#robotics Logs

Jan 12 2018

12:10 AM mrdata: rue_, IBM-029 punch cards would put FORTRAN code beginning in position 7, labels in positions 1 to 5, and position 6 was for a continuation
12:13 AM mrdata: if needed
12:13 AM mrdata: later versions of FORTRAN i think changed that spacing
12:13 AM mrdata: but i didnt learn them
12:13 AM mrdata: university classes had me lesarn Pascal
12:13 AM mrdata: but i moved on to C soon enough after
12:25 AM rue_: I dont think pascal was a bad language, it just didn't live
12:26 AM rue_: it generated MUCH nicer assembler
12:26 AM rue_: C was all like >#%$!#%@#$^%!#%$@#$^%!#@%!@%$#@$%@#%< and pascall was all like >..<+>.-.<
12:30 AM mrdata_ is now known as mrdata
01:20 AM mrdata: and perl was worse than C in that regard
02:47 AM rue_: :) my fft code is working!
02:49 AM mrdata: \o/
02:50 AM rue_: not on the avr yet
02:50 AM mrdata: but, soon to be on the avr?
02:50 AM rue_: but I totally rewrote the 1967 code
02:50 AM rue_: we will see!
02:51 AM rue_: it looks like, if the input value is capped to 32 units, that the output caps at 255 units
02:51 AM mrdata: huh
02:51 AM rue_: so, there can be a 32:1 digital AGC attenuator from the ADC
02:51 AM rue_: just check the peak value of the array and scale accordingly
02:52 AM rue_: aaanyhow
02:52 AM rue_: will still require an input filter
02:52 AM rue_: low pass
02:52 AM rue_: nyquest
02:53 AM rue_: but I'm putting in calculated sine waves and getting nice spikes out
02:54 AM mrdata: nice
02:54 AM mrdata: thats good news
02:55 AM mrdata: how about mixtures of calculated sine waves
02:55 AM mrdata: how about mixtures of calculated sine waves with interesting phases
02:55 AM * mrdata must sleep and get over his cough
02:55 AM rue_: because I dont have the formulas to know what the bins are (in radians) I'm a little in the dark
02:55 AM mrdata: nn
02:56 AM rue_: as the sinewave freq/sample is ?? virtual hz
02:56 AM rue_: I have image comming up
02:57 AM rue_: http://ruemohr.org/~ircjunk/images/fft.png
02:58 AM rue_: it sounds like fft's are usefull for a bunch of things I dont know about yet
02:59 AM rue_: I typed out the whole origional FORTRAN program
03:00 AM rue_: I couldn't help myself
07:42 AM anonnumberanon: veverak, let me know if you find a cool lidar
07:42 AM anonnumberanon: i want to play with one
07:42 AM anonnumberanon: that's cheap
07:46 AM veverak: sure
07:46 AM veverak: :)
07:48 AM anonnumberanon: i see that this freaking thing has not moved priced since 2 years ago when i built our done
07:48 AM anonnumberanon: https://www.robotshop.com/en/hokuyo-urg-04lx-ug01-scanning-laser-rangefinder.html
07:50 AM anonnumberanon: we ripped one out of a neato vacuum cleaner robot but didn't have time to use it
07:50 AM anonnumberanon: I probably still have it.
07:59 AM mrdata: rue_, you printed the frequencies table, can you print the phases table, too?
08:58 AM rue_: that was a power table
08:58 AM rue_: yes, I can print the raw one
08:58 AM anonnumberanon: is it legal to make own self-driving car and ride it home?
08:59 AM rue_: is it illegal?
08:59 AM anonnumberanon: not sure
08:59 AM anonnumberanon: are you supposed to have at least one driver inside a moving car?
09:00 AM anonnumberanon: if(drivers < 1) ticket;
09:00 AM rue_: I bet there aren't laws that say so
09:00 AM rue_: who they gonna ticket?
09:00 AM anonnumberanon: that's a very good point
09:00 AM rue_: and for what, absent driving?
09:00 AM anonnumberanon: YOU DID NOT DRIVE, CITIZEN
09:01 AM anonnumberanon: YOU'RE GOING TO JAIL
09:01 AM rue_: "your honnor, I would like to submit that by the ticket itself, I wasn't there when it happened" "Case dissmissed!"
09:02 AM rue_: mrdata, I dont think its a good idea for me to paste the whoel array
09:02 AM rue_: 0.00 0.00
09:02 AM rue_: 0.00 4.91
09:02 AM rue_: -15.07 0.00
09:02 AM rue_: 0.00 -20.91
09:02 AM rue_: 16.31 -0.00
09:02 AM rue_: 0.00 -251.84
09:03 AM rue_: tho I thought the idea was to throw away the phase information
09:04 AM rue_: cause, itsn't the complex result amplitude and phase?
09:14 AM anonnumberanon: no, you were there in the vehicle, but you were a passenger, at the time you were tramming the z axis of your new mill that was in the trunk of the car, while watching Star Trek Discovery
09:16 AM rue_: oh crap, my values were wrong
09:16 AM rue_: I corrected it and the difference is amazing
11:29 AM mrdata: rue_, the phase information will confirm more about the accuracy of your fft implementation
11:30 AM mrdata: for instance, from the time domain set of samples i can see the phase of the sine
11:32 AM mrdata: = 0
11:32 AM mrdata: so the phase table confirms that, good
11:34 AM mrdata: (sort-of; i'm not sure i'm relating the coefficients correctly)
11:37 AM mrdata: but if you were to shift the time-domain signal's phase, say forward 1/16 th of a cycle, then we should find thsat in the phase table
11:37 AM mrdata: after fft is calculated
02:22 PM Snert__ is now known as Snert_
03:55 PM i-make-robots: https://www.marginallyclever.com/2018/01/best-way-program-arduino-microcontroller-linux-command-line/
04:45 PM To_Aru_Shiroi_Ne is now known as ToAruShiroiNeko
05:31 PM ace4016: here you go rue_ http://robotics.sciencemag.org/content/3/14/eaar3276.full
05:33 PM ace4016: more details on the muscles. sounds like all you need are some conductors, potato chip bags, canola oil, polydimethylsiloxane, and hydrogel (?)
05:36 PM Ubuntivity: Hello
07:22 PM rue_: ace4016, loads to a blank page for me
07:22 PM rue_: i-m...
07:22 PM rue_: I'm gonna sleep 2 hours brb
07:23 PM Tom_shop: extended the Z offset
07:24 PM Tom_shop: you don't need sleep, it's the weekend!
08:03 PM ace4016: rue_, interesting; i wonder if it's region locked? or maybe i sent you to wrong level of link
11:08 PM * Tom_L sets off rue_bed's alarm
11:15 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Mill_Steel/Assembly/Z_Axis_standoff_new.jpg
11:15 PM Tom_L: 6" Z offset now
11:18 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Mill_Steel/Assembly/Y_Axis_limitswitches1.jpg
11:18 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Mill_Steel/Assembly/Y_Axis_limitswitches2.jpg
11:18 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Mill_Steel/Assembly/Y_Axis_limitswitches3.jpg
11:19 PM Tom_L: hopefully Z will be a little easier than X and Y switches were