#robotics Logs

Dec 24 2017

12:17 AM rue_shop3: 17c...
12:20 AM rue_shop3: I think those boiler tubes need a cleanout!
01:00 AM rue_shop3: so I need to work out how to control my automatic solder dispencer
01:11 AM rue_shop3: 2khz
01:35 AM rue_shop3: 45c
01:35 AM rue_shop3: this is not how a healthy full throttle behaves
02:01 AM rue_shop3: 52c
03:33 AM AchiestDragon: hate doing this in software ,, ok so walking gate motion , low level 3 servos per leg 6 legs 18servos in total ,, now on one leg servo 1 has to move say 100 steps servo two 30 and servo 3 only 2, step = pwm value change
03:34 AM AchiestDragon: the time for the move is set by the servo with the most movment ie the rate of change of the serov to make a move of 100 in this case
03:36 AM AchiestDragon: the positions updates for the other servos need to make there change at a rate proportional to the step time ie not just instantly move to there new positon they need to gradualy move there to keep the movment in sync
03:37 AM AchiestDragon: it was easy doing this in hard logic where i could create a pwm driver that you set the positonal change and time to actualy get there indipendanly for each
03:40 AM AchiestDragon: to get close to doing it in software im going to have to configure at least 18 indiviual (real time) timers plus the aving to update the values over i2c for each
03:42 AM AchiestDragon: what i am going to miss is i also had the timer clock controlable freq so i could just adjust that to increase or decrese the rate of change of all motors in sync
03:43 AM AchiestDragon: so allowning me to vary the speed of the walking sequence to provide smooth accelaration and deceleration , on the fly
03:46 AM AchiestDragon: the actual walking motion was then calulated and a table created , buffered so that it created a step sequence buffer in a fifo form that was just fed to the pwm state machine
03:48 AM AchiestDragon: that would let you have the foot stay parralell to the body duing walking ie not make a curve
04:02 AM AchiestDragon: as i have said it chould be done by clockwork ,, and may well have been but was desroyed as little of them survived the revolutions https://www.youtube.com/watch?v=bY_wfKVjuJM for clockwork mecanisams pre industirial revolution
04:11 AM AchiestDragon: in those days thy had streets of "common" people each making small parts in bad working conditons ,so they could be sold to the upper class rulers as novaly devices to show off to there frends ,rulers that where worse than the king from the farytail chity chity bang bang if you ever seen that film as a kit , not far from the trouth of what they were like
04:13 AM AchiestDragon: not that its much diferent now
05:18 AM veverak: AchiestDragon: your task is easy
05:18 AM veverak: you always know actual servo posititions
05:18 AM veverak: you take a goal positions ofr a set of servos as input
05:19 AM veverak: execute a loop at specific frequency, and at each tick, calculate where the servo can go in specified period
05:19 AM veverak: from it's actual position to it's goal position
05:19 AM veverak: you send that as command
05:20 AM veverak: or, to be more precise, you calculate the % distance to their goal the servos can achieve, get the smallest % distance and use that for each servo
05:20 AM AchiestDragon: no you dont in the case of rc pwm servos , you know where you told it to goto but not if its got there , or how long its goingto take it to get there unless your able to read its actual positon
05:20 AM veverak: no
05:20 AM veverak: you can assert that you know
05:21 AM veverak: ;)
05:21 AM veverak: and you know how long it can take minimally
05:21 AM veverak: that is good enough for reasonabe algorithm
05:21 AM veverak: AchiestDragon: if you set position to 0 angle and wait a few secs
05:21 AM veverak: you can assert that you know that it is at 0
05:22 AM AchiestDragon: if you know the time it takes to move x number of steps then yes sort of but only if you also know the time taken and the servo load
05:22 AM AchiestDragon: ie no load on the servo and the time is different than with a lot of load on that servo
05:22 AM veverak: you can approximate that
05:23 AM veverak: I have a code written, that does this logic
05:23 AM AchiestDragon: but you should not have to approcimate it
05:23 AM veverak: get time window, get set velocity, calculate possible distance traveled
05:23 AM veverak: that velocity is configured as 1/3 of maximal velocity of that joint
05:24 AM veverak: AchiestDragon: you can derive some way to change the velocity based on load you expect, or whatever
05:24 AM AchiestDragon: what i am doing though is having positon sensors in the way of pots on the leg joints anyway so then it has pid feedback and should be easyer to keep things in sync
05:25 AM AchiestDragon: but i want to include fallback code maybe rougher bu so it can be used without those
05:25 AM veverak: ah, Isee
05:25 AM veverak: yeah, if you have position sensors that is better :)
05:27 AM AchiestDragon: i was looking at the code for hexy https://github.com/mithi/hexy that i thought may make a good base but after reading though it looks like i would need to totaly rewrite it
05:28 AM veverak: well
05:28 AM veverak: if you have code that is based on the position sensors
05:28 AM veverak: I would simply fake the position data in a manner that should be safe
05:28 AM AchiestDragon: that dose not even incude keeping the leg parralel with the body on move
05:28 AM veverak: and use same code
05:29 AM AchiestDragon: have to say ive done less python than c or c++ but it seems more sutable to write it in python than c or asm
05:31 AM veverak: I wanted to use C++ code
05:31 AM veverak: I've read the surce
05:31 AM veverak: I freaked out
05:32 AM veverak: I want to burn it with fire and rewrite it
05:32 AM veverak: ^^ this happens a lot
05:32 AM AchiestDragon: i have the option atm that i need 18 adc inputs so prob going to use some microcontroler for that , since i may be able to find a micro with 10 adc inputs there should be enough extra io to offer 10 pwm outs on the same , using a pair of micros as both pwm and pid should make doing some of the low level pwm drive on those
05:34 AM * weyland|yutani asks himself if you give a NN complete control over a bipedal robot whould it be able to teach itself walking properly by trial and error?
05:34 AM AchiestDragon: one good thing about python over c is python was written with object orentated stuff as native so not a mindfucked extention like on c++ and also multy thread processing is native in python also
05:35 AM veverak: wat
05:35 AM veverak: what is problem with OOP and C++ ?
05:35 AM weyland|yutani: well depends what NN is used or what kind of learning systems we use
05:38 AM AchiestDragon: anyway using python and a pi3 main control should let me actualy calculate the leg step positions real time rather than having to rely on a precalculated database
05:58 AM AchiestDragon: oop and c++ well half the problems are dew to having to define half the type structure in seperate hedder files , and almost again in the code , its just a pita to keep track of
05:58 AM veverak: got a 10000 lines of code
05:58 AM veverak: do not had to separate it even single time
06:53 AM AchiestDragon: ho great ,, 10k dual gang pots with d shafts in 5 off were cheaper than single gang pots , good so not quite wired in parrallel but 2 outputs diode couppled though 1n4148's so it should reduce pot track noise when warn
06:58 AM AchiestDragon: and the ones wth split nurled shafts may be good for volume controls etc but if you need to mount them by using a screw to hold the shaft they tend to snap
07:09 AM AchiestDragon: https://drive.google.com/file/d/1F8DY25BYNNutFgxRKgE0iOvHGx9w4Z2a/view?usp=sharing
07:10 AM veverak: fuck this shit
07:10 AM AchiestDragon: still a few mods to do like actual foot ground contact switches /sensors
07:10 AM veverak: the code written by some academic people is full of shit
07:10 AM veverak: literally
07:12 AM AchiestDragon: what like every line commeted ie a=1 # set the value of a to 1
07:13 AM AchiestDragon: and another 3 paragraphs of comments saying why and how
07:13 AM veverak: well
07:13 AM veverak: just made a patch that replaces cout << "..." << endl;
07:13 AM veverak: with proper LOG commands
07:14 AM veverak: https://github.com/raulmur/ORB_SLAM2/blob/master/src/PnPsolver.cc
07:14 AM veverak: https://github.com/raulmur/ORB_SLAM2/blob/master/src/PnPsolver.cc
07:14 AM veverak: void PnPsolver::qr_solve(CvMat * A, CvMat * b, CvMat * X)
07:14 AM veverak: explain to me what that method does
07:16 AM veverak: (not to mention that to see new/delete in C++11 is somehow ... )
07:17 AM AchiestDragon: ho yea the alternate methods like rather than r= a*b its for x=1 to a do { r=r+b}
07:17 AM veverak: anyway
07:17 AM veverak: the theory behind the code is nice
07:17 AM veverak: the code itself is full of bullshit
07:17 AM veverak: that is unreadable
07:17 AM veverak: thus something that I can't rely on
07:19 AM AchiestDragon: theres 2 sorts of education coders ,, the ones that code with the skill that makes there code do the job but not clear how it does it and the better sort that code in a way that makes it readable so students can learn from it
07:23 AM veverak: well
07:23 AM veverak: this calimed that their code is practically usable
07:23 AM veverak: I object
07:23 AM AchiestDragon: ie the prof that codes using methods to attempt to baffle others so they look at the code and think wtf , in an attempt to show others there masters of the language ,, like you said
07:24 AM AchiestDragon: when a far more readable method could of been used with no code performance impact
07:24 AM veverak: ...
07:28 AM AchiestDragon: theres other coding style issues also not covered , in published source especialy gpl stuff comments are often removed , this is to keep source code files size low
07:30 AM AchiestDragon: when codeing i was tought you should comment your code so when you come back to it 3 or 4 years later to do a bug fix or update you can remeber what the f you where thiking when you wrote it that way
07:31 AM veverak: not exactly
07:31 AM AchiestDragon: and then put out a version stripped of the comments for publication / distribution ,,
07:32 AM veverak: why?
07:32 AM veverak: you should comment everything that is not clear from the code, and when comment something, you should rather figure out how to rewrite the code so it is clear what it does
07:33 AM AchiestDragon: well 1 so anyone else reading the code would prob think wtf and want to rewite it if asked to mod it ,, rather than asking the author to do the changes
07:34 AM AchiestDragon: 2 so that a 3rd party may have the code uncommented and it will take them weeks longer to fix than you could if you had the version with comments
07:34 AM AchiestDragon: especialy true if the code is pure uncomented ASM
07:37 AM veverak: wat
07:37 AM veverak: that's why I am asking why would you remove the comments
07:37 AM veverak: it does not make sense
07:37 AM AchiestDragon: idealy yea you should comment code where posible , if its for your own use or open , i tend to despise the fact that some gpl coding standard prevent it
07:38 AM veverak: wat?
07:38 AM veverak: :)
07:38 AM veverak: I can't imagine why would anybody do that
07:39 AM AchiestDragon: linux for example comes with a distribution currently taking a full dvd of object code , ie 4.5gb , the souce for all that last time i saw that as iso's stood at a 5 dvd set , and that was after unneeded whitespace and comments removal
07:41 AM veverak: ah
07:41 AM veverak: I should not
07:41 AM veverak: written something to it
07:41 AM veverak: it is obviously managed by a bunch of asholes
07:41 AM veverak: they have 42 push requests, starting form February
07:41 AM AchiestDragon: as with comments that may be arround a 24 dvd set
07:41 AM veverak: without response
07:42 AM veverak: *from
07:44 AM AchiestDragon: the best gpl offers are api function descriptions , that describe the undividual function calls if there intended to be called from externaly
07:46 AM veverak: fuck this
07:46 AM veverak: codebase at bad state and the maintainers ignore the push requests for over a yer?
07:46 AM veverak: not a realiable piece of software, NEXT
07:54 AM AchiestDragon: yea only 12 commits all year
07:55 AM veverak: I mean, I've seen pus hrequests that took long
07:55 AM veverak: but there was conversation
07:55 AM veverak: here, most of them are ignored
07:55 AM AchiestDragon: 299 issues , and 42 pull fix requests pending
07:57 AM veverak: the number of forks is interesting
07:57 AM AchiestDragon: Couldn’t load network graph. Too many forks to display. lol yea
07:58 AM AchiestDragon: in that case i would look for the fork that was still activly beeing maintained but not easy to find with that error
07:59 AM veverak: yep
07:59 AM AchiestDragon: guess that says something about the original coding style if most forked it and rewote sections
08:00 AM veverak: exactly
08:00 AM veverak: well
08:00 AM veverak: I still got a tip for anothe rSLAM
08:00 AM veverak: will see about that one
10:15 AM robotustra: @rue_shop3> 100 gal of antifreeze?
10:15 AM robotustra: use oil
10:25 AM * Tom_L chips rue_mohr out of a block of ice
10:29 AM robotustra: 21 century in canada - ho heat in the hause
10:29 AM robotustra: no
10:30 AM Tom_L: this is rue we're talking about
10:30 AM robotustra: yes
10:30 AM robotustra: can't heat a house?
10:32 AM Tom_L: http://ruemohr.org/%7Eircjunk/projects/heatplant/
10:32 AM Tom_L: that's how he does it
10:32 AM AchiestDragon: i have a hp dl740 its got 8 xeon cpus only reason i still have it is incase the heating fails , it makes a good 1kw room heater other than the fan noise
10:33 AM robotustra: http://www.instructables.com/id/Sheet-Metal-Fireplace/
10:35 AM robotustra: Tom_L, what is I2C used fro in this boiler?
10:35 AM Tom_L: temp sensors
10:35 AM Tom_L: probably
10:35 AM robotustra: ah, I see
10:35 AM robotustra: what type?
10:35 AM robotustra: 35?
10:36 AM Tom_L: i don't remember
10:36 AM robotustra: celsius?
10:40 AM Tom_L: if rue_mohr used oil, he'd burn is place to the ground
10:41 AM robotustra: oil in the closed contour is more efficient than water
10:42 AM robotustra: and it does not freese
10:42 AM robotustra: it boils at higher temperature
11:34 AM AchiestDragon: been thinking , as to servo calibration once its got pid sencors shuuld be easy and quick i dont real want to spend hours mesuring the angles of each joint in order to work out the usable scail it should be quick and simple
11:36 AM AchiestDragon: so if i use some 3d printed templates that i can place on the leg like limit stops then all i have to do is move each joint till it is in contact with it and read the value for that limit
11:37 AM AchiestDragon: that can be done manualy wihtout enabling the servos initaly , once set it should not need readjusting unless theres a pot fail
11:38 AM AchiestDragon: then the servo pwm range can be auto done to the actual pid values , should make inital setup quicker and more repeatable , as you just move the template form leg to leg untill all joints limits have been registered
11:47 AM AchiestDragon: and it should be a less prone to error method than manualy mesuing the angles and reading the value for each
11:56 AM AchiestDragon: one year , before i had central heating there was a small room that i used as a computer room with a desk and 2 chairs , if the door was keeped closed the heat from 2 desktop pc's used to keep that room at a warm temp without the use of any other heating , although it was in the days of crt monitors and a 19 and a 21" crt monitor also adds to that
11:57 AM rue_bed: cats, always going in and out
11:57 AM rue_bed: I have 2 computers and a laptop in the room
11:57 AM AchiestDragon: saved spending an hour or so cleaning out the coal fireplace and going out to the coal bunker for more coal
11:57 AM rue_bed: but its a badly insulated place
11:58 AM AchiestDragon: fit a draft proof cat flap
11:58 AM rue_bed: the wood boiler is a temp solution till I build a new place
11:58 AM rue_bed: I cant let the house get cold anyhow, it molds up
11:58 AM rue_bed: already had a problem with an unused room
11:59 AM AchiestDragon: it rarely gets as cold in the uk as it does in canada -15c and its almost breaking records
12:01 PM AchiestDragon: one of the advantages of living on an island in the atlantic gulf streem it is usualy warmer here for its latitude than most of the rest of europe
12:02 PM AchiestDragon: if theres snow on the ground for more than a week its unusual
12:12 PM AchiestDragon: insulation matters , you know those sringy bead type things they use in door ways usualy restraunts so they dont have to actualy open the door and it makes a space devider , make some of those out of strips of bubblewrap , there quite effective at keeping drafts out and still allowing cats in , as long as the cat dosent think its some toy to shread to peaces
12:13 PM AchiestDragon: bubble wrap also makes for good quick and dirty method of double glazeing if you dont have that
12:14 PM AchiestDragon: use like full lenth net curtains
12:25 PM rue_mohr: your using a warp bubble for glazing?
12:25 PM rue_mohr: isn't that like using a piledriver for finishing nails?
12:25 PM AchiestDragon: thermal curtains
12:26 PM rue_mohr: ah, this is half a trailer that was put here and modified by a poor family that didnt care and didn't know a damn thing about what they were doing
12:26 PM AchiestDragon: not now but have done in the past
12:26 PM rue_mohr: there is no repairing it, its junk, have to tear it down and make a new building
12:28 PM AchiestDragon: if you make something perminent you should look at ground source heat pumps , there comulsary in sweden and norway on new builds , basicaly they just use a heat reansfer pump to suck heat from underground where its warmer and are quite energy and environmentaly efficiant
12:28 PM rue_mohr: if it didn't have aluminum wiring, if it didn't have r5 insulation, if the roof were properly supported, if it hadn't been built over the septic system, if the addition were self supported, if if if it would be repairable for less than building new
12:29 PM AchiestDragon: unless you live on top a a few hundered foot of permafrost
12:29 PM rue_mohr: nope, not here, they aren't worth it, the maintenance is too costly, put solar cells on your roof and use electric heat
12:29 PM rue_mohr: or wood
12:31 PM AchiestDragon: exes father has ground source he also has solar for both power and water , will ask next time i see him how much mantainace he has to do , hes had it for around 10 years now
12:32 PM AchiestDragon: hes remote so no gas or mains water but does have a mains power feed , and phone line
12:33 PM AchiestDragon: most of his probs from what i recall have to do with water supply
12:34 PM AchiestDragon: the stream amlost dries in summer and frezes up in winter , allong with some of the collection tank pipes
12:35 PM robotustra: < AchiestDragon> it rarely gets as cold in the uk as it does in canada -15c and its almost breaking records
12:35 PM robotustra: do you have water ipes outside the buildings?
12:35 PM robotustra: pipes*
12:35 PM AchiestDragon: he does
12:36 PM robotustra: who is "he"?
12:36 PM AchiestDragon: about 100 yards to the streem , mosly burried ecept for where it connects ot the tank
12:37 PM AchiestDragon: the exes father its a farm he lives on
12:37 PM robotustra: pipe+water+frost = problems
12:37 PM AchiestDragon: in some remote welsh vally at least a mile away from the nerest street lamp
12:38 PM AchiestDragon: yea
12:38 PM AchiestDragon: unless pipe is at least 1m below ground
12:38 PM robotustra: I lived in a country where pipes below the level of soil freeze
12:39 PM robotustra: depper than 1m 20 cm
12:40 PM AchiestDragon: ground source heat pumps usualy envolve drilling a bore hole down a maybe 100 meters , its warmer there regardless
12:41 PM rue_mohr: and then pumping brine thru it to a heat exchanger
12:41 PM rue_mohr: the brine needs to be presureized/checked/topped up and the heat pump is basically a refrigerator
12:41 PM rue_mohr: so lots of maintenance every yeat
12:41 PM rue_mohr: year
12:41 PM AchiestDragon: although here in the uk we can get way with digging a series of trenches about 2m deep and laying out a pipe up and down each , then filling it back up
12:42 PM rue_mohr: or solar farm, and the solar cells just sit there
12:42 PM rue_mohr: for the most part,t hey dont really even need washing off
12:43 PM AchiestDragon: think he just uses deionised water , works well , no yearly maintanance if you get an heat exchanger that works that way
12:43 PM AchiestDragon: rather than some sails man trying to sell you expencive chemicals every year
12:43 PM rue_mohr: usually the system is given a bit of pressure, to what I know that needs to be checked, maintained
12:45 PM rue_mohr: hmm, conductive oogoo anyone?
12:46 PM AchiestDragon: i will ask when i see him and let you know , but he did reserch it himself rather than getting some company to do it for him so a diy fitting , and he does the mantancance so dont think he spends much $$$ just maybe a service check once a year
12:47 PM rue_mohr: most of the systems out here end up inopertive and people dont want to get them serviced/fixed, so they sit there running on the electric backup heaters all the time
12:47 PM AchiestDragon: and even thats got to take less time than creating a winter supply of logs for the fire
12:47 PM rue_mohr: I think the heat pumps are a common failure too
12:47 PM rue_mohr: yep, I'm lucky as I get a lot of construction scrap wood
12:48 PM rue_mohr: and in the winter, nature literally throws trees at you
12:48 PM AchiestDragon: should build a better wood house rather than using it for fire wood
12:48 PM rue_mohr: in feb I'll have the mortguage paid off
12:49 PM rue_mohr: I'm gonna spend money for a year
12:49 PM AchiestDragon: nice log cabin made from thick tree trunks
12:49 PM rue_mohr: then go back to penny pinching to build the new house
12:49 PM rue_mohr: logs dont actaully insulate that well
12:50 PM AchiestDragon: neather do 500mm to 1m thick stone walls
12:50 PM rue_mohr: I'm going to try to twist the project, build a new 1200 square foot shop with a house on it
12:53 PM rue_mohr: huh, the 'follow santas journey' link on google is a bad url
12:53 PM rue_mohr: OOPS
12:53 PM AchiestDragon: you know best solution is a couple of 40 foot metal shipping containers welded side to side with the joing side covers removed , dig a hole deep enough so they are just below ground , fill in the sides with dirt , build a set of stairs down to them and allow dranage to a point lower and away from it , cover the top with 1.5 m of dirt and you got a warm underground bunker , add a flu for a woodburner in one conrner and utities and
12:53 PM AchiestDragon: your away
12:54 PM rue_mohr: na, too narrow
12:54 PM rue_mohr: and they would rust away
12:56 PM AchiestDragon: not in your lifetime , ensure you dont scratch the paintwork when burrying them and ensure theres way too much paint on them to keep water out as long as posible
12:56 PM rue_mohr: "burry them, but dont scratch the paint" ok
12:58 PM AchiestDragon: yea fill cose to it with soft compost type dirt or moss not stony or gravely dirt
12:58 PM rue_mohr: you know this is not good advise right?
12:59 PM rue_mohr: that is why I have this place now
12:59 PM AchiestDragon: depends on what planning permission will let you get away with
12:59 PM rue_mohr: casue some twitt didn't do something properly
01:00 PM AchiestDragon: read up on it you could pad it with old tires and make a tire wall that is used sometimes but again depends on the countries planning regs and codes
01:00 PM rue_mohr: a shipping container out here is $10000, lock blocks are $100ea
01:00 PM rue_mohr: http://www.groundx.ca/uploads/2/0/6/6/20667724/5603503.jpg
01:02 PM rue_mohr: geez look in the background, SOMEONE didn't play with legos as a kid
01:07 PM AchiestDragon: ha yea dont bury one , just seen other site saying why not to
01:10 PM AchiestDragon: still think if you get free wood then you should be able to do somthing like this over a few years , and prob need a sawmill http://www.battlecreekloghomes.com/wp-content/uploads/2015/12/Custom-Barn-Builder-and-Contractor-TN-750x500.jpg
01:13 PM AchiestDragon: and somthing like thats going to have a high resale value when finished
01:16 PM AchiestDragon: whats the av winter temp there -10 to -30C
01:26 PM AchiestDragon: not shure of the suitabilaty of this method in your environment but it may be a solution for you http://buildingwithawareness.com/the-pros-and-cons-of-straw-bale-wall-construction-in-green-building/
01:28 PM AchiestDragon: you plaster clad the inside and outside so the bails keep dry , and then overclad the outer again with a render to finish it neatly , again supply of hay and building regs dependant
01:31 PM AchiestDragon: ho yea its xmas eve , watch out for low flying slightly drunk raindeer
02:53 PM Tom_L: rue_mohr, during your spending year plan to build a nice CNC.... mkay?
02:53 PM Tom_L: i can model it if you want
02:53 PM Tom_L: with available parts
03:20 PM Tom_L: ( within reason )
04:19 PM Tom_L: AchiestDragon they use drones now
04:50 PM AchiestDragon: i bet weather limits opperation , especaly when high winds andhigh city blocks , once you pas the wind null side of a building and encounter the crossroads with a 40 mph wind gust blowng down even good delivery drones are going to end up beeing unable to respond with enough power and splattering your piza over the side of some building
04:54 PM AchiestDragon: the local search and rescue hele (full size pioloted one ) cant withsand 80mph max wind , and in 60 to 70mph gusts you can realy hear the engines struggle to keep the thing stable , fuck knows what gyros and controls they use to hold its position but theres a lot more rotor power and stabilaty than you can get from a drone , by a long way
05:04 PM SpeedEvil: Actually it's easier for drones, in some ways as their disk loading can be way higher.
05:04 PM SpeedEvil: Racing drones go at >>80mph
05:05 PM AchiestDragon: but with a payload that reduces power and adds more wind resistance in the fact that its carring a not verry airo dinamic stack of piza's
05:06 PM AchiestDragon: or packs of bear
05:06 PM AchiestDragon: beer
05:08 PM AchiestDragon: raing drones are smal light and overpowered compaired to normal drones , the ones that would be needed for delivereis are going o need some beefy motors and bateries
05:08 PM AchiestDragon: ffs typo bad day again
05:24 PM AchiestDragon: and then given time there will start piza wars where the piza delivery is high stakes custom , and competitors start to use jamming and other methods to take out compeating companies drones in
05:26 PM AchiestDragon: mid flight , maybe crashing them into things like cars leving the company with damige litigation and lost of customers
07:09 PM rue_shop3: Tom_L, its been a thought
07:14 PM justanotheruser is now known as justan0theruser
07:21 PM rue_shop3: the ambulances are going crazy
07:21 PM rue_shop3: I hear a 3rd one in like the last hour
07:21 PM rue_shop3: wonder whats up
07:22 PM robotustra: race on ambulances - cool
07:22 PM rue_shop3: anyhow, I have this 700vdc circuit I been wanting to play with...
07:35 PM Tom_L: play carefully
07:36 PM Tom_L: i was quite surprised last night after trying that Z axis. i figured it would have more drag than it did considering it hasn't been milled flat yet. Things _do_ align pretty close though
07:38 PM Tom_L: not real sure what i can work on with it until i do get things milled flat
07:39 PM Tom_L: i don't want to cut the Y rails until i get the screw back from being shortened
07:39 PM Tom_L: and i can't mount the X axis stuff because he's got that out squaring it up
10:46 PM rue_mohr: yep
10:46 PM rue_mohr: I'm working on vco's to make adc's from
10:48 PM Tom_L: get the boiler back up?
10:48 PM rue_mohr: its hot
10:48 PM rue_mohr: I have to watch it more carefully is all
10:58 PM Tom_L: goin to ma's tomorrow?
11:46 PM anonnumberanon: Merry Christmas! May you build the most roboty robots!
11:49 PM anonnumberanon: My new kit: https://imgur.com/a/Vvq2l