#robotics Logs

Aug 21 2018

04:12 AM weyland|yutani: rue_mohr, well it could also be infested IoT devices, NAS, network cams there was a huge botnet a years or so back which consisted of cheap network cameras and NAS devices
04:16 AM SpeedEvil: I still have a dozen cams which are vulnerable to Mirai.
04:16 AM SpeedEvil: (which are not conected to the internet)
04:18 AM weyland|yutani: well the mirai code is in the wild maybe those spammers use it who knows
08:48 AM rue_mohr: but its not
08:49 AM rue_mohr: you need to look at the machines that are attacking, most of them are commercial webservers
02:17 PM weyland|yutani: https://www.cs.utah.edu/~rajeev/pubs/isca16.pdf
03:43 PM BitEvil is now known as speedevil
10:20 PM gaze__: hey guys, I have shaft that's free to rotate and free to translate along it's length. I want to be able to push and pull on it while allowing it to rotate freely. Any opposing force will be very low. It's paramount that there be zero backlash.
10:20 PM gaze__: I feel like I should use a pair of preloaded bearings
10:21 PM Tom_L: linear bearings
10:21 PM gaze__: I just wonder if I should spring for a pair of angular contact bearings or if preloading off the shelf vanilla single race bearings actually buys you anything
10:21 PM gaze__: well no, I don't want linear bearings
10:21 PM gaze__: I want to push and pull on the shaft
10:21 PM orlock: how would you preload the off-the-shelf bearings?
10:21 PM gaze__: belville washers or the like
10:21 PM z64555: push/pull in which direction? along its length? perpendicular?
10:21 PM gaze__: along its length.
10:22 PM Tom_L: push pull and rotate sound ideal for a linear bearing
10:22 PM gaze__: How do you push and pull on a shaft along it's length with a linear bearing
10:23 PM gaze__: a linear bearing allows free translation of a shaft along its length and no rotation.
10:23 PM Tom_L: it will do both
10:23 PM orlock: Are you talking about a round shaft?
10:23 PM orlock: or some other shape?
10:23 PM z64555: not if you rotate it 90 degrees, I'm thinking?
10:23 PM z64555: although this is the electrical guy talking
10:23 PM Tom_L: if he's rotating it, it must be round
10:23 PM gaze__: yeah, it's a round shaft.
10:24 PM Tom_L: supported in a linear bearing, it will rotate and move back and forth
10:24 PM Tom_L: or a bushing
10:24 PM gaze__: well it's in a misumi ball bushing... so it's free to rotate and translate
10:24 PM gaze__: but I want to control the translation
10:24 PM Tom_L: put a round stick in a hole. you can turn it and you can push it back and forth
10:24 PM orlock: the question would be how the fuck you stop a round shaft rotating
10:25 PM gaze__: I don't want to stop it from rotating.
10:25 PM gaze__: I want to allow it to rotate
10:25 PM Tom_L: i'm afraid we need more data
10:25 PM orlock: so why isnt it able to rotate?
10:25 PM gaze__: and I want to control it's linear translation
10:25 PM Tom_L: full assembly drawing in solidworks would be in order :)
10:26 PM Tom_L: or catia if you prefer
10:27 PM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/Mill_Steel/Assembly/Shims/Cover_slide_bracket.jpg
10:27 PM Tom_L: now i have to figure out how to hold that to machine it
10:27 PM orlock: so you need something captive on the shaft to allow for linear movement, that doesnt constrain the rotation
10:27 PM orlock: ?
10:27 PM Tom_L: a slot with a fork to move the shaft back and forth
10:27 PM * z64555 looks at a linear bearing finally
10:27 PM z64555: oh that thing
10:27 PM Tom_L: not enough data.
10:27 PM gaze__: orlock: I'll start from the beginning. I have a shaft which is inside bushings that allow free rotation and translation of the shaft.
10:28 PM gaze__: orlock: I want to permit the shaft to rotate freely while controlling it's linear translation.
10:28 PM Tom_L: so push it from the end
10:28 PM gaze__: okay now we're getting somewhere.
10:28 PM orlock: yeah, i dont see the problem?
10:28 PM Tom_L: or have a slot with a fork
10:29 PM Tom_L: i don't either really
10:29 PM orlock: it grips it by the shaft!
10:29 PM Tom_L: there are at least half dozen ways to do it
10:29 PM gaze__: I want minimal friction and zero backlash
10:29 PM orlock: i mean, you will need to machine the shaft in some way?
10:29 PM gaze__: yeah.
10:29 PM Tom_L: push it with a thrust bearing on each end
10:30 PM gaze__: thrust bearing sounds pretty reasonable
10:30 PM gaze__: do you know of a good cheap source of thrust bearings
10:30 PM Tom_L: use air bearings for near zero friction
10:33 PM Tom_L: vbx.com
10:33 PM Tom_L: vxb.com sry
10:39 PM orlock: what's it for?
10:45 PM _unreal_: gaze__, so you need to be able to index this shafts rotation AND liner motion?
10:45 PM _unreal_: one or both? if one which is priority number one
10:50 PM orlock: pushrods
10:50 PM orlock: linear movement with free rotation
10:51 PM orlock: he's working on an electric cam!
10:53 PM rue_mohr: easy
10:54 PM rue_mohr: linear guide, thats round, with a keyway for a gear
10:54 PM rue_mohr: oh I suppose I should be working on the motor driver, where was I?
10:54 PM rue_mohr: step sequencer, hmm, ok, I'm in a codine mode anyhow
10:54 PM rue_mohr: er, coding...
10:55 PM rue_mohr: hmmm
10:56 PM rue_mohr: damn there isn't an avr with just a few more pins then a t13
10:56 PM rue_mohr: waaaait waait waiiit
10:56 PM rue_mohr: state machine
10:56 PM rue_mohr: damnit, why didn't I think of it
10:56 PM rue_mohr: 1 bit input,
10:56 PM rue_mohr: 2 bits feedback
10:57 PM rue_mohr: 4 bits output
10:57 PM rue_mohr: mix the....
10:57 PM rue_mohr: 2 bits output....
10:57 PM rue_mohr: 8 states
10:57 PM _unreal_: gaze is dead
10:57 PM _unreal_: sigh
10:57 PM rue_mohr: gaze?
10:58 PM rue_mohr: 4 bits wide...
10:59 PM rue_mohr: damn those 5x7 leds, not being 5x8
10:59 PM rue_mohr: so I can use an 8x8 matrix...
10:59 PM _unreal_: * gaze__ has quit (Ping timeout: 252 seconds)
11:02 PM rue_mohr: so then I need a 3 to 8 line decoder
11:02 PM rue_mohr: 8x8 led matrix
11:02 PM rue_mohr: latch... hmmmm
11:03 PM rue_mohr: ooooh there was a latching decoder somewhere
11:03 PM rue_mohr: I wonder if the polarity was right
11:05 PM _unreal_: rue_mohr, https://pastebin.com/NgtZDLZN
11:12 PM Tom_L: 74HC137
11:20 PM gaze__: oops sorry got disconnected
11:21 PM gaze__: I don't wanna index rotational motion at all
11:21 PM gaze__: I just want it to freely rotate
11:21 PM gaze__: I want to control linear displacement
11:24 PM rue_mohr: ok, toolchanger, somehow
11:24 PM gaze__: eh it's a coil winder
11:25 PM rue_mohr: you want to do a cnc coil winder?
11:25 PM gaze__: but the wire diameter is 0.004"
11:25 PM rue_mohr: sweeeet
11:25 PM gaze__: yeah.
11:25 PM rue_mohr: so, lathe, and a wire guide
11:25 PM rue_mohr: "lathe"
11:25 PM rue_mohr: how big coils?
11:25 PM rue_mohr: can you use a 1/4" dril chuck like mine?
11:25 PM orlock: ahh
11:25 PM gaze__: 2 inch diameter and 3/8" wide
11:25 PM orlock: not a pushrod
11:25 PM gaze__: and basket wound.
11:26 PM rue_mohr: wait
11:26 PM rue_mohr: coil 2" dia thats 3/8" long?
11:27 PM gaze__: 3/8" wide
11:27 PM rue_mohr: 2" is id?
11:27 PM rue_mohr: ok, ok just a sex
11:27 PM rue_mohr: sec
11:29 PM rue_mohr: https://d2vlcm61l7u1fs.cloudfront.net/media%2F4a8%2F4a8cbd52-e650-4660-a83d-9e6d33c186a0%2Fimage
11:29 PM gaze__: oops sorry, that's a typo. 1.1" final diameter, 0.6" diameter of the former, 0.5" inner diameter
11:29 PM rue_mohr: ok THAT was hard
11:29 PM rue_mohr: you want r2 to be 2"/2
11:29 PM rue_mohr: and h to be 3/8?
11:29 PM gaze__: h=3/8", d2=1.1", d1=0.5"
11:30 PM rue_mohr: ok
11:30 PM gaze__: where d_i=2*r_i
11:30 PM rue_mohr: :) a normal circle then
11:30 PM gaze__: but the windings aren't circular.
11:30 PM gaze__: the windings weave back and forth along the circumference
11:31 PM gaze__: https://www.youtube.com/watch?v=FIOocMoRsYQ
11:31 PM gaze__: it's like this
11:31 PM rue_mohr: but just to be clear, d_i!=sin(2*ri^e^-2)
11:31 PM rue_mohr: ok
11:31 PM gaze__: but... this machine isn't sophisticated enough to produce the wind I need
11:31 PM rue_mohr: thats easy
11:31 PM rue_mohr: ok, got two stepper motors and an arduino?
11:31 PM gaze__: oh yes.
11:31 PM rue_mohr: ok, got two stepper motor drivers?
11:31 PM gaze__: yes
11:32 PM rue_mohr: got a keyboard and a mouse?
11:32 PM gaze__: no.
11:32 PM rue_mohr: wait... wait... thats not related....
11:32 PM gaze__: I have no idea how I'm doing this right now
11:32 PM gaze__: *disconnected*
11:32 PM rue_mohr: ok, a) make a lathe with a drillchuck
11:32 PM rue_mohr: b) make a wireguide
11:32 PM rue_mohr: so, lets focus on A
11:32 PM rue_mohr: we can attack mechanics first
11:33 PM rue_mohr: erm, I'm gonna live another 10 minutes
11:33 PM rue_mohr: how fast can you follow instructions?
11:34 PM gaze__: so I have a version of this where I've mounted a drill chuck on the stepper and I have a wire guide attached to one of the steppers
11:34 PM gaze__: wire guide stepper waggles back and forth while drill chuck stepper turns
11:34 PM gaze__: it's not repeatable enough
11:34 PM gaze__: I have a pretty elaborate thing in mind
11:34 PM rue_mohr: why not?
11:35 PM rue_mohr: what makes it not repeatable? can I see a pic of it?
11:35 PM rue_mohr: is it dropping steps?
11:36 PM gaze__: it's on the other coast at the moment. I'm in california, my workshop is in connecticut
11:36 PM rue_mohr: thats a long walk
11:36 PM gaze__: yeah
11:36 PM rue_mohr: you dont have a pic of it online?
11:36 PM gaze__: nope
11:37 PM rue_mohr: ok, can you describe why its not reliable
11:37 PM rue_mohr: sounds like the wire guide is sloppy
11:37 PM gaze__: It's possible. I'll probably replace it with a watch jewel or something
11:37 PM gaze__: right now it's just a bent piece of wire
11:38 PM rue_mohr: is it a radial arm or a linear positioner
11:38 PM gaze__: if not a watch jewel, I have some of those PCB drills and some drill rod. I'll probably cross drill a hole in the tip of some turned down drill rod and harden it
11:39 PM orlock: gaze__: have you looked on thingiverse?
11:39 PM gaze__: it's a radial arm.
11:39 PM gaze__: I want to switch to using a linear positioner
11:39 PM rue_mohr: ok, I'd think you want a linear system
11:39 PM gaze__: I have these really nice little misumi ball bushings
11:39 PM gaze__: which allow free rotation and linear translation
11:39 PM rue_mohr: like a threaded rod that pulls a carriage across with the guide on it
11:40 PM rue_mohr: it kinda sounds like you got it worked out then
11:40 PM gaze__: and I'm gonna have the stylus hang off a rod that goes through the ball bushings
11:40 PM gaze__: hence free rotation.
11:40 PM gaze__: but what I don't have nailed down
11:40 PM gaze__: is how to control the displacement of the rod that goes through the ball bushings
11:40 PM gaze__: I want to push and pull on that rod while letting it rotate freely
11:40 PM rue_mohr: carriage
11:40 PM gaze__: while having backlash that's below a half thou
11:40 PM rue_mohr: pull a carriage and put the guide off that
11:40 PM gaze__: because the wire is 4 thou
11:40 PM gaze__: okay.
11:41 PM rue_mohr: there are a million ways to do low-force antibacklash
11:41 PM rue_mohr: use the head positioner from a cdrom
11:41 PM rue_mohr: thats all anti-backlash
11:41 PM rue_mohr: usually every second gear is split
11:41 PM gaze__: I ripped apart a cdrom drive for this. Microstepping a cdrom drive stepper kinda sucks
11:42 PM rue_mohr: or the helix drive from a floppy, they use a spring thing on those
11:42 PM gaze__: and full stepping does't give the needed resolution
11:42 PM rue_mohr: whaaaat
11:42 PM rue_mohr: with all those reduction gears?
11:42 PM rue_mohr: their like 50:1
11:42 PM gaze__: mine had no reduction
11:42 PM rue_mohr: whaaat
11:42 PM gaze__: yeah.
11:42 PM gaze__: it was real weird
11:42 PM rue_mohr: pull one apart that uses s dc motor, and swap in a stepper
11:42 PM gaze__: hmmm
11:43 PM gaze__: alright
11:43 PM rue_mohr: #80?
11:43 PM rue_mohr: 4 thou?
11:43 PM rue_mohr: not an RF coil
11:43 PM gaze__: 38 awg wire
11:43 PM gaze__: it's the power transformer for the tektronix 556 and 547
11:43 PM rue_mohr: oh, 38 is still pretty big
11:43 PM gaze__: well the HV transformer
11:43 PM rue_mohr: :S oh
11:44 PM rue_mohr: so, you have to stop every so often for insulating layers anyhow
11:44 PM gaze__: 38 awg is 0.1 mm/0.004"
11:44 PM gaze__: pretty tiny
11:44 PM rue_mohr: 556 and 547...
11:44 PM rue_mohr: whats on my deck...
11:45 PM rue_mohr: 535, haha mines older
11:45 PM gaze__: I like the old tek scopes
11:45 PM gaze__: they break a lot so you learn a lot fixing them.
11:46 PM rue_mohr: I was amazed by the delay line in it
11:46 PM rue_mohr: its gt a screwed tube tho, I'm not gonna take time to fix it
11:46 PM gaze__: http://w140.com/tekwiki/images/4/4a/556-crop.jpg
11:46 PM gaze__: this is the one I'm fixing at the moment
11:46 PM rue_mohr: ma was at a thriftstore and said she found a tek scope, she couldn't tell me more, so i asked her to grab it for me
11:47 PM gaze__: nice!
11:47 PM rue_mohr: ah yea, thats really similar, mine only has one side of that
11:47 PM gaze__: yeah.
11:47 PM gaze__: the 556 is unique in that it's basically two scopes occupying one tube and one box
11:47 PM rue_mohr: know anyone who wants it as a fixer-upper?
11:47 PM gaze__: there's nearly perfect left-right symmetry
11:47 PM gaze__: the tektronix museum in oregon
11:48 PM rue_mohr: oh is that a fully independent dual beam?
11:48 PM rue_mohr: like, seperate deflectors and everything?
11:48 PM gaze__: yeah.
11:48 PM gaze__: exactly
11:48 PM rue_mohr: would they be willing to pay shipping to them?
11:48 PM gaze__: not sure
11:48 PM rue_mohr: huh
11:49 PM rue_mohr: if not, you want the module?
11:49 PM rue_mohr: the module dosn't weigh 80lbs :)
11:50 PM gaze__: what module do you have?
11:52 PM rue_mohr: cant find a number
11:53 PM rue_mohr: just sayd type53/54c dual amplifier module
11:53 PM rue_mohr: I guess yours is a 55
11:53 PM rue_mohr: but I'v never noticed how much those screens react to uv light! wholy!
11:53 PM gaze__: oh yeah they're really sensitive
11:53 PM gaze__: how sure are you that it's the CRT that's dead?
11:55 PM rue_shop4: hardly any of the other scopes do that, not even the heathkit
11:55 PM rue_shop4: gaze__, one of the horiz defl is boofed
11:56 PM rue_shop4: amplifier circuit I think, prolly a dead tube
11:56 PM rue_shop4: (not the crt tho)
11:57 PM rue_shop4: there was a crack when i powered it up
11:58 PM rue_shop4: I couldn't see a bad tube, but I didn't open it and take a good look
11:58 PM rue_mohr: its history includes the USAF
11:59 PM rue_mohr: they got their stuff all etched into it