Back
[00:02:16] <anonnumberanon> I tried to do things but the weather stopped me in my tracks.
[00:02:40] <anonnumberanon> Gonna be -5C all throughout the day too... :(
[00:05:24] <Snert> ok ppl....ready for a serious question?
[00:07:01] <Jak_o_Shadows> always
[00:08:34] <Snert> prepare for a squirt of text.......
[00:08:44] <Snert> I need a linear gray code strip. For an extensor for a robot forearm.
[00:08:44] <Snert> So I can read what position it's extended to.
[00:08:44] <Snert> .
[00:08:46] <Snert> So print 2 code strips on laser transparency sheets.
[00:08:48] <Snert> (one normal, one mirrored - using 2 transparency sheets)
[00:08:50] <Snert> Put them back2back. Glue them together with what?
[00:08:52] <Snert> (what kind of glue won't yellow or haze ???)
[00:08:54] <Snert> .
[00:08:56] <Snert> Then sandwich the code strip between 2 microscope slides.
[00:08:58] <Snert> Mount the code where I need it so I can know the extensor's position.
[00:09:00] <Snert> Anyone think that the light will diffuse sideways through the microscope slides
[00:09:02] <Snert> and screw up the optical detectors?
[00:09:43] <Jak_o_Shadows> why 2 strips?
[00:10:07] <Snert> they are mirrored ...that way more solid black.
[00:10:29] <Snert> you know, blacker black.
[00:10:40] <Jak_o_Shadows> ah.
[00:11:55] <Snert> might not be actually needed but similiar examples I've seen do the mirrored thing.
[00:12:13] <Snert> but those were gray code wheels, not linear strips.
[00:22:01] <rue_bed> just a reflective encoder
[00:22:54] <Snert> let me find the example I'm going on.....
[00:26:59] <Snert> damnit cant find it now.
[00:27:48] <Snert> but it's 2 laser transparencies back to back like a wheel and optical detecrots and the light does pass throo the wheel.
[00:28:06] <Snert> so it's not reflective.
[00:28:16] <Snert> but would reflective be enough?
[00:29:13] <Snert> seems harder to squirt the light correctly if the photodetectors are in the way.
[00:45:32] <anonnumberanon> Snert, what's your project?
[00:46:03] <Snert> extensor for a robot arm... forearm.
[00:46:55] <Snert> maybe 6" of travel.
[00:48:28] <Snert> 2-56 or 4-40 threaded brass rod driven by a mini motor.
[00:48:40] <anonnumberanon> just one extensor or many?
[00:48:54] <anonnumberanon> like, a robot muscle?
[00:49:00] <Snert> maybe 2....not fingers.
[00:49:10] <Snert> so not 5 or 6
[00:49:36] <Jak_o_Shadows> my pots arrived
[00:49:39] <anonnumberanon> You gotta think of the productivity of the project. Time/money, etc.
[00:49:41] <Snert> just 2 square metal tubes one slides inside the other.
[00:49:43] <anonnumberanon> what kinda pots?
[00:50:02] <anonnumberanon> if you're only making 2, might as well not spend too much time engineering them?
[00:50:35] <anonnumberanon> dunno, it's really different if you're doing it as a hobby and unique piece that is not really portable to another project later on
[00:50:53] <anonnumberanon> I'm giving opinions you probably don't want to hear lol sorry.
[00:50:59] <Snert> no biggee.
[00:51:14] <Snert> I'm not happy with what I can get for $2500.00
[00:51:26] <Snert> I don't expect mine to cost less.
[00:51:40] <Snert> but it will be what I want made from high quality shit.
[00:52:26] <anonnumberanon> Oh waw, with that money it has the potential to be really nice.
[00:52:53] <anonnumberanon> that's a good budget for buying tools to make it exactly how you want
[00:53:07] <anonnumberanon> and then you can remake it for cheap if your investment is only in tooling
[00:53:37] <anonnumberanon> How small form factor are you trying to get to?
[00:54:13] <Jak_o_Shadows> The rotary kind. 100K
[00:55:26] <anonnumberanon> what is that?
[00:55:32] <anonnumberanon> factory assembly arms?
[00:55:52] <Jak_o_Shadows> Gonna be a stewart platform
[00:59:02] <rue_bed> ball screws?
[00:59:25] <anonnumberanon> oh iv seen combat robots in a tv show that had that sort of stuff
[00:59:37] <anonnumberanon> one of the guys on the show is in this channel
[01:00:12] <Jak_o_Shadows> Dunno what I'm going to do about the joints
[01:00:26] <Jak_o_Shadows> Want a spherical joint and a UV jont
[01:01:37] <anonnumberanon> rue_bed, can we use excavator actuators for rapid motion of lightweight robotic machines?
[01:02:07] <anonnumberanon> I was trying to see how fast an excavator's pneumatics actually are the otherdays.
[01:05:58] <Snert> you talking hydraulic arm?
[01:06:21] <Snert> thats gonna be big
[01:07:11] <Snert> wheelchairs and bed lifts have some good screwjack type extensors and hardware.
[01:08:19] <Snert> http://www.linak.com.au/corporate/pdf/english/data%20sheet/linear%20actuator_la31h_data%20sheet_eng.pdf
[01:10:21] <Jak_o_Shadows> Does aybody have any ideas for a cheap alternative to
http://www.aliexpress.com/item/Bolt-Through-Ball-model-plane-accessories/1238898532.html?spm=2114.01020208.3.1.ugpthS&ws_ab_test=searchweb201556_1,searchweb201644_3_10001_10002_10005_10006_10003_10004_62,searchweb201560_2,searchweb1451318400_6148 ?
[01:13:22] <Snert> cheaper than 34cents each?
[01:14:37] <Snert> check hobbyking I guess.
[01:14:56] <Jak_o_Shadows> The shipping is killing me
[01:15:40] <Snert> those sure look like 500-600 size heli ball rod ends.
[01:16:13] <Jak_o_Shadows> I don't need exactly them: Something similarish though
[01:16:19] <Jak_o_Shadows> I need 10 UV joints
[01:16:26] <Jak_o_Shadows> Well, preferably 5 UV joints and 5 spherical
[01:17:29] <Snert> since you have to bend over on shipping anyways....spend more and get not plastic.
[01:17:35] <Snert> http://www.hobbyking.com/hobbyking/store/RC_PRODUCT_SEARCH.asp?strSearch=ball+joints
[02:51:15] <anonnumberanon> jhylands, is this junk? I don't actually have time to build my own.
http://www.ebay.com/itm/T962-Reflow-Oven-Infrared-IC-Heater-Visual-Operation-Micro-Computer-Setup-New-/260799459468?hash=item3cb8dbf08c:g:fgYAAOSw0e9UwJfC
[02:53:22] <anonnumberanon> https://www.youtube.com/watch?v=5eQB83aufkg
[03:44:07] <hypodyne> howdy
[03:45:01] <deshipu> hi
[04:34:15] <RyanS> why do brushless rc motors never specify torque?
[04:48:52] <deshipu> never?
[04:49:41] <RyanS> just Kv, Amps, turns
[04:51:11] <deshipu> if you have the power and the speed, you can calculate the torque
[06:14:09] <LiohAu> SpeedEvil hey
[06:17:41] <deshipu> http://paste.sheep.art.pl/a310a825-6c30-481d-a99f-56c529f21246/%2Binline
[06:43:32] <jhylands> anonnumberanon, I have no idea
[06:43:47] <jhylands> according to the video guy, it has a lot of bad reviews
[06:43:54] <jhylands> I've never looked into them
[06:45:26] <jhylands> might be worth reading this:
http://www.instructables.com/id/T962A-SMD-Reflow-Oven-FixHack/
[06:47:21] <jhylands> https://github.com/UnifiedEngineering/T-962-improvements
[06:49:04] <jhylands> So it looks to me like it would be a good purchase if you're willing to spend a little time and effort to fix the obvious flaws
[09:14:23] <rue_house> hahahah
[09:14:39] <rue_house> from the USA, does not ship to canada, comes complete with an EU plug on it
[09:14:41] <rue_house> hhahahaah
[09:36:22] <adam789654123> how clever
[09:36:38] <adam789654123> perhaps they only accept yen?
[10:14:11] <rue_house> my van cant go anywhere in this snow
[10:16:02] <robopal> thats ok
[10:38:34] <deshipu> I think this might be the stupidest mistake I made on a pcb so far
[10:39:23] <rue_house> execting it to come out right the first time?
[10:40:07] <rue_house> thats why etching it yourself at home is nice, didn't just pay $50 for having $60 of baords made that are all wrong
[10:40:41] <rue_house> iirc I'm at 5 or 10c/sq inch at home
[10:40:43] <deshipu> well, it's $14 and it's repearable
[10:41:02] <deshipu> I basically shifted all servo sockets by one
[10:41:43] <deshipu> so it's power-signal-ground instead of ground-power-signal
[10:42:33] <deshipu> but since they are in a row one next to the other, I just need to plug them in differently, and provide ground for the plug that sticks out
[10:42:57] <deshipu> still embarassing
[10:44:21] <veverak> happens :)
[11:51:25] <Snert_> http://www.wired.com/wp-content/uploads/2016/01/CES-unveiled-07.jpg
[13:12:05] <maiden> hello, back from the dead ;)
[13:13:37] <robotustra> maiden: hi
[13:15:54] <maiden> have been overloaded with work, but now finally i am on parental leave and have some time for robotics :D
[13:16:30] <maiden> have just got the basics for my load cells to work :D
[13:17:13] <maiden> now that i have verified that it works good enugh, i will start to design the feet with integrated load cells :D
[13:20:30] <robotustra> what is load cell?
[13:21:46] <robotustra> судды
[13:21:46] <femtogram> maiden: what do you do for work?
[13:21:52] <robotustra> cells
[13:22:25] <robotustra> he is doing embedded stuff
[13:22:43] <robotustra> like gps trackers
[13:22:47] <femtogram> ahh, i see
[13:25:32] <maiden> http://www.ebay.com/itm/121635444344?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
[13:25:56] <maiden> and
[13:25:58] <maiden> http://www.ebay.com/itm/351107338114?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
[13:27:15] <maiden> i am going to measure 4x points in z, 2x in x and 2x in y, they work for both positive and negative load.
[13:28:29] <maiden> with this in the feet i can adapt the foots angles to uneven terrain and detect impacts as well as getting the center of mass
[13:29:57] <maiden> see
http://www.jsk.t.u-tokyo.ac.jp/~report/proceeding02/icra02/nishiwaki_forcesensor.pdf
[13:30:15] <robotustra> nice
[13:30:48] <robotustra> I did once a load cell but I rip it from usual kitchen scale
[13:32:20] <maiden> yea, thought of that too, but i need several with the same specs ;)
[13:32:28] <robotustra> http://ic.pics.livejournal.com/maholet/24765393/1211/1211_900.jpg
[13:32:42] <robotustra> but this one only for 5 kg
[13:33:22] <robotustra> http://pics.livejournal.com/maholet/pic/00005rzx/s640x480
[13:33:36] <robotustra> and your ADC is much better
[13:33:40] <maiden> sweet
[13:33:48] <robotustra> I had only 1024 points
[13:34:32] <robotustra> may be I also order some sensors like this
[13:34:33] <maiden> they are built to be used in pairs, but you can use them one by one as well, but then you do not get the same resulution
[13:35:01] <robotustra> I can live with that
[13:35:23] <maiden> see:
https://www.youtube.com/watch?v=0d_xWLeBPRk
[13:36:35] <maiden> btw ordered a delta 3d printer, will get it in beginning of feb.
[13:36:54] <robotustra> nice
[13:37:04] <robotustra> I'm waiting for my lifepo4
[13:37:17] <maiden> now i get a bit more freedom with my designs ;) but it will steal a lot of time from me :p
[13:37:36] <maiden> http://www.ebay.com/itm/331397170359?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%Folger%20Tech%20Kossel%202020%20RepRap3AIT
[13:37:44] <robotustra> how many sensors are you going to use per leg?
[13:38:13] <maiden> as i wrote before, 8/foot
[13:38:29] <maiden> 4x z, 2x y, 2x x
[13:38:34] <robotustra> hm
[13:38:59] <robotustra> ah, side sifts too?
[13:39:04] <robotustra> ok
[13:39:37] <robotustra> I would use 5
[13:40:51] <robotustra> or you can't use them push-pull?
[13:43:30] <maiden> i can use them for positive and negative load
[13:44:22] <robotustra> are you going to measure twist of foot?
[13:45:13] <robotustra> I would rather see your foot design
[13:47:47] <maiden> so i will begin with somthing like a rectangular box, where the top is 3 sided and the bottom are 3 sided, then the load cells is what keeps the foot together, while it can "slide" freely in x and y (while fixed to the load cells)
[13:48:12] <maiden> to be able to measure x and y load
[13:48:34] <maiden> i do not have a fancy 2axis load cell as in the article i linked to.
[14:06:43] <robotustra> maiden: you want to make one?
[14:07:06] * robotustra music offtop https://www.youtube.com/watch?v=BI7wnIfBwxk
[14:10:04] <maiden> 86robotustra, no, i want to use as much cheap, off the shelf parts as possible
[14:11:27] <maiden> its faster and cheaper that way, when i got something solid or need to improve, then i will go for more custom parts.
[14:11:59] <robotustra> I mean that 8 sensors -is too much
[14:12:06] <SpeedEvil> maiden: Make the feet out of wii balance boards, and connect them over bluetooth
[14:12:18] <robotustra> for 1 solid state body
[14:12:44] <robotustra> 3 points to read reaction from floor
[14:13:07] <robotustra> 2 - to measure sifts
[14:13:23] <robotustra> 1 the momentum
[14:14:21] <maiden> how would you mount the momentum sensor?
[14:16:02] <SpeedEvil> maiden: what sort of robot are you trying to make?
[14:16:02] <deshipu> very sturdily
[14:27:06] <maiden> 1:1 ish humanoid
[14:27:27] <maiden> biped to start with
[14:28:59] <SpeedEvil> ah.
[14:29:18] <SpeedEvil> You're expecting quadruped humans?
[14:29:43] <SpeedEvil> Or conversion bureau
[14:29:44] <femtogram> see: centaur
[14:31:35] <SpeedEvil> hmm
[14:31:36] <SpeedEvil> https://en.wikipedia.org/wiki/Exercise_physiology
[14:31:45] <SpeedEvil> ' an adult human male when jumping up from a squat can mechanically generate 314 W/kg.
[14:31:49] <SpeedEvil> (of muscle)
[14:32:28] <maiden> oh,
http://www.takanishi.mech.waseda.ac.jp/top/research/wabian/
[14:32:37] <maiden> nice page :D
[14:33:12] <SpeedEvil> The above would give you at least ballpark for watts you'd need if you simply replace all the muscles
[14:33:59] <maiden> i am just happy if it can walk a bit before tipping over ;)
[14:35:53] <SpeedEvil> maiden:
http://www.ebay.co.uk/itm/231799374944 - one per foot, triggered on contact
[14:36:06] <SpeedEvil> Obvious range limits though
[14:36:08] <maiden> and to get it to walk, you do not need that much energy if you have clever design, see:
https://www.youtube.com/watch?v=KLepY1AsaRk
[14:36:22] <SpeedEvil> maiden: yes, of course not.
[14:36:26] <Snert_> great page - but that was a decade ago. I wonder how they're doing now on the subject.
[14:36:30] <SpeedEvil> maiden: it is an interesting upper limit in principle.
[14:36:44] <maiden> nail gun?
[14:36:48] <SpeedEvil> :)
[14:38:34] <maiden> SpeedEvil, how about having an motor loading a strong spring over time and then release it to get a lot of instant power? :p
[14:38:58] <SpeedEvil> Err - no.
[14:39:03] <maiden> :p
[14:39:08] <SpeedEvil> I mean actually drive nails into the ground to secure the foot
[14:39:13] <maiden> well thats what solenoids is for ;)
[14:39:28] <SpeedEvil> Solenoids are unfortunately terrible.
[14:39:28] <maiden> aha, thats a great idea!
[14:39:48] <maiden> i so love the EMP pulses :D
[14:39:57] <SpeedEvil> The power output of a motor is (neglecting lots of stuff) weight * strokes/revs per minute.
[14:40:05] <SpeedEvil> For a constant design of motor
[14:40:34] <SpeedEvil> (an electromagnetic motor limited by heat production)
[14:40:55] <maiden> built a footboll kicker with a home built kicker, the caps where huge!
[14:41:14] <SpeedEvil> If you can increase the RPM of the motor, you increase the power output almost free.
[14:41:25] <SpeedEvil> The downside is that there is never quite a free lunch
[14:41:37] <anonnumberanon> https://scontent.cdninstagram.com/hphotos-xpf1/t50.2886-16/12407100_1756297151271192_1093029065_n.mp4
[14:41:40] <anonnumberanon> lol
[14:41:41] <SpeedEvil> Because the energy stored in the motor goes as rpm^2 too.
[14:41:58] <SpeedEvil> Which means that reaction times rise for non-planned changes in speed
[14:42:46] <SpeedEvil> It would be quite interesting to have a 'x-prize'
[14:44:03] <SpeedEvil> Which would - for example - award a prize for the linear actuator capable of doing 1kW output in 5kg at the fastest cycle rate (that cycle rate to be sustained for 100Kcycles
[14:45:17] <maiden> sweet
[14:46:51] <SpeedEvil> (1kW in 5kg is not remotely challenging alone)
[14:56:32] <robotustra> maiden: I'm busy now, but I'll draw you how to make momentum sensor
[14:56:37] <robotustra> later
[15:25:20] <maiden> time to go
[15:25:46] <maiden> will use 6 sensors as you suggested, i think i have it figured out now.
[15:25:57] <maiden> thank you and bye for today!
[15:27:26] <robotustra> л
[15:27:28] <robotustra> ok
[15:27:50] <maiden> 3x and a T shape that pushes down on the z sensors, then 2x x sensors on the upper and lower right side of the T, then one y sensor at the bottom of the T shape
[15:29:11] <maiden> the ancle joint will attach to the middle of the T shape.
[15:29:45] <maiden> simple and straight forward ;)
[15:30:26] <maiden> the math will be a bit funny with the x sensors not aligned, but thats easily solved.
[15:31:08] <maiden> oh, silly me
[16:26:03] <anonnumberanon> maiden> how would you mount the momentum sensor?
[16:26:15] <anonnumberanon> with double sided tape that has a layer of foam
[16:29:50] <SpeedEvil> there is no such thing as a momentum sensor.
[16:29:58] <SpeedEvil> Do you mean an accellerometer?
[16:30:18] <woddy> SpeedEvil, uhhm..
[16:30:19] <anonnumberanon> maiden,
http://www.amazon.com/CS-Hyde-Conformable-Double-Sided-Adhesive/dp/B004V40SUS/ref=sr_1_7?ie=UTF8&qid=1452031503&sr=8-7&keywords=3m+double+sided+tape
[16:30:29] <woddy> SpeedEvil, "indirect measurement"-mechanism sensors
[16:30:31] <woddy> of course
[16:30:37] <woddy> can ultimately be about momentum
[16:30:49] <woddy> direct physical property will not be momentum
[16:30:59] <woddy> but based on inference and some calculation in signal processing system
[16:31:03] <woddy> you will get the momentum
[16:31:05] <SpeedEvil> sure.
[16:31:11] <woddy> so why would you say that :D
[16:31:25] <SpeedEvil> Because it's not sensing momentum.
[16:31:39] <woddy> SpeedEvil, well at university
[16:31:46] <woddy> aI learned to call a sensor by what one wants to achieve to measure
[16:31:53] <woddy> and call it direct measurement if what you want to measure is what you measure
[16:31:55] <woddy> and indirect mesaurement
[16:32:00] <anonnumberanon> you went to a bad school
[16:32:04] <woddy> if what you want to measure is infered from another physical property that you actually measure
[16:32:06] <woddy> anonnumberanon, no
[16:32:12] <woddy> its merely a definitory matter
[16:32:20] <SpeedEvil> Indirect momentum sensor I would have no problem with.
[16:32:24] <anonnumberanon> LOL
[16:32:30] <woddy> anonnumberanon, not english language anyway
[16:32:49] <woddy> this is normative
[16:32:51] <woddy> so meh
[16:33:01] <woddy> so thats "<anonnumberanon> you went to a bad school" retarded
[16:33:20] <woddy> also in any textbook its like that
[16:33:25] <woddy> also in lecture notes from other universities
[16:33:27] <anonnumberanon> the normative statement that is correct to say is: "I use an accelerometer"
[16:33:28] <woddy> basically the definition I gave you
[16:33:31] <woddy> is used commonly
[16:33:35] <woddy> in .. where I'm from
[16:33:45] <woddy> anonnumberanon, you don't know what normative means
[16:33:48] <woddy> obviously
[16:33:52] <anonnumberanon> you could also say "I use an accelerometer to measure momentum"
[16:34:00] <deshipu> what are you talking about?
[16:34:34] <anonnumberanon> deshipu, people deciding to call IMUs "momentum sensors" and the people defending them
[16:34:41] <deshipu> momentummeters? ;)
[16:34:56] <anonnumberanon> 3m
[16:35:03] <anonnumberanon> that's what i said use 3m tape
[16:35:19] <anonnumberanon> but there are 4 Ms in that word so that tape prolly wouldn't work
[16:35:26] <deshipu> you can't measure momentum, at best you can calculate it from other things you measure
[16:35:33] <woddy> har
[16:35:36] <anonnumberanon> it's funny cause maiden may miss this convo totally
[16:35:36] <woddy> and so it goes on
[16:35:39] * woddy departs from circlejerk
[16:36:05] * anonnumberanon smirks at SpeedEvil
[16:41:20] * SpeedEvil wishes nitinol were cheaper
[16:42:03] <maiden> miss what :D
[16:47:31] <maiden> i will measure forces and calculate a angular momentum (Torque) on the joint based on the placement of the sensors with respect to the joint
[16:48:14] <deshipu> how do you measure forces? by deformation of some kind of spring?
[16:49:11] <maiden> load cells, i take inspiration from this paper:
http://www.jsk.t.u-tokyo.ac.jp/~report/proceeding02/icra02/nishiwaki_forcesensor.pdf
[16:49:38] <maiden> i linked to the load cells earlier, its the same that is used in many bathroom scales.
[16:49:47] <maiden> they are dirt cheap
[16:50:03] <deshipu> sorry, I got disconnected
[16:50:16] <maiden> sorry to hear that ;)
[16:50:36] <maiden> http://www.ebay.com/itm/121635444344?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
[16:50:42] <maiden> http://www.ebay.com/itm/351107338114?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT
[16:50:55] <maiden> the load cells and an amplifier
[16:51:24] <deshipu> ah, the ZMP method of motion planning
[16:51:33] <deshipu> Asimo does that
[16:52:15] <maiden> i will probably just input rectified sensor data into a neural network :p
[16:53:00] <maiden> but will play around a bit with ZMP as well
[16:54:09] <maiden> fond an awesome robot before that had implemented a humanoid/biped that used sensor data and a neural network to walk and used 12 pic processors to calculate everything.
[16:54:34] <deshipu> "naural network" doesn't mean much
[16:54:51] <deshipu> I mean, there are so many different ways to do it
[16:56:53] <robotustra_> maiden:
https://www.twiddla.com/2467459
[16:58:52] <maiden> http://home.ctlnet.com/~mechadon/mechadon.htm
[16:59:50] <robotustra_> this legs do not measure torque
[17:00:02] <robotustra_> and shift force
[17:00:25] <robotustra_> not all torque
[17:00:36] <maiden> no, i know ;)
[17:00:41] <maiden> but its still very cool
[17:00:48] <robotustra_> only around X and Y
[17:00:56] <robotustra_> but not Z axis
[17:11:27] <robotustra_> maiden: how do you connect X cell?
[17:11:44] <robotustra_> you should make a slider under
[17:13:49] <maiden> so the T shape is only attached to the load cells
[17:14:22] <robotustra_> maiden: one X cell should be on the sliding bar
[17:15:22] <maiden> sliding bar?
[17:16:07] <robotustra_> I wander how do you connect X cell
[17:16:16] <robotustra_> on the top of T
[17:16:36] <maiden> on the side
[17:16:47] <maiden> i am going to use pipe for the T shape
[17:19:26] <maiden> and i guess straps and 3d printed mounts/fixtures will be easiest to attach the cells to the T shape.
[17:22:20] <maiden> the foot is attached to the load cells while the T shape is attached to the ankle, only the load cells a holding the foot together with the rest of the robot.
[17:22:57] <maiden> should be fine, the load cells is specified for 50kg each
[17:23:49] <maiden> and I hope to be able to get a total weight of ~20kg,
[17:31:29] <robotustra_> whole robot?
[17:32:07] <robotustra_> at the step time the force could be much more than 50 kg
[17:43:21] <Snert_> that's gonnabe 1 fine foot. Hope she can build it
[17:43:44] <robotustra_> she?
[17:43:57] <robotustra_> maiden?
[17:44:27] <Snert_> ok...well whatever sexless automation that purports to be of a female name.
[17:45:11] <robotustra_> as soon as I know maiden is a guy
[17:45:20] <robotustra_> imho
[17:45:38] <robotustra_> as soon as I got from his videos
[18:01:26] <RyanS> why do i rarely see brushless motors used on wheeled robots? too fast?
[18:04:10] <Snert_> $
[18:05:21] <Snert_> alot of the motors used come from wheelchairs...vacuum cleaners...what have you.
[18:05:36] <Snert_> and so those aren't usually brushless.
[18:09:20] <RyanS> hmm, rc people still use brush motors on rock crawlers, supposedly they are 'smoother'at low RPM. Although 2 pole motors are fairl jerky
[18:21:48] <Hyratel> RyanS, well there's a lot more math that goes into picking a brushless motor
[18:21:59] <Hyratel> all a brushed motor's performance is self-contained
[18:22:15] <Hyratel> you can use a brushed motor with any controller and its behavior will always be the same
[18:22:37] <Hyratel> but with a brushless, it's a Motor+Controller pair always
[18:22:48] <Hyratel> you have to pick a suitable motor and a controller that works the way you want
[18:23:01] <Hyratel> like
[18:23:25] <Hyratel> a 5200RPM 1/2 HP 12v brushed motor won't give a damn what kind fo controller you use
[18:23:45] <Hyratel> you give it 12v, it'll always be 5200 RPM unloaded
[18:24:14] <SpeedEvil> I picked brushless for my wheelbarrow drive (that I have not constructed yet)
[18:24:29] <Hyratel> but brushless, you ahve to know what your pole count is, what your kv is, what your controller outputs/expects
[18:24:34] <Hyratel> Inrunner/outrunner
[18:24:37] <Hyratel> yeah
[18:24:38] <SpeedEvil> http://www.ebay.co.uk/itm/New-GIANT-Garden-Trolley-Wheelbarrow-Shelf-Interlinear-HEAVY-DUTY-X-LARGE-/200598327118
[18:24:53] <SpeedEvil> you don't need to know the pole count
[18:24:57] <Hyratel> there's a lot more understanding required to apply a brushless to a problem
[18:24:58] <SpeedEvil> in many cases
[18:25:20] <RyanS> yeah, im converting my nitro truck to electric to use as a rover.. i want good run time, 15km/h max
[18:25:41] <Hyratel> RyanS, how big?
[18:25:46] <Hyratel> what diameter wheels
[18:25:56] <Hyratel> and 10mph is Dang Quick
[18:26:02] <SpeedEvil> yeah
[18:26:21] <RyanS> 1/10, 2.5kg, 100mm wheels
[18:26:54] <SpeedEvil> 10mph is 'I should be able to run this fast, really' territory
[18:26:54] <RyanS> whats a good rover speed?
[18:26:56] <Hyratel> https://socratic.org/scratchpad/268e0caa73575a4eabde
[18:27:27] <Hyratel> here's my notes on a 10" wheel
[18:28:16] <SpeedEvil> What do you want to do with this?
[18:28:21] <SpeedEvil> RyanS:
[18:28:24] <RyanS> yeah, i took a look
[18:28:47] <SpeedEvil> 175W motor is _ridiculous_ for your app
[18:28:51] <SpeedEvil> (RyanS)
[18:31:34] <RyanS> well, i want to build a wheelchair motor rover and use an autopilot like with APM Rover perhaps. the r/c car is a test platform for that
[18:32:15] <SpeedEvil> probably scale speed of the final thing is reasonable then
[18:32:35] <SpeedEvil> So if it's 4* as big you want to go, and you want it 10MPH finally, then 2.5MPH
[18:32:42] <Hyratel> what's your shaft RPM on your wheelchair motor
[18:32:50] <RyanS> 370
[18:36:32] <RyanS> they can drive a wheelchair to about 7km/h at 150kg. im shooting for 5km/h for a 50kg rover
[18:38:34] <SpeedEvil> A 50kg rover will be - compared to a 2.5kg vehicle - about 3 times the linear size.
[18:39:11] <SpeedEvil> So, 5/3 kph, or so would be sane.
[18:39:36] <SpeedEvil> to get the scale speed comparable - terrain looks to go by at thesame rate, steering comparable, ...
[18:40:08] <RyanS> yeah so 3-4km sounds ok
[18:41:37] <SpeedEvil> ^the above wheelbarrow I'm looking at 30m/minute or so.
[18:41:55] <SpeedEvil> And wondering if that may not be too fast
[18:42:22] <RyanS> would brushless provide better range, battery life?
[18:42:29] <SpeedEvil> yes.
[18:42:34] <SpeedEvil> to a limited extent
[18:43:05] <RyanS> probably not worth it
[18:43:09] <SpeedEvil> gearbox efficiency and picking a motor optimal for the speed may be more important
[18:43:33] <SpeedEvil> I'm using brushless because I want several hundred watts peak
[18:44:39] <RyanS> 1m/s .. 3.6kmh sounds ok
[18:47:18] <RyanS> im using this
http://www.societyofrobots.com/RMF_calculator.shtml
[18:51:26] <Hyratel> 110 lb final size?
[18:51:29] <Hyratel> pretty hefty
[18:55:15] <RyanS> ?
[18:55:56] <RyanS> no im working out the rc one
[19:07:45] <RyanS> Apparently this is suitable
https://www.pololu.com/product/1107/specs
[19:23:05] <Snert_> RyanS: I got nuthin against pololou motors but aliexpress will get ya a greater choice.
[19:23:57] <SpeedEvil> And way more excitement about delivery time
[19:26:17] <Snert_> 3 weeks for a used absolute gray code encoder...for $18.00 - I was happy.
[19:26:45] <Snert_> normally a $300.00 encoder.
[19:33:15] <Snert_> see? 40Gbytes of enhanced sexual
[19:33:20] <Snert_> http://www.aliexpress.com/item/Enhance-sexual-40mm-50RPM-dc-12V-motor-powerful-high-torque-gear-box-motor-gearmotors-Operating/32249938131.html
[19:36:10] <Hyratel> RyanS, you'll want the one with a built-in encoder -
https://www.pololu.com/product/2827
[20:18:31] <rue_house> hahah nie link
[20:18:47] <rue_house> they must use programs that put words in from top selling products
[22:15:49] <rue_bed> ok
[22:15:59] <rue_bed> i need to wake up
[22:16:34] <rue_bed> eat, I should eat
[22:16:38] <rue_bed> what to eat
[22:25:25] <Jak_o_Shadows> food
[22:25:26] <Jak_o_Shadows> Bread
[23:13:05] <anonnumberanon> you're not going to bed soon?
