Yeah, but that comes with big disadvantages as you undedrive servo and can’t get full force in normal, non violent spike situations.
Sounds like we need another “Torque Spike Prevention” filter or something along these lines, which when force ramps up quickly to certain configurable torque level will soft clip the signal.
Perhaps it is already existing Slew Rate Limit , except description is not very specific on that
Slew Rate Limit
This effect limits how fast the torque is being changed. “Off” (default setting) means, that the rate of change is unlimited and the rate will be as fast as the Simucube 2 servo drive and motor will be able to accomplish. Limiting the slew rate can be useful to soften the driving feel. If the Driver has experience of other DD wheel systems, limiting the slew rate can be useful to make Simucube 2 feel less active.
Theoretically you could use the notch filter with a wide enough band to attenuate sudden torque spikes. The challenge is identifying the frequency ranges that a wall hit will produce. Most of these unwanted spikes in torque are in the very fast frequency range. It would be good if multiple notch filters could be implemented in the software. That would allow the user to use them to tune out multiple unwanted frequencies. Lowering the torque slew rate helps but then that effects everything else which might interfere with details. What might help is a tunable map of torque slew rate vs frequency. You can tune it so that you have max slew rate at low frequencies (<10-20 Hz) and then ramp down the slew rate at higher frequencies where all of the sudden torque spikes happen. Just an idea. In this way, you retain the high slew rate for all of the low frequency details but then you ramp it down at higher frequencies to help tone down all of the sudden torque spikes when you hit a wall or someone crashes into you.
Spikes are just high frequency content in FFB, so reducing the torque bandwidth limit or the slew rate will make them less sharp. Other types would need to have a look-ahead examination of the torque, i.e., the torque output to the wheel would have a lag in order to pre-determine the upcoming torque.
Just use the Automatic Hands-Off detection feature in Simucube 2 and let go of the wheel before you hit the wall - thats what real Drivers do.
Yes, the real solution would be a predictive algorithm but that would require some sort of internal plant model and coordination with the game so I can see how that is not a feasible solution. As you mentioned, I currently let go of the wheel when I’m headed into a wall. I still think a tunable frequency-based slew rate map would be cool though. I would definitely use it.
How static force reduction works?
I envision something similar but for rapid increase in torque value above certain torque threshold.
Unfortunately it’s not always practical to let wheel go in time or “before” as you suggested, as most of the times things like these come completely unexpected.
You could always wire a second switch into the loop that maybe is on the dash that allows you to open the loop without hitting the estop, In doing so you can use the foot estop for emergencies with the main switch on and then the main switch off and on for set-up and other things that require it…
You could ALSO if you really wanted to wire up a Normally Closed Momentary Foot Pedal and temporarily cut power by having it e-Stop ONLY when you have your foot on the switch…
So is it possible to wrap it up into user friendly filter or not?
I am honestly not sure how much of a practical solution one can make out of all that information posted.
Nice read though, now I know what square wave is.
Also curious about this. Seems like I’m making.blind guesses as to which frequency the noise is at, right now I’m just using Beano’s settings because I personally can’t discern where the unwanted spikes are at. Also seems like it could occur at multiple frequencies. Would be cool to be able create multiple filters at different frequencies. Anyway, guess it’s not a big deal in the overall scheme of things.
UPDATE: Jed from Simtech Racing quickly contacted me and we solved this problem offline. All is well now. Thank you Jed.
Just a word of caution for anybody thinking of ordering their SC2 from Simtech Racing. I ordered my SC2 Pro through them and received it on August 22nd (I was part of batch 2). When I placed my order on June 4th, the simtech racing website stated that a motor mount was included with the SC2 Pro. I verified with Jed that a mount would be included in the package before it shipped (the website says its not included anymore). However, when I received the package (I’m in the US) the package did not come with the mount. I immediately sent Jed an email asking about the mount. Since then, I’ve sent him 4 emails (both directly and through the website) and have yet to hear back or receive a mount. I got tired of waiting and not being able to use my SC2 Pro and bought my own mount form another seller. I normally don’t write things like this but the after-sale customer support from Simtech Racing has been terrible. I paid for a mount that I never received and have yet to hear back from them. I tried to give them a chance to make this situation right but I’m tired of waiting and this is my last resort. Hopefully my problem is not their regular business practice.
