I have voted crash-detection too 
Not sure Mika.
I never tried to make iRacing crash. I was always happy if I didnāt have issues!
Yes me as well.
Bloody scary when it is a car crash and you get stuck in a wall!
I think Mika is talking about iRacing program crashing though is he not?
I agree to give priority to safety.
This is just my personal opinion,I think that features that affect everyone should be prioritized as major updates.
Features that affect a specific device will be included in subsequent minor or next updates
Im sorry my English is not good.
The safety issue here is that the bumpstop at the limits of the steering range set in True Drive doesnāt always work. Trying to patch that by creating a crash detection algorithm using some magic number heuristics is only going to cause unexpected behaviour and be one more thing that needs tweaking every release wasting devs time.
Fix the hard bumpstop in TD and you solve the safety issue.
5 Likes
if you have oculus S a virtual sound driver is created, and if in iRacing you configure to use these virtual sound devices it will cause it to crash.
It is a much larger change to make full motor current available for the bumpstop. If that change is made, the FFB will also feel different - so different that users will need to change their profiles.
I think we will experiment with doing this only when the crash detection is triggered.
My choice is ābetter documentationā, i think that any of the other options are useless and/or can create more problems, in my personal case
Maybe I donāt understand the problem fully. But if I have set the bump stop at 900Ā° and the game sends FFB signal to continiue rotation past the 900Ā°, cant the firmware just stop with the force application in the direction past the bump stop?
Lets say the game crashes and sends out FFB that conitinues to rotate the wheel to the left. Once the bumpstop range is hit, surely the firmware can just stop generating forces that would rotate the wheel to the left. Or is it not that simple?
2 Likes
couldnt you just have it so if the wheel rotates x degrees beyond the bump stop, which is not likely to be intentional to push that far over the wheel goes to safe shudown mode, gives you the beeps and you cycle the e stop?
No need for forces.
2 Likes
It does that already in the servo drive, but it has to take to account the maximum possible rotation angle of 2800 degrees lock-to-lock, and this is mainly meant to prevent wildly spinning.
Wasnāt the whole crash detection idea spawned because one userās cable got ripped, because the wheel didnāt stop spinning?
To me it seems we have two use cases for sim crash detection:
- sim crashes and tries to rotate the wheel past the bump stops -> correctly implemented bump stops would fix this
- sim crashes and frantically moves the wheel between bump stops -> we have the e-stop for that already, no?
In other topics you talk about removing filters to keep the code base clean and maintainable. Trying to detect whether a sim has crashed seems like going in the opposite direction 
safety features are somewhat separate from any effect processing, so they can be tested separately
Mika, how is it possible that TD allows the rim to rotate beyond the limits that we set?
Iāve never experienced the rim spinning well out of control by itself, luckily, but Iād like to think it would hit the stop and not go any further.
Iāve got a well made USB cable that could most likely survive a stretch beyond the 1200 degrees that Iāve set in ACC but the thought of it going much beyond that would worry me.
By experience, the wheel can take lots of inertia and overpass the 100% opposite force that is working as Bumpostop of any direct drive wheel
Bumpstop is no more than a spring force
1 Like
Thanks @Alfye20
I thought the encoder would be able to detect that the bumpstop limit has been passed and then shut down torque?
Yep and its working as it should, auto shut off power but the wheel keeps rotating by itself
Ok, so the constant spinning in the video is just the result of the inertia and if you didnāt grab the rim, it would eventually stop?
Thing is, thereās still enough rotation there to rip up a USB cable from what I could tell.
RBR ffb test send constant force to the left for 2 seconds, then to the right another 2 seconds, then auto center the wheel.
The constant spinning is caused by a bad config in TD (too much inertia on profile) and too weak forces for bumstop, so 100% force send by RBR can overpass the SC2 bumpstops.
If i dont stop the wheel, it could be rotating like a minute, no power on the motor and a 350mm rimā¦
Wow, I bet it could rotate quite some distance in that one minute!
If you were to increase the bumpstop to its maximum strength, do you think that would be able to stop the rotation, even with the inertia settings you use?