1. Do not share user accounts! Any account that is shared by another person will be blocked and closed. This means: we will close not only the account that is shared, but also the main account of the user who uses another person's account. We have the ability to detect account sharing, so please do not try to cheat the system. This action will take place on 04/18/2023. Read all forum rules.
    Dismiss Notice
  2. For downloading SimTools plugins you need a Download Package. Get it with virtual coins that you receive for forum activity or Buy Download Package - We have a zero Spam tolerance so read our forum rules first.

    Buy Now a Download Plan!
  3. Do not try to cheat our system and do not post an unnecessary amount of useless posts only to earn credits here. We have a zero spam tolerance policy and this will cause a ban of your user account. Otherwise we wish you a pleasant stay here! Read the forum rules
  4. We have a few rules which you need to read and accept before posting anything here! Following these rules will keep the forum clean and your stay pleasant. Do not follow these rules can lead to permanent exclusion from this website: Read the forum rules.
    Are you a company? Read our company rules

Showroom My 3 DOF seat mover + GS-4 + Simvibe

Discussion in 'DIY Motion Simulator Projects' started by Avenga76, Jun 29, 2014.

  1. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    20,464
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    144,620Coins
    Ratings:
    +10,741 / 52 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    Fortuitous timing. I am having a similar issue with one of my 60:1, deforming the brass gear. Fast reversals punish these motors and irrespective of the error issue with high speed motors I think we need more caution in refining the PID and other settings for high torque gearboxes.
  2. Avenga76

    Avenga76 Well-Known Member

    Joined:
    Jun 29, 2014
    Messages:
    518
    Occupation:
    Network Engineer
    Location:
    Auckland, New Zealand
    Balance:
    3,260Coins
    Ratings:
    +853 / 6 / -0
    My Motion Simulator:
    2DOF, DC motor, JRK
    I have my motors and post out of my rig so it is really easy for me to test either the 4,300RPM or 5,300 RPM motors.

    If you can make any suggestions for PID the I will test them.
  3. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    20,464
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    144,620Coins
    Ratings:
    +10,741 / 52 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    Up until now I have not bothered with Integral action but when I do put my sim back together I do intend playing around to try and soften the reversal of the motors.
  4. Avenga76

    Avenga76 Well-Known Member

    Joined:
    Jun 29, 2014
    Messages:
    518
    Occupation:
    Network Engineer
    Location:
    Auckland, New Zealand
    Balance:
    3,260Coins
    Ratings:
    +853 / 6 / -0
    My Motion Simulator:
    2DOF, DC motor, JRK
    I tried a bit of Integral action last night but it didn't work well when turning the POT by hand. I couldn't get the motor to stop when move the POT back to the middle neutral position. Maybe it will work better when the motor is driving the POT.

    Bumping my video to the current page

  5. bsft

    bsft

    Balance:
    Coins
    Ratings:
    +0 / 0 / -0
    Fellas, with the JRKS, Use P and D only, do not use I. I does almost nothing to soften the reversal, is more for really twitchy motors, but to be honest, I have never had to use I . @eaorobbie can enlighten us more when he has a chance.
    Stay working with P and D and the motor power settings.
  6. Avenga76

    Avenga76 Well-Known Member

    Joined:
    Jun 29, 2014
    Messages:
    518
    Occupation:
    Network Engineer
    Location:
    Auckland, New Zealand
    Balance:
    3,260Coins
    Ratings:
    +853 / 6 / -0
    My Motion Simulator:
    2DOF, DC motor, JRK
    Yeah, up until now I have only been using P and D. Last night was the only time that I have tried any I. The PID settings don't really work the same when you are turning the POT by hand because the POT is not reacting to the movement of the drive shaft.

    The main thing that has worked for me is limiting the duty cycles. From my early tests on this new motor it looks like I can run about twice the maximum duty cycles on the smaller motor so I don't know if it is better to run the slower motor at a higher duty cycle or the faster motor at a lower duty cycle. I will only be able to tell that once I get the motors back in my rig and test how the motion feels with the new motors.

    I think the one thing we can take away from these test is that the faster the motor is then the more motor driver errors you get and the more you have to limit them. So for example I with the 5,300 RPM I have to limit the duty cycles about 50% before I can do a single reverse with out an motor driver error. But with the 4,300RPM motor I can do a single reverse at 85% duty cycle, and the 3,600RPM motors that most people use would be able to do a full reverse at 100% duty cycles.

    The other thing I have found is that rapid reversals in quick succession will also cause motor driver errors. For example the 4,300RPM motor can do a single reversal at 85% duty cycles but after 4-5 quick reversals it will start giving motor errors so to get that motor running well I would probably need to limit it to ~66% duty cycles. The same with the big motor, it can do a single reverse at 50% duty cycles but with rapid direction changes you need to drop it down even lower to something like 33%.

    So there is a point where these motors are too fast for the JRK to handle, it is now just a matter of finding where that point is and figuring out if it is better to use the bigger motor at lower duty cycles or the slower motor at higher duty cycles. I suspect it will be the smaller motor.

    If anyone else has any idea on how to minimise the impact of reversing a fast motor then let me know because now is the perfect time to experiment while I have both motors to test.
  7. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    20,464
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    144,620Coins
    Ratings:
    +10,741 / 52 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    It may be the Integral may not be part of the solution, but with these big torque motors there is the need for a bit different approach to JRK settings, or they deform the brass gear leading to excessive slop.
  8. Avenga76

    Avenga76 Well-Known Member

    Joined:
    Jun 29, 2014
    Messages:
    518
    Occupation:
    Network Engineer
    Location:
    Auckland, New Zealand
    Balance:
    3,260Coins
    Ratings:
    +853 / 6 / -0
    My Motion Simulator:
    2DOF, DC motor, JRK
    Yeah, I think it might be that we have to use unusual settings for these high torque motors to get them to behave. my next round of testing will be with the gearbox attached and the POT connected back to the shaft so the feedback loop will be working again.

    I won't connect my rig up to the other shaft so I will still be able to trash the motor without shaking the rig to pieces.

    Any idea of what a good starting point for the I would be on a bigger motor. I figure I might as well test it while I have the rig in pieces, even if I do go back to just using P and D
    • Agree Agree x 1
  9. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    20,464
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    144,620Coins
    Ratings:
    +10,741 / 52 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    I have to say I was a bit shocked to be told that 90 newtons is the limit for these brass gears, I certainly have not been able to find reference to it on the Motion Dynamics site but that was the email outcome from the manufacturer when I sent my gearbox back.

    Unlike the 25:1 boxes, for big torque gearboxes there needs to be a controlled reversal to manage the torque spike, rather than just the fasted reversal that can be accurately sustained.

    I think the aim is a difficult balance to slow the reversal without killing the motion experience completely. The reason I had in mind to fiddle with the Integral settings, which could be totally wrong, is that it can be thought of as a kind of timing control, countered by D. There may well be better ways to do that other than the Integral. Certainly design is the obvious one. If the angle by which a force is applied increases then the torque applied decrease.
  10. JRoque

    JRoque New Member SimAxe Beta Tester Gold Contributor

    Joined:
    Jun 5, 2014
    Messages:
    21
    Location:
    Miami, FL USA
    Balance:
    282Coins
    Ratings:
    +20 / 1 / -1
    Hi,

    Could the errors be related to voltage peaks during fast motor reversal? For a quick test, drop your input voltage to something lower, ie: 9V, and bump up the duty cycle. Lowering your input voltage will also slow down your motors so it's not a perfectly symmetrical test but it will give you an idea if the issue with with BEMF from rapid motor reversal.

    If the errors are substantially improved, while the motors don't seem too slow, you might want to try to attach a pair of 1.5KE22A TVS to your motor terminals like this:
    mt1--><--mt2, where mt = motor terminal
    Essentially the TVS face each other with their anodes connected to each motor terminal. There are more complex (and better) diode networks you can use but the above provides basic clamping at the motor terminals when voltage reaches 22V.

    Of course, your errors could be coming from something completely unrelated, like overshooting the pot due to the additional mass (and inductance) of the larger motor.

    JR
    • Informative Informative x 1
  11. Avenga76

    Avenga76 Well-Known Member

    Joined:
    Jun 29, 2014
    Messages:
    518
    Occupation:
    Network Engineer
    Location:
    Auckland, New Zealand
    Balance:
    3,260Coins
    Ratings:
    +853 / 6 / -0
    My Motion Simulator:
    2DOF, DC motor, JRK
    That is a good idea. I don't have a variable voltage power supply so I am stuck with 13.8V

    I haven't tested for transient voltage spikes lately but when we first had these problems we hooked the motors up to a oscilloscope and we saw massive spikes in current up to 70A. I can't remember saying any spikes in the voltage though. The JRK's have protection for transient voltage up to 40V.

    My multimeter is too slow to see anything. I just tried it now and the highest voltage I saw was around 8-9V.

    My thoughts is that it is the current that is causing the errors, not the voltage.

    One of my friends has a variable power supply and a scope that I can use so I might buy another JRK and setup a little experiment on the bench.

    If you overshoot the POT then you get a different error, you get a feedback disconnect error. In my video above you can see when I am turning the POT by hand I turn it too far several times during the test and get the feedback disconnect error. This doesn't happen on my rig because I left plenty of room before I hit the absolute max setting. If it does overshoot then it will give you the feedback disconnect error and slow the motor down to whatever you had set your max duty cycle outside of feedback range to. The motor driver errors are very different.

    These are the motor driver errors that I get.

    • Bit 2: Motor driver error
      This error occurs when one of the motor driver’s fault conditions are triggered, and the motor driver shuts down the motor and reports the error to the jrk’s microcontroller. This error also occurs when the jrk is connected to USB and motor power becomes disconnected. When this error occurs, the jrk will try to automatically recover from it by toggling the appropriate lines on the motor driver. The jrk 21v3’s motor driver fault occurs on under-voltage, over-temperature, or over-current conditions. The jrk 12v12’s motor driver fault occurs when it detects that motor output A is shorted to ground or VIN.
    And this is the error you get if you overshoot the POT.

    • Bit 5: Feedback disconnect
      This error occurs when the feedback is above the Absolute maximum or below the Absolute minimum (these parameters can be set in the configuration utility). The absolute maximum and absolute minimum can be set using the configuration utility. Additionally, when using the Detect disconnect with AUX option in Analog Feedback Mode, the jrk periodically tests to see whether the feedback potentiometer is disconnected and generates this error if it finds that it is (Section 3.c).
    • Informative Informative x 1
  12. Avenga76

    Avenga76 Well-Known Member

    Joined:
    Jun 29, 2014
    Messages:
    518
    Occupation:
    Network Engineer
    Location:
    Auckland, New Zealand
    Balance:
    3,260Coins
    Ratings:
    +853 / 6 / -0
    My Motion Simulator:
    2DOF, DC motor, JRK
    Okay. So I have installed the new motors and my first tests are extremely positive.

    I can run the motors a lot harder now. I have the max duty cycle set to 500 and the max acceleration set to 300. Before on the old motors the most I could reliably use was 250 duty cycles and 100 acceleration, and even then it was prone to motor driver errors so sometimes I would even have to go do to 200/100 on bumpier tracks.

    The new motors are way more lively, like extremely so. It feels like twice as fast now and it is quick enough to react to small bumps and everything. Feels like it should do now. I could get this type of speed out of the old motors but they would do a chain of motor errors every second corner.

    Now I do still get the odd motor error now and then if I go through a period of really heavy activity but this is rare and way better than the old motors.

    Now to the first downsides that I have felt.

    It just feel less powerful. Like it has the speed now but not as much torque as the old motors. I think if I could get the old motors running at this speed reliably then that would be the bext of both worlds.

    The second downside is that the motors get warm now. I did about an hour of testing and the motors were warm to the touch, not hot, just warm. The old motors could run all night and still feel cold to the touch. I will have to monitor this to see how they are on longer stints.

    But all up it is a massive improvement and with a bit more tweaking I think I can get these motors running perfectly. I will do some more testing and report back.
    • Like Like x 2
  13. welen123

    welen123 Member Gold Contributor

    Joined:
    Jan 10, 2015
    Messages:
    98
    Location:
    HongKong
    Balance:
    35Coins
    Ratings:
    +57 / 0 / -0
    My Motion Simulator:
    2DOF, DC motor, Arduino, Motion platform
    very good job! i like it!
  14. Avenga76

    Avenga76 Well-Known Member

    Joined:
    Jun 29, 2014
    Messages:
    518
    Occupation:
    Network Engineer
    Location:
    Auckland, New Zealand
    Balance:
    3,260Coins
    Ratings:
    +853 / 6 / -0
    My Motion Simulator:
    2DOF, DC motor, JRK
    Thanks.

    So I did some more testing.

    To test the heat buildup I did 35 minutes non-stop in iRacing with the Skippy and Lime Rock, about 30 laps.

    The motors were warm to the touch but never got any hotter.

    I played around with the PID. I found that the motors are very sensitivity to P. I run both the new motors and these ones at P=0.5 and D=0.6. If I increase it to P=0.6 then I start to get the odd motor driver error and if I go to P=0.7 then I get more motor driver errors.

    I am happy with the performance at P=0.5 on the new motors. I get about twice the speed and no motor driver errors with the new motors.

    I am really pleased with these experiments. It seems to prove that my theory of the faster motors having problems reversing on the JRK's was correct.

    I will keep these new motors on my rig because they run twice as good as my old ones but I am going to build up a test rig on my bench so I can experiment with either a more powerful controller or ways to get the JRK's to work with bigger motors because I would like to find a way to get my bigger motors running at full speed because they have a ton more torque.

    BTW. Here is a quick size comparison so you can see the difference between the 4,300RPM motors and the much bigger 5,300RPM motors.

    [​IMG]

    [​IMG]
    • Informative Informative x 1
  15. noorbeast

    noorbeast VR Tassie Devil Staff Member Moderator Race Director

    Joined:
    Jul 13, 2014
    Messages:
    20,464
    Occupation:
    Innovative tech specialist for NGOs
    Location:
    St Helens, Tasmania, Australia
    Balance:
    144,620Coins
    Ratings:
    +10,741 / 52 / -2
    My Motion Simulator:
    3DOF, DC motor, JRK
    Very informative, it certainly does look like very high RPMs are a problem for the JRKs.

    I will be looking forward to your further bench tests with the big motors.
  16. Avenga76

    Avenga76 Well-Known Member

    Joined:
    Jun 29, 2014
    Messages:
    518
    Occupation:
    Network Engineer
    Location:
    Auckland, New Zealand
    Balance:
    3,260Coins
    Ratings:
    +853 / 6 / -0
    My Motion Simulator:
    2DOF, DC motor, JRK
    Thanks.

    I am very pleased with how these experiments went. It has easily doubled the speed and intensity of my simulator, if not more because I can run at a much higher duty cycle and get rid of the brake duration pause that I needed to get the bigger motors to work.

    I am not sure if I will be able to ever get the bigger motors to work with the JRK's, I might have to look at a different controller. I think they just draw way too much current, even though both my 4,300RPM and 5,300RPM motors are rated at 160W I think this is a gross underestimation of how much these motors really draw.

    When I have it all set up on the bench I will be able to do more thorough testing. On the rig we had a lot of background electrical noise with everything else going on on my rig so on the bench I will be able to isolate everything and hook it up to a clean power source for my tests.

    If my JRK tests fail I will experiment with some other controllers and/or Arduino.

    Ideally somewhere down the track when I find a suitable controller I will swap back to my faster motors but at the moment I am just amazed by how much harder I can drive these smaller motors, easily doubled the speed and intensity.

    BTW, Thanks guys for listening to my rants about big motors and JRK's. It has been really great having people to discus my findings with.
    • Like Like x 2
  17. Alexey

    Alexey Well-Known Member

    Joined:
    Sep 23, 2014
    Messages:
    452
    Occupation:
    Electronics Technician
    Location:
    Adelaide, Australia
    Balance:
    8,060Coins
    Ratings:
    +620 / 2 / -0
    My Motion Simulator:
    3DOF, DC motor, Arduino
    While the rating of the motor might be 160W it's the stall current that can end up being over double the rating of the motor. The 200W motion dynamics motors have a stall current of 30A and I have personally seen them draw that whilst looking at a current clamp.
  18. Avenga76

    Avenga76 Well-Known Member

    Joined:
    Jun 29, 2014
    Messages:
    518
    Occupation:
    Network Engineer
    Location:
    Auckland, New Zealand
    Balance:
    3,260Coins
    Ratings:
    +853 / 6 / -0
    My Motion Simulator:
    2DOF, DC motor, JRK
    Yeah, I saw mine draw 70A when I had it hooked up to an oscilloscope. I have been thinking of some type of current clamp or current limiting device but I don't have the electronics knowledge to be able to engineer one. It was one of the things is was going to research when I start my bench tests. If you can recommend a way of limiting the current then I can test it out.
  19. Alexey

    Alexey Well-Known Member

    Joined:
    Sep 23, 2014
    Messages:
    452
    Occupation:
    Electronics Technician
    Location:
    Adelaide, Australia
    Balance:
    8,060Coins
    Ratings:
    +620 / 2 / -0
    My Motion Simulator:
    3DOF, DC motor, Arduino
    The only sensible solution would be some way to drop the duty cycle as the current rises or is reaching the output maximum of the controller. I really can't think of a way to limit the current that wouldn't involve you buying gear that's more expensive than just buying beefier controllers. Your problem is like buying a shoe that is too small for your feet and then cutting off your toes so that you can fit your foot in. By limiting the current you are reducing torque.
  20. Avenga76

    Avenga76 Well-Known Member

    Joined:
    Jun 29, 2014
    Messages:
    518
    Occupation:
    Network Engineer
    Location:
    Auckland, New Zealand
    Balance:
    3,260Coins
    Ratings:
    +853 / 6 / -0
    My Motion Simulator:
    2DOF, DC motor, JRK
    Yeah, I think it is just pushing the JRK too far. I have limited the max duty cycles and acceleration duty cycles but it slows the motor too much. The smaller motor works much better because I can run it at a higher duty cycle.

    I have been looking at other controllers for a while. I like the look of the Roboclaw 2x60A 120A peak controller but it doesn't work directly with SimTools so I would have to use an arduino or something like that. But then some other people have said that the Roboclaw is no good. Other people use other big controllers but I have no idea what ones are good and what works well with SimTools.

    Now that I have the smaller motors installed in my rig I can test a new beefier controller on the bench and swap it over when I am happy with the new controller and my bigger motors.

    I just need to do some research and pick the right controller so I can build a little mockup test motion simulator on the bench.

    If anyone has any experience with beefy controllers, 60A+ then can you let me know.