Yaw velocity output possible?

[Turn]

Is there a way to get accurate yaw speed out of the Converter? Yaw angle is easily enough obtained because this data comes right out of the Extractor as a raw input value but I want rate of change of this value, positive and negative. I'm stumped.

I'm really new to this, so please excuse me if this is super easy.

Thanks for any help.

[yokoyoko]

I'm really new to this, so please excuse me if this is super easy.

Hi Turn,

I think you want to have this motion?

It depends on the game you want to use. In rfactor and iracing I got the yaw acceleration by default. (I'm @ work now, will edit the used effect number: I guess 36, but not sure now)

All Simbin releases (and others) with the offered yaw effect - are related to the "map" and not to the "car" center of axis (hope this is explained correct). You need the derivation of this to get the acceleration.

I talked to sirnoname about this the last days - perhaps there will be "yaw acc" in simbin titels in the future. Until then you can test to get this work with the "grafic driver".

I hope you can understand my bad english.

Best regards
yoko

[Turn]

Hi Yoko. Yes, as shown in the video, that's yaw motion. iRacing only offers Yaw (effect 29), not yaw speed:

New Bitmap Image.jpg (36.67 KiB) Viewed 4058 times

The only Extractor game plugin I can see which offers a Yaw Speed effect is KartRacing as shown below (effect #32):

New Bitmap Image2.jpg (33.08 KiB) Viewed 4058 times

Yeah, I realize taking the derivation of Yaw effect #29 with time gives Yaw speed and have plotted some Excel graphs showing this works using CSV file outputs of test laps. What I'm looking for is a direct way to get this data out of X-Sim.

[yokoyoko]

Hi,

Use car velocity y = 32

This is the "yaw" movement about the axis you want!

Greets yoko

edit:

Yes, as shown in the video, that's yaw motion

Yeah, I realize taking the derivation of Yaw effect #29 with time gives Yaw speed and have plotted some Excel graphs

->sorry for that - I wanted to backup the definition. Nice idea with excel

[Turn]

No, that's not yaw speed either. Effect 32 is lateral velocity (side to side). Coordinates rotating with the car as it turns.

(I'm talking about iRacing, so let's stick to that for now for the sake of minimal confusion, but I'm guessing it's similar for other games...?)

Check this out, this is pretty cool. Here's a graph of Converter output of Effect 32 "Car velocity Y", as it's named, as a car transitions from a controlled turn into a spin. I'm driving on grass during this test to be able to control the spin at will:

spin Converter graph-Car velocity Y.jpg (21.89 KiB) Viewed 4044 times

-the car is turning to the left the whole time during the graph, i.e. all the way from point 1 to point 3 (CCW looking down on car)
-during time 'A' the car is in a controlled constant turn to the left (no skidding)
-notice the flat part. This is the car's lateral speed towards the center of the circle about which it turns (not it's angular velocity).
-now at point 2 I floor the gas...
-The car whips around, still turning CCW, but much faster, and simultaneously starts sliding outward to its right.
-This is stretch 'B'. See how the lateral velocity has gone from a small positive value to a larger negative one in the spin? Instead of laterally moving towards the center of an imaginary circle its flying outwards away from it while in a skid.

I'm not trying to prove anything, this is just what I discovered while hunting for the elusive yaw speed and thought it would be helpful to share. I did at least confirm Effect #31 'Car velocity x' is indeed forward velocity. I'm still not sure what the heck Effect #33 'Car velocity z' is supposed to be. For consistency I'd like to believe it's car velocity in the up/down direction, but why in the world is it a constant non-zero value under constant slow braking, then?! That makes no sense at all.



Still need help, please!

[yokoyoko]

Sorry still trying translate/understand your problem...
Will post screenshots - no probs on using the effect here - could be this is just luck (it worked and i moved on). The movements are ident. to the expected movements.

Edit: what is ccw?

[yokoyoko]

Hi Turn,

simple abbreviations can be hard for none native speakers. (maybe all technical guys are lousy in languages, don' t know)...

I looked @ my convertor:

a.PNG (31.4 KiB) Viewed 4027 times

b.PNG (189.23 KiB) Viewed 4027 times

If I turn a whole lap in iracing - very calm in 1-2 gear - there is no movement in the "realtime result analyzer" - in the moment the car starts to get loose there is a "peak"...

c.PNG (96.15 KiB) Viewed 4027 times

this fits your discovery - to be honest ->I just did not think far enough (i got the result I wanted and expected and tada... but...)

I got the Peak just in one direction, the "neg. peak" is the result of the car swinging back.

at least confirm Effect #31 'Car velocity x' is indeed forward velocity. I'm still not sure what the heck Effect #33 'Car velocity z' is supposed to be. For consistency I'd like to believe it's car velocity in the up/down direction, but why in the world is it a constant non-zero value under constant slow braking, then?! That makes no sense at all.

Why don't you expect a negativ value for braking (in x)?

[Turn]

Yoko, why are you adding Effect #32 Y values in your g-force plugin setting?

Untitled.jpg (12.44 KiB) Viewed 4023 times

This causes a constant turn to produce increasing math output values.

Also, what's the purpose of using the Logarithmic scaler in the math section afterwards? I'm just wondering what your logic is as I'm a beginner at this.

Aside from that, to answer your question (I think you meant to say 'z' not 'x'), I would expect the car to nose dive under breaking. I would also expect the center of gravity to not move much vertically, though; just a rotation about it. Anyway, I don't want to worry about that too much, I want to focus on this yaw speed question: how to get yaw speed out of the Converter somehow for iRacing (or other games) where there appears to be no such Extractor Effect listed.

[yokoyoko]

This causes a constant turn to produce increasing math output values

I'm no pro - just researching on xsim since a month. But as 32 is the only effect used for this axis - is there any difference if I use add, overwrite? ... Bobbuilt said use "add" when I first tried to profile (so I use this and I#m very pleased with the result - see video).

What are you trying to do in iracing?

Also, what's the purpose of using the Logarithmic scaler in the math section afterwards? I'm just wondering what your logic is as I'm a beginner at this.

I'm using wiper's - and want to have a more linear movement out of the circular movement of the motor.

I think you meant to say 'z' not 'x

there are too many coordinate systems... is a point of definition. ENU-System --> x=forward backward + roll , y=left right + pitch , z=up down + yaw , if I remenber correct.
Sorry but I can't find a definition for x-sim (so I think ENU as world standard is correct?)

I would also expect the center of gravity to not move much vertically

- doesn't this depends on the height of your rollcenter???
What about your values for effect 31 on braking?

Sorry for not really helping you on your main problem.

[Turn]

That's okay, Yoko, I'm enjoying the conversation! I didn't realize that was your rig in the video - sorry. Very nice build! You did a great job - and all this in only a month? Amazing. The bottom line is: if the effects work for you personally and feel right to you, then it's a success regardless of what Effects are used and how.

For coordinate system language I think it's easier to agree on the common "right-hand rule". Hold your right hand out like you're going to shake somebody's hand: fingers point +x (longitudinal/forward), thumb points +z (up), palm points +y (lateral/sideways). Rotating about each axis gives roll (x), yaw (z) and pitch (y). Rotation sign is determined by pointing your thumb in the + axis direction and curling your fingers. This all matches iRacing/X-Sim outputs. Sorry if you already know this.

- doesn't this depends on the height of your rollcenter???

Thank-you! I never considered rollcenter being different than center of gravity. That helps a lot. Allow me to get geeky for a second (as if I haven't already ). Consider a typical front-engine with rollcenter behind CG. Coordinate system attached to rollcenter (not CG). Vel-z is that component of total vel in its z direction (which changes with car's pitch...and roll even). That's why I was so confused. Slam on the brakes and it spikes positive because the rollcenter's z-axis (up) swings to point more down the track. Etc, etc. Wow. That clears up all my weird test results. Thanks!

I'm using wiper's - and want to have a more linear movement out of the circular movement of the motor.

Ah, I see. Cool.

Of course, none of my z-velocity talk matters with braking/acceleration maneuvers because their translation into vertical forces pale in comparison to road bumps and dips. I estimate those effects are felt 5x-10x more. So, when you slam brakes on, not only are you being thrown forward onto your shoulder belts, you're also being thrown up vertically out of your seat (a bit). All that work for nothing, only to satisfy my curosity... Darn curosity

Sorry for side tangent. Back to yaw... All I'm trying to do is figure out how to simulate the back end breaking loose as accurately as physically possible. Still much work to do.