Back
[01:42:22] <rue_shop3> ok
[02:21:46] <rue_house> screw part numbers, I should sort the transistor drawr set by votlage and current ratings
[02:22:35] <rue_house> with maybe NPN in the front and PNP in the back
[02:22:57] <rue_house> then when I'm looking for a transistor all I have to worry about is what gains are available
[02:27:56] <rue_house> matrix gets kinda big
[02:28:03] <rue_house> voltage / current
[02:28:11] <rue_house> NPN PNP P channel and N channel
[04:36:50] <rue_shop3> today I learned about the half totem fet driver
[06:22:18] <orlock> jumper wires - tool or component?
[06:23:51] <vicarion> or just kinky sex toy?
[06:24:16] <orlock> i meant breadboard style
[06:27:37] <deshipu> robotustra: yeah, lots of those kits out there
[11:56:36] <rue_house> nobody operates mosfets over 12V
[11:56:49] <rue_house> so, nobody has ever had to make a high side level shifter
[12:05:33] <emeraldgreen> don't people just use optical/capacitive decouplers ?
[12:11:26] <emeraldgreen> IRF even has special half-bridge irf2110 series drivers that have integrated decoupling
[12:15:06] <rue_shop3> if your doing fast switching, you cant use an opto
[12:15:09] <rue_shop3> they aren't fast enough
[12:16:43] <rue_shop3> funny how you never see high side drivers for P channel
[12:17:15] <rue_shop3> those high side drivers are bootstrapped, they cant run from 0 - 100% duty
[12:17:30] <emeraldgreen> hmm, ok
[12:17:59] <rue_shop3> I'm working thru it tho
[12:18:26] <emeraldgreen> When I did amaturish power electronics stuff I just tried to pass power to the drivers through their own ferrite coil decoupler
[12:18:39] <emeraldgreen> that's an overkill tho
[12:19:03] <rue_shop3> if your duty is 0-50% there are lots of options
[12:19:14] <emeraldgreen> I was a bit reckless then, now I don't do HV power electronics myself :)
[12:19:46] <emeraldgreen> Also, in battery powered robots there is no need in HV power stuff
[12:20:39] <emeraldgreen> Not to say it can't be done by a good DIYer, but it should be done right
[12:21:18] <ace4016> the need for HV power depends on application
[12:21:30] <ace4016> if you have a component that needs HV power, what do you do then? :P
[12:21:37] <ace4016> doesn't always mean high current either
[12:21:59] <emeraldgreen> ace4016 never happened in years
[12:22:10] <emeraldgreen> I agree about application though
[12:31:03] <rue_shop3> so I think that fast steps from marlin are getting missed by my driver
[12:32:41] <jhylands> I'm using a high side switch for my power board for Roz
[12:33:09] <jhylands> but its not something you switch on and off quickly - its actually switching the power supply for the robot
[13:00:50] <anonnumberanon> sup
[13:03:27] <anniepoo__> the direction of tyhe gravitational potential vector?
[13:21:17] <codepython777> anyone here?
[13:21:55] <Snert> ask away maybe someone is here that knows
[13:22:35] <deshipu> what do you mean by "here" exactly?
[13:23:13] <Snert> ---------------------------------------------------------------------------------------------------------> over here.
[13:23:31] <Snert> ... and what do you mean by what do you mean?
[13:23:46] <Snert> and this word, exactly, how much precision do you want.
[13:27:50] <codepython777> how can one use GD32F103 to control a brushless motor?
[13:33:18] <jhylands> you would use an ESC and send pulses to it like you would a servo, presumably
[13:41:51] <codepython777> jhylands:
http://ifixit.org/blog/7821/swagway-teardown-hoverboards/ - what ESC are these guys using?
[13:47:46] <jhylands> it looks like they built a custom ESC
[13:48:08] <jhylands> section titled "Motor Control Board"
[13:48:25] <jhylands> "the main focus of the control board is the six half-H drivers, which drive two 3-phase motors"
[13:50:13] <codepython777> jhylands: can i drive those wheels with a esc from outside?
[13:53:08] <jhylands> are you looking to build one of those boards, or fix one that is broken?
[13:54:08] <codepython777> jhylands: I just want to use that wheel in another project
[13:54:18] <codepython777> wheel + motor
[13:54:40] <jhylands> well, its just a brushless motor, so you need an ESC like from a quadcopter or something that can drive that much current
[13:56:13] <jhylands> how long did the thing run before (per battery charge, I mean)
[14:01:32] <codepython777> jhylands: what other sensors does the wheel have. If I want to use it in my project, I'll have to figure out how to get all those outputs into something
[14:07:41] <jhylands> well, the write-up says it has five leads for hall effect sensors
[14:08:12] <jhylands> looking at the picture, I would guess you have ground, power, and 3 outputs from the actual sensors
[14:09:12] <jhylands> you would want to hook those up to a logic analyzer or something, so you could see what the overlapping waveforms look like
[14:09:27] <jhylands> you typically only need 2 to make a quadrature encoder
[14:10:02] <codepython777> i see
[14:10:09] <jhylands> do you have a logic analyzer?
[14:10:24] <jhylands> they probably run at 3.3 volts, given that it is an ARM chip driving the whole thing
[14:11:06] <jhylands> with the five smaller wires, red is power (+3.3 volts), black is ground, and yellow/green/blue are the sensor outputs
[14:11:18] <jhylands> they may be digital or analog, depends on the hall effect sensor
[14:11:25] <jhylands> probably analog
[14:11:42] <jhylands> in which case your control algorithm will be more complicated
[14:12:34] <jhylands> the 3 larger wires (also yellow/green/blue) are the drive wires for the motors
[14:12:43] <jhylands> I assume they are larger, in any case
[14:12:51] <jhylands> you can trace them to their location on the control board
[14:13:41] <jhylands> if the hall sensors are analog, a logic analyzer isn't going to help you much
[14:14:18] <jhylands> basically, you'll want to hook them up to your ARM chip, and read the voltages as you rotate the wheels by hand
[14:14:52] <jhylands> they may be directly related to the phase wires controlling the motor, in terms of when they have to be powered or not
[14:14:56] <jhylands> I don't really know
[15:29:27] <deshipu> steppers are a pain
[15:51:30] <vicarion> obviously you need pneumatic muscles instead
[15:53:29] <deshipu> um, pneumatics are bascially just on.off
[15:53:41] <deshipu> you can't do precise movement with that
[22:20:47] <rue_shop3> yea I still need to do jhylands project too
[22:20:52] <rue_shop3> the spool and motor are finished
[22:21:00] <rue_shop3> I just need to screw it down and apply power
[22:21:13] <rue_shop3> jhylands, would you like some 3d printable files?
[22:21:23] <jhylands> why?
[22:21:29] <jhylands> I already have that part done
[22:21:30] <rue_shop3> dunno
[22:22:02] <rue_shop3> I suppose if I screw it to a price of wood and do the demo I can move on to my power supply
[22:22:23] <rue_shop3> I'm all tangled up on how to drive the transformer again
[22:23:14] <jhylands> rue_shop3,
https://app.box.com/s/ewm7phmm42fqc94e90h257se3ib3hene
[22:23:33] <rue_shop3> yea
[22:23:43] <rue_shop3> I made one with a 4" dia
[22:23:47] <rue_shop3> from some 4" pipe
[22:23:55] <rue_shop3> 3d printed endcaps and stands
[22:24:31] <jhylands> this is just a proof of concept right now - once I get it working, I'll build a full-sized version that can hold 60' of wire
[22:24:49] <rue_shop3> oh, I'll use my eye-destroying demo board
[22:24:57] <rue_shop3> yea, I'm about to prove it for you
[22:25:05] <rue_shop3> how many grams retraction force you want?
[22:25:22] <rue_shop3> I hope its more than 10, thats the resolution of my scale
[22:35:08] <rue_shop3> jhylands, well?
[22:35:11] <rue_shop3> 20g?
[22:35:27] <jhylands> sorry, was looking at something else
[22:35:34] <jhylands> I have no idea
[22:36:01] <jhylands> the retraction force has to be enough to spool up the line when it is loose
[22:36:27] <jhylands> its not really about numbers, its about experimentation (at least for me)
[22:38:39] <rue_shop3> at 1A it does 130g
[22:38:55] <Jak_o_Shadows> por queno los dos
[22:38:59] <rue_shop3> 1A is te limit of my stupid supply
[22:39:24] <Jak_o_Shadows> Anyway, shed should be getting reassembled later today.
[22:39:45] <Jak_o_Shadows> Can finally do the improvements - vacuum dust collection system, cyclonic separator, etc.
[22:49:46] <rue_shop3> hahaha
[22:50:00] <rue_shop3> the only thing I have that can hold up the scale while I film is the 4' robot arm
[23:03:59] <rue_shop3> jhylands, ok I made the video
[23:04:08] <jhylands> ok
[23:04:13] <rue_shop3> when I reduced the votlage to limit the retract speed
[23:04:31] <rue_shop3> I messed up and it didn't retract fast enough to not tangle the string
[23:04:41] <rue_shop3> so, I'm gonna call that my fault
[23:04:59] <rue_shop3> could you hang around till I get it uploaded?
[23:05:06] <jhylands> probably
[23:05:16] <jhylands> if it doesn't take a long time - its approaching midnight here
[23:07:27] <rue_house> I didn't say anythngm but I diddled the current from prettymuch zero to 1A, and then brought the voltage limit down to demonstrate how the retract speed would be reduced
[23:07:35] <rue_house> copying...
[23:08:14] <rue_house> I updated my limnux machine and now my card reader dosn't work I have to go right from the camera
[23:08:37] <rue_house> uploading
[23:08:38] <rue_house> ...
[23:11:12] <rue_house> 38%..
[23:11:39] <rue_house> 47%...
[23:12:58] <rue_house> 67%
[23:13:57] <rue_house> 80%
[23:15:31] <rue_house> arg PROCESSING!?
[23:16:53] <rue_house> https://youtu.be/IXXLvoa9BXk
[23:17:02] <jhylands> yeah
[23:17:05] <jhylands> always processing
[23:18:39] <jhylands> That's pretty cool
[23:18:58] <rue_house> I think I suck at taking videos
[23:19:50] <Hyratel> your scale's buttons make a smiley
[23:19:55] <Hyratel> c:
[23:20:14] <jhylands> I don't think you have nearly enough drawer units
[23:20:15] <rue_house> yea
[23:20:16] <Hyratel> rue_house, so what are we looking at
[23:20:24] <rue_house> its a chineese smiley scale
[23:20:33] <Hyratel> oh, CC mode for driving a motor?
[23:20:41] <rue_house> jhylands, no, sadly the size upgrades cause choas
[23:20:45] <rue_house> yea
[23:20:54] <rue_house> cc to cause constant force
[23:21:00] <Hyratel> so it's a constant force spool
[23:21:10] <rue_house> with the 1A my supply can do, I could get about .1 to .15kg
[23:21:13] <rue_house> yea
[23:21:23] <rue_house> and if you limit the voltage, you limit the retract speed
[23:21:50] <Hyratel> what I want to know is, is there a good way to make a CC supply that ISN'T resistive-mode?
[23:21:51] <rue_house> jhylands, for the record I worked out a new way to sort the transistors, by votlage and current, screw part numbers
[23:22:04] <rue_house> resistive mode?
[23:22:07] <jhylands> heh
[23:22:13] <Hyratel> LM317
[23:22:19] <Hyratel> linear regulators
[23:22:28] <rue_house> yea, switchers
[23:22:31] <jhylands> Well, I have to say I'm impressed
[23:22:49] <rue_house> jhylands, ok, if you want the 3d files, just shout and I'll put them somewhere
[23:22:58] <jhylands> no, I don't need the 3D files
[23:23:02] <rue_house> ok
[23:23:10] <jhylands> I've been designing 3D cad stuff for almost 20 years
[23:23:26] <rue_house> all fine, I'm obliged to offer
[23:23:29] <rue_house> its not my project
[23:23:30] <jhylands> but the fact that the direct drive motor works is very cool
[23:23:41] <rue_house> well, technically, the motor holder is part of my electronic scale
[23:23:43] <jhylands> I really didn't think it would have enough torque at low RPMs
[23:24:02] <rue_house> thats also only a 1A supply
[23:24:09] <rue_house> I"m working on a 5A supply
[23:24:33] <jhylands> I have a 5 amp supply, at least I think it can go to that
[23:26:15] <rue_house> I'm building one
[23:27:08] <jhylands> yeah
[23:27:09] <jhylands> http://canada.newark.com/extech-instruments/382260/power-supply-switch-mode-16-4v/dp/61M1150?ost=Extech+382260
[23:27:15] <jhylands> that's the one I have
[23:30:01] <rue_shop3> https://hackaday.io/project/296-bench-power-supply
[23:30:03] <rue_shop3> mine, sofar
[23:31:38] <jhylands> rue_shop3, what motor are you using?
[23:31:49] <jhylands> do you have a link, or is it just a random motor you had?
[23:31:53] <rue_shop3> its the motor from a cannon inkjet printer
[23:32:02] <rue_shop3> I chose that so you could try the same motor
[23:32:09] <jhylands> not a stepper, right?
[23:32:15] <rue_shop3> brushed dc
[23:34:18] <rue_shop3> bigger motor with more poles is fine
[23:34:25] <rue_shop3> just need to watch the inertia
[23:35:32] <jhylands> I have a bunch of straight non-geared motors
[23:35:54] <rue_shop3> play
[23:37:31] <jhylands> https://app.box.com/s/omb9zxna7fu9rwgfn2s90wc9qq051uvo
[23:37:59] <rue_shop3> those are a bunch smaller than that one I had
[23:38:05] <jhylands> yeah
[23:38:20] <rue_shop3> the promax is close, but I suspect its a high rpm, low pole motor
[23:38:27] <rue_shop3> or maybe not
[23:38:29] <jhylands> there's an electronics surplus place about ten minutes drive from here, he has hundreds of different sized DC motors
[23:38:40] <rue_shop3> sounds perfect
[23:39:35] <jhylands> so that's all you have - you're running a 1-amp fixed current, with 6-9 volts DC
[23:39:43] <jhylands> no other circuitry involved?
[23:40:45] <rue_shop3> no, its just eth cc power supply
[23:40:55] <rue_shop3> it was .5A for the 50g
[23:41:21] <jhylands> well, I'll play around with that here while I'm building the sprung pulley system
[23:42:44] <rue_shop3> I like the low latency of passive systems
[23:48:38] <jhylands> well, I need to hit the sack - thanks for the demo
[23:51:40] <rue_house> np