- 577
- NC
Ok I have been modding my All-Fanatec sim rig for some time. It's not very stock at this point lol, all of the components have many changes. I came across a concept I thought I would share, one that improves car control and therefore lap times. This can help any sim wheel of any brand, and it is my suggestion to keep this in mind if you are a DIY modder such as myself. There are myriad details as to how to embody the belownoted, but the concepts work on any wheel which is my point.
First of all, (1) linearity as many of us know is important. You can read about this in many places, it has been discussed for years. This can be checked fairly well using Tucker's Wheelcheck program over at iRacing. You want a straight line from 0,0 to 100,100 when you compare FFB requested and delivered on a graph. Some wheels do little at first when sent weak but important signals. That is not good, especially if you modify a sim wheel and gain in one place but lose linearity. Since the signals telling the driver how the car is doing can be delayed/attenuated leading to "late" corrections which then hurts car control and lap times.
Some modded wheels suffer from this woe, seeing gains in one area, but with a penalty to linearity. That's a shame if a wheel started out linear. It's also common to clip or respond inadequately to stronger signals. An example of a wheel with a bit of both common linearity issues is shown below (a Thrustmaster plot from another forum). It is in pretty decent shape in the middle ranges IMO. Some wheels also exhibit a dead zone or initial low FFB action too in terms of slop, such as some gear driven units which need to cut objectionable clacking sounds during reversals.
Another major factor which I have been emphasizing in my own efforts is (2) Drag or Friction. So my suggestion and reason for this post has to do with this second factor along with the first noted above. So, (1) Linearity and (2) Drag.
Drag / Friction can delay response since this force has to first be overcome before wheel movement occurs. It also leads to heat and wear and tear since the motor(s) and drive components have to constantly deal with drag / friction. Think of it like trying to ride a bike with one of the brakes lightly engaged. It's not as much fun, and you'd work up a sweat prematurely.
Some time ago I posted a chart showing stock drag torques for both Thrustmaster T500xx and Fanatec CSWs. I showed the drag I had gotten down to and noted that below some point "Magic" starts to occur. This won't show up on a linearity test. You can get a sense of drag if you look at the coastdown after a step test signal ends however. High drag wheels stop moving soon after the drive signals cease. Since things like inertia vary amongst wheel options, this can complicate comparisons. However it can be valid if you mod since you are keeping things like that constant.
In my case I maintained and slightly improved upon the Fanatec linearity (which is quite good). I did not want to harm that with my mods. And I massively and immensely cut drag. So I marked up my old graph (below) so that you can get a sense of drag levels. I've been pretty much at zero for many months and cutting it more does not seem likely to help course times / immersion much if at all.
So, with drag being dealt with and maximum force levels increased quite a bit there is a large range of sensations available to the Sim Racer. This helps immersion / realism. You can feel new things that were there all along. Including bad things such as FFB inadequacies.
Anyway WHY have I kept seeing consistently seeing lap time and split improvements when I compare low and high drag levels? It comes from being able to countersteer more rapidly, and being able to sense when to do so more easily. So you can start a countersteer before the car gets out of shape. And complete it quickly. I like a "loose is fast" car setup moreso than a "Mr Plow" arrangement.
My reaction time is about .2 seconds which is decent. Senna was consistently under 0.1 seconds which is simply astonishing. He picked his parents very well lol.
Well, this wheel concept allows countersteer corrections to take place about 0.13 seconds quicker on average. So...if I raced Senna and he could react .1 seconds faster than me but I could complete a countersteer quicker than him (I'd give him the stock wheel haha) then I'd have more of a chance. No this does not mean that you complete each corner over a tenth faster. Rather it means that you'll make fewer mistakes over a series of laps hence lap times and car position benefits. The car stays more stable without giving you a heart attack.
A right hand turn I have been messing around with dives after turn-in and so it is easy to get oversteer. Now I am used to it so I can anticipate things to an extent. So I tried driving some laps while videoing the wheel in the same turn for the necessary countersteer action. Car in the same position on track. Wheel set to stock drag and later with *very* low drag. Went back and forth a few times. Every Single Time the low drag wheel could complete the intended driver input faster than the stock Fanatec wheel. Not once is it slower or with lower wheel speed. I counted video frames to get time splits and found an average improvement to the duration of the steering correction of 0.13 seconds in various turns The car stays more stable and can be positioned more accurately.
Below is a pair of screen grabs after turn-in and just before (one video frame) a quick countersteer was needed. I have the LEDs modded set to stay on, switchable so they can be a wheel speed reference. I ran out of time but the Low Drag wheel hits higher wheel speeds which shows up as longer LED blurs which is why I note it.
This next shot below shows the Stock-Drag wheel hitting the 8:00 point in a CCW correction. Meanwhile with Low Drag mods the other shot on the rights shows the wheel heading back to its initial position at 3:00. Again, synchronized in time. Both wheels hit 8:00 as their maximum.
And here is a shot from when the countersteer correction was completed by the Low Drag wheel, the stock wheel is still on the way back.
You can't really measure the driver / sim-wheel oneness with a program like Wheelcheck. So that is why I thought I would share some video clips from a series of laps on a familiar course. On a new course, it is likely there would be more benefit due to higher feel along with the quicker corrections. In this case on a familiar corner I got the low drag wheel off to a well timed start due to knowing what to expect as opposed to feeling it and then reacting.
Sorry for the long post or article. I hope that some of you might find benefit to very linear / low-drag wheels in your own mod efforts since they will help you find course time while increasing enjoyment and device longevity. If you are not a DIY type you can also benefit since you can consider such factors when you purchase a product.
In conclusion, things you can do to retain and increase linearity are very important IMO. And even moreso is cutting Friction / Drag which is not talked about very often. But it works for sure as I am at around one year with much reduced and later nearly eliminated drag. I also have some data logs of wheel speed and other things and may post those when time permits.
First of all, (1) linearity as many of us know is important. You can read about this in many places, it has been discussed for years. This can be checked fairly well using Tucker's Wheelcheck program over at iRacing. You want a straight line from 0,0 to 100,100 when you compare FFB requested and delivered on a graph. Some wheels do little at first when sent weak but important signals. That is not good, especially if you modify a sim wheel and gain in one place but lose linearity. Since the signals telling the driver how the car is doing can be delayed/attenuated leading to "late" corrections which then hurts car control and lap times.
Some modded wheels suffer from this woe, seeing gains in one area, but with a penalty to linearity. That's a shame if a wheel started out linear. It's also common to clip or respond inadequately to stronger signals. An example of a wheel with a bit of both common linearity issues is shown below (a Thrustmaster plot from another forum). It is in pretty decent shape in the middle ranges IMO. Some wheels also exhibit a dead zone or initial low FFB action too in terms of slop, such as some gear driven units which need to cut objectionable clacking sounds during reversals.
Another major factor which I have been emphasizing in my own efforts is (2) Drag or Friction. So my suggestion and reason for this post has to do with this second factor along with the first noted above. So, (1) Linearity and (2) Drag.
Drag / Friction can delay response since this force has to first be overcome before wheel movement occurs. It also leads to heat and wear and tear since the motor(s) and drive components have to constantly deal with drag / friction. Think of it like trying to ride a bike with one of the brakes lightly engaged. It's not as much fun, and you'd work up a sweat prematurely.
Some time ago I posted a chart showing stock drag torques for both Thrustmaster T500xx and Fanatec CSWs. I showed the drag I had gotten down to and noted that below some point "Magic" starts to occur. This won't show up on a linearity test. You can get a sense of drag if you look at the coastdown after a step test signal ends however. High drag wheels stop moving soon after the drive signals cease. Since things like inertia vary amongst wheel options, this can complicate comparisons. However it can be valid if you mod since you are keeping things like that constant.
In my case I maintained and slightly improved upon the Fanatec linearity (which is quite good). I did not want to harm that with my mods. And I massively and immensely cut drag. So I marked up my old graph (below) so that you can get a sense of drag levels. I've been pretty much at zero for many months and cutting it more does not seem likely to help course times / immersion much if at all.
So, with drag being dealt with and maximum force levels increased quite a bit there is a large range of sensations available to the Sim Racer. This helps immersion / realism. You can feel new things that were there all along. Including bad things such as FFB inadequacies.
Anyway WHY have I kept seeing consistently seeing lap time and split improvements when I compare low and high drag levels? It comes from being able to countersteer more rapidly, and being able to sense when to do so more easily. So you can start a countersteer before the car gets out of shape. And complete it quickly. I like a "loose is fast" car setup moreso than a "Mr Plow" arrangement.
My reaction time is about .2 seconds which is decent. Senna was consistently under 0.1 seconds which is simply astonishing. He picked his parents very well lol.
Well, this wheel concept allows countersteer corrections to take place about 0.13 seconds quicker on average. So...if I raced Senna and he could react .1 seconds faster than me but I could complete a countersteer quicker than him (I'd give him the stock wheel haha) then I'd have more of a chance. No this does not mean that you complete each corner over a tenth faster. Rather it means that you'll make fewer mistakes over a series of laps hence lap times and car position benefits. The car stays more stable without giving you a heart attack.
A right hand turn I have been messing around with dives after turn-in and so it is easy to get oversteer. Now I am used to it so I can anticipate things to an extent. So I tried driving some laps while videoing the wheel in the same turn for the necessary countersteer action. Car in the same position on track. Wheel set to stock drag and later with *very* low drag. Went back and forth a few times. Every Single Time the low drag wheel could complete the intended driver input faster than the stock Fanatec wheel. Not once is it slower or with lower wheel speed. I counted video frames to get time splits and found an average improvement to the duration of the steering correction of 0.13 seconds in various turns The car stays more stable and can be positioned more accurately.
Below is a pair of screen grabs after turn-in and just before (one video frame) a quick countersteer was needed. I have the LEDs modded set to stay on, switchable so they can be a wheel speed reference. I ran out of time but the Low Drag wheel hits higher wheel speeds which shows up as longer LED blurs which is why I note it.
This next shot below shows the Stock-Drag wheel hitting the 8:00 point in a CCW correction. Meanwhile with Low Drag mods the other shot on the rights shows the wheel heading back to its initial position at 3:00. Again, synchronized in time. Both wheels hit 8:00 as their maximum.
And here is a shot from when the countersteer correction was completed by the Low Drag wheel, the stock wheel is still on the way back.
You can't really measure the driver / sim-wheel oneness with a program like Wheelcheck. So that is why I thought I would share some video clips from a series of laps on a familiar course. On a new course, it is likely there would be more benefit due to higher feel along with the quicker corrections. In this case on a familiar corner I got the low drag wheel off to a well timed start due to knowing what to expect as opposed to feeling it and then reacting.
Sorry for the long post or article. I hope that some of you might find benefit to very linear / low-drag wheels in your own mod efforts since they will help you find course time while increasing enjoyment and device longevity. If you are not a DIY type you can also benefit since you can consider such factors when you purchase a product.
In conclusion, things you can do to retain and increase linearity are very important IMO. And even moreso is cutting Friction / Drag which is not talked about very often. But it works for sure as I am at around one year with much reduced and later nearly eliminated drag. I also have some data logs of wheel speed and other things and may post those when time permits.
Last edited:
Enough of the repetition already. Christ you're beginning to sound like a friggin' broken record/player. If you want to send a shot across my bow I guess I'll send one back.
Sure, not much weight could be removed but, apparently only an ounce or 2 WILL net you real world gains...
