Here is a suggestion which I have found to work *extremely* well for over a year if you are motor modding a Fanatec CSR-E or CSW. This might help other
true DIY modders such as folks like me, or commercial projects such as the one in this thread when the design is improved at some point in the future. I've had some email correspondence with some of you not actively posting about their efforts, there are some very good ideas out there. Some do not want to post as it can lead to the "ugh" feeling which I have noticed personally. That is a shame. Meanwhile the Physics of it all is what it is regardless.
One drive pulley is better than two. If you can, use
ONE. If you absolutely must use two drive pulleys, don't do it in the stock manner, there are more effective strategies.
Besides being cheaper (one beefier motor instead of two), a one pulley rig allows the belt drive to function more efficiently. It is certainly a part of how I got my wheel working so well. Chosen wisely, your single motor can have both more torque and less inertia / friction than two motor concepts. More of the motor's drive makes it to the wheel rim.
Let's look at the stock belt drive for a moment.
With this concept two identical motors (at pulley "B" and "C") are equally driven to create the FFB forces. The two motors are independent hence there will be three belt tensions presented to the drive. The highest will be seen by motor "B". This is because to it's left it sees the total tension both motors create ignoring preload. While on the other side it sees the tension created by motor "C". Motor "C" sees less belt tension if you follow the same thought process. Well things like belt install tension, the belt "wrap" angle, and in-use tension affect how much grip the belt and pulley can create.
What happens when you drive this two drive pulley setup hard and belt slip occurs? Well it will slip at pulley "C" all the time. Since that pulley sees less tension and a bit less wrap angle than pulley "B". Reverse the directions and then pulley "B" slips and slides.
So what I do instead is use ONE drive pulley. This results in around a 150-160º wrap angle. You wind up needing much less belt tension to minimize slip, which helps cut drag. Motor shaft bending and bearing loads are slashed. More force reaches the wheel rim instead of it being wasted with drive inefficiency and higher thermals. Choosing your motor carefully you wind up with much lower inertia than the Buhler approach. I probably have about 5-10% the Buhler motor inertia loads. Remember that the motors turn around 17 times faster than the wheel rim, therefore the impact of their rotational inertia is similarly magnified. Which again helps slash drag and wheel correction speed. You get grip with much lower belt tension requirements. Note that belt drives lose some efficiency right as the belt enters the pulley (rubber distorts/stretches), and then some when the belt leaves the pulley. If you run two drive pulleys at the necessarily higher belt tensions needed to cut slip at pulley "C" you wind up with well over double the belt ingress/egress drag. Since there are TWO entries and TWO exits all of which take place under higher belt tension than a one pulley drive requires.
This is all born out in the theoretical calculations in Engineering texts and in simulations such as MITCalc. And in practice. Let's suppose I drive ONE pulley in a two drive pulley setup to a high load with no slip. The driven pulley is locked in place. Note the amperage consumed (which is directly proportional to torque). Now add drive to the other pulley and observe the amperage as you twist the knob on your lab power supply upward. There are a number of variables however in some of my tests you'd start getting the second pulley ("C") to start slipping when it is fed with about 1/3 the current being fed to pulley "B". Clearly pulley "C" is an issue. In the stock Fanatec and now Killer-B installs, the same exact problems exist and so this is handled or covered up by jacking the belt tension higher than is actually needed to hit a given FFB force at the wheel rim. Leading to needlessly high drag, more heat load, and slower than necessary countersteer.
If one simply wants a more reliable Fanatec, you can DIY a one motor mod with higher torque than stock, less inertia / drag and higher wheel speed (this all helps track speed due to faster countersteer) for an inexpensive price. Like 30-50 bux or thereabouts. Sometimes a lot less. You can do way more torque than stock too of course, and also slash drag massively but that is another story involving additional basic Engineering concepts.
Now some may not like this post however the theory is good and it really works well in practice. Maybe some of you can benefit from it? Or you can at least know that devices can always be improved. I still find little things that help wheel performance even after my lengthy mod process. For instance the lower drag and more reliable zero-contact steering shaft harness shown below. I DIY though (which few do) and so I don't have to wait for someone to start selling a product, nor see that some have some easily resolvable tech or frustrating forum issues that are part of the deal. I'm more focussed on brushless / servo / other drives at the moment in any event, fun stuff.