Sim Racing servo ffb systems : OSW & Bodnar

Hi Adrian, I have attached my settings in the picture below, let me know if you see anything funny
I'm gonna check again tonight but I don't feel any detail or dynamics anymore when at full force. It's just constant power.

 

Or
D) I'm messing up my settings, always a possibility haha

But regarding the V8, i'm already at full force when slowly driving out of the pits at Mid-Ohio.
I have put my slider also on 30NM in iRacing to see if it still clips. And it does. So I'm guessing the V8 needs more then 30NM
Unfortunately I'm currently at work, I will try and make a video this weekend to show you what's happening. Thanks for the input!

 

Cheers Kat

I'm not familiar with settings on the OSW ( I use a bodnar )
I'm sure you've checked with fellow OSW users on the OSW thread on Iracing

I'm very surprised that clipping is even an issue at all if have thought you'd have a load if power there + plenty of headroom

 

Have you noticed that iracing steering strenght starts from 8500Nm and gets lower when you adjust the slider. Altough the Nm gets lower the force goes higher.
So how does it really work?

 

That sounds like the value in the slider (e.g. 8500Nm) represents the max virtual steering wheel torque that corresponds with max (i.e. 100%) torque from your ffb wheel. So hypothetically if we say that 85Nm was the actual maximum virtual steering wheel torque on your virtual car when racing some car on some track, this would represent 1% of the max virtual steering wheel torque you set (8500Nm). As a result, you would only feel up to 1% of the max torque output possible from your ffb wheel. If you were to set the slider to 850Nm, now that 85Nm max virtual torque would represent 10% and you would feel only 0-10% of your ffb wheels torque output range. Set the slider to the sweet spot of 85Nm and you have match them 1:1 and take full advantage of your ffb wheels full torque range possible (i.e. 0-100%). If you continued to set it lower to say 8.5Nm, now you have massively clipped your ffb. The feeling would be that slight and gentle turning of the car will result in 100% constant (i.e. clipped) ffb output from your wheel.

Ofc, the actual value you should use depends on the car in question with each car having a different max virtual steering torque you can expect on track. I don't know about iracing but this is why in rfactor 2 there is a "car-specific ffb multiplier", allowing you to set the optimal ffb matching for each car.

Whilst i haven't driven in rf2 for a long time, it has been my experience that the default car-specific ffb multiplier has always been too low causing ffb clipping issues, most likely to help make weak older wheels feel better than that would otherwise. I wonder if iracing is the same….set this value too low by default.

There is more info in thread linked in my sig if you're interested.

 

Sweet! Thanks Paul.

 

If I recall correctly, logs should be auto saved to the documents folder.

 

Thanks Paul and here's your graph:



It's bizarre to see it so damn non-linear/clipped and could very well explain the observations made in the SRG comparison video (regardless of whether this is more realistic as Berney claims or not).

It looks as if anything above 30% of max ffb output produces the same torque which is very odd.

 

Very well could be. Would you mind double checking with them off to see if the result is different with 100% torque at 300ms to the one you already provided for either the large or smaller rim?

In the meantime, here is a plot of your AFPro's distance/time graph from the data you provided:



What is interesting here is that (despite effectively the same result from torques greater than 40% suggesting some serious clipping issue going on), apart from the initial acceleration, between 100-300ms the velocity becomes constant (as shown by the linear curve/slope) implying no further acceleration. This is interesting because a torque output should cause continuous acceleration (that is up until it's speed limit).

From the 100% torque, the constant velocity between 100ms-300ms is 857 degrees/second = 2.43 RPS = 146 RPM. (though i haven't provided a graph for the smaller rim, it reaches a max constant velocity of 149 RPM between 100-300ms for 100% torque…pretty much the same).

Interestingly from the picture Led566 provided/posted of an AFPro's RPM/time graph, we see a peak RPM of around 190-200 RPM that reaches steady state in 200ms:



So i'm curious if your settings are causing the funky results in your graphs.

Another interesting data set would be to use 600ms as well to see if the constant velocity after 300ms truly remains constant and/or at same rate for the 40-100% torque range.

For now, bed time. 3am in HK.

That may be so but without trying a bodnar or OSW, you could find yourself driving even better and not necessarily due to them having a higher peak torque. But please don't take that as me saying the AFPro is therefore necessarily a bad wheel. For how much little extra it costs to buy the AFPro vs the CSWv2 (here in europe at least), the price to performance ratio is amazing. Even compared to the bodnar, it's amazing but if the data is correct the bodnar does seem to perform better (based on observations by people who have compared the two). The question however is why. I'm currently unconvinced that there is no way to make it perform similar in characteristic as the OSW and bodnar wheels, even though it's a stepper motor.

 

it will not perform as well as a servo system will, because it's (the AF) not a servo system

I have no idea why people expect it too unless they've too been caught up in the hype

anyone that moves from a plastic mainstream wheel to the AF should still be blown away however, losing those nasty belts /pulleys & plastic cogs is a nice plus point

 

Perhaps your right but I'm not quite convinced yet that that's the reason. Just saying "because one is a servo system and the other is not" does not validate your claim. If you can give reasoned mechanistic explanations as to why you think so, I'm all ears. Until then, I'll wait for the final verdict. As Berney said, there technically should be no reason the stepper cant be made to perform characteristically the same/similar as the OSW/bodnar motors and the more I've researched the more I'm coming to the same conclusion based on how the different electric motors work. There is still a good chance it is due to the motor not receiving enough voltage or current to overcome the inductance (which is characteristically higher for steppers hence there lower peak RPM...the data do far seems to support this being plausibly the cause of tge issue at heart and if that's the case, it should be solvable with more power). Berney even said he would try via software first to see if it resolves the discrepancy and if not then perhaps allow the psu to supply 36V instead of just 24V to the motor.

We'll have to wait and see.

In the meantime, becareful of personal biases possibly clouting your judgement. We've all been guilty of it from time to time for understandable reasons.

 

Adrian, had an chat with John Beeson today with a few last questions before i make a decision and there is a new version being tested at the mo.
It will include new electronics in the control box to make use of the new firmware adjustments and an updated servo the same make and model with a different encoder.
Price remains the same and testing should be complete this week and ready to order!

 

Ok, thanks Paul.

Maybe but don't think it's likely. We have an observational differences in a sim between the motors that should be capable of similar performance characteristics (independent of peak torque/torque range) and good empirical data in the form of you and Led566's graphs that seem to confirm/agree with it.

 

wow I didn't know that , worth waiting a week I'd say, I'll be after the specs of the new electronics
I'm happy with mine but will update mine to new version if it's justified
(if Leo makes parts available to existing customers)
keep us informed

-------

dripper

me just "claiming"/saying "an f1 car is faster than vw golf" dosnt valadate this "claim" either

 

the analogy isn't quite the comparable. For starters, a servo motor can be made out of any motor (including a stepper converting it from an open loop to a closed loop and many use this type of servo instead of a non stepper servo because it's cheaper for the same performance provided its run below its peak RPM just before the torque curve falls drastically. There's a good chance the discrepancy in performance is due to this in which case indeed the stepper is performing worse but the issue should be rectifiable by simply bumping up the voltage supply to the motor. If it turns out not to be the case, I'll happily eat my words but not before the evidence confirms so.

We have very good evidence that F1 cars are faster. I don't think you can say the same for the stepper motor not be able to match performamce vs the motors in the bodnar/OSW if it's not configured to its best possible performance for a fair comparison. Whereas the best F1 will always beat the best vw golf.

 

might as well eat them now then lol , only joking , don't take me to seriously I don't mean to be rude

a stepper dosnt have the speed , agility , resolution , precision of a servo
I did have a really boring looking key fact sheet which was expected industrial grade ( Bodnar OSW ) servo system performance which I think is what your after, I'll post it if I can find it, with a simple explanation ( I've forgotten a lot not that I ever tried to remember it)

pretty much any type of motor can be Put into a system in the place of a servo ( a servo system with a non-servo motor )
but to a certain degree this would be pretty pointless as the whole system is then compromised
and why have high grade components when they are being let down by such a weak link ?

the AF dosnt comprise of high end components ( or a servo motor ) of a proffesional grade ( like the OSW & Bodnar systems )
however the AF system is running the stepper -it simply isn't in the same league

p.s the AF is a great ffb wheel though don't get me wrong, +we shouldnt turn our noses up at a good stepper which I'm fairly certain the AF is,
it actually came closer in comparrison to the high end systems than what I thought it would to be honest but not sure if the reviewers were being kind ( a few have gone up for sale on the Iracing forums and in a few PM's an OSW user said he was "underwhelmed " by the AF, bear in mind he uses an OSW.

I think people think I'm anti simX, I'm not at all

p.s the colours in those graphs look lovely

 

here we go, I may have posted this before, there are 6
( I seem to remember x6 but I could be wrong ) servos working precisely here,
the servo system ( for each servo motor)

http://youtu.be/SOESSCXGhFo



there would be no bashed cans due to heat fade ( even if the servos ran very hot ) the servo is very precise with every positional instruction -they could run like this for very long periods in harsh environments & need zero maintenance

if for instance a stepper motor was fitted into these systems ( swapped for the servo motors) ........bash city it would be a comedy show

of course race sim ffb dosnt have to be as precise as this, the servo systems ( Bodnar OSW ) are a massive overkill for sim racing,
I like overkill though, they transmit everything that a sim has to offer & do so with ease & plenty of capability headroom so of which may be used in the future ( future proof )

they are just a tad expensive though which is the only downside

 

If our requirements of spped range can be fulfilled by the stepper motor, then it should be able to compete with the non stepper servos (when both motors are compared in servo and non servo and non stepping driving mode). It's only limitation really is the speed but so long as the requirements fall within what it's capable of, then it should be absolutely fine to use instead of a non stepper servo. And when I say used instead of a non stepper servo I mean used in a closed loop fashion (I.e. converting a stepper into a servo motor). For racing sims, RPM's should be low enough for steppers to provide comparable performance. As it stands, good reason and evidence to believe this is the problem atm for the incompatable perfoanve characteristic but should be fixable.

Sure, I'd be interested in looking at that.

You understand that "industrial servo motors" do not use stepper motors because they only produce torque for a small band of RPM. For industrial uses, higher RPM'S are required hence the non stepper motor is used. But so long as the RPM range required lies within the stepper motors performance envelope, it should be a perfectly good and cost efficient substitute. Sim racing requirements should fit within this performance envelope provided the stepper has enough juice supplied to it.

Lastly, the Mige, Lenze and bodnar servos are simply AC synchronous motors with the addition of an encoder and servo controller. The motorsport a much larger performance envelope than stepper but it's nothing more special than that. Again, provided the two motors are compared to one another in the fair performance envelope region, the performance should be similar.

 

we have solid factual evidence re.servos ......go take a look into almost any medium to large engineering company
-do tell them that any motor can be used instead of a servo motor whilst your there just for their entertainment lol

next time your in Leicester , nip round again , I'll get my engineering pal lee to nip in (he comes round a lot ) you'll have enjoyable chat with him about it ( he works in developing industrial robotics & has to select & order control / servo systems a lot etc )

believe me I've been lectured many times but usually prefer to switch off

do also bear in mind that your knowledge is very very basic with many misunderstandings ( not having a dig here )
I understand your thirst for the correct knowledge though etc as it's more you field
( my field -business is unrelated & puts a lot of other types of technical loads on my brain lol so I try my best to keep out of complex robotic talk & stuff like that , although I take interest but keep it to surface information only)

p.s yes sim racing dosnt use anywhere near the capabilities of a high grade servo system
thus my overkill mention

 

I suppose as long as Berney uses results from wheelcheck in his marketing on the website then he will have to take the good with the bad.

i.e.

'Game provided instruction to movement time of less than 2ms (1ms to instruct + 1ms for game to read back instruction is a round trip 2ms)'

Which is strange as David Tucker already said the timing at this level is not accurate anyway... but I digress...