#robotics Logs
Jun 13 2017
#robotics Calendar
12:01 AM rue_house: ok, I changed the motor on the compressor for one I got at work today, its a perfect fit
12:04 AM rue_house: just dosn't have an overload, but its 1.5HP, so as long as its on a 15A circuit, whatever.
12:04 AM z64555: lol
12:09 AM rue_house: well, atleast the channel was buzzing today
12:11 AM z64555: just a few dits and dahs. :D
12:12 AM rue_house: today my camera gimbal controller arrived
12:12 AM rue_house: I have NO idea how long ago I ordered it
12:12 AM z64555: neat
12:13 AM rue_house: I cant even find it in the aliexpress lists
12:16 AM rue_house: oh and the 12mm peizo elements arrived
12:32 AM rue_house: I wonder if a person can feel signal from a 12mm peizo
12:33 AM rue_house: white led + 12mm piezo, hit with pen = light!
12:34 AM z64555: ah, putting out a pretty good voltage, then
12:34 AM rue_house: 2.something there ateast, 3-9ma or so
12:35 AM z64555: what was your idea with the one-element integrater on the piezo again?
12:35 AM rue_house: but I wonder if I pulse from it can be felt
12:35 AM z64555: it involved a transistor, but I forgot the pinout
12:35 AM rue_house: mosfet
12:35 AM rue_house: cause a mosfet has a cap
12:35 AM rue_house: that integrates charge
12:36 AM rue_house: and provides you a resistor to tell you how much charge
12:36 AM rue_house: and it works
12:36 AM rue_house: its not just an idea, I'v done it
12:36 AM rue_house: whats impressive, is that the drift rate is a few minutes
12:37 AM rue_house: but I'm wondering if a known differentiator rate is better, taken into a microcontroller and integrated there
12:37 AM rue_house: I may have to play with that one
12:37 AM z64555: a uC would have less drift, at any rate
12:38 AM rue_house: but would the integration be as accurate
12:38 AM rue_house: it would be time sampled
12:38 AM rue_house: the fet is continious
12:39 AM z64555: it would be as accurate as small as the time sample
12:39 AM z64555: err, if that makes sense
12:39 AM z64555: smaller time sample = more accuracy on sampling
12:40 AM z64555: blah, still not quite right
12:40 AM * z64555 works on his engrish
12:42 AM rue_house: sorry, falling asleep
12:44 AM z64555: The smaller the time sample, the closer the discrete-time signal is to the analog signal
01:11 AM rue_bed: and the smaller the voltage difference, so the greater the percentage error in the reading
03:53 AM Jak_o_Shadows: My PNP transistors rocked up today
03:53 AM Jak_o_Shadows: and the pot knobs
03:54 AM z64555: woo
04:50 AM SpeedEvil is now known as Guest51178
05:35 AM Jak_o_Shadows: Ok. Recovered the mapping from buttons to pins
08:14 AM BitEvil is now known as SpeedEvil
08:51 AM squirrel1 is now known as veverak
09:17 AM rue_house: Jak_o_Shadows, PNP? in a joystick?
09:22 AM theBear: not all joysticks are standard pc 15pin D interface remember ? you know, way waaaay back all them years ago when such things still existed :)
09:23 AM theBear: bumped into my old old arcade-grade/components portable/home joystick (well top half of, as it always been since i had it) that's been waiting and waiting for the right time to be used again... must be pretty close now i think
09:26 AM rue_house: PNP transistors
09:27 AM rue_house: not plug and play
09:27 AM rue_house: hmmm plug and play PNP transistors...
09:28 AM theBear: heh, at least we got P'n'P transistor/comp. identifiey testey thingies now, makes a lot of things easier in life
09:28 AM rue_house: and they will tell you if it happens to be a jfet or a npn
09:35 AM theBear: http://i.imgur.com/cpOjD26.jpg
09:35 AM theBear: or even a pnp <grin> and which pin is which and how much of the most important specs it has
09:36 AM rue_house: theBear, bad ass chain saw!!!!
09:36 AM mbrumlow: welp I got 4 robots on line
09:36 AM mbrumlow: http://robot247.io
09:37 AM rue_house: site dodn't work
09:37 AM rue_house: its blank
09:37 AM mbrumlow: really?
09:37 AM mbrumlow: what browser ?
09:37 AM rue_house: iceweasel
09:37 AM mbrumlow: ohhhhh
09:37 AM theBear: wtf is a iceweasel ?
09:37 AM rue_house: not everyone uses todays release of chromium
09:38 AM mbrumlow: well too bad for now I guess, I used polymer for this site :(
09:38 AM mbrumlow: I am not really a website guy.
09:38 AM theBear: it works in a recentish ffox it seems, slow tho, and obviously waiting on various stuff that slows up the loading/initial display :)
09:39 AM theBear: mbrumlow, heh, between what i just mentioned and the fact it works in at least 2 browsers (i assuming) you already done better than most "website design/programming professionals" out ther e:)
09:39 AM mbrumlow: I tested in chrome and firefox
09:39 AM mbrumlow: but I will get some more test to see what I can do better.
09:39 AM theBear: there ya go ! you doin ok :)
09:39 AM rue_house: I also run debian, so I have nothing newer than 3 years old
09:39 AM * theBear just looked at the actual page and noticed what it was
09:39 AM theBear: that's kinda cool
09:40 AM mbrumlow: I have the entire backend on github
09:40 AM rue_house: (its debian policy to be no less than 3 years out of date)
09:40 AM mbrumlow: thanks theBear
09:40 AM mbrumlow: I need to get better instructions but /help in the chat window should show you how to get to /controls
09:40 AM mbrumlow: which will show you the hotkeys active.
09:40 AM mbrumlow: for moving the robots
09:41 AM theBear: rue_house, i recently did my first ffox update in at least that long, and i can certainly say, a couple of very minor things seem to be improved or fixed, and that the disappointingly as-fashion-demands redesigned gui aspects can be fixed with a minimal fiddle :)
09:41 AM rue_house: are those your office robots?
09:41 AM mbrumlow: yeah
09:41 AM mbrumlow: I reworked a bunch of stuff.
09:41 AM mbrumlow: The intel edison is the one that has the most trouble.
09:42 AM mbrumlow: frozenbubble is just a demo of how flexable the software is.
09:42 AM rue_house: theBear, I block the ad servers at the hosts file
09:42 AM mbrumlow: that is just a VM running a game.
09:42 AM rue_house: 127.0.0.1 ad.doubleclick.net
09:42 AM rue_house: 127.0.0.1 pinterest.com
09:42 AM rue_house: 127.0.0.1 images-cdn.onsidelines.com
09:42 AM rue_house: 127.0.1.1 tpc.googlesyndication.com
09:42 AM rue_house: 127.0.1.1 s0.2mdn.net
09:42 AM rue_house: 127.0.1.1 s1.2mdn.net
09:42 AM rue_house: ...
09:43 AM mbrumlow: rue_house: you see adds on my webiste? I hope not.
09:43 AM rue_house: no, just saying re browsers and ad compatability
09:44 AM theBear: rue_house, i do it at my ns proxy/cachey thinger between me and the wide world of net, at least the worst ones, and of course a handful of others that offend me morally, like various apple "spy" servers that would be preying on friends with apple anythings visiting and using my wifi, which i just feel is kinda uncool, ya know, since they paid a premium for a supposedly better hard and software machine
09:44 AM rue_house: I have apache on my machine with a 404 page that serves a cat rolled up in paper
09:45 AM rue_house: these days I see lots of cats rolled up in paper
09:46 AM rue_house: mbrumlow, oh hey, wait, I'm also blocking a huge list of ipv6 isp addresses, are you on an ipv6?
09:48 AM mbrumlow: I am on ipv6
09:48 AM mbrumlow: wait
09:48 AM mbrumlow: not for that domain.
09:48 AM rue_house: http://paste.debian.net/971315/
09:48 AM rue_house: any of those?
09:49 AM rue_house: I think most of those are chineese
09:49 AM theBear: rue_house, oh, those ones you collected personally (or via some auto-collector maybe ?) ?
09:49 AM rue_house: going thru ssh logs of password attacks
09:49 AM rue_house: I would just isolate the network and block the whole thing
09:50 AM rue_house: my ssh attacks are few and far between now
09:50 AM theBear: mmmm, i got a list, but just right now a bit hard to get to, unless, hmm
09:51 AM theBear: hmm, lotta same higher-class networks as your list, but a glance (literally) shows no matches
09:52 AM mbrumlow: 2604:e880:0:1e::2/64 <- is the only one I use right now
09:52 AM theBear: mostly numbers i hand-picked from passing tail'ed logs, often mail or web attacks, i find ssh in recenter years tend to be distributed amongst ones assumes a smurfnet
09:52 AM theBear: only one what ?
09:52 AM theBear: i don't think i can access v6 anyway
09:53 AM rue_house: theBear, that why I blocked the network, they cant move a machine outside of that
09:53 AM mbrumlow: I use only one ip in some huge range.
09:54 AM rue_house: yea
09:54 AM rue_house: some of those networks are isp's who gave everyone exploitable routers
09:55 AM rue_house: and guess what happened?
09:55 AM theBear: rue_house, what i shoulda made clearer is that the ones i paid attention to (just cos i was looking at the right time, no special treatment) seem to be VERY widely varied addresses for the distributed attacks, from one assumes windows machines (we can all fill in the gaps <grin>)
09:55 AM rue_house: one day I need to steal a botnet...
09:55 AM theBear: rue_house, ahh, never thought of it from isp-provided-blah pov
09:55 AM theBear: rue_house, hehe
09:57 AM rue_house: off to work, bye!
09:57 AM theBear: toodles :)
10:05 AM mbrumlow: I have a 3d printer now. so I hope to make a multi leg robot (a real one this time)
10:05 AM mbrumlow: and put it on line too.
10:07 AM theBear: nice work, get 'er done !
10:12 AM z64555: how many legs do you plan to start with?
10:13 AM theBear: two, then drink until leg-less <grin>
10:14 AM mbrumlow: lol
10:14 AM mbrumlow: I was thinking of 5
10:14 AM mbrumlow: not sure if that is a good start.
10:14 AM mbrumlow: but I figure the more legs the more stable at first.
10:16 AM theBear: mmm, look at a baby cow or giraffe vs a human baby, and that's only 4 vs 2
10:16 AM theBear: and also not only unscientific, but blatently misleading
10:27 AM z64555: don't use 5
10:28 AM z64555: Probably better to start off with 4, since it can be both in axial and in rectangular geometries
10:29 AM mbrumlow: I will keep that in mind.
10:29 AM z64555: 5 isn't recommended because there isn't an easy path solution
10:30 AM z64555: 6 is another common leg number, and I've seen them in both configurations, as well
10:31 AM z64555: You need at least 3 feet in contact with the ground, always, unless your legs can move fast enough to catch the carriage before it tips over
10:32 AM z64555: or have another mechanism on the body to keep its balance on 2 or fewer legs
10:34 AM z64555: As a visual aid, you can draw a polygon between the contact areas of the feet. Then, form a prism by making edges that are normal to the gravity vector
10:36 AM z64555: If the bot's center of mass is within that prism, it's stable
10:36 AM theBear: i like that little trick
10:36 AM theBear: handy
10:37 AM z64555: yup
10:37 AM z64555: The closer the CoM is to the centroid normal of the contact prism, the more stable the bot is
10:38 AM z64555: or, said in another way, the more distance away from an edge (while still instide the area) the better
10:40 AM theBear: indeed, which is kinda intuitive/basic balance concepts, just that your trick lays it out in kinda lowest common denominator simplicity
10:40 AM theBear: so simple you don't need any thinking really (not that i can advocate not-thinking, i see a lot of it, and it all looks bad :] )
10:41 AM z64555: thanks
10:46 AM z64555: The trick can also be extended to include velocity, momentum
10:46 AM z64555: I'll leave that for discovery. :P
10:52 AM * theBear 's head happily extrapolates the various extensions without hassles
10:52 AM theBear: mmmm nice
12:47 PM theBear: ooh crap, sorry if i been hassling yer server or robot portals or anything, i just noticed that the page was still sitting there idling, and for bonus points sounded like it was kinda sleeping/waking for some reason while the tab wasn't the visible one, as if it were trying to be quiet until i got back to it, just failing now and again :)
01:45 PM z64555: yep... this electric screwdriver is trash
01:46 PM z64555: battery holder reinforced the tabs, so it was impossible to take out without the use of two slotted screwdrivers
01:46 PM z64555: Thus, the original batteries were still in there. Coupled with 5+ years of disuse, the batteries exploded, got salt everywhere and corroded the contacts
01:46 PM z64555: and also corroded the switch
01:50 PM z64555: it has a 12:1 planetary gearbox
01:51 PM theBear: mmm, opposite of how galvanised steel/iron works, corrodes as far as it conducts on one terminal
01:52 PM z64555: no, it's the same principle
01:52 PM z64555: but the battery acid acted as a catalyst
01:53 PM z64555: or, at the very least, a better electrolyte than humid air
01:53 PM theBear: same i guess, just opposite results (kills something you kinda want, not something consumable)
01:54 PM z64555: speaking of which, I noticed that the new cars are going back to negative ground
01:54 PM theBear: they were pos ground ? only pos ground cars i know of are ancient/weirdo ones
01:55 PM z64555: they were positive ground up through the 90's
01:55 PM z64555: I think they went to negative ground when electric/hybrid cars started to show up again
01:55 PM theBear: yeah ? hmm, jap and landrovers (at least some) definately weren't, and my 03 corolla isn't
01:55 PM theBear: isn't pos ground i mean
01:56 PM theBear: and pretty sure dads volvo was neg chassis... and every car stereo i ever poked at which must be some from the 90s
02:01 PM veverak: rue_house: you said your controller can work with up to 8 servos?
02:01 PM veverak: so stewart platform should be doable
02:03 PM z64555: well, this motor might still be usable
02:03 PM z64555: as well as the gearbox
02:15 PM logicmoo is now known as dmles
02:15 PM dmles is now known as dmioles
02:38 PM dmioles is now known as dmiles
04:17 PM anniepoo: cars stopped being pos ground in the 60's. IIRC VW bugs were pos ground until early 60's
04:46 PM Snert_: right. But why did they ever think that pos ground was the way to go?
05:01 PM LoRez: ttp://lajagclub.com/the-argument-for-positive-ground/
05:01 PM LoRez: grrr... add an h
05:02 PM LoRez: TL;DR corrosion on "leaky cloth covered wires" was reduced with +GND
05:03 PM LoRez: Same reason telcoms used -48V. avoid galvanic effects in telephone cabling.
05:04 PM Snert_: I see.
05:05 PM Snert_: so evidently, wiring insulation got better and better galvanic techniques made it possible to switch to neg grount.
05:06 PM LoRez: according to that article, the inuslation and NPN transistors in transistor radios made the switch to -GND a thing.
05:07 PM Snert_: yes. interesting.
05:09 PM Snert_: without npn why, we'd all just be in a pnp world.
05:09 PM SpeedEvil: From memory, common transistors were typically PNP in that age
05:09 PM SpeedEvil: germanium, PNP was more common
05:10 PM SpeedEvil: I may be misremembering, I never paid much attention, it seems like a very questionable reason over the other ones
05:14 PM LoRez: First one made was an NPN. Bell Labs
05:15 PM LoRez: Apparently not... 1947 a PNP was made
05:16 PM Snert_: somebody said "what happens if we dope it inverse? pnp was born.
05:17 PM Snert_: but I bet the math bore the theory out long before, actually.
05:17 PM LoRez: https://en.wikipedia.org/wiki/History_of_the_transistor is interesting
05:26 PM SpeedEvil: I mean practical transistors, PNP was used lots more than NPN
05:26 PM SpeedEvil: As the properties were better for germanium transistors in the 60s-70s
07:34 PM rue_house: veverak, 8 servos per controller, 4 controllers per serial bus
07:34 PM rue_house: high resolution PERFECT timing
07:37 PM rue_house: you will not get high resolution
07:37 PM rue_house: well, my stm32 goes higher, but only does 1 servo right now
07:38 PM rue_house: over 15 bits accuracy with the stm32
07:39 PM rue_house: SpeedEvil, if you look at all the old books, its all PNP
07:39 PM rue_house: their world revolved around pnp transistors
07:39 PM SpeedEvil: yes, I know
07:39 PM rue_house: makes it a pain to use the cirucits, between the transformer coupling, the pnp and the split supplies
07:40 PM rue_house: my but is getting stuck in my chair
07:40 PM rue_house: I can feel it fusing on
07:40 PM SpeedEvil: err - you simply swap the supplies, the polarity of any electrolytics/diodes, and swap npn for pnp, and you're done
07:41 PM rue_house: heh, yea
07:42 PM rue_house: hey, you got a newish smartphone?
07:43 PM rue_house: what size phone is a google cardboard meant for?
07:46 PM rue_house: awe now my eyes are starting to snap shut, arg
08:08 PM Tom_L: google cardboard?
08:08 PM Tom_L: wtf is that?
08:09 PM Tom_L: i have an s6 & s8
08:09 PM Tom_L: about the same size really
08:09 PM Tom_L: s8 has wraparound screen
08:10 PM Tom_L: maybe a bit longer
08:23 PM rue_house: you dont know about google cardboard?
08:24 PM ace4016: an interesting take on affordable VR headset
08:25 PM rue_house: https://www.aliexpress.com/item/7-Inches-High-Resolution-1024x600-Screen-Display-LCD-TFT-Monitor-with-Remote-Driver-Control-Board-2AV/32563394364.html
08:25 PM rue_house: mmmmm soooooo tempting...
08:33 PM z64555: :|
08:34 PM z64555: nobody does 4:3 anymore, it seems
08:36 PM rue_house: ok I need to go string the weed eater
08:36 PM rue_house: hopefully it will snag me into doing lots of tings and keep me from going to sleeop
09:29 PM zhanx: so i had to spend my robot money on new boots. but.. i still found somethings to use in the robots
09:35 PM z64555: robot boots?
09:36 PM zhanx: no, if you remember i have had several new right knees and they wont give an a bionic one. so new boots are lighter so as to put less stress when i walk
09:38 PM rue_shop3: :/ the gear just fell off my extruder motor
09:39 PM rue_shop3: well, thats is to say, half of it did
09:40 PM anniepoo: oh noes!
09:41 PM rue_shop3: I smart, I printed two
09:41 PM rue_shop3: but I also have another printer
09:59 PM rue_shop3: ok , new gear is on...
10:02 PM z64555: zhanx: ah
10:02 PM z64555: no robot boots :(
11:19 PM anniepoo: WHOO http://partyserver.rocks/anniepoo/theshell.jpg
11:19 PM anniepoo: The shell is FINISHED!
11:19 PM anniepoo: http://partyserver.rocks/anniepoo/theshell2.jpg
11:19 PM anniepoo: I am SOoooo cranked!
11:20 PM * z64555 applauds
11:20 PM z64555: http://media.giphy.com/media/xLJE238iO1Cta/giphy.gif
11:21 PM z64555: I think that's the right one, internet's a bit slow at the moment for me to check
11:21 PM rue_shop3: the sense of touch goes from about 20Hz to 1.5Khz, with a peak at about 250Hz
11:22 PM rue_shop3: but, there is a strange alias of the freq going on, its like, you can feel it, but there is no way you could be feeling that actual freq
11:23 PM rue_shop3: anniepoo, nice!
11:23 PM anniepoo: put a speaker against it, and measure the wave off the speaker
11:23 PM anniepoo: 8cD I'm unreasonably proud of my shell
11:23 PM rue_shop3: I'm holding a piezo
11:23 PM rue_shop3: you should be, its been a lot of work
11:23 PM anniepoo: that outta work
11:23 PM rue_shop3: what finial patten you gonna go with?
11:23 PM anniepoo: yes, it's been a crazy lotta work
11:23 PM anniepoo: pattern?
11:24 PM z64555: rue_shop3: harmonics, perhaps
11:24 PM anniepoo: I decided on uniform brown
11:24 PM rue_shop3: they are uaually striped from the layers of adding shell
11:24 PM anniepoo: yes, but Pomatia's a robot snail
11:24 PM anniepoo: I did think about giving her a pattern
11:24 PM z64555: could just add some black or darker lines of brown for texture
11:24 PM rue_shop3: http://christchurch247.co.uk/wp-content/uploads/2014/07/snail-shell.jpg
11:25 PM anniepoo: I like how it looks now
11:25 PM rue_shop3: you know you can get more complex than that :)
11:25 PM z64555: hm, lemme look at it agin
11:25 PM rue_shop3: we can draw this out to take YEARS
11:25 PM anniepoo: lol
11:25 PM rue_shop3: :)
11:25 PM anniepoo: well, it's done
11:25 PM anniepoo: now thinking about the body
11:25 PM rue_shop3: https://s-media-cache-ak0.pinimg.com/736x/08/ce/83/08ce83a5c1b76c5730fad35ce979a99b.jpg
11:25 PM z64555: could stratch the lines to make it more defined
11:25 PM rue_shop3: .ooo i gpt dizzy
11:26 PM z64555: but is otherwise good and usable :D
11:26 PM anniepoo: that's pretty
11:26 PM anniepoo: but I think I'm happy with what I have
11:27 PM rue_shop3: http://cdn.earthporm.com/wp-content/uploads/2014/10/painted-snail-shell-23.jpg its all good till some assho** with a paint can outruns you
11:28 PM rue_shop3: https://peopledotcom.files.wordpress.com/2016/11/snail1.jpg?w=2000&h=1333 "WHERE IS IT!???? I STILL CANT SEE!"
11:29 PM anniepoo: lol
11:29 PM anniepoo: poor snail, bet that's not good for it
11:31 PM anniepoo: (first pic)
11:31 PM anniepoo: kinda like the bright red with polka dots
11:31 PM rue_house: so, i ahve a problem
11:31 PM anniepoo: just one????
11:31 PM rue_house: I want to use a peizo to sensse pressure from an operator
11:31 PM anniepoo: ok
11:31 PM rue_house: and I want to use a peizo to vibrate an operator to tell them how much pressure the recieving end is experiencing
11:32 PM anniepoo: ok
11:32 PM rue_house: but a) they are not compatible modes on the same peize
11:32 PM rue_house: b) they would screw with each other if put side by side with two peizos
11:32 PM anniepoo: mount each one on a small steel plate
11:33 PM rue_house: users fingertip...
11:33 PM z64555: so uh
11:33 PM anniepoo: mount each plate on a block of pink insulation foam
11:33 PM anniepoo: 8cD
11:33 PM rue_house: aka, force feedback grip ball
11:33 PM z64555: I guess people's fingers can't sense how much pressure they're pushing
11:34 PM rue_house: z64555, they can, but I need to tell the user how much force the robot is pushing
11:34 PM anniepoo: not quite understanding the morphology here
11:34 PM z64555: ah, ok
11:34 PM z64555: it's a user interface for the robot, then
11:34 PM anniepoo: haptic interface
11:34 PM rue_house: telecontroller type A: you move the joint, in real space, and a motor pushes back against you for force feedback
11:34 PM anniepoo: these are often designed with small permag motors for feedback
11:35 PM z64555: I'd use an LED
11:35 PM anniepoo: I'd use haptics
11:35 PM z64555: haptics interfere with the piezo
11:35 PM rue_house: telecontroller type B: you push against a static plate, thats vibration amplitude changes with the force feedback, but the plate never actaully moves
11:35 PM z64555: You already get a degree of haptic from pushing down on the button
11:35 PM anniepoo: ok
11:36 PM z64555: but that's just a reference from min/max
11:36 PM anniepoo: whats the piezo for?
11:36 PM rue_house: this isn't about knowing your force, its about knowing the force on the 'gripper' of the robot
11:36 PM z64555: yes
11:36 PM anniepoo: ok, two sep. problems
11:36 PM rue_house: piezo is for force sensing and vibration feedback
11:36 PM rue_house: yes and no
11:36 PM anniepoo: 1. measure the force on the gripper
11:36 PM z64555: that's not the prob
11:36 PM rue_house: the gripper dosn't matter
11:36 PM rue_house: thats easy
11:36 PM anniepoo: 2. provide that feedback to user
11:37 PM z64555: it's the UI that he's having trouble with
11:37 PM rue_house: I need the bidirectional interface with the human
11:37 PM anniepoo: ok
11:37 PM z64555: The input sensor cannot dual-function as a feedback device
11:37 PM rue_house: you dont need to move a joint
11:37 PM z64555: and the buzzer feedback device interfers with the input
11:37 PM rue_house: you can have the user apply force, not position, apparently they are interchangable
11:37 PM z64555: unlesss...
11:37 PM z64555: rue
11:37 PM z64555: filter
11:37 PM anniepoo: actually, it can
11:38 PM anniepoo: ok, I'm lost again
11:38 PM z64555: the output to the buzzer can be fed into a filter on the input
11:38 PM anniepoo: not sure how this buzzer's involved
11:38 PM anniepoo: feel like I've missed some critical info
11:38 PM anniepoo: what are we making?
11:38 PM z64555: the filter will then ignore any AC of that frequency
11:38 PM rue_house: z64555, the force signal from the peizo is in the order of like 2hz, but its a really sensivtive dc level
11:38 PM rue_house: anniepoo, 12' servomecha
11:39 PM anniepoo: ok
11:39 PM anniepoo: so I move a joint, I want operator to get appropriate force feedback proportional to force on the joint?
11:39 PM rue_house: yup
11:40 PM rue_house: but, part of the thing, is that unlike...
11:40 PM anniepoo: how's any of that related to vibration?
11:40 PM z64555: his feedback mode is vibration
11:40 PM rue_house: there are two substitutions
11:40 PM z64555: unlike haptic feedback on a joystick, which alters the position
11:40 PM rue_house: one is that you dont need to move the controller, you just varry the force against it
11:40 PM anniepoo: well, you can apply force to controller without moving it
11:41 PM anniepoo: think about race car game wheels
11:41 PM rue_house: two, is that you dont get force pushing back on you, you get a vibration of varrying amplitude
11:41 PM anniepoo: they do this
11:41 PM rue_house: yea, I jknow
11:41 PM anniepoo: they give you force feedback
11:41 PM rue_house: http://ruemohr.org/~ircjunk/robots/armcontroller/p1070088.jpg
11:41 PM rue_house: it dosn't need to be one of thsoe
11:41 PM rue_house: sorry this keybaord is raelyl not working for me layley
11:42 PM rue_house: its like its getitng smaller
11:42 PM z64555: I have a similar issue, my s key doesn't register sometimes
11:42 PM rue_house: the spacing on the model M must be better
11:42 PM rue_house: its like this thing is overall about an inch too narrow
11:42 PM z64555: not sure if that's do to loss in feeling of my left ring finger, or if there's some junk in between the key
11:43 PM rue_house: anyhow, yea
11:43 PM rue_house: I have the force feedback thing for a servo all worked out with the stm32
11:43 PM rue_house: I need to get more servos going at the same time, but I'd like to have a controller like this set up
11:43 PM rue_house: for gripper anyhow
11:44 PM rue_house: then I can motivate myself to finish the mecha gripper
11:44 PM rue_house: which I think I worked out the last few issues with
11:44 PM anniepoo: haptic feedback devices are often a ball on the end of a delta robot
11:44 PM rue_house: http://ruemohr.org/~ircjunk/robots/sparrow1/p1080426.jpg
11:44 PM anniepoo: http://gearmedia.ign.com/gear/image/article/771/771358/gdc-2007-novint-falcon-haptic-feedback-pc-controller-20070307071900430-000.jpg
11:45 PM rue_house: I need a whole hand
11:45 PM rue_house: well, a thumb and 2 fingers
11:45 PM anniepoo: lots of servos?
11:45 PM rue_house: http://ruemohr.org/~ircjunk/robots/sparrow1/p1030901.jpg
11:45 PM rue_house: ... you not been folling me the past few years? air muscles
11:46 PM anniepoo: air muscles works
11:46 PM anniepoo: this is the mechas hand?
11:46 PM rue_house: yea
11:46 PM rue_house: thats how big the hand is of a 12' robot
11:47 PM rue_house: http://ruemohr.org/~ircjunk/robots/sparrow1/p1040015.jpg
11:47 PM rue_house: http://ruemohr.org/~ircjunk/robots/sparrow1/p1040018.jpg
11:47 PM rue_house: a point of amusment for me, my head clears its crotch
11:47 PM rue_house: :)
11:47 PM rue_house: its muuuuch smaller laying down
11:48 PM anniepoo: is that a mockup?
11:48 PM rue_house: yea
11:48 PM rue_house: and no
11:48 PM rue_house: the idea is 'thats the robot' and that I ahve to upgrade it to the working design
11:48 PM rue_house: the pipe is the same size as the steel I have
11:49 PM rue_house: 8-|
11:50 PM rue_house: "but I hve a problem cause I cant weld the ABS into the shapes I need to make steel versions of...UNLESS I MADE SOME KINDA HUGE MEAN 3D PRINTER...."
11:50 PM rue_house: I'm talking the kinda machine you dump a barrel of pellets into
11:55 PM anniepoo: so the final version will be steel pipe?
11:55 PM rue_house: yes and no, not as much as I thought origionally
11:55 PM rue_house: but all steel yes
11:56 PM anniepoo: consider steel studs
11:56 PM rue_house: I'm hoping less than 1.5 tons
11:56 PM anniepoo: I'm making (slowly) a powder coating oven from them
11:56 PM rue_house: na, steel studs are tinfoil
11:56 PM anniepoo: I bought some commercial ones
11:56 PM rue_house: :)
11:56 PM rue_house: have a sprayer?
11:56 PM anniepoo: def. not tinfoil
11:56 PM anniepoo: yes
11:56 PM anniepoo: have a small setup
11:56 PM anniepoo: making a closet sized oven
11:57 PM rue_house: can you shoot me images of it so I can make one?
11:57 PM anniepoo: I can send you CAD drawings
11:57 PM rue_house: ?
11:57 PM rue_house: the sprayer, not the oven
11:57 PM anniepoo: oh, hmm...
11:57 PM anniepoo: can't tell much from the outside
11:58 PM anniepoo: Kevin bought his own, we think that one's better
11:58 PM rue_house: do you ever have to strip it dow to cleam it?
11:58 PM anniepoo: hang on
11:58 PM anniepoo: no, never
11:58 PM anniepoo: just blow it out with compressed air
11:58 PM rue_house: ah
11:58 PM anniepoo: much easier than paint
11:58 PM rue_house: is it as messy as toner?
11:58 PM anniepoo: it is if you spill it
11:59 PM rue_house: does it have the same melting point as toner?
11:59 PM anniepoo: I don't know
11:59 PM anniepoo: what's mp of toner?
11:59 PM * rue_house glances at the HUGS pile of partly empty toner catridges
11:59 PM rue_house: iirc about 150c
11:59 PM anniepoo: this is around 450F
11:59 PM anniepoo: and needs baked 20 mins
11:59 PM rue_house: oh, degrees vs degrees