|
|
Post by paulg100 on Mar 22, 2016 12:04:31 GMT
did you try fabis auto PID calibration yet? its pretty easy to use.
|
|
|
|
Post by mhensen on Mar 24, 2016 21:45:20 GMT
I did now several times.. But I have some trouble wit it as It sometimes move to unsuspected directions hen I press stop.. So I am not done with it..
I can feel the oscillation on the way down much better then the way up.. "
I still think it is the oscillation that causes the shaking of the platform..
So I think I will retry the fabis tuning with only motors 1 and 2 turned on as that is where I mostly feel the vibration.. (with my hand on the emergency stop!!) So I am still puzzled in the whole PID circus and I have trouble with the fabis tuning due to sometimes a single motor can start kicking, showing a spike in the charts.. perhaps I stop too early with upping the P as I get scared sometimes with the random kicks!!
So if there is anybody reading and have had similar issues please help me to resolve these :-)
|
|
|
|
Post by paulg100 on Mar 25, 2016 7:08:24 GMT
if you didn't see it before this explanation I found very helpful.
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.
Manually I left P set then entered some big value first for I (then reset to zero) and then for D and you can start to see the effect it is having on movement and it becomes much clearer. Looking from others settings and what Fabis autotune produced Its not unusual to have values 100+ so jack it up and see how it affects over and undershoot and affects correction to get back to setpoint.
After playing with BFF in joystick mode and changing these values I was able to start getting a feel for what is going on. I had to stop then as my jitter/noise issue was causing to many issues.
Also it was helpful to google "PID ziegler Nichols" to get some better understanding of how things work. Theres quite a few methods for working out PID and some look fairly easy. Just havnt had the time
to play with it much yet.
Also I left any top bracing of my sim at the moment so that the arms can rotate 360 without a problem, so I don't have to worry about arms running outside values.
Other than that hopefully someone else can help here...
|
|
|
|
Post by mhensen on Mar 25, 2016 8:02:28 GMT
I intend to shut down all motors except 1 and 2 as these are my troubled ones.. and will manually increase de P to see where it all gets shakey..
Then with that P value, lowered just a bit, I will start the fabi autotune to see what the oscillation is and try and see if that will work out better..
As the oscillation is more on the way down I think I will revert the motion so that the tuning is on the way down, the way up is always smooth!! hopefully I can smooth out any oscillation.
I can feel the motion on the way down like riding over ripples or how you might call it, like riding curbstones..
|
|
|
|
Post by mhensen on Mar 25, 2016 22:04:46 GMT
Still can't figure things out.. now want to first find the oscillation point of the P value, but I can raise it but don't feel any up to 45 ? can I continue to go higher? won't I break anything or is impossible to go higher and higher safely..
still the tremble in the front is creating oscillation in the upper frame.. please advice??
|
|
|
|
Post by paulg100 on Mar 26, 2016 6:56:33 GMT
yes you can go higher just keep going until it starts to overshoot/oscillate.
|
|
|
|
Post by mhensen on Mar 26, 2016 13:41:12 GMT
Ok guys,
I need help :-(
it is dark, wrong direction of the light but it shows the problem.. PID I have had all over the place but at P-60 it was ok for the front motors but the others where oscilating (tested by turning of engines 3 to 6) but at P-30 all motors run good but the front is shaking..
Tried all kinds of values on the I ( from I-0 to I-254)
F002 and F003 set to all over in the range from 0.1 to 0.5 but again, online speed from 50 to 70% Change seating position to bring more pressure to the front or lessen the front.. I tried the autotune pid from Fabi, but I never can find the oscillation in the charts as I saw in the video (the first post in the configuration tool thread).
After a move I down I always can feel M1 and M2 searching for the position.. (slowly) this while the other motors are stable..
It feels like I need to have 2 different PID's.. one for the front and the other for the sides...
I still think I need to cut the levers 30/40mm shorter.. from 180 back to 150/140? Any suggestions appreciated,,
The shaking is approx. 4 to5 times a send in the move.
Who has seen this before and or has any idea? It is driving me mad and puzzled!!!
|
|
|
|
Post by carbon on Mar 26, 2016 20:39:50 GMT
All your motors should tune the same (or so close it does not matter)- are all inverters set the same?. Maybe back to basics - You could try, (don't know if you have already done this?) swap the suspect motor/gearbox at the front with a motor/gearbox at the rear and see if the fault follows. Ie. could be a motor/gearbox or mechanical/feedback problem?
|
|
|
|
Post by mhensen on Mar 26, 2016 22:09:57 GMT
The VFD's have all been set up at the same time and resetted and reconfigured at the same time..
I will do some more tests in the coming days..
Fist I will try and isolate the problem and find if it is 1 or 2 that is giving me troubles.. Also will try and reroute the signal cables of the 2 motors in the back of the cabinet.
also could try and rewire the thanos board and fool motors 1 and 2 and wire them to 3 and 4.. ( I will turn of 1,2,5 and 6) just to see if it isn't the board giving me trouble.. It is so weird that is has been motors and 1 and 2 all the time..
tronicgr : perhaps you can tell me if it is possible that only 1 and 2 ( the front engines) could be because of something troubling on the board?
I have changed a chip but that was on the directional side as motor 1 wasn't moving correctly.. could it be that I also have managed to get another chip busted? Or might it be something else..
Edit: I think I will swap the 2 LM chips first .. If motor 1 is still the troubled on it should transfer to engine 5, if it is engine 2 the problem should resolve..(in case it is the chip that is damaged)
Otherwise I will try and rewire the sensor cable of 1 and 2 to be clear of other cables, making sure it isn't noise it is picking up.. If that doesn't change anything I will swap engines.. although I don't believe that there should be any issue as the move is 1 up and 2 down I have no problem.. only when both move down
I can't imagine it is the frame itself that is causing this problems, as I don't feel any binding or measure any discrepancies.
This problem is the same as I had from the beginning.. and the weird things is that I can up the P almost 2 as high the for the other motors before I have oscillation..
I also noticed that on a move to a position it feels and shows that motors 1 and 2 still slowly searching their last final position while the others are already at the correct position.. left/right/left/right.. it is minimal but it shows. This would be correct as the motors 1 and 2 need a higher P then the others
So perhaps this is all bogus rambling but this is what I think of the situation.. So any advice is much appreciated!!!! I want to race:-) So please any ideas and/or suggestions are welcome !!!!
this is stil the issue
|
|
|
|
Post by tronicgr on Mar 27, 2016 13:14:31 GMT
I think that the motors are a little faster going down resulting in braking. Let me explain...
The PID gives target to move a little bit down but because of the backlash of the gearboxes they don't respond fast enough so the PID sets the max speed to the motor to make it move. The motor finally moves but overshoots the target position resulting in hard braking or even oscillation.
What I'm getting here is that you should try the following:
1. Set PID to 20,0,0
2. Set max motor speed to 10% and slowly increase it until you notice the oscillation (applies to both offline and offline speed). Mark the speed.
3. Try again with PID 25,0,0 and mark the speed.
4. Mark down the speed for PID 30,0,0
Now comparing the 'breakdown' speed it should be the same for all PID settings.
Another tip you could use to get around the shaking during movement is to switch off the LCD menu ( DIP switch #1 down) as it takes some processing power to update the LCD disturbing the PID. For this reason in the latest firmware, I disable the LCD updates automatically during online mode displaying a message on the LCD.
To explain to you what's happening in plain words imagine this scenario:
You are in a car going downhill trying to keep constant speed but your car has a idiomorphy: if you press the gas pedal it accelerates to 100% of the speed but if you leave your foot off the pedal the car applies 100% of the brakes on its own... There is no way to keep constant speed but 'pulse' it to try to average the speed resulting in tremendous vibration for the passengers in the car...
Another setting that might worth looking into is the 'step' setting in the menu that sets the 'home position' of the motors, if that is not 99% it might produce the issues I see on the video.
Thanks
Thanos
|
|
|
|
Post by mhensen on Mar 27, 2016 14:26:26 GMT
thanks Thanos,
tronicgr : I've tried your options: but even at the lowest of lowest settings I have the vibrations in the front.. VFD's are all the same. Even downgraded the firmware to V2.0 as I don't know where to search anymore. I think I need to reduce the backlash by making the levers shorter?! Is that a valid option?
I now have F002 at 0.35 and F003 at 0,65 (acc/dec) to make the stop slower, it helps a bit
Did the fabi tuning with just the 2 fonts but I got P80 on the outcome. this as the first sign off oscillation. way outside the 20-30 range.. can this be?
TROUBLE PART!!
EDIT 2: now the VFD seems to be ok, Motor nr 4 had gone in error 30.1 with a bang..I think it came from the motor.. Now the VFD is back on but in autotune without rotation in the second cycle I have beside the whistle noise a ticking noise.. I think the motor is broken! Now the VFD is nagging me error 38, low speed overload.. So trouble all over..:-( I think I have to go to a repair company to let them check the motor.. I don't think I just can change the motor with some other.. then I have 5 of the same brand and 1 other?? I don't think that the PID would work out the same..
EDIT: Seems to be working again..
I will try that, but the problem is that I managed to blow one of the VDF's today?! don't know what happened and don't know it the motor is busted or the VFD. I'll have to contact my supplier next Tuesday.. I have an error 30.1 in my hitachi wj200. some internal error in the circuit between cpu and main circuit.. the error showed me 29A at a 65hz frequency.. but the problem is that I cannot reset the error and don't know how to resolve :-( |
|
|
|
Post by mhensen on Mar 27, 2016 21:17:57 GMT
Ok Now I have an additional issue.. I think I managed to blow my motor nr 4! The VFD is ok, but running autotune on the WJ200 gave me the normal whining in cycle 1 but in cycle 2 it was combined with crackling sound..
EDIT: Not sure if there is a condensator in it.. Will bring it to a specialist!! I suspect a condensator is blown as I heard a load pop when the VFD went into error, even through my headphones.. And the fuse went off.
Is it wise to have it checked by a specialist or can I try to replace the condensator myself, or might it possible to have more issues and thus a specialist is much better.. ?
BTW: a P of 80 indicates to me that I am asking too much of the motors.. so I'll shorten then to 140 or 150 mm .. that is 40/30mm of the current length.. that should lower the PID a bit.. and also reduce some angle of the backlash.. or is this wrong thinking! |
|
|
|
Post by tronicgr on Mar 29, 2016 14:59:40 GMT
I think you should reduce the length of the arms, that will help with backlash and load less the motors. Especially if the gearboxes are low ratio. What is the ratio of the gearboxes you are using again?
Thanks
Thanos
|
|
|
|
Post by mhensen on Mar 29, 2016 16:15:54 GMT
I use 1:60 on 1400 rpm motors.
I think 140mm should be enough for an oculus race rig
|
|
|
|
Post by mhensen on Apr 20, 2016 6:43:19 GMT
So I have been quite for a long time but I should be getting back in business..
I have had some major issues with one of the motors and after having it checked it seemed that there had been a short in the motor it self. Caused by a fragment of metal that was sitting in the edge where everything was blackened..
This also took out the VFD.. So I had to buy a new motor, different brand but with the same specs, but luckily I have had a replacement of the VDF under warranty..
I also shorted the levers by 40mm so my levers are now 140mm long.. this should be better too.. and hopefully it helps in the fight against the jittering..
So all in all .. Need to place the new inverter and mount all of the stop triggers and we are good to go again! and hopefully this time without any issues!!
|
|
