I posted in the Simucube 1 thread but has anyone else experienced pretty crazy oscillations while in the garage and/or pits? Doesn’t happen in every car.
Thanks for the response, NuScorpii. I am familiar with the Windows 10 calibration routines, and tried them many times while trying to troubleshoot PC2 with no luck I have heard others say multiple +X assignments are also what they have without problems. It took me days of trying to finally get PC2 controls assigned and working. It started working Ok without specific reasons, and after weeks, began not working AGAIN without warning or reason.
I’ve tried everything in PC2, but glad I’m not having trouble so far in AMS2. In PC2 I tried Steam integrity checks, Steam reinstalls, Windows file integrity checks, swapping USB ports between motherboard and other ports. Installed a specific dedicated USB3 card, etc. etc. without any luck. Only thing I have not tried are different USB cables with magnets. I may do that at some point. But I’m sure with AMS2 working well, that I could immediately close AMS2, try PC2, and it would not work correctly. Not sure why, but I’m now using ACC and may use AC more, and after more AMS2 updates, I’ll be using that also. I don’t have low-latency broadband available where I live, but I’m hoping by the end of the year that Elon Musk will have his Starlink system open for subscriptions for broadband low-latency satellite. And I’ll try out Iracing. All those plus maybe a few others for the occasional diversion, and I’ll be happy! Thanks again for your response! 
In PCars2 it was due to braking system bug.
http://forum.projectcarsgame.com/showthread.php?66720-Christiaan-s-Pure-FFB-(Custom-pCARS2-FFB-File)
Considering the engine is the same, this is most likely the root of the problem.
The problem lies in the manner how braking force is applied to the tyres, which causes the tyres to act against each other and forces acting against each other will start to oscillate under that struggle.
That link talks about AMS 2 FFB how it works. They release this article just today. Check it out guys very useful.
1 Like
Thanks you…today I will continue the tests on the ffb and I will follow your advice
1 Like
I have been trying all these posted truedrive/automobilista ffb combinations, but nothing really got me to a point of being completely happy.
To me it feels like the original ffb for PC2 was clearly build with only/mostly belt wheels in mind, which makes perfect sense.
This unf. can be felt very clearly here in AMS2 too. Like on the MCR S2000 as an extreme example.
It is as if the delay & smoothing of a belt wheel has been countered in that code/params.
I have had that same issue in iracing on the Porsche Cayman car.
The recon filter seems to preserve the extremes of the ffb signal, so it does not work well for cars made for belt wheels, since the belt driven wheel would have NOT preserved the extremes all the way through. It would be lost in the belt.
I was able to fix this for the Porsche Cayman in iracing by using irffb, which smoothes as well as upsamples at the cost of a very slight delay.
Unfortunately irffb does not work for AMS2.
For AMS 2 smoothing alone would help a huge deal on these cars that are set up using code that still iis not written or configured correct for a direct drive wheel.
It would be really nice to have that smoothing option in TrueDrive. Simply how much % from current position of wheel to where it really should be & then it should apply that delta at the rate as provided in hz.
This could be a pre-filter before everything else is then calculated. A source feed smoother.
(I suppose this filter would work well for any other game not made with a DD wheel in mind too)
But since we don’t have that, I’ve played around with TrueDrive & AMS 2 ffb settings & finally arrived at something I think is 2nd best(but not nearly as good as that smoothing filter would have been).
With these settings my ffb now feels very good in AMS 2, finally!
It will only get better when AMS 2 has upgraded all code & configs for DD wheels, at which point I will switch back to my more normal settings in TrueDrive.
One big difference between AMS 2 and iracing is that AMS 2 has a much higher FFB rate, so upsampling is not really required.
So the solution that works extremely well for me is different; at the core are 2 changes in TrueDrive only:
- To NOT use the recon filter at all. Not even minimum. So set to “OFF”.
- To set slew rate limit to absolute minimum (0.10).
(So if you want to try yourself, just make these 2 small changes to whatever you have now.)
This combination seems to get closer to the intended behavior of the wheel, by not reinforcing the extremes of the ffb(as it seems the recon does), while also trying to limit the noise from the motor by limiting the slew rate. At the same time has all the small details.
All other settings should be set as you prefer otherwise, both truedrive & in-game.
I have not driven all cars using this yet, but it has clearly improved the 5 cars that I have, so I would now say AMS 2 is at the very top of sims when it comes to FFB, even before targeting DD wheels specifically.
I plan to only use this “fix” until real DD ffb is 100% implemented in AMS2, as it does feel shitty to have to emulate belt-feel in a DD wheel! Specially when it cannot even be done because of a lack of a simple smoothing filter.
By the way this would have never been possible with only the “simple mode”, so I hope the idea of removing “advanced” will NEVER happen.
2 Likes
Have you noticed the new oscillation in the pit that last for most of the time on track unless you turning into a corner? It came back when the game updated today. 
I was using your settings and was pretty happy with them
you can make us screens please
1 Like
Hey katoche69
sure,
This for simucube 2 pro, so adjust accordingly.
I like the Overall strength to not be at full, only for safety reasons 8)
2 Likes
I wouldnt use LFB try it with it on zero . In the article i posted above Reiza Studio themself talk about not using it on direct drive wheels not needed.
Hey Johan,
There seems to be different thoughts about it.
The most important to me was to try it all, which I did.
This felt better to me, so it’s by purpose it’s not at zero.
But actually you are incorrect.This is from that article:
So not really saying not to use it, but saying when you should use it & in which range if you do.
But try yourself & pick what you think is better 8)
2 Likes
I found a bit of LFE good too in the range 10-15%. What I find odd in your settings is having the Slew RL at 0.1. Doesn’t that make it really rubbery and effectively filter out high frequency effects?
I postred a lengthier description just prior to those settings images, not sure if you read them.
It is to counter the rather extreme of having reconstruction filter “off”, so it does not produce too much noise from the motor. Also I think it helps to take out just the most extreme spikes.
I don’t think it filters out high frequency as much as spikey, big ffb difference from update to update.
So I think the high freq. stuff is still there. Feels so to me too.
They really need to have a option to remove the menu spring all together.
Zero menu spring does not equal to have no menu spring. I think this is causing problems with calibrating the wheel.
I am now having NO FFB issue after i tried to recalibrate.
You can avoid issues by enabling the e-stop during axis-assignment / calibration.
1 Like
With E Stop you have no sense of where the Bump Stops are.
You have to have bump stops for the calibration part.
1 Like
Fair point, it’s still useful for axis detection though.
With the new update yesterday there is no more harsh wheel movement when assigining left and right. Tty it.
50 bump stop strenght 10 degrees and 99 % bump stop effect.