Another Playseat 2DoF -> 3DoF

I will have a sqidge at this when i ger home from work. Easier to see on a puter screen.

@noorbeast - not sure I quite understand where a . should be put in?

Did a bit of googling, and on the Pololu forum one of the staff commented on a similar problem:

From my days with RC this got me to thinking that maybe the big motors fools the JRK to believe it is a short when reversing.
Googling a little more brought my back to this forum
@Vef445 - had the same problem, http://www.xsimulator.net/community/threads/time-to-open-my-own-thread-my-2dof.7208/page-2
and @Avenga76 too. They found out that 4000+rpm motors gave this error with JRK's. The solution (for them) was to add a little "Brake duration" on the motor tab - http://www.xsimulator.net/community...r-gs-4-simvibe-gauges.5596/page-12#post-75672
"So yeah, I think the key points are that if you are going to run bigger motors like mine than limit the duty cycles and add a brake duration so the controllers can handle it"

I'll try that when I get home tonight. My motors are 4800rpm so it makes sense that this could be the problem.

ah, 4800 rpm motors, ok then.
And I looked at the PID, theres nothing in D
as @noorbeast suggested, run 0.39 p , 0 I, 0.5 D

Played with the settings last night, but can not get the results I want.
Either it is slow and sluggish but without error codes, or quick and crisp but throws out error codes (motor driver)

I bought the ZYT-90 with 60:1 from motion dynamics because I believed they would be a perfect match with JRK's (wanted a plug-n-play setup) guess I should have done some more reading
But I've been thinking... first off, the connectors on the JRK's are really small, I will solder them off and solder the 14 AWG cables directly to the board.
Secondly - what if we add a little capacitor between the motor connectors to avoid the big spikes in current on acceleration/reversal, to sort off "clean" the signal. Together with a huge capacitor on the 12v supply to make sure it gets plenty off input power.

I've been googling to find successful JRK settings for the ZYT-90 motors, but fail to find any (will continue searching). In theory a setting that works on identically the same engine (ZYT-90) with 60:1 gearbox should be the same (in JRK config, not Simtools obviously because of different sim designs).

If these improvements does not help I will configure it with the settings that work without error codes (slow and sluggish) and mount the PlaySeat, so I can try it and see if maybe it is "good enough" and try to source smaller motors

Please post screen shots of the utility tabs on both motors and we can advise.

@flyingdingy , apart from PID settings ( I am not sure what you have now), try these settings on the MOTOR

also , can @Avenga76 please advise of settings help?

Here are my latest settings.

For tuning fast motors I would start off with really low settings. Like try Max Duty cycle and acceleration at 150-250 then try that. If that works then try going up to 350, then repeat until you find the sweet spot.

If find that the PID is also very important, again, start off low and then slowly increase them.

These faster motors are really pushing the limit of the JRK's so it is just a matter of experimenting and finding out how close you can go to the limits.

I would say the first step is turn all your settings down so you can prove that you motors can work at a slow speed then work on increasing the settings until you get a nice balance.

If you check out my build thread I have tons of videos which show exactly what is happening when you get these errors. I did a lot of testing with controlling the motors by hand and demonstrating the motor errors with quick reversals.

My big motor vs small motor video shows the problem quite well

Thanks a million for taking the time and helping me! I will do exactly what you advice, hopefully tonight when I get home.
Last night I pulled the JRK's off my SIM and replaced the heatsinks with alot larger ones (the heatsink I had on them got hot just from moving the engines with no load) and I have soldered one of the JRK's with the cables directly onto the board. This way I can test if it actually makes a difference (since the other JRK still has the blue connectors)
If the newly modded JRK performs better I will re-solder the other one as well. (I have soldered on high current connectors on the cables instead, supporting 90 amps (XT90 from hobbyking) and are using 14AWG cables) but to be honest I doubt it.
But either way I had to pull them off to replace the heatsinks so wanted to try to see if it makes a difference.

As a general question - would the Motomonster/Arduino combo be a better choice for me with these motors? Will the JRK get the full potential out of them?
For $50 (and free shipping) I get 4 x MM and 2 x Arduinos. Running 2 x MM on one motor gives 60A cont and 120A peak... I am no stranger to Arduinos (use them on my DIY drones) and if they are that much stronger than the JRK and 1/8 of the price... I think the JRK's are easier to sell than the motors. Pluss with those motors I know I have plenty of power. Guess I should have read more before I ordered a year ago

http://www.xsimulator.net/community...bridge-60a-continuous-and-120a-peak-12v.6537/

Thank you. Yes, I've read trough your thread, impressive!
The problem you had is exactly the one I am having.
I was a little sorry to learn that you ended up changing your motors and had a better experience with the smaller motors.
It appears that the 4800rpm motion dynamic motors are closer to 5000rpm... what I really need is to find a way to reduce the current spike during reversing as that seems to be the biggest problem.

As expected, soldering the wires directly to the JRK did not make alot of difference.
It does however give about 20-30% less motor drive errors with the exact same settings that is on the other JRK that is using the terminal connectors. (40 motor errors from one lap at Assetto Corsa while the other JRK has about 60-80 errors on one lap at the same track and car, but I guess that margin of error can be related to something else, not exactly the same driving, but it seems rather consequent.)
So for people using JRK's I actually would recommend soldering directly to the JRK, even though the difference is more or less negligible.

The old cooling:


Soldered directly to the JRK (first picture is WIP)

"New" setup - the one in the middle is a "spare" that is planned to use on the traction loss


I fiddled with the motor settings and PID's and got to the point where I felt it could be working and completed the mounting of the PlaySeat to give it a try

I do not have a video now, the clip my GF shot with my cellphone was so dark that it is no point in publishing it
But it is fun, it rocks me pretty good. I need to fine tune everything.

To get it to work I had to set the Derivative Coefficient to 1.8 to stop the motors from overshooting. When they did not overshoot as much the motor drive errors also was reduced alot.

But I must admit, I have already ordered 6xMM's and 2x Arduino Uno R3's
I will run 2 MM's bridged on each motor.
I have not yet decided if I should try to sell the JRK's or keep them for maybe a second SIM down the road (on which I will take care to order max 4000rpm motors)....

I recorded another video about PID setting which might help you with your motor driver errors.

I show a few different PID settings and there effects on the motors, the second example is an aggressive setting and you can see the motor driver errors happening then.

I have the CTC at 70mm.
I've pre drilled holes at 30 - 50 - 70 and 90 so it is easy to change.
The sim is pretty good balanced, there is practicly no load on the levers when I am seatet. If I rock my body gently back and forth I can feel that I am on the pivot point.
But the MDE's comes even without the levers attached/no load.
I will look at your movies and play a little more with the settings.
But I am pretty determined to try the Arduino/MM combo when it arrives, linking two MM's gives 120amp of driving force so at least I dont have to worry about that (next on the worry list then is the PSU's , but that will be taken care of by a 1 fahrad condensator)

As of now there is only one thing I am actually wondering about, the rest is just to play more with the settings....
I am not 110% sure that my simulator moves the correct direction...

In SimTools - output testing - Dof Output :
which direction should the SIM move when I move the different sliders left and right?

i.ex. if I drag the Sway slider all the way to the left, is it correct that the SIM tilts to the right? (opposite of the roll)
is there a "guide" that tells which direction the sim should tilt when moving the different sliders?

Not really, DOF output testing to make sure a particular force works, problem is, it might move right in output testing, but could move wrong because of the game plugin.
Thats where overall, its best to set up a axis assignment and go into game to test if its moving the right way.
Sway is opposite the car steers, yes, so turn to the right, seat leans to the left, turn left , seat leans right.
Roll and the actual camber of the track, so an oval track with a bank will lean you into the corner.

Here is a guide to how the sliders work

http://www.xsimulator.net/co-ordinate-system/

That is correct, sway moves the opposite way to the roll axis, so if you move the slider to the left then the rig will move to the right.

As bsft said, it is easiest to test it in game because you might have to set the axis differently in game. Sway should lean your seat towards the outside of the corner and roll should move the direction of the bank, so go find an oval and if the back is tilted down to the left then the seat should move to the left.

I have the MDE's happen with no load also, this is shown in the big vs small motor test I posted earlier, then the other video I show some of the PID settings that I need to run to get rid of the MDE's.

When I did my initial testing I saw spikes of 70A per motor so I think the JRK's are just too small for these motors.

In terms of power supply, I opted to run three 12v 9Ah UPS batteries which take all the load then a 13.8V 40A power supply to keep the batteries topped up. That works really well because the average load on the motors isn't that high so the power supply is plenty big enough to keep the batteries charged and the batteries can easily take the big spikes.

@Avenga76 what did you measure the average load on the motors?

Man, now you are asking the hard questions.

My old motors were on average about 20-25A for both motors combined from memory, but the peaks were up around 70A.

Running the batteries allows me to run a much smaller power supply because I don't need to worry about those large peak values. Plus I have the added benefit of being able the run on batteries alone, I haven't tested how long I can run off batteries but I often use just batteries if I am testing or doing a short session.

I think my current power setup should be powerful enough to had another motor for my traction loss and still not be over my 40A power supply.

I haven't tested it with my new motors yet, I installed a shunt inline after my batteries so can use that to test the overall amps

Gauges and shunts come from Ebay for about 7-9$ depending on the Amperage rating of the shunt.