The eKretz CSW V1/V2 comparo/review

  • Thread starter eKretz
  • 31 comments
  • 24,041 views
1,603
United States
IN
Eric_J2358
My FANATEC CSW V2 Review

Let's get this out of the way straight out of the box. A lot of you guys know me as the guy who introduced a few mods for the CSW and CSR Elite. I'm no pro reviewer, nor a writer, so hopefully you won't be scratching your eyes out after reading all of this, heh. I'm going to be giving a comparison here between my original stock CSW V1 (I installed a set of stock motors especially for this) and the new CSW V2 beta wheel I've been testing for the last month or so. I won't get into a comparison of the modded V1 wheel and the V2. A friend will be posting up his review of that comparison in the next week or so. If you own a stock CSW V1 and don't want to be bothered with modding, and if you have the means ($$) and are wondering if it would be worth it to upgrade to the V2, get off your butt and order one! Now, we'll get further into this in a bit, but in my opinion, this wheel is an improvement in almost every possible manner over the original stock V1.

Before we delve in too deeply, let me remind you again that the unit I have been testing is a BETA unit, and there may be (in fact, I have been informed that there already are) changes made that may alter the outcome of this review in the case of production units.

A quick list of the major improvements and differences are as follows:

1.) The move from a pair of small DC brushed motors in the V1 to a larger stronger single brushless motor in the V2. The new brushless motor is almost completely enveloped in a much nicer and more effective looking heat sink than the V1 had that also has a new rear fan (instead of the fan that used to be next to the motors in the V1) blowing directly through its fins from the rear of the base. Brushless motors in general are FAR easier to cool than their brushed counterparts, since the wiring is pressed tight up against the motor can rather than wrapped around the armature. The heat can go straight from the wires to the motor can and be carried away by the heatsink, rather than having to radiate across an air gap, travel through the magnets and then to the motor can and heatsink. This should also allow the V2's motor to cool down from the same temp as the V1 even faster. The motor measures 2.24" - or 57mm - in diameter and is 2.62" - or 66.5mm - in length. It has a .20" - or 5mm - motor shaft, vs. the .125" - or 3.175mm - shafts on the V1 motors. This is a great step up for the V2 wheel in my opinion, and I'm very glad to see it.

2.) The new motor uses a larger power supply along with the stronger motor to make more maximum torque - the V1 used a 24VDC 5 amp power supply, the V2 uses a 24VDC 7.5 amp power supply - and to good effect. My stock V1 was good for about 5 Nm of torque when cold at stall, the V2 I tested was good for more like 7.3 Nm under the same conditions!

KJ2T5954.jpg~original


3.) The elimination of the V1's optical encoder wheel - which has been replaced by magnetic hall sensors in both the motor and on the steering shaft. No more loss of center! The magnet on the back of the steering shaft is also nicely rounded, as well as the wiring through the shaft being encased in a thick vinyl sheath - which should also eliminate the wire cuts that some users suffered.

4.) The belt drive has been redesigned to give a different reduction ratio that helps negate the effects of belt tension and freewheeling friction by using a smaller ratio and larger ball bearings, including ball bearings in the motor. Higher tension = less damping by the belt drive = more felt effects of higher frequency.

5.) The plastic limiter nut that rides on the steering shaft's threads inside the base that mechanically limits rotation to 900 degrees appears to have been made from a different type of plastic than the V1 unit I own. I noticed many micro-cracks in my V1 unit's limiter nut, what is known as "crazing" in the industry. I am guessing that this will eliminate that issue, but have no official confirmation on that.

6.) Cosmetic changes - the chamfer around the base's front plate has been machined after anodizing or masked before anodizing, giving it a contrasting silver outline. The formerly clear top plate of the V1 has been replaced by a smoked/tinted version, and the paint color on the side covers has been changed from a medium dark gray to black.

CSW-V2_07.jpg~original


KJ2T5936.jpg~original


KJ2T5945.jpg~original


KJ2T5942.jpg~original


7.) Steering play that some folks experienced with the V1 has been addressed - according to Fanatec they have tightened up tolerances on the mating parts and as you can see in the photos, they have added an o-ring to the steering QR mount to take up/dampen any remaining play. My personal V1 never had this issue, so I can't address the difference made by this change, but it should help.

8.) Future Current-Gen console compatibility will be addressed with branded steering wheels that will allow the V2 to be used on the respective consoles that they are branded for. According to Fanatec, they have a deal with one console maker already, and are working on getting the second. No word yet on whether this will be possible with the V1.

So, first impressions. When I first grabbed the box off the porch I was itching to get it opened up and check it out. The presentation is much the same as the V1, with some slight wording adjustments, and a slightly larger box. Upon opening it, the same nice black bag as with the V1 presented itself, including the nice bright shiny 'F' logo peeking out at me. I'll leave the unboxing photos to other reviewers. The larger power supply was packaged alongside the wheel base, along with the accessory cords and quick start sheet. I would like to see Fanatec offer a little better manual - as I have seen quite a few people on various fora that seem lost when trying to figure out why their wheel is doing something simple, like vibrating when they hit the accelerator. I kind of liked the contrasting ring and tinted top cover, if for nothing other than to give a quick visual identification of which base I'm looking at when watching a video of someone else racing online. I turned the snout of the QR by hand immediately, since Fanatec had already gone to great lengths to let us testers know about the reduced freewheeling friction. As a quick comparison, on a well-adjusted stock CSW V1 you can get this down to about .5 - .75 (1/2-3/4)Nm, and on the V2 I measured it at about .2 -.25 (~1/4)Nm. So about 1/2- 1/3 the freewheeling friction on the V2. This should help theoretically with feeling a bit more of the very light forces, and increase the dynamic range of the base. (The ratio of the very heaviest force it can produce to the very lightest). In practice, I haven't noticed it to be that important when actually racing, but there is a fellow who constantly tells all who will listen about how great this is. For me at least, it made no difference in enjoyment, immersion, or lap time. That said, I will still certainly count it as an improvement, because I'm sure there are some circumstances where it could help immersion or control by better reproducing some effects on some cars.

After toying with it a bit I bolted it on to my wheel stand and fired up some Assetto Corsa. I wanted to try it cold - without having raced my modded wheel or the stock V1 first, so I could get some initial unbiased and uninfluenced perceptions. I hadn't touched a sim wheel for at least a week prior. The wheel goes through a startup calibration similar to the V1 but slightly different - it makes a bit of buzzing/tuning sound as it spins, dialing in the parameters of the drag and position as it spins, as I understand. It goes through it's routine nice and easy, no snapping to center like on some of the V1 firmware routines. The wheel travels first all the way clockwise, then all the way anti-clockwise before coming to center and flashing the firmware version on-screen of the tuning menu display (which, BTW is NOT the same as the V1 - since this wheel uses completely different electronics, by necessity it also uses different firmware). Before entering the sim, when the wheel is on and has completed its start-up routine, there is a slight bit of a damped feel when rotating the wheel - as I understand it from what Thomas has said, they introduced this damping to prevent excessive oscillation in the wheel. He has also said it could be deactivated, but I have no info on how to do that at the moment.

Upon first hitting the track, in a Zonda R at Mugello, I tried letting the steering wheel go down the straight to see what happened - pretty good oscillation. Once I gripped the wheel again it settled right down. My stock V1 behaves much the same. (I have just been informed by Fanatec that AC has just issued a patch regarding this oscillation and will be continually working on tweaking the game code for the wheel, since they really like it). I used the following settings on the V2 wheel - SEN OFF, DEA 000, LIN 000, FOR 100, SPR 000, DPR 000. Most sims I use the same settings except for the SEN, which I alter for some things like F1 cars for instance (280 - 360). The wheel felt pretty lively, and I slipped right into the groove with it, almost as if I were using my stock V1 again, but noticeably peppier. The FFB was really pretty good, as I've come to expect from AC. I had no problem feeling when the front wheels were starting to slip, as the force in the wheel gradually reduced as it started to happen. There is a little more feel available in the V2 than the V1 in my opinion, and I enjoyed racing with the wheel quite a lot. I was able to feel the subtleties of traction loss pretty well and I would say a good bit better than the V1 due to the higher force available.

I switched to Race Room Experience and had the same impressions there - the wheel felt really good. I was really impressed by the smooth feel of the wheel - The V1 had a small amount of cogging if one paid very close attention, but this V2 wheel has none that I could feel - and trust me, I was ready to nitpick if there was any, since Fanatec claimed there was none before I got the wheel. They were not wrong, it is VERY smooth. Some of the effects seemed a little light at times, but a few adjustments to the tuning menu cleared that right up. One thing I really like about the Fanatec wheels is the ability to boost the response of the wheel in certain areas of the linearity curve by using the tuning menu settings.

I next did a bit of running in F1 2013 - the feedback in this has always seemed a bit muted to me when not going off-line (hitting kerbs or grass of course wakes you up in a hurry), but I think maybe that's more an effect of the type of car and running on what is effectively a super smooth track. I doubt that there are a lot of imperfections on F1 tracks, since they'd surely upset the cars horribly. Regardless, the wheel performed well, and I was easily able to run right with my best lap times right off the bat. Good laps rewarded me with lap times hovering within a few tenths of a second of each other - the wheel is definitely consistent.

Then I switched to the V1 and WOW! I had gotten used to the V2's stronger forces and to me it felt like I was using a rubber band wheel in comparison at first! I slowly got adjusted to the weaker V1 wheel and ran the same cars at the same tracks in the same sims as with the V2. I switched back and forth between wheels several times mid-race. Here's what I thought:

The V1 still has pretty decent feel, and if you don't drive it back to back with the V2 it feels pretty good. The fall-off when you begin to lose traction is noticeably less, so you really need to pay more attention with the V1 to really use the tactile info to your advantage. There is noticeably more cogging if you are really attentive and moving the wheel very slowly against the motors at higher power. It's still not a very large amount, but back-to-back with the V2 it is noticeable. The V2's damping that can be felt when not in-sim seems to vanish when racing - I couldn't feel it at all on-track. The V1 obviously has none unless dialed in on certain sims. The V2 felt noticeably peppier in terms of response speed on high frequency jitters and rapid twitches - however, I'm not sure if this is really due to the tighter belts and/or differing reduction ratio or just the fact that the wheel is more powerful. All in all, I feel that the new wheel is a very big improvement over the V1, and if I had a stock V1 I would really want to upgrade to the V2. It really does feel better and the reliability improvements are there. The increased power, for me, is welcome. The nice thing about having that increased power is that you feel it much better when that power drops off when you lose traction, and also for me with more power it's harder to overcorrect when recovering from a slide - the wheel holds you a little better from overshooting the correction rotation past where you need to be.

I originally had intended to do comparison Wheelcheck graphs with the V2 so I could compare to the V1 but when I tested the wheel, I found that the damping mentioned earlier made it underperform in the Wheelcheck program, so I won't be including those graphs as they don't accurately reflect the wheel's performance. I have sent an email in to Fanatec asking how to disable this if possible and if I hear back I will re-run the tests and post the graphs. I did manage some extended sessions and testing in terms of power reduction compared to the V1. My original V1 was awful in this respect in my opinion. I really felt the power reduction a lot when the wheel had gotten warm/hot - and at the worst times I measured nearly a 40% loss of power in max stall torque when the wheel was at its hottest.

This doesn't happen with every car in every sim, of course, but in the cars I like to drive - which are cars like the Z4 GT3 and Zonda R in AC and vintage F1, etc. they have higher FFB levels due to suspension setup, fat slicks and downforce. This equates to more motor heat. In cars like a Ferrari 458 or most other types of sport cars or race cars with lower downforce levels or less aggressive suspension settings, the heat and subsequent power loss isn't so much of a factor. A lot of people will adjust FFB levels down to a level that they can run fast laps - but not me. I like to feel the cars as I think they were intended to feel. If I see Ayrton Senna's hands bouncing all over the place while trying to control his McLaren on a certain track, I want mine to be too! For me, this is a great part of the fun of sim racing. I have been working with my hands all day every day for most of my adult life, so I never get tired arms from racing with a FFB wheel like I see some folks report. I can race all day long without wearing out.

I measured wheel torque on the V2 by using a stout wire wrapped along the circumference of the formula wheel rim, and measured the linear force the wheel was able to apply to the wire at max stall torque (by setting SEN to 90 and pulling the wheel past its soft electronic stop). I then later checked the V2 unit after several hours of long racing - when cold, the wheel was able to apply right around 12 pounds of linear force to the cable, which equated to about 7.3 Nm. After the very long sessions and extended running, I'm happy to report that in the car and track combos I tried so far the V2 wheel almost never lost more than a very small amount of power - in most cases on the order of a single digit percentage - as you'd expect from a warm electric motor. This is considerably better performance than the V1 wheel, and if I'm honest, I couldn't really notice the reduction as it was so gradual and came on slowly with the heat buildup. Bravo and much improved over the V1.

Some further thoughts on the bullet points

The original CSW V1 used a double reduction PJ belt drive that ran the wheel at about a 20:1 reduction to the motor speed. Getting this ratio was tough, and some of the pulley sizes bordered on the bare minimum that the manufacturer recommended using (motor pulleys @ 20mm). In my modded wheel, the reduction pulley was a very bad trouble spot for slippage also - the steering shaft belt needed to be super tight to prevent excess slippage there, which put a LOT of stress on a very small bearing (607Z - 19mm O.D. x 7mm I.D.) and a quite small aluminum shaft/axle (7mm diameter). My personal V1 wore this aluminum axle to the point that the bearing was a sloppy fit quite quickly, although I must add the caveat that it was a modded wheel and running far in excess of stock belt tension.

The new CSW V2 uses an altered design that is still a double reduction belt, still using the PJ size belt, but utilizes the better belt wrap of larger pulleys to gain some much needed belt traction and reduce slippage. The reduction ratio from steering shaft to motor has also been reduced from the 20:1 of the V1 to a much more manageable and freewheeling-friction-friendly ~ 7.5:1. Fanatec accomplished this by using a larger diameter motor pulley (now 1.25" or 31.75mm vs. the V1's .79" or 20mm) and a larger diameter secondary reduction pulley (now 1.73" or 44mm vs. the V1's 1.07" or 27mm). They also changed the ball bearing and axle on the reduction pulley to a much larger size - the bearing is now a 16003Z, 35mm O.D. x 17mm I.D. with a 17mm axle diameter. This allows considerably higher belt tension without affecting freewheel friction and without premature wear to the axle. This isn't super noticeable to me in either wheel when they are powered up and racing, but can be felt in the freewheeling friction and a little bit in the V1 as a slight muting of smaller FFB effects if the belt is overly tight.

CSW V1 reduction pulley size:

IMG_20140303_231506.jpg~original


New V2 reduction pulley:

KJ2T5953.jpg~original


CSW V2 motor/pulley:

KJ2T5968.jpg~original


The new V2 wheel has tossed out the failure-prone optical encoder wheel in favor of 2 magnetic hall sensors - a very good move IMO. The position of the motor is important because the brushless motor controller needs to know the motor armature position, direction and speed at all times in order to accurately control the motor. That said, if the position of the wheel were to be reported by the motor sensor, its accuracy would be reduced by a factor of the belt drive reduction - in this case approximately 7.5. So Fanatec have used two sensors - one on the motor and one on the steering shaft itself. So we get the best of both worlds - accurate motor control and low cogging, and accurate wheel positioning. The positional accuracy is specified by Fanatec as 4,096 pulses of the sensor per revolution, and I posted on iRacing a chart that shows all the various wheels and what this equates to in more real life numbers, which I'll repost here:

"Let's put it in more realistic/relateable terms - for a 280mm steering wheel, the positional resolution of each system at the wheel's circumference (outer periphery) is as follows:

Logitech G27 @ 798- 936 ppr = .043" - .037" or 1.09mm - .94mm

Fanatec CSR @ 1,064 ppr = .0325" or .83mm

Fanatec CSW V1 @ 1,638 ppr = .021" or .53mm

Fanatec CSW V2 @4,096 ppr = .0085" or .22mm

Accuforce @ 16,000 ppr = .0022" or .056mm

Bodnar @ 10,000 ppr = .0035" or .088mm

For me at least, there would seem to be a point of diminishing return as far as positional accuracy/resolution. Human hands are only capable of a certain resolution of fine movement on their own, and I'd expect that when racing and making rapid movements while holding a wheel it's diminished considerably from when able to slowly move very carefully. Looking at the numbers above, it seems to me that somewhere between the V2 and the Accuforce would be plenty, if not excessive."

The wiring through the center of the wheel was a failure point for a fair few V1 wheels - the edge of the hole in the steering shaft had sharp edges sometimes which would cut into the wires, eventually causing a short or complete wire failure. I would have personally preferred to see a slip ring implemented in this spot, but Fanatec have instead sheathed the wires in a vinyl tube to prevent them being abraded and keep them better centered in the shaft. (Trying to keep costs down I presume). As the magnet for the hall sensor now resides where the sharp edge used to be, and it is nicely radiused, the failures should hopefully be a thing of the past.

Old style sharp edged V1 steering shaft and subsequent wire damage:

IMG_20140327_180558.jpg~original


New V2 magnet with rounded edges and wire sheath:

KJ2T5965.jpg~original


In wrapping up, I'd like to say that I think Fanatec have done a very good job on this new version of the wheel with respect to the V1. Pretty much all the complaints I had with the old wheel seem to have been addressed. Only time will tell whether the V2 will be as bulletproof as it seems so far. I am pretty sure that the motor is made you-know-where, but didn't tear down far enough to know for sure, since this wheel isn't my property. However, being a brushless motor, this shouldn't be as much of a problem as with the brushed motors in the CSW V1. There is much less likelihood that these motors will fail - about the only thing that can go wrong with a brushless motor is a short in the armature wire insulation, which is pretty unlikely in this day and age of manufacturing hopefully (fingers crossed, anyone?) The controller for the motor itself will hopefully be as reliable as the V1's PCB, which I personally never saw an issue with on any wheel I have owned or modded when used at the stock 24V. To put it simply in one sentence, the V2's FFB reproduction seemed like the V1 but smoother, more precise and on steroids, with a heck of a lot more endurance. I will end with the immortal line from character Ferris Bueller, which goes, "If you have the means, I highly recommend picking one up."

Edit to add:

I have been informed that the motor dampening should not be active any time the wheel is receiving a FFB signal, so I'm not sure why the Wheelcheck steplog tests do not look right. Fanatec have taken a look at the info I've sent them but haven't gotten around to figuring out what is going on yet. If they do so, I will update again.
 
Last edited:
Thanks appreciate this a lot Eric and also well done Fanatec! Have got to give them props for this, have complained about Fanatec in the past so feels only right to give them props when they do something good!
 
The wiring through the center of the wheel was a failure point for a fair few V1 wheels - the edge of the hole in the steering shaft had sharp edges sometimes which would cut into the wires, eventually causing a short or complete wire failure. I would have personally preferred to see a slip ring implemented in this spot, but Fanatec have instead sheathed the wires in a vinyl tube to prevent them being abraded and keep them better centered in the shaft. (Trying to keep costs down I presume). As the magnet for the hall sensor now resides where the sharp edge used to be, and it is nicely radiused, the failures should hopefully be a thing of the past.

I found that detail interesting. I have one of the very first CSR Elite wheels and the steering wheel cables are already sleeved in vinyl. Any idea when Fanatec made this change? I imagine a number of CSW V1s were revised in this manner.

Interior shot of my CSR Elite:

csr_elite_shaft.jpg
 
Actually I think most of the Elites already were, but not the CSWs. They use different cabling and have a different number of conductors also. I have never seen a CSW V1 with the vinyl sheath, (and I have seen a few) but that doesn't mean that they aren't possibly out there.

Edit: a guy at iRacing contacted me and said that his CSW V1 purchased in July this year has the vinyl sheath over the wires.
 
Last edited:
For the damping it seams that you have to set DRI to 5 to remove the artificial damping (off means full built in damping and then it decreases from 1 to 5).
 
For the damping it seams that you have to set DRI to 5 to remove the artificial damping (off means full built in damping and then it decreases from 1 to 5).

Someone said the same thing on iRacing, I'm just going to copy/paste my reply from there:

"Yes, I'm already aware of that, but it only seems to work at lower speed, unless they are limiting the wheel in some other way. I've spun the wheel when powered off and it will go pretty far, spins pretty freely. Powered on with DRI off it will stop immediately. Turning DRI to 5 makes the wheel easier to spin, but I conducted the same test powered on and with DRI at 5: the wheel will spin almost from one stop to the other with a gentle spin. However if I spin it vigorously it slows down and comes to a stop much earlier. It also reacts differently when spun clockwise and anti-clockwise when powered on with DRI at 5. When powered off no such difference."
 
When you spin the wheel do you do so with the BMW wheel on or Formula (or other wheel) ?

Just curious tested on my Buhler modded V1 here and the Beamer wheel sure help to spin it further compared to Formula wheel. Also I wonder is the DRI setting of 1 - 5 for the V2 only or V1 with higher FW then I´m running on mine (037)?
 
Formula wheel here. The results I spoke about in my last post above are only applicable to the V2, which uses different firmware than the V1.
 
Great review, because of being technically in-depth on the one hand, but also precise in terms of what is felt in a direct comparison with the old wheel. I like how you factually put in things like the higher dynamic range of the base in terms of torque resolution and honestly say that you think it doesn't really make a difference.

It is also a bit of an eye opener for me that the V1 had - in my opinion - such poor mechanical quality when knowing about FFB motor overheating and internal wire cutting. Good to know how things have been done differently this time round.

Too bad you did not put in the current Wheelcheck graphs, because this could have been an opportunity to test out Fanatec's response; whether or not they respond at all and how their improvement will affect the result. I would almost want to ask to add the current graphs in there anyway for the sake of accuracy.

Overall a very informative read and thank you for putting the time into this.
 
Thanks guys, I can probably add the graphs in, since it's starting to look like they may possibly be as accurate as we're going to get them. It seems that the Wheelcheck test doesn't really put a magnifying glass on the difference in feel between the wheels. The difference in the graphs between V1 and V2 are subtle, but they are there if you know what to look for.

As far as dynamic range, it actually does help with wheel feel that it's been increased, but the increase in maximum power does a lot more to help than the decrease in freewheeling friction, IMO.
 
It remembers me some correlation graphs between scientific accuracy measurements on loudspeakers and users quality perception. Once you reach a good enough restitution level then there are no more homogeneity between what oscilloscope and ears measure . So maybe once there is enough speed and torque to send to the user enough informations then it is more about anyone taste than fidelity.
 
So maybe once there is enough speed and torque to send to the user enough informations then it is more about anyone taste than fidelity.

At that stage we start moving into the realm of special rigs to optimize the stability and ensure we feel microeffects from our wheels, then it'll be high quality gold plated USB cables to ensure the signal quality towards the wheel is optimal, and heading into multi-thousand-dollar power solutions to ensure that the motors receive smooth and unblemished AC power...
 
Not sure that hands are as accurate than ears :)

Ears and brain can hear the brownian noise, so I think it is easier to build a sim wheel :)
 
Awesome review!! Mines is scheduled to ship out late January. Should I change from the gt2 wheel to the f1 wheel?
 
I think that depends on your preferred driving. I really like the formula rim but I mostly drive F1 cars.
 
I don't drive a lot of Formula cars but I bought the Formula rim as well as the BMW rim for the sole reason that the McLaren 12C GT3 uses a Formula rim and that is my favourite of GT3 cars, so I'll use it for that.

It's nice to even just have the option to drive a Formula car with the right rim, would feel weird steering one with the larger rim, although in saying that it's weird to drive a Formula car while in the same driving position as the other cars anyway but you know what I mean.
 
I got the F1 rim as well as the GT rim just to be able to mix it up a bit. I love racing my GT's but every now and then take on a formula car. Also saw a 280mm rim online for a good price for when I feel like a smaller rim.
 
Did you ever figure out how to run wheelcheck properly and compare to other wheels? i have no clue how to benchmark, graph. I wanted to test my 911 gt2 and compare to my V2 when it gets here.
 
Back