[Solved or not ...] Fanatec CSL DD FFB freezes

I read that some Fanatec users avoid issues by opening Windows Task Manager and closing the Fanatec App and FanatecService before launching their game. Might be worth a try before attempting to roll back Fanatec software and firmware.
Good shout. And don't let it start with windows in the first place.

I usually kill FanatecService after I used the app - because it annoys me that there is a process running full blast.
Maybe that's the reason why I see the bug so rarely - ie only when I forgot to close it.

I now definitely do remember that I had a race on AC EVO ruined, so at least "my bug" is not an LMU only thing.
 
I will try that. Thanks for the tip.

I don't have the app running, but Fanatec service is running at start of PC if i remember well.

It's probably possible to disable it completely, by the way. And even remove it completely.

My other games work so well without issues, that is really bothering me on LMU.

AC, AC Evo, AC Rally, rFactor2 are flawless on FFB.
 
There is really only one way—one and only one way—to fix it (assuming your problem is like mine, which I suspect it is): download a much older version of the Fanatec driver and, if you wish, an older version of FanaLab as well.

I just did a few tests.

Disabling definitely the Fanatec Service doesn't do anything ans it seems to be useless by the way.
Without anything from Fanatec launched, only last firmware i always had the issue with inconsistent FFB, spikes and losses.

I couldn't resist tonight ...
I finally did the retrofit to Fanatec 464 474 driver and Fanalab, also disabled definitely the Fanatec Service, and the issue vanished !!!
I will do more tests along the way but it seems to have solve the issue. And i hope it will stay like that.

Now i got back my confidence on the FFB.
This wheel is a Beast, i still set 90% FFB on LMU and 100% on wheel settings. Strong and sharp FFB. Awesome.

Thanks a lot for your help, i was afraid of the retrofit, but i shouldn't.
 
Last edited:
That’s great news for you—it certainly was for me a few months back, after a lot of hard searching to find this solution—and it’s great news for the community, too. It proves that this method works. That said, if you’re willing to accept the *feeling* of slightly weaker FFB, try my settings all together; the quality will be far superior while still delivering convincing force. One setting that might seem daunting—much like reverting to older firmware, software, and drivers—is "LINEAR." It is the only way to ensure the wheel force responds linearly to the raw signal coming from the game. While the "PEAK" setting might feel impressive at first, you lose out on FFB realism, even if it remains enjoyable. Also, you need to use all the settings I’ve listed; otherwise, you won't get the full benefit of the realistic FFB. The sensation of weaker FFB is simply because you’re used to "PEAK," but you’ll adapt to "LINEAR" quickly once you experience the improved quality. However, if you lower the in-game FFB intelligently (75% is a good overall value for rF2; 70% is excellent for LMU, considering the wheel and clipping issues specific to that title), you’ll need to set the "FOR" setting to 120% to achieve a comfortable level of force. Don't panic—as I explained in another post, this takes the raw game signal without altering it, rescales it (adding 20%), and then applies your "LINEAR" setting in conjunction with your wheel force setting (100%). The other settings are important—specifically, no filtering, no in-game spring, no in-game damping, and the other settings I mentioned. I recommend trying this out for at least three one-hour sessions; if—after those three sessions, or even a bit longer—you still don't like it, you can simply revert to your previous settings. It doesn't cost you much to try, yet the potential payoff is a massive improvement. It’s all about optimization while prioritizing realism above all else. The FFB is so satisfying in terms of pure realism—and the strength is objectively excellent—that I think it’s completely unnecessary to switch to a Simucube or another ultra-high-end wheel (like Fanatec’s top-tier models). Even a slightly higher-end Fanatec wheel—like the 15Nm model, which features slightly better technology and much more torque than ours, along with the same overall quality and software—would likely be incredible if configured perfectly using the settings I shared. There’s really no need to go beyond that unless you have a very high income; in fact, I might look into getting one second-hand myself in the future.
 
474. Sorry for the error.

Firmware on wheel, base and induction stayed the same. And i won't change anything until i see issues, at this point.

The issue was at the interface between the game and the hardware, so only the driver. See it as a sort of gateway, pass through.

I kept all my previous and newest settings, and do not have the linear/peak setting anymore, sadly.

But from what i read i think they kept something similar to Linear and added the "Full Force" setting.

When Full Force is used by the Game it probably is equivalent to the Peak setting but with like "exponential" values now available for intensity, which is a much better option, imo.

FFS at 0 = Linear
FFS at 100 = Peak
FFS at X value = Exponential effect

Exponential is often also called "sensitivity".

LMU doesn't use the Full Force effect, imo, like DPR or SPR.

And the Podium Fanatec DD is surely the best overall DD "ever". Everything else is too expensive, too cheap, less powered, too much powered.

Fanatec surely got one of the best compromize with their Podium DD, i think, from what i know/read. I should have choose this option coming from my CSL Elite "Belt Drive", instead of the CSL DD "Boost".
 
Last edited:
Found the new Fanatec App 1.6.1.3 today, let's see how it goes.

About the Fanatec Services. There's actually 2. These Devs are nuts.
The "Fanatec Wheel Service" in Windows Services is there to keep the wheel from suspending. Not touching that. It's not using CPU.
I was killing the other process named "Fanatec Service" that spawns with the Fanatec App. Which is for telemetry afaik and fully uses one CPU core probably active wait.
 
464, yes. Sorry for the error.

Firmware on wheel, base and induction stayed the same. And i won't change anything until i see issues, at this point.

The issue was at the interface between the game and the hardware, so only the driver. See it as a sort of gateway, pass through.

I kept all my previous and newest settings, and do not have the linear/peak setting anymore, sadly.

But from what i read i think they kept something similar to Linear and added the "Full Force" setting.

When Full Force is used by the Game it probably is equivalent to the Peak setting but with like "exponential" values now available for intensity, which is a much better option, imo.

FFS at 0 = Linear
FFS at 100 = Peak
FFS at X value = Exponential effect

Exponential is often also called "sensitivity".

LMU doesn't use the Full Force effect, imo, like DPR or SPR.

And the Podium Fanatec DD is surely the best overall DD "ever". Everything else is too expensive, too cheap, less powered, too much powered.

Fanatec surely got one of the best compromize with their Podium DD, i think, from what i know/read. I should have choose this option coming from my CSL Elite "Belt Drive", instead of the CSL DD "Boost".

If you successfully rolled back the drivers to 464 then the linear/peak setting should have been restored. Based on your description, it appears to me that you uninstalled the Fanatec App, installed Fanalab, and the FullForce drivers are still installed. Not necessarily a bad thing if the Fanatec App was the root cause of your LMU FFB issue.

Before FullForce was introduced, the CSL DD with Boost Kit had 6 Nm of hold torque and 8 Nm of peak torque in Peak Mode and 6 Nm of hold torque and 6 Nm of peak torque in Linear Mode. The FullForce update increased the hold torque to 8 Nm, so now it is always in Linear Mode with 8 Nm of hold torque and 8 Nm of peak torque.

The FullForce drivers have oscillation issues in other games, which is why I'm not eager to update until Fanatec sorts everything out.
 
One more thing, it's my understanding that the increase in hold torque was a coincidence to the introduction of FullForce. FullForce is the vibration gimmick that imitates "texture".
 
Hello, for being sure that everything is really right, i just re-did the uninstall/reinstall process.

In fact it's 474 drivers, not 464 as said before (Post corrected).
And i didn't removed USB devices properly the previous time.

I used this complete and accurate tutorial this time :


And now i have the right setting with FFS : Linear / Peak selector.

After last install you just need to do a firmware update but select the "Manual" option.
It will use the firmware based on actual drivers.
So now my firmwares are properly retrofitted (Wheel and Base, the WQR stayed the same).

Everything is OK so far :)

Just noticed a "lighter" FFB in Assetto Corsa Rally, which probably use the Full Force i don't have anymore :(
 
PLS warm up tyres before you judge! Best tracks are Road Atlanta & Portimao because of the load&weight transfer
Ingame FFB 35% & 0 Smoothing (Filter & NM are set in FanatecApp) for all cars.
You finally can catch oversteer by instinct!
Thank me later...
 

Attachments

  • CSL DD 8NM Profiles for LMU.zip
    10.7 KB · Views: 4
Last edited:
Thanks for that, but i was running on a 2 hour race, and got FFB "cuts" all over the session.

Cold, warm, on wets, it happens randomly and without even loading steering or else (No clipping).
The randomness makes me think it is a software "framerate" issue with inconsitencies and blanks.
It seems LMU exagerates the problem, but the retrofit show clearly the issue is from Fanatec Side.

I tried almost all settings one by one, nothing worked. The retrofit did it instantly.

I was always able to catch over/under steering before and after the bugs described in this post.

When freeze/cuts from FFB happens, the car doesn't even flinch, even on Porsche turn at Le Mans !!!
There is something weird going on making the game still playable.

It's like stutters on image frames, you see it, it's annoying, but game is still highly playable at 144 fps.
You still loose continuity of images.

Same for FFB here.
 
Important !

The solution I proposed, which involved using an older driver and firmware that worked perfectly before and predated the Fanatec App and, in particular, the Full Force update, does work perfectly.

However, I've found a much better solution that allows me to use the latest Fanatec App software and firmware:

I simply realized that a recent update, released in September, fixed the major, game-breaking bugs introduced with the first major Full Force update for CSL DD.

It actually works on my GT DD Pro 8nm (which is exactly the same wheel as the CSL DD, both mechanically and in terms of software and electronics; it's the PlayStation-compatible version of the CSL DD).

The advantage of upgrading to the latest Fanatec app software along with the latest firmware is that the major full force update no longer causes any problems, and it also brings essential improvements:

They've fixed the issue where, to get a completely realistic and linear force feedback (FFB), you had to use the linear mode, which was limited to 6 nm.

Now there's no linear setting; the FFB is 100% linear, but now with a maximum torque of 8 nm.

Furthermore, I think the FFB is slightly improved by the firmware optimization.

This just produced a well-known bug in Fanatec IN RF2, which I'd never personally experienced, and it's very easy to fix in the controller file (set the steering resistance line to 0).

Actually, the solution I originally proposed was excellent—before the September update came out, it really was the right move.

And with the updates, doing it the way I just described will make it even better.
 
For rFactor 2, gotto controller file, ONLY CHANGE one line, this line :

"Steering resistance saturation":0,

(Only for fanatec wheel, and only if an issue with the ffb)
 
Thanks, good to know.

I also realized something new.
I may only have issues on LMU because of Discord.

LMU is the only game i play with a group of friends, and my discord is active on a salon/chat.

I wonder if it's Discord generating spikes of CPU loads which affects the game and by side effect the FFB within Fanatec software.

I already noticed Discord uses a lot of "Cycles" in the Task Manager. That would explain the randomness of FFB spikes/loss which were probably linked to Discord loading the CPU on a certain period less noticeable.

I can't verify that now, but it is something remarkable.
And more when you read that Discord is working on a "Game mode" :

We often blame LMU but sometimes the truth is somewher else.

Like me, lots of drivers play LMU while streaming and being on Discord. They probably also play solo on other games, and so don't use Discord too.

So the "It works on other games" is not a valid argument quite often probably.

My issue started almost a day after the last update : 21/09/2026

I was probably using this version, because that was my first option : make the last update.

I will give it a try later on. Maybe on the next update.

At this point i am a bit disappointed and will see how it works for the next days, on driver 474.
 
Last edited:
I wonder if it's Discord generating spikes of CPU loads
In my experience a FFB drop also happens if an app steals the focus
Like when you press the Windows key or Alt-Tab.

Discord might steal focus if smth happens there, so...
I woulndt really know because I avoid it for the other performance reasons you discovered
 
I will be fixed tomorrow if the bug gets back on my Friday night race with Discord opened.
I stay on 474 driver for the moment.
 
Well, you know what, same problem a bit less strong but still there.
It really seems linked to Discord causing some sort of clipping for some reason.
It can also be linked to my overlays.

I will get back to last Fanatec drivers :rolleyes: and do some more tests.
 
Last edited:
Did a few tests :

Back to very last Fanatec app.
It's not Discord or OBS (Happens with or without them)
It's not my overlays ( TinyPedal, HyperTrace, same ).
It's not a CPU or a GPU temp issue ( Both at 60°C )
It's not clipping : Even at 25 % FFB in game i get the "stutters" but it's more visible than in FFB
( FFB is inexistant but the wheel noticeably still "stutters" visually randomly )

I just did a few things.
Reactivated the "High performance profile" via the Chris Titus Tools, and lowered to 32 the number of sounds (Seen on some reddit posts).

I will test it farther, but there seems to be an improvement. 20 Minutes at COTA in LMP3 (+LMP2 and GT3) and no FFB/wheel stutter at all.
 
Last edited:
Back
Top