|
Post by clyevo on Sept 20, 2015 14:57:05 GMT
I am able to power up the analog from the 24v terminal, i am quite sure i didnt change anything. Maybe i was still recovering from the heart attack due to the burning chip and didnt even realize i didnt switch on the 24v supply. Luckily i didnt need the chip. Anyway thanks tronicgr
|
|
|
Post by clyevo on Sept 27, 2015 12:33:10 GMT
Some help with setting up the 1) Ki and Kp of the inverter, i should be doing this with full load on platform, but i need to be sure before setting up the platform 2) PID settings of thanos board I am sure there is relation between both this settings I cant get fabi software tuning to work so i need to tune it manually. Can someone suggest PID settings and way to tune it manually. This is Ki and Kp settings from manual
|
|
|
Post by tronicgr on Sept 27, 2015 20:00:16 GMT
The PID settings for the VFD inverter are just to keep the speed constant under variable load, no need to mess with them until you have you platform on the arms.
On the AMC1280USB side: The P (proportional) setting sets how big the reaction to the error between setpoint position and current position is. The I (integral) setting sets for how long it can tolerate being outside setpoint position before it kicks the motor to get it right on target. Its useful end there is a constant load on the motor that prevents the motor from reaching the target as it gets underpowered near the target setpoint. The D (Derivative) setting sets how soon it will cut off the power of the motor near the setpoint position so the motor will not overshoot (like your motors do on the video you posted.
Also TURN OFF the LCD screen update (DIP Switch 1) while do the autocalibration utility or using it with motion software as the LCD updates can mess up with PID timings sometimes as the updates of the LCD are not 1000 times per second like the PID but more like 20 times per second.
Thanks Thanos
|
|
|
Post by clyevo on Sept 28, 2015 3:52:00 GMT
At that video the setting on my amc1280 is PID 24,0,0 and on inverter is at default like in picture Of course the lcd screen was on
I ll try connecting the load, but how can i prevent mechanical injury due to arm moving over the maximum potentiometer (sensor) 180degrees range I noticed if i messed with PID sometimes the arms make 360 degree turn and i am scared it will damage my platform if i do calibration with platform on the arms. Any tips to prevent this while connected to platform like how do i tune PID manually. I mean which one do i increase first (P,I or D) and which one do i increase or reduce next.
|
|
|
Post by tronicgr on Sept 28, 2015 19:55:03 GMT
You increase P first until it goes as fast possible to the setpoint with little overshoot.
Then increase D until you see no overshoot.
The I should be only increased if you have a load on the platform and does not reach the setpoint (constant error deviation).
Thanks Thanos
|
|
|
Post by clyevo on Sept 29, 2015 13:03:41 GMT
You increase P first until it goes as fast possible to the setpoint with little overshoot. Then increase D until you see no overshoot. The I should be only increased if you have a load on the platform and does not reach the setpoint (constant error deviation). Thanks Thanos Thanks, that is helpful my current PID without load is 7 - 0 - 0 the recommendation by fabi software is D is 0, so i dont know if my P is too low, but there is no oscillation with this settings but the movement a bit slow i will adjust it more later
|
|
|
Post by tronicgr on Sept 29, 2015 15:27:24 GMT
Your P is too low. To be able to reach the full speed of the motors, should be at least 15. Then increase the D to minimize the oscillation (overshoot).
Thanks Thanos
|
|
|
Post by clyevo on Sept 30, 2015 2:12:35 GMT
Your P is too low. To be able to reach the full speed of the motors, should be at least 15. Then increase the D to minimize the oscillation (overshoot). Thanks Thanos Thanks, i ll try that But even though the degree of oscillation is small, the mechanical relays make so much noise in quick succession, i am quite annoyed by it. I am constantly thinking about buying the ULN2803 chip for sinking the REV FWD terminal. The problem is my understanding is that ULN2803 is similar to solid state relay in which it uses semiconductor instead of true conductor like mechanical relay -> so i am scared it will end up like my solid state relay - in which the problem was that the relay triggered however the high resistance of the closed circuit reduced the voltage of the 24v FWD REV input and in turn does not activate the voltage sensitive sensor at the inverter. Do you think i should try ULN2803, and does vendor/brand/varieties/model of the chip makes any different?
|
|
|
Post by tronicgr on Sept 30, 2015 15:25:53 GMT
You can try the ULN2803. I had many using them with no problems in sink only vfd inverters. They all work the same no matter the brand. Its just a Darlington transistor array. I haven't managed to blow up any yet, but I did blow up some udn2981 in the past while testing. Still never damaged anything else in the boards except these power chips, thus the reason I use ic bases to easily replace them if they burn for any reason (even static electricity sometimes).
Thanks Thanos
|
|
|
Post by clyevo on Oct 6, 2015 6:38:13 GMT
Ok now that i got all the electronics part and software parts (ian bff and x-sim) working well. To my dismay, I just noticed that one out of the 12 of the rod end bearing/heim joint which is suppose to be right handed thread is left handed. So what happen if i left one arm off just for testing the platform meaning i use 5 arms - 5dof. Maybe sacrificing heave/sway. If so which axis i can get off without -> 1,2,3,4,5 or 6??
Please help me. I wound need to wait another month to get the same rod end bearing from china.
Edit : i think its not possible without reconfiguring the motor position and joint positions, what do you guys think?
|
|
|
Post by Mitch on Oct 6, 2015 20:15:14 GMT
Due to load sharing of the rods, my upper platform was very easily movable by hand into many positions with only one rod disconnected - without support of one rod the platform will likely collapse. Can you get another similar temporary rod end from Amazon or eBay quickly? One of us could probably source something locally and get it to you in less than a month.
Also the number of arms is not the same as dof.
|
|
|
Post by clyevo on Oct 7, 2015 3:03:23 GMT
Due to load sharing of the rods, my upper platform was very easily movable by hand into many positions with only one rod disconnected - without support of one rod the platform will likely collapse. Can you get another similar temporary rod end from Amazon or eBay quickly? One of us could probably source something locally and get it to you in less than a month. Also the number of arms is not the same as dof. yeah due to custom handling the items will still reach me approximately 1 month. you mean the platform can be supported by 5 rods but without the motor active (offline)? you mean the platform with 5 rods will collapsed when motor is active (online)? I just want to test the max range of motion my sim can go, to determine whether the rods will touch my gearbox/motor. And also so that i can determine the dimensions of a wooden box at the middle of the base to keep the electronic boards, inverter and pc. But i am not sure if the rods can support the platform even in offline mode and unsure how it affects the maximum ROM. Thanks.
|
|
|
Post by Mitch on Oct 7, 2015 4:03:48 GMT
It depends on what kind of rods you used and their buckling strength. Your platform is light, you will not have any issues supporting the platform with only 5 rods. It will just be inherently unbalanced and want to fall. I would not operate the motors with only 5 rods attached and supporting the 6th point by hand. If rotation stops on one motor for any reason it will pull it from your grip - too dangerous.
Why not just support your platform at neutral with rods disconnected, move the motors and arms into a configuration and you want to check and then reconnect rods by loosely connecting bolts to hold position.
|
|
|
Post by clyevo on Oct 7, 2015 6:32:19 GMT
Hurm, in that case, i ll just hold the project for another few weeks till parts arrived. The customs is unpredictable, sometimes they released the item sooner depends on their workload. So hoping for that.
|
|
|
Post by clyevo on Oct 15, 2015 3:08:19 GMT
tronicgr, i was thinking could it be that the delay of the relay electromechanical (rated about 10ms-20ms per cycle) are causing the increased oscillation since it is odd my machine oscillates above P of 7. My logic is that the motor overshoot because the signal send to stop it is delayed. I dont know if it matter my acc/dec inverter settings is 0.5s
|
|