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

I just did mine by measuring.

I have an iPhone 6 Plus and it has a really nice level built in which I used for checking if some of my other bits are square. I didn't try it on my arms though.

I went for movement at the center of the bolt rather than angle because the angle would change depending on how long your arms are. My arms are 85mm CTC and I have 96mm of travel from top to bottom. I don't know what that works out to in angles. I will try the level tonight.

I think @Avenga76 your motors may have a BIG stall current .
Ive found that seeing other problems people posted.
The rated current is fairly low and the stall current is HUGE.
Sometimes suppliers lie in order to get a product sold.

@bsft Yes, that is exactly what I see when I hook it up to my oscilloscope.

It peaks over 70A in stall or when reversing. This is why I limited the duty cycles. It is also why I went with my battery configuration so it would handle the HUGE stall currents.

The JRK seems like it can handle a few of these huge spikes but too many of them in short succession makes them trip out.

The weak part of the system is now the JRK's. This is why I was thinking of those Roboclaw 2X60A controllers because they would be able to handle the stall currents.

I am fine running with the JRK's because they work most the time but it would be good if in the future I could swap out to the Roboclaws and unleash the full power of my rig.

Rather "unbelievable", only 160 Watt and 70A+ in use...you know I have winches which are rated at 750W (theoretically), they are running at ~40 - 60A max. current and peaks don't know (Arduino, MotoMonster). Not to forget in mind more weight more current. A problem could be the very fast returns of the levers (oscillation). Nevertheless your motors could be over the specs, too: P = U * I = 13.33A. If Roboclaw will not be available for us, I could imagine to build a "Quad MotoMonster Driver". If two vnh2sp30 chips can be driven parallel as one driver than it works with four chips as well = 60A current, 120A peak. Only theoretically at the moment, but IMO worth a testing. Cost: 4 x Motomonster = 48$, one Arduino 15$, wire, cooler etc. Again, my suggestions are only if you are going to like an experiment and the RoboClaw will never come to us...

I agree with @Pit : "quad motomonster" driver per motor could be a nice and cheap solution

I have send a request to pololu, first about roboClaw and second about Quad MM.

Yeah I was quite amazed by the 70 amps also.

The peaks are very brief, we installed a shunt and hooked it up to an oscilloscope.

The average is quite low, only around the 13A mark so the continuous running is only about 160W.

My battery charger only needs to run at around 17-23A to keep the batteries topped up while I race so the overall draw isn't too bad. it is just these short peaks in very bumpy areas.

I just did a whole night of racing and everything held up fine and it felt great but I did have 2223 motor driver errors on one controller and 2224 on the other.

That number might look high but they happen really quickly, You might get like 50-100 by just going off the track once.

And it does make it look worse than it was because I was just racing in GSC with some friends in a private server so I would pull off the track to chat and that would add another 50 errors to the count. If I know I am going to go off track I usually just hit my kill switch.

For normal track racing it is totally fine. I don't get hardly any errors while on track and if I do it will only be one or two and they are so quick that you don't feel them.

It that whole racing season I didn't feel any of those errors happening, they were too quick.

I am also using the settings that Dave suggested and I think removing the current limit has made it worse.

Ignore the No Power errors because I have my kill switch engaged.

I had the same problems with ZYT90 motors from UK and JRKs, on the paper the motor specs were looking just fine, but I had a lot of motor driver errors, JRK's were choking up making motors jitter, after lots of PID testings, JRK settings, changing the motors positions, I eventually gave up on that combination..

I blame the motors, because as soon as I remove those and put back wiper motors, everything was working fine, small wipers were moving me without any problem with JRKs. So it's got the be the big motors.

I still have the equipment though, jrk's and motors, they've been lying around for months.. I need to make a move again and do something with it

Thanks for that info.

I agree that it would be the motors, I am a big guy though. 6 ft tall and 130KG plus my GS-4 so there is a lot of weight to more around. I just don't think the little wiper motors would be up to the task of moving me around.

I have tweaked them to the point where I am not getting any problem while on track and it is throwing me around like anything. I did a couple of hours racing tonight and I didn't notice the problem at all and I am pretty sore afterwards. The combination of the seat mover + GS-4 + belts is pretty extreme. It reminds me of when I used to race IRL. after a race I would have sore ribs and that is what I am getting with this combo so I am more that happy.

Interesting stuff @Avenga76 and @prodigy about the motors.
It does explain why some motors with JRK combination do not seem to be suitable.
Ive had a lengthy discussion with motion dynamics and he explained the rated current and stall current of motors can be a long way off.
The motor suppliers do not account for the fact that we overdrive them.
However @Nick Moxley , you run your motors ridiculously hard, do you have any thoughts?
@eaorobbie may have some ideas as well

Ok one suggestion, since the motors have the power to achieve the motion we can do two things.

First your PID is way out, noticed in the pic above , pid period exceeded in red , your either running the Jrks too quick which will cause errors and silly things to happen like motor errors too. This needs to be addressed too. When running oversized motors you need a couple of tricks.

In the motor tab derate the Duty and Accel down near 300-350, run the amps at 0 (unlimited).
Now revisit the PID and tune the error down, while testing its motion in steps.
Once happy test in game if motors seem to not cope with weight shifting as in a little soft , increase duty a little
and if they seem to not be reacting with some speed you can tickle the Accel up a little but beware not to open the flood gates or you might end back up here.

200W - 320W motors I have tested to run flat out.

my old winch motors were 25amp and drew 48amp on direction changes , ran flat out on all settings , but with a mild , not aggressive PID

Sorry a bit of a brief description but Pololu docs do cover this well now, never use too.

your PID and motor setting may not be good.
I suggest you spend time with PID and motor settings to try to rectify this
and use Robs suggestions
http://www.xsimulator.net/community/marketplace/guide-to-setting-pid-and-motor-settings-on-jrk.124/

I found that .1 difference on my P would cause issues. So unfortunately the PID needs to be rather fine tuned when running thing's maxed out or cranked right up. Add in a car with some questionable physics/suspension or a Funky clutch pedal glitch like in Assetto that can cause some MAJOR oscillation, and thing's can prove even more challenging to diagnose.


FWIW, I had a buddy who is 280lbs in my rig when i first built it (BARELY fit in my seat), and My PGsaw motor's had Zero issue moving him around in a replay. sure thing's weren't as fast as with my Skinny ass in the seat, but buddy's also near 6'10" compared to my 5'10" so Major COG change.

I am not sure if that PID exceeded is just because I have my kill switch engaged.

I will do the test that you described tonight and see if I get better results but in my initial testing I go these errors even at low PID.

That is the guide that I used to set my PID. a couple of pages ago I included all my PID testing and how I came to those figures.

Can you please take a look at them and see where I went wrong.

I have since lowered my D a bit because I get less overshoot when I am sitting in the seat and I did my initial testing with me not in the seat.

Hello,
i've read a lot in here. Thanks for all the Informations. But I'm a little bit confused now. My last rig failed in case of weak motors. I collect some money for about half a year now and thought this zyt90 Motors are perfect, and spend all the money in order 2.
Now after reading all this, i'm a little bit afraid if i spnd my money wrong.
Can someone confirm that this motors will work with one MM per Motor with a 47 Psu per motor? I hope my rig won't fail again incase of wrong parts...

I haven't had any experience with those motors @prodigy said he had some. Maybe he could shed some light.

I think @Pit uses MM's so he could probably help you. also @RacingMat

@tombo : please open a new thread dedicated on your question!
give us all the information (motor reference, specs, link) and we'll see

@Avenga76 : to tune your PID, the "Simple Digital Scope for Motion Sim Testing" from @RufusDufus could help you
http://www.xsimulator.net/community/posts/66666/

Okay I did some testing on the PID Period Exceeded. It is not because I am running my JRK's too quick.

It was because of a setting that Dave told me to change yesterday.

The manual says. The indicator labeled “PID period exceeded” at the top of the window is provided as a warning for when the analog sampling takes more time than the specified PID period.

I lower that back down to 512 and that PID Period Exceeded error goes away but it doesn't change anything.

When I first got my rig running used that article and I spent a day testing this and posted my results above with tons of screenshots and I listed my methodology on how I came to those numbers. So if you all think I have it wrong then please tell me what I did wrong.

I have recreated that test using the JRK graph and I am very happy with how to PID is looking what I am driving. This is using my original PID setting which I think are pretty good. As I said I spent a full day testing this and this is the best I could come up with.

This is with a pretty extreme motion profile. The scale is of 0-100% of my motor position and the time span is 5 seconds so it is probably more zoomed in the @RufusDufus test (I can't tell what time scale his is so mine might look worse if it is more zoomed in)

This is a braking zone followed by a sharp hair pin corner. this is using 70% of my travel so it is a pretty extreme movement but the motors are following the movement very nicely



A right then left chicane



A gentle sweeper



I have recorded a video of the graph while the AI races with me in the seat. I am very happy with how quickly the simulator responds. It is capable of this kind of motion with out any motor errors, It is only when I go off track.

Please watch it and give me your feedback

I still think the feedback numbers are not helping
Again, Absolute max should be about 2-5% higher than max , and absolute min should be 2-5% lower than min
So it should be more like
2625
2500
1600
1520

The idea of absolute max and absolute min is to allow a little extra travel on the motors when they reach there target.
Going off track produces spikes, so you might have it right on the main road, but going off road may be sending the motor past that point and causing an error.