|
|
Post by fabi on Jan 5, 2015 20:49:11 GMT
my English is bad and the video is difficult to understand.
I probably wrong Startup with the use of the software.
Video can be directly with the startup configuration would be more meaningful? Hello thanks a lot for the videos. Please download the latest version I've just published (see at the begining of this thread). I've fixed the issues of 2,3,4,5 DOF not working. Now it should also work with 2 DOF config. Let me know. Thx Fabi |
|
|
|
Post by riton39 on Jan 5, 2015 20:57:01 GMT
ok ! I'll test! thank you! videos it's easier to see the problem  |
|
|
|
Post by fabi on Jan 5, 2015 20:57:05 GMT
hi Fabi, I agree that for the PID setup I have to seat on the simulator but now for me it's not possible until the overshooten problem not solved. i think that the problem is when the position is overshooten the amc try to reach the new position but keeping the same direction ( CW or CCW depends of the actuator) .that means a complete turn to get the new position and not change over the direction. Hello Gilles please download latest version (begining from the thread) There I disabled the buttons for manual move until heave evaluation has been completet successfull. That means the plattform starts moving from there. Also I've set Min to 20000 and Max to 40000 so we are somewhere in the 30% mid range for "jumps" to be a bit more save. Let me know Thx Fabi |
|
|
|
Post by vicpopo on Jan 5, 2015 21:12:00 GMT
Thanks Fabi !
Will give a try in the next minutes and give a report !
|
|
|
|
Post by riton39 on Jan 5, 2015 21:14:47 GMT
it is the same problem, it still does not work.
edit:
the minimum is now possible to 0 and the maximum is 65535.
"The feedback limit" takes into account
the maximum?
I feel that with 10% feedback limit and maximum 65,535 my cam goes up ....
|
|
|
|
Post by vicpopo on Jan 5, 2015 23:35:25 GMT
Thanks Fabi ! Will give a try in the next minutes and give a report ! Ok made a few tests and the results.I start the Pid auto tune from the mid postion each time.What I not yet completly understood is to start the auto tune procedure .I pushed the up button in the manuel menu and then launched the auto tune after. 1- no overshooten position 2- always crash for the osciloscope when adjusting the left cursors ( look pic) 3- made three times the procedure for'auto tuning and for each test never get I or D values , only P term ( from 30 to 45).I don't know why , maybe I stop the auto tune test before great oscillations , I don't know ! look at the pics from,sensors feedback positions. For all these tests I decreased my max motor speed to 50% to keep the sim safe in case of emergency. Pics will follow.    |
|
|
|
Post by riton39 on Jan 6, 2015 7:36:53 GMT
Can you explain "Tick dashed" "Highlight actor position axis" and "Hide actor position axis"?
thanks.
Can you tell me how many minimum peak it take?
on the video you indicate 8 not counting the first, right?
there is a minimum? Approximate or it's got to be precise?
edit: now I can no longer make and use "evaluate heave position platform"
both versions no longer works.
2 DOF or 6DOF, I was unable to validate "evaluate heave position."
Yesterday, I arrived with version 1.0
and now impossible with 2 versions |
|
|
|
Post by riton39 on Jan 6, 2015 8:44:01 GMT
I have a note:
to take the test, you must increase P up getting oscillation.
initially the torque is not enough.
it is necessary for some variator
to have a minimum speed to have a minimum torque.
I arrive at the position to P = 20
maybe the speed setpoint should be more important?
|
|
|
|
Post by vicpopo on Jan 6, 2015 19:58:10 GMT
Hi Fabi ,
I tested your last update and now the overshooten problem never happened.It's a good think ! Hope this point solved.
Always sometimes osciloscope crashed but seems to be a little better or it's me who adjusting cursors with more accuracy ;-)
But I made the auto tune process several times and I get only a term P from 40 to 80 and always I = D = 0 .I measure the period like you made in yout video tuto and always the values in autôtune window still stay to XXX .I ask myself if the calculation is ON !?
For the rest I noticed anything wrong.Just forThanos maybe because when increasing the P to achieve oscillations I noticed that the motors 2 4 6 was shaking and the motors 1 3 5 no oscillations.My setup on my 6 dof is like i suppose everyone to pos 135 neg 246.
Could you give me,feedback for that ?
|
|
|
|
Post by riton39 on Jan 6, 2015 22:29:43 GMT
the dropbox link to page 1 changed name, it's a new version? different from the 1.01?
|
|
|
|
Post by tronicgr on Jan 6, 2015 22:46:10 GMT
the dropbox link to page 1 changed name, it's a new version? different from the 1.01? Yes, its now 1.0.0.3 |
|
|
|
Post by riton39 on Jan 6, 2015 22:49:23 GMT
oh ! ok I'll test tomorrow,
it's time to go to sleep in France !!
|
|
|
|
Post by tronicgr on Jan 7, 2015 6:39:02 GMT
Hi fabi, Can you check the calculations for the I and D ?? I did a measurement and I got these values:  <Bug>: I get again freezing of the oscilloscope if I move one of the sensors to the lower position ( near zero). <bug>: if I click OFF on the oscilloscope the values jump to the start but does not clear the screen, leaving a mess... <request>: Can you change the default value of the "frequency of the Burst Feedback" to 16 on the start of the program? (I could enforce it in firmware if not...) It seems that if its set to 100hz, the counter that moves the motors up or down, is slowing down too much, perhaps skipping steps as well. So to help people understand the procedure: 1. Do the sensor evaluation, 2. In the PID Tuning tab, click on the button that states that you understand the risks of using that PID tuning code. 3. Open the oscilloscope and press the ON button ( it will display smooth waves as long the timeout is not reached, set the timeout of the 6DOF parking position to 60 or 90 seconds from the Controller config tab. 4. From the manual motion tab, press the up or down, to get the motion going. 5. Back to PID tuning tab, click on increment P value. And keep increasing until you see oscillation on the motors. 6. When the motors start oscillate press the button to stop the motion. 7. Click on two of the recorded peaks on the oscilloscope, and set the correct number of peaks between the two points. 8. You should see the correct PID calculation on the PID tuning tab. Thanks Thanos |
|
|
|
Post by riton39 on Jan 7, 2015 7:46:07 GMT
I have the same problem.
the first time I got to spend "evaluate current position Platform heave"
but then in auto tuning, he would not confirm when I clicked on validation.
I tried a second time restart the app and now it hangs on the manual step "evaluate current position Platform heave"
the same in 2DOF or 6 DOF
no way to go further.
may be the reading of the position?
it awaits potentiometer value anyway?
may be there should be one DOF, 1 axis, 1 potentiometer.
2 axis potentiometer 2 ...
|
|
|
|
Post by riton39 on Jan 7, 2015 7:58:57 GMT
when I pass 6DOF to 2DOF the AMC display always shows six potentiometers.
the app shows 2DOF.
|
|
