#robotics Logs

Dec 04 2013

00:18 Fiesta why a Canfield joint? it has good strength in that direction?
00:18 Fiesta in that direction of force
00:20 GargantuaSauce it has a full hemisphere of motion without twisting
00:20 GargantuaSauce so you can put hoses/wires up the middle
00:21 GargantuaSauce and yeah its parallel kinematics give it decent strength in all directions
00:23 Fiesta i'm kinda amazed that it's new, it looks like any other kindof parallel mechanism
00:24 GargantuaSauce yeah i found that surprising too, and now that i understand the kinematics they're actually quite simple
00:25 GargantuaSauce the calculations aren't though, which might be a factor...hard to pull them off for high speed control with an older microcontroller i'd say
00:26 Fiesta ah
00:26 Fiesta totally different topic: why do we use mercury/alcohol in thermometers? Arn't there reversible chemical reactions that have a greater change in volume in response to temperature?
00:28 Fiesta or is phase change (ie: mercury/alcohol) actually the best when it comes to volume changes in response to temperature.
00:29 GargantuaSauce they do not employ phase change
00:29 GargantuaSauce it's just the liquid expanding/contracting with temperature
00:29 GargantuaSauce i imagine it's pretty hard to come up with a chemical reaction that produces a linear change in volume over as nearly wide a range of temperature
00:30 Fiesta doesn't have to be linear, particularly in a thermometer you can just draw the degree lines where you want
00:30 GargantuaSauce sure, but most instruments are
00:31 Fiesta i get your point about expanding/contracting though. Coz now that i think about it hotter temps make the marker go up not down.
00:31 Fiesta i want it for a mechanical mechanism, something that will move the most in response to temp changes.
00:32 GargantuaSauce bimetal strips
00:32 GargantuaSauce tear apart a few old home thermometers
00:32 GargantuaSauce they have mercury switches too :)
00:32 MrCurious1 dont eat mercury
00:32 Fiesta "a few" haha. i actually want millions
00:33 GargantuaSauce okay that sounds kind of wonky, what's the application
00:33 Fiesta why i'm looking at liquids is that it may allow me to just have a film of it.
00:33 Fiesta it's for a tactile matrix.
00:34 Fiesta i'm thinking of howto make a tactile matrix outof a liquid filled bag draped over an LCD screen. Thinking of liquids that will respond to light.
00:34 Fiesta or heat.
00:34 Fiesta very early-stage brainstorming
00:34 GargantuaSauce well you'd need to be able to cool as effectively as you can heat
00:35 GargantuaSauce and i don't think there are transparent peltiers
00:35 Fiesta one idea was perhaps a combo of a Dilatent fluid mixed with a suspension of quartz crystals. The quartz crystals will vibrate triggering the dilatent fluid around the pixel area.
00:36 GargantuaSauce seems to me there are MEMS pumps these days
00:36 Fiesta not sure howto trigger the crystals though... not too keen to google anything about crystals lest i get lost in new age woowoo
00:36 GargantuaSauce perhaps an array of those piped to pouches
00:37 Fiesta MEMS too complex/expensive for my style. I like to mostly deal with cheap-ass macgyvered things
00:37 GargantuaSauce you could have a very dense array that way i think, potentially fabricated using the same techniques as semiconductors
00:37 GargantuaSauce you aren't going to macgyver a tactile array
00:38 GargantuaSauce unless we're talking like....16 elements max
00:38 Fiesta yeah, it could be done with a semiconductor fab. But i left mine in my other pants
00:38 Fiesta But GargantuaSauce... my visiooooooon, you donp't appreciate my visiiiiiiiion
00:38 Fiesta haha
00:39 rgantuaSauce feels his blood pressure skyro
00:39 Fiesta no but really i may be able to get a lot for nothing. i won't rave about it until it gets done though dont' worry
00:40 Fiesta 1 sec, i'll give you my initial inspiration, it's very impressive.
00:40 Fiesta ...scratch that i can't think of what to search for to re-find it. i'll Just explain what it was:
00:41 Fiesta i saw a team make a tactile sensor outof a camera and a stretchable film covered in dots.
00:41 MrCurious1 hey! i got vision too!
00:41 Fiesta by pressing on the film, it deformed the dot pattern, the computer used typical stereo techniques to figure out the shape of the intruder
00:42 Fiesta it was multi-touch, and could tell not only what shape the pressing object was - but in what direction it was pressing. it was essentially a multi-touch directional force sensor. Incredibly impressive stuff for such low price components
00:42 Fiesta did i explain that well enough for it to make sense?
00:43 GargantuaSauce isnt that exactly the opposite of what you're proposing?
00:43 Fiesta yes
00:43 Fiesta i want to make an actuator outof the principle
00:43 Fiesta or rather a matrix of actuators
00:44 GargantuaSauce i think if you go with something heat-based it will be very unresponsive
00:44 Fiesta i want to make a much more complex version of this https://www.youtube.com/watch?v=lvtfD_rJ2hE
00:45 GargantuaSauce and without active cooling you'll have no way to 'clear' an element without waiting a very long time
00:45 Fiesta yeah
00:45 GargantuaSauce ah yeah i was looking for that video
00:45 Fiesta i'm more keen to go with light, or perhaps even magnetic fields like with e-ink.
00:46 GargantuaSauce wtf would you make physically respond to light!?
00:46 Fiesta that thing in the video has over 900 actuators, which would be a pain in the ass to work with.
00:46 GargantuaSauce i would start with a solution like theirs
00:46 GargantuaSauce a bunch of spring-return solenoids
00:47 Fiesta mentally lazy
00:48 MrCurious1 air bladders and valves?
00:48 GargantuaSauce as opposed to consisting of technologies that don't exist
00:48 Fiesta i'm thinkin of new things. new new new
00:48 Fiesta well yes, why would i invent something that already exists
00:49 Fiesta i'm not sure how that MIT one actually works, it may already be valve based. I've had a gut-full of valves i'm looking for somethign else
00:49 MrCurious1 says its done with motors
00:49 MrCurious1 in the description
00:49 Fiesta urgh, what a pain in teh ass that woudl be
00:50 Fiesta wirign each one
00:50 Fiesta wiring
00:50 MrCurious1 solenoids?
00:50 MrCurious1 custom coils?
00:51 MrCurious1 or smart pegs where they have the motors in them, and catch power from contact with the honeycomb
00:51 Fiesta i'm not sure there's any liquid that reacts mechanically to light :-/
00:51 MrCurious1 also convert it to a hex from a grid
00:51 GargantuaSauce well i'm off to bed...have fun simultaneously breaking new ground in physics and materials science and applying it to a novel user interface
00:51 MrCurious1 how about to sound...
00:51 Fiesta thx Gar
00:51 Fiesta have a good night
00:52 Fiesta sound would be difficult to focus onto pixels
00:52 MrCurious1 perhaps a pizzo element bouncing particles, simulating gas and expanding the presure under the block
00:52 Fiesta doable, but hard
00:52 MrCurious1 individual speekers
00:52 Fiesta but that's kinda what i was thinking with the suspension of quartz crystals mixed with a dilatent fluid. It'd be vibration based, or sound if you will.
00:53 Fiesta 900 individual speakers? nah man, too crazy
00:53 MrCurious1 pizzo elements
00:53 MrCurious1 could wire em using a grid like led's?
00:53 Fiesta g'damn, how powerful would those crystals need to be to be bouncing particles hard enough to increase pressure under the block?!
00:54 Fiesta a dilatent fluid however will lock in resposne to vibration, that could be useful
00:55 Fiesta man is there any sort of thing that is kidna like... a quartz crystal that vibrates in response to light?
00:55 Fiesta short of a crystal with a solar panel attached :(
00:57 MrCurious1 suppose you could use pizzo reeds to control air flow in
00:58 Fiesta hmm...
01:01 Fiesta sudden realisation that i could use an LCD panel as a projector shadow mask. Then blast a super-powerful heat-lamp through it
01:01 Fiesta still i'd need cooling on the other side like Gar said
01:09 Fiesta https://en.wikipedia.org/wiki/Solar-powered_watch hnnnngggg wikipedia you're toying with me giving me these types of article titles
01:09 Fiesta oh whoops
01:32 Fiesta urgh, disconnected inbetween typing something wrong and wanting to correct it
04:59 shiftplusone Would anyone happen to know of a good robotics base that doesn't have a lot of nonsense on it? Just the base and motors without any controller board would be good.
05:07 Centauri heh, build one
05:10 shiftplusone Not really sure what the best... materials, I suppose... are. I am open to suggestions. It needs to have two motors and carry probably about 0.5kg. Where would I start?
05:14 Centauri I built mine from Plexiglas and lexan sheets in a hexagon
05:14 Centauri and stacked
05:14 shiftplusone Would you happen to have a photo handy?
05:29 Fiesta bluhhhh... finding useful photomechanical materials is hard, it's hard and no-one understands.
05:30 Fiesta everything seems based around light-induced heating, and thus gives super-duper slow actuation. and thus i hate everything and i will bite you all.
05:30 Fiesta well not you, you're special
05:39 Centauri shiftplusone: I have plenty.. hang on
05:40 Centauri shift: next from here: http://www.centaurihome.net/gallery2/main.php?g2_itemId=162
05:40 shiftplusone Thanks
05:40 Fiesta this is probably why phtotosynthesis is so complex
05:41 shiftplusone Why are the wheels so massive? Does it provide something or was it just easier that way?
05:42 Centauri faster speeds
05:42 Centauri I get 2 FPM by having wheels that large in diam
05:43 Centauri the hex base is 6 inches on a side, as a reference
05:43 shiftplusone ah
05:43 shiftplusone What's stopping it from tipping forward or back?
05:43 Centauri the circumference fo the wheels is very close to exactly one foot
05:43 Centauri there is a castor wheel
05:43 Centauri its a 3 point system
05:44 shiftplusone Ah, okay, I was just on the first image.
05:44 Centauri keep going and you will see the evolution over 9 months
05:44 Centauri there are some 10-20 pics
05:45 Centauri it ended up being 4 stacks deep
09:56 robotustra hi
18:44 anonnumberanon GargantuaSauce, have you made more videos of your hexapod?
18:44 GargantuaSauce no just the epilepsy one
18:49 GargantuaSauce will have new stuff to show sometime between soon and eventually
20:02 anonnumberanon GargantuaSauce, oh, awesome, I can't wait then.
20:03 anonnumberanon Around christmas I have this project of doing everything I humanly can to fix a laptop's motherboard.
20:42 realigner anyone know of an FRC specific IRC channel?
20:44 Tom_itx frc?
20:44 Tom_itx apparently not
20:44 realigner First Robotics Competition
20:44 Tom_itx you could ask about it in #seattlerobotics
20:44 Tom_itx some of them do first
20:44 realigner perfect, thank you
20:57 MrCurious cyber-week?
22:25 GuShH http://www.lathes.co.uk/petrie/
22:42 MrCurious milling in the 3rd world?
23:01 MrCurious solved the mystery of the failing to wake from slleep chromebook
23:02 MrCurious appears the 2nd ram socket is largely decorative now
23:29 MrCurious any robo-action in the world today?
23:45 MrCurious GuShH: do anything interesting today?