#robotics Logs

Sep 16 2018

12:17 AM rue_mohr: flav0r, use grbl
12:21 AM rue_mohr: or tell me what your trying to do and I might step dwon from trying to tell you that you need to implement a physics engine
12:25 AM _unreal_: rue_mohr, I just ordered more shit from amazon :)
12:25 AM _unreal_: sigh 1:30am here
12:25 AM _unreal_: So I just finished half the wiring of my control box
12:25 AM _unreal_: Got the power supplys wired up. waiting on special connectors from amazon. blade splitters
12:25 AM rue_mohr: I'd love to help, but your all over the map and I cant follow
12:26 AM _unreal_: all over the map?
12:26 AM _unreal_: I'm working on what I can as I have time
12:26 AM _unreal_: dude if you saw the pile of dirt I moved today at my fathers house. OMG...
12:27 AM _unreal_: tomorrow were "going to hang the new fence" "ya right".... and get the new rain gutter on the front of the house.....
12:27 AM _unreal_: knowing my father he will take for ever and hardly get anything done
12:28 AM _unreal_: I also found the correct cap and fixed the 12v psu for my cnc control box.
12:29 AM _unreal_: If I get time tomorrow I'll try to wire up up the phase control's
12:30 AM _unreal_: rue_mohr, I do have a question for you maybe you can answer? FMD2740C
12:30 AM _unreal_: Can I tie the control grounds all together and drive +5v for active on HIGH?
12:30 AM _unreal_: and do I need resistors?
12:30 AM _unreal_: the manual gives SHIT for details
12:32 AM _unreal_: rue_mohr, ?
12:33 AM _unreal_: leave a message then I'm heading to bed
12:33 AM rue_mohr: ok, you havn't shown me what your trying to connect
12:33 AM _unreal_: LOGIC
12:33 AM rue_mohr: it could be opto-isolated
12:33 AM rue_mohr: it could be cmos
12:33 AM rue_mohr: it could be ttl
12:33 AM rue_mohr: I dunno
12:33 AM _unreal_: FMD2740C
12:33 AM _unreal_: Thats the motor controller
12:33 AM rue_mohr: ok
12:34 AM _unreal_: I already posted that model number
12:34 AM rue_mohr: chip does not exist as far as I can tell
12:34 AM _unreal_: I'm just trying to figure out how to connect it. My logic source currently is TTL
12:34 AM _unreal_: for driving the steps
12:34 AM _unreal_: ?
12:35 AM _unreal_: https://kenswu.zzvps.com/Function/Manual/DS_FMD2740C_EN.pdf
12:35 AM _unreal_: first google link
12:35 AM rue_mohr: ah , its opto
12:36 AM rue_mohr: so, if you want yea, you can tie all the + signals to 5V and drive the - lines low with your avr
12:37 AM _unreal_: I need to go the other way
12:37 AM _unreal_: the pdf also seems to show ? 1k resistors on the optos
12:37 AM rue_mohr: all over the mpa
12:37 AM rue_mohr: yes, you can connect all the -s togethor and dirve the +'s with your avr, no resistors needed
12:38 AM _unreal_: ok? shows a 1k in the PDF though?
12:38 AM _unreal_: I think its a 1k
12:38 AM rue_mohr: they have a 270 ohm internal resisotr, you can add 1k if you want
12:39 AM _unreal_: so 1k isnt necessary
12:40 AM _unreal_: ok I'm heading to bed yawn......
12:40 AM _unreal_: god I have a ton of parts on my bed
02:17 AM flav0r: BACK !
02:18 AM flav0r: to answer your earlier question :
02:18 AM flav0r: i inherited this thing
02:18 AM flav0r: what the guy is trying to do is turn something (lol) and then keep it in syc with an identical unit … hence my question a couple nights ago about rotary encoders and what not
02:19 AM flav0r: now im just looking at what i’ve been handed and looking through it wondering if its just poorly thought out thus far and poorly implemented in respect to code and hardware
02:20 AM flav0r: also i think a physics engine might be more enjoyable
02:37 AM mrdata_: hmm
02:38 AM mrdata_: this sync problem has many applications
02:38 AM mrdata_: gear cutting comes to mind
02:38 AM mrdata_: (probably because gears are on my mind this week)
02:39 AM mrdata_: in fact,
02:39 AM mrdata_: i'm doing some math around it, https://imgur.com/a/haEoOr4
03:07 AM flav0r: hmm maybe sync was the wrong word
03:07 AM flav0r: they jsut have to be in the same position
03:08 AM flav0r: so i dont think they need to move at exactly the same time and the same amount
03:08 AM flav0r: just os long as they both go to the same pos
03:08 AM flav0r: but im really wondering about that approach to turn the driver on and off
03:08 AM flav0r: btw … that pic looked cool dude
09:07 AM VansFannel: Hi!
09:07 AM VansFannel: How can avoid an u-shaped obstacle with a robot? What technic or algorithm do I have to use? Thanks
09:14 AM polprog: what does the shape of the obstacle has to do with avoiding it
09:17 AM VansFannel: polprog: your robot can go inside the U, and it has to know how to go out. The problem is that between robot's position and its goal there is a u-shaped obstacle, with the U pointing to the robot. The robot has to move to its goal, but it can fall inside the U-shaped obstacle.
09:18 AM polprog: any pathfinding algo should do
09:19 AM VansFannel: I have to use reactive paradigm and potential fields.
09:22 AM VansFannel: Something like this: https://www.researchgate.net/figure/Simulation-of-the-QVFF-algorithm-with-a-holonomic-robot-and-a-U-shaped-obstacle_fig3_261710818
09:27 AM VansFannel: Or this one: https://www.researchgate.net/profile/Nadeem_Javaid/publication/237054021/figure/fig2/AS:393251568799749@1470769888583/Dead-end-scenario-of-Artificial-Potential-Field-method-U-shaped-obstacle.png
09:31 AM VansFannel: https://answers.ros.org/question/303462/move-a-robot-to-its-goal-avoiding-u-shaped-obstacle/
11:03 AM rue_mohr: I dont think many of us are up on they university theroy stuff
11:04 AM polprog: for me, pathfinding algo is not immediately trivial
11:04 AM polprog: id have to read up on it