Sim Racing servo ffb systems : OSW & Bodnar

hi Paul

no strain, just referred to when I ran for a while at higher settings, never suffered clipping either.
I do like holding a steering wheel that feels as though it's connected to something beastly though lol -like powerful racing car
I think I have the balance of settings just right now.
we tune into what ever we have -the t500 is a great wheel for the price & worth customising too if someone doesn't want to spend out on a higher end wheel

 

I am assuming that they are using a reasonable stepper motor, i.e. one that is suited for this application. If you look at many torque/speed curves for stepper motors you will see that the torque does not drop off until the speed is beyond what we do for sim racing. A torque/speed curve is usually generated form 0- many thousand RPM with little to no drop off early on. Remember most normal driving would be sub 300rpm.

 

hmm, you make a good point. Confused what might be causing the observed difference then (provided it's not software and/or the motor driver of the AF-Pro but rather the motor).

I don't remember hearing it in the comparison video but was there any mention of comparison with the peak torque variable held constant? (i.e. set the bodnar and OSW to 13Nm max).

edit: I just found this and maybe it's as simple as this non linearity of the ffb from the AF-Pro? I think the kollmorgen motors in the bodnar and OSW have perfectly linear responses.



This is what MockRacer had to say about this in his review:

Does anyone know if this is factually true or not?

I mean it would be simple enough to do via the PC wheel driver by simply checking the rotation speed of the wheel and if it exceeds "X" rpm it applies a reducing coefficient to the torque signal output. But would you really want to do that to your product? Is/was it really done for safety reasons?

But regardless of whether there is a "speed limit" of sorts on the wheel or not, it's non-linear response curve is very interesting if we're to compare it to the bodnar wheel that has a completely linear response (based on the following product description quote as I can't find a step2 linearity test graph for the bodnar online):

This could very well explain the consistent observational differences in Barry's comparison review.

So either the non-linearity is caused by an artificial speed limiter (or a more accurate description of it's function would be a torque limiter based on the speed) or we're seeing the natural fall-off of torque with increase in speed of the stepper motor. But based on how the ffb response curve appears to make up two distinct linear regions, i think it would suggest the former (i.e. artificial speed limit) being more likely (and not the stepper motor).

edit: i sent Mr.Bedo (reviewer) a question regarding the artificial speed limiter but if anyone knows anything in the mean time, please share.

 

The speed reduction is a safety measure for the motor itself.
What happens with a stepper compared to servo is that when speed increase so inductance increase and this means that, in order to get max torque (max Ampere) at a certain speed you need to have more Volt.
Accuforce is limited to 24V (if I remember well) that in turns means that Accuforce must limit the speed when inductance reach a certain level.


This is better explained here (for French speaking people):
http://www.racingfr.com/forum/index.php?showtopic=49522
This thread also means that in a few months we'll also have an OSW with a stepper motor instead of a servo

 

Ah-ha. So the problem does not reside with the motor but rather the power supply providing inadequate voltage?

 

I'm surprised no one has mentioned the Accuforce Team Edition in this or the other thread yet. Why doesn't SimExperience stop being pansies and just sell the Accuforce "Team" to regular consumers?...

source --> http://www.racedepartment.com/threads/simxperience-accuforce-wheel.81922/page-2#post-1677713


P.S. Yes, I'm sure the Team Edition provides 2x the torque of todays best direct drive wheels., lol...right.... just like the regular ("pro") model has been "Designed to compete head to head with today's best force feedback systems" and has been tailored "to meet the needs of the most demanding sim racers and sim drivers by exceeding the specifications of existing high end wheels." LOL! Right... Who hired the iRacing marketing guy? Lol.



More infor regarding the Team model:

source --> http://www.racedepartment.com/threads/simxperience-accuforce-wheel.81922/page-5#post-1816829

 

I agree, it sounds odd having to "exaggerate" forces. Nothing should be exaggerated. If the motor is capable of quicker speed and more torque (e.g. Bodnar, OSW, and SimX's own "Team" model) then it's as simple as that and should be able to run even more cars at 1:1. Nothing should be getting exaggerated. It sounds like to me they're trying "tricks" to give the user an experience that seems (keyword "seems") to overcome the limits of the stepper motor and the lower torque. Just what I'm thinking, I could be totally wrong here.

P.S. Do you or any of you guys know anything more regarding the better motor "Team" model?

 

That's just the thing. Mr Bedo made a good point in that the linearity graph shows a response curve made up of what clearly appears to be two distinct slopes. Now if this "speed limiting" effect (or more accurately it should be called an "acceleration/torque limiter" when the motor exceeds a certain RPM) were naturally caused, one would expect a continuously reducing gradient (i.e. plateauing) and yet we do not. Which is certaintly a good indicator of it being caused artificially. I proceeded to ask how certain he was of this and he said not only does the indirect evidence in the graph demonstrate so but that SimX confirmed the functions exists (for safety reasons) and that it's being controlled in the driver.

Ok, "perfectly linear" maybe too absolute a description. I simply meant that it's so linear that it can be thought of as essentially being the same as perfectly linear. (you know…e.g. 97% ≈ 100%). I couldn't find a wheelcheck.exe step2 linearity test on the internet for the bodnar, the only source i could find claiming the linearity was leo bodnar's description page where it states/claims:

And in general, it's not an implausible claim. Heck, even the stepper used in the AF should be able to but is being artificially prohibited from doing so.

 

Berney made a few assumptions and they have not all turned out to be correct.

The assumption biting him in the ass here is that sim racers want the wheel to feel like what real racers think is realistic. In some fairness, he was told this by sim racers in a poll during development of the wheel. So he went down the path of giving users what they told him they wanted based on having no idea what the difference between the options would actually feel like.

It turns out that people prefer the raw data coming out of the sim and played on the hi fidelity instrument that is the OSW and bodnar, rather then the other option using dolby noise reduction.

Thats just how it is and I understand the feedback he got from people about what they thought they wanted but hey, you are asking a bunch of people who have no idea what any of it would feel like to give their preference. They all got it wrong.

There is an easy fix and Berney is endeavouring to do something about it so good on him.

What I dont particularly agree with is the feeling eh gives off that 'all our customers are wrong but we will do what they want anyway.' The mathematical models are not wrong, the other wheels just play the music given to them. He is the only one applying any complicated mathematical modelling to the signal to get other effects. This also makes some of his statements a bit strange that his wheel feel just like a real wheel yet anyone can basically do anything with it to make it feel like just about anything. Chances are most people dont run it like a real wheel with so many options - and they would never know it either.

BTW, he may have some issues with the stepper but for the servos it is a very easy to do thing that they do. We are not changing the voltage with the servos, we only change the allowable amount of current to be increased so that we can get more torque from the motor. There is no risk to doing this, the max amps given for the motor is based on a particular usage scenario quite different from what we do, i.e. much more demanding.

From memory at least three people have taken torque readings from their small mige's that they run with more current and they remain linear to within a couple of percent from 1-20Nm.

There is certainly spin from Berney to make all their decisions and the results of the recent review look better for them but if their customers are happy then thats OK.



Oh yeah, and about the AF Pro, it says twice the forces of the best DD wheel out there, clearly he was talking about the AF at only 13NM so he only has to deliver 26... (tongue in cheek)

 

I'm going around in circles now (lol) but by what Berny is saying, it sounds as if the problem is in fact caused by torque fall off above a certain rotation speed (which is classically occurs at a much lower rpm for steppers). And so perhaps my assumption that the two distinct slopes making up the linearity response curve is indicative of something artificial was wrong there, but that there is still something causing it (which seems to be a lack of voltage).

I also though what you said here at first but then found this:

Which makes a lot of sense.

 

well said metalnwood

 

TBH, when he says exaggerated forces I read that as being the natural FFB coming from the game and we are talking iracing here.

From the long threads at iracing my observation was that what people like is different than they planned for and he is basically saying that what you want is wrong, it's exaggerated, it's not real but if you want it we can do it. We can bump up the voltage to give it more speed and make it more responsive. I.e. if you want something 'unrealistic' we will deliver it but he is pretty much carrying the line that they have the perfect 'realistic' system and making no bones that the OSW and bodnar is not but they will make the AF behave like that if users want it.


So I dont think they have any linearity issues at all, they are jsut pointing out how they have to deviate from their 'perfect' model to give users the 'unrealistic' feeling they want.

 

Correct.

Now this is a MIGE:


This is the Accuforce:


This is the CSW V2:


Speaking about reactivity the Accuforce seems only marginally better than the CSW.
Reactivity, speed, linearity of the MIGE solution is on a different league.

 

Thanks for the info.

 

Thanks for those graphs! Do you happen to that same data for the big MiGe and the T500RS (or even the TX/T300RS)? I would love to compare the CSW V2 with the TM wheels, but more importantly for me personally, the big MiGe VS the small MiGe.


A couple OSW questions for you all (regarding 2 different "small" Miges" and the "big" Mige).


1. The small MiGe at stock rated power (voltages, amps, etc.) is 20 Nm, and 30 Nm for the big one. So, how does the small MiGe perform relative to the big MiGe when both are running at 30 Nm (over-amped small MiGe VS stock big MiGe)?


2. Apparently there's a model of the small MiGe which comes rated, stock, with a quite-a-bit higher amperage. The "regular" small MiGe is the 130ST-10010 but the 130ST-10015 specs say that it's rated for more amperage, on top of that, it also shows a higher RPM than than the "regular" small MiGe. Other than that though, everything is identical - the same rated 20 NM, the same rotor intertia, everything. So why isn't the "130ST-10015" the standard-go-to "small" Mige rather than the "130ST-10010"?

 

Is the 130ST-10010 cheaper? If so, you probably have your answer.

 

Being a track day driver as well as a sim racer I can see the desire (need) for exaggerated forces in a sim wheel. Unless the sim has quality full motion we are missing a key source of feedback when sim racing: accelerations from the inner ear.

 

Exactly, I have OSW and it's really easy to notice eg. when you turn the wheel too much and the front wheels looses grip

 

Which servo motor did you go for jokeri ???

 

Small mige.