a mashup of code from 3 people
10:54 PM Tom_L: hindsight is 20/20
10:54 PM rue_: and I wonder if it was
10:55 PM rue_: the 'data shuffle by bit reversal' is a mess
10:55 PM rue_: but I get the impression that fortran had no shift operator
10:56 PM rue_: or integer types
10:56 PM Tom_L: time to sleep
10:56 PM rue_: I'ma poke code for awhile
10:56 PM Tom_L: 2am?
10:57 PM rue_: its amazing to think how much more powerfull an avr is compared to the 1967 machines
10:57 PM Tom_L: were they tube based?
10:57 PM rue_: 60k instructions/sec vs 16M
10:57 PM rue_: it looks like it might have been done on one ofthe first solid state machines
10:57 PM rue_: before they called it solid state
10:58 PM Tom_L: when did bell invent the transistor?
10:58 PM rue_: (still fill a room)
10:58 PM Tom_L: i think it was bell..
10:58 PM rue_: long before, I think they had problems with consisant manufacturing for a loooong time
10:59 PM Tom_L: then there was.. i forget the term but those loop memory cells
10:59 PM rue_: core memory
10:59 PM Tom_L: yeah
10:59 PM Tom_L: iron core
10:59 PM rue_: ferrite, it was static ram tho
10:59 PM Tom_L: ever see a 18" hdd platter?
11:00 PM Tom_L: you can _see_ the bits on it
11:00 PM rue_: I seen some big ones
11:00 PM rue_: :)
11:00 PM rue_: havn't seen a drum drive tho
11:00 PM Tom_L: me either
11:00 PM rue_: when i was a kid I took apart 2 8" floppy drives :(
11:01 PM rue_: I have a few parts from them
11:01 PM Tom_L: my buddy had/has? a what were they 9 or 18 track tape drive
11:01 PM Tom_L: tried to give it to me and i declined
11:02 PM Tom_L: also had a huge hp? raid drive
11:02 PM Tom_L: the tape drive was about the size of a refrigerator
11:03 PM rue_: this document was contracted by the airforce, it might have been destined for radar
11:04 PM rue_: reffers to a CDC 3300
11:05 PM Tom_L: well it was printed by the USAF
11:05 PM rue_: http://60bits.net/images/CDC3300Brochure.jpg
11:05 PM rue_: hmmm
11:06 PM Tom_L: you had to be pretty important to sit in that middle chair
11:07 PM rue_: the one they are all avoiding?
11:07 PM rue_: :)
11:09 PM Tom_L: gnite