Differential settings - please :-)

[BrunoBæ]

Can I get some help here?
I have been experimenting with diff settings in the gearbox for the Classic Gen1 model1 F1 car.
And when I enable the Clutch LSD then I can NOT feel much difference when I lower/raise the power or coast ramp.

If I instead enables the Geared LSD then there absolutely is a difference if I change the correlating Bias Ratios Power/Coast - but because of the abrubt change in car behavioring entering/exiting a corner I would much prefer to be able to use the Clutch LSD mode.

Result

 

[Shadak]

Hi guys these are my conclusions after testing it out in the classic F1 gen1 mod2(?).
In the Clutch mode!

High Power Angle = Acceleration heavy Oversteer
Mid Power Angle = Acceleration slight oversteer
Min Power Angle = Acceleration slight understeer/neutr

High Coarse angle = Decelleration Neutral
Mid Coarse angle = Decelleration Neutral
Min Coarse angle = Decelleration slight understeer/neutr

Ill check the video out to see if the behaviour correlate.

CatsAreTheWorstDogs: Its easy (somewhat) to feel the impact of the power settings - while the coast settings is difficult to feel.

EDIT: Not much to see in the video Shadak
OK after reading your explanation it makes more sense

I dont want to make this confusing even more .. BUT
if you want to do more tests, would you mind trying something a bit more linear? Like Ultima race for example. Classic F1 seem to be a bit broken and are supposed be patched this weekend, plus they have a massive high rpm spike.

Try preload at 0NM, clutches at 4.
Set low coast to make as low impact as possible, around 30.

Now try power at 25-30 and then 85.
I dont want to dispute your findings but lower angle should allow more locking (afaik low ramp measn easier to engage the "lock") which then make sthe car power out of the corner in an oversteer manner. With high angle (harder to start locking) it will behave more like an open diff at throttle, giving understeer effect.

Test it in a longer corner, 2 or 3 gear. Switch between no throttle and almost full throttle for short periods and see what the car does.
I just had a go again just to make sure i wasnt imagining things and at low angle (30), on throttle the car rotates noticably more.

Again, im not an engineer so would be nice if someone with an appropriate knowledge could chime in. But from the physics/mechanical perspective this is how it makes sense to me.

 

[David Wright]

I dont want to make this confusing even more .. BUT
if you want to do more tests, would you mind trying something a bit more linear? Like Ultima race for example. Classic F1 seem to be a bit broken and are supposed be patched this weekend, plus they have a massive high rpm spike.

Try preload at 0NM, clutches at 4.
Set low coast to make as low impact as possible, around 30.

Now try power at 25-30 and then 85.
I dont want to dispute your findings but lower angle should allow more locking (afaik low ramp measn easier to engage the "lock") which then make sthe car power out of the corner in an oversteer manner. With high angle (harder to start locking) it will behave more like an open diff at throttle, giving understeer effect.

Test it in a longer corner, 2 or 3 gear. Switch between no throttle and almost full throttle for short periods and see what the car does.
I just had a go again just to make sure i wasnt imagining things and at low angle (30), on throttle the car rotates noticably more.

Again, im not an engineer so would be nice if someone with an appropriate knowledge could chime in. But from the physics/mechanical perspective this is how it makes sense to me.

I have a degree in Mechanical Engineering. I have researched car physics for the GPL 65 and 69 mods and the Power & Glory mod for GTR2. I do have some familiarity with limited slip differentials. My claim to fame was finding out that the 67 F1 cars almost certainly didn't use clutch pack differentials.

You are correct in what you say about ramp angle and locking from the theoretical point of view. It is also true of the madness engine as used in PC2. I haven't actually tested the AMS2 car in question using the clutch pack diff. I will have a go later today.

 

[BrunoBæ]

This isn’t a knock against the game BTW. The behavior I experience is manageable and predictable. My gut just tells me that some of what I experience may be "controller compensation" and not exclusively just the sim's modeling of the driveline.

Thats OK Bill.
But conserning this then I have to say that Im not "disturbed" by any FFB settings when I feel the car behavioring in/out of corners - because I use a non FFB wheel.

I dont want to make this confusing even more .. BUT
if you want to do more tests, would you mind trying something a bit more linear? Like Ultima race for example. Classic F1 seem to be a bit broken and are supposed be patched this weekend, plus they have a massive high rpm spike.

Try preload at 0NM, clutches at 4.
Set low coast to make as low impact as possible, around 30.

Now try power at 25-30 and then 85.

OK my good man I will try to follow your directives
But then I need to get "compatible" with your "Ultima race".
Is that a car?
By "compatible" I mean if my findings should have some relevance then I need to at least use this car so much thet I can run repetive laptimes.
So it will take me a few days.

CatsAreTheWorstDogs: Ill be back

 

[BrunoBæ]

Im baaack
OK this was one of the easiest cars to get a reasonable feeling for.
Took me a few laps.

I used the default setup with wings(2/5) - but later adjusted to(2/2).
As can be seen on the right in my pic is Diff settings for Power/Coast and Pre(0).
My assessment of the behaviour can be seen to the right of laptimes.
nu-unu = Power(neutr)-Coast(underStNeutr)
nos-UU = Power(neutrOsteer)-Coast(veryUnderSt)
os-unu = Power(Osteer)-Coast(underStNeutr)

CatsAreTheWorstDogs: Just for fun I positioned myself on the TT LB just behind the mandatory aliens

@Shadak I will return with some more testing in that car - trying to use some higher clutch settings.
To make the over/understeer more pronounced.
Because I have a feeling that my findings isnt 100% consistent. Maybe

 

[BillGuy]

Thats OK Bill.
But conserning this then I have to say that Im not "disturbed" by any FFB settings when I feel the car behavioring in/out of corners - because I use a non FFB wheel.

I’m also a member of the non-FFB club. Something tells me that we might be the only two people on the club’s roster though. In my case, the unique aspect of my wheel (ECCI Trackstar 6000) is that it has a fixed and very low range of physical rotation at roughly 190 degrees.

 

[BrunoBæ]

I’m also a member of the non-FFB club. Something tells me that we might be the only two people on the club’s roster though. In my case, the unique aspect of my wheel (ECCI Trackstar 6000) is that it has a fixed and very low range of physical rotation at roughly 190 degrees.

Haha. OK maybe there are a few others in our club. But they hide themselves under some stones
Conserning your Tracstar then it was the king of wheels before all the FFB wheels really took over.
Conserning your 190 degr then my much more simple SRW-S1 (and cheap) can be set to 2 different rotations - 180 and 360.
But problem for me is that because I use fingers for thr/br then I cant even use the lowest range fully - because if you turn a wheel more than about +- 35-45 degr then your fingers/hands get twisted too much to keep control of thr/br.
So I have to undercalibrate to about max 80-90.
But I have got accustomed to it (somewhat ).

Hehe the only benefit is that I mentally use it as a great excuse when somebody are faster than me on as example a leaderboard.
If I just could use a normal wheel then everybody should see how good I really is

 

[Leonardo Chaves]

AMS 2 is using the Pcars 2 interface with diffs so... here's what remember.
On/Off determines if that part/option of the diff is active or not.
Power and coast are angles and are reversed compared to most games which use percentage, higher numbers, closer to 90 degrees is more open (90 being fully open) lower values more locking.
Preload is like most games lower numbers are looser.
i don't really recall what clutches number and viscous lock do.

They really outta turn on tool tips.

A guide.

 

[hotak]

AMS 2 is using the Pcars 2 interface with diffs so... here's what remember.
On/Off determines if that part/option of the diff is active or not.
Power and coast are angles and are reversed compared to most games which use percentage, higher numbers, closer to 90 degrees is more open (90 being fully open) lower values more locking.
Preload is like most games lower numbers are looser.
i don't really recall what clutches number and viscous lock do.

They really outta turn on tool tips.

A guide.

Yeah, it's exactly as you say: 0 angle = 100% lock, 90° = no lock, for intermediate values the value is influenced by the number of clutches: the more clutches you use, the more locking you'll have (both power and coast) at the same angle.
In all my pcars2 playtime i always left clutches alone and just played with angles and preload, i think more clutches should also give you more "smoothness" because the force from - for example - a right wheel lock, would need to transmitt over more slipping surfaces before reaching the other wheel.

Here you find a full description of how differentials work in pcars2/ams2 including also other kinds of locking (viscous, geared)

[Magic inside!] Need Help With Differentials

Even with the help texts I'm confused about which ones I would want turned on/off and why. Could someone please explain. The diff setup in PC1 was so much easier to understand.

forum.projectcarsgame.com

 

[BillGuy]

Do you guys think that having the three different kinds of differentials available on some cars is just a byproduct of where the game’s development is currently at?

I’m no diff expert, but I honestly can’t recall ever hearing of a car that was intended to have different kinds of differentials swapped in and out.* In AMS 2 terms, that would mean a car should have either a spool OR an LSD OR a geared diff and that’s it. For nearly all cars, you shouldn’t really have the ability to change back and forth between them. Does this agree with your understanding and expectations?

I only bring this up because I see the occasional mention of people experimenting with the different diff types. It seems at least somewhat likely that this could end up being a pointless endeavor if the game’s diff choices are brought into line with the real world.

*The only exception to this I’m aware of are cars that routinely race on both road courses and oval tracks where spools are used for ovals and LSDs for road circuits.

 

[BrunoBæ]

I’m no diff expert, but I honestly can’t recall ever hearing of a car that was intended to have different kinds of differentials swapped in and out.*

You are completely right.
I guess AMS2 have just got on with this completely unrealistic monstrosity because they got it "for free" by getting access to use the Madness engine.
No other racing game than pCars2 have this.
At least what I know of.
Both iRacing, rF2 and Raceroom use only a clutched diff.

CatsAreTheWorstDogs: Thats also what I personally use in AMS2. Because of some kind of realism

 

[Shadak]

I think there could be scenarios where Geared LSD could help. As it inherently doesnt restrain differentiation (mainly due to no friction present), it can for example help with turn in on some cars.

But it doesnt like when 1 wheel has no or very little tration, eg. wheel off ground. And its also not super smooth on transition when the bias is too different. So it also has disadvantages and personally Ive not had to use it in ams2 yet. Maybe when I really get into setting up cars.

 

[Leonardo Chaves]

Yeah, it's exactly as you say: 0 angle = 100% lock, 90° = no lock, for intermediate values the value is influenced by the number of clutches: the more clutches you use, the more locking you'll have (both power and coast) at the same angle.
In all my pcars2 playtime i always left clutches alone and just played with angles and preload, i think more clutches should also give you more "smoothness" because the force from - for example - a right wheel lock, would need to transmitt over more slipping surfaces before reaching the other wheel.

Yeah, i think i forgot the clutches thing because like you i mostly adjusted the angles if i felt any adjustment was beneficial.

 

[David Wright]

Do you guys think that having the three different kinds of differentials available on some cars is just a byproduct of where the game’s development is currently at?

I’m no diff expert, but I honestly can’t recall ever hearing of a car that was intended to have different kinds of differentials swapped in and out.* In AMS 2 terms, that would mean a car should have either a spool OR an LSD OR a geared diff and that’s it. For nearly all cars, you shouldn’t really have the ability to change back and forth between them. Does this agree with your understanding and expectations?

I only bring this up because I see the occasional mention of people experimenting with the different diff types. It seems at least somewhat likely that this could end up being a pointless endeavor if the game’s diff choices are brought into line with the real world.

*The only exception to this I’m aware of are cars that routinely race on both road courses and oval tracks where spools are used for ovals and LSDs for road circuits.

It could be the state of development. It is certainly possible to restrict the choice of differential - just check out the V8 Supercar or the Shifter Kart.

It can be relatively easy to change the type of differential in the real world The Porsche 962 originally ran with a spool differential but the cars that run in historic racing today typically have a limited slip diff. In the late 70s Lotus ran their F1 car with either a ZF cam and pawl limited slip diff or a clutch pack LSD. Racing Minis in the 60s were also run with either the ZF type or the clutch pack type. However in general a car will use one kind of LSD, though it won't necessarily be a clutch pack LSD,. Which is why I find Bruno's opinion puzzling. Cars are not restricted to clutch pack limited slip diffs and sims shouldn't be either.

What you can debate is the degree of freedom you have in the setup screen which I believe is your point. I am generally in favour of restricting LSD type and settings to those used in RL. Of course finding what they are in RL isn't easy. The availability of extra options never really upset me in PC2. If the developer gets the fixed option right that's perfect but if they don't you are stuck.

 

[hotak]

I’m no diff expert, but I honestly can’t recall ever hearing of a car that was intended to have different kinds of differentials swapped in and out.* In AMS 2 terms, that would mean a car should have either a spool OR an LSD OR a geared diff and that’s it. For nearly all cars, you shouldn’t really have the ability to change back and forth between them. Does this agree with your understanding and expectations?

I only bring this up because I see the occasional mention of people experimenting with the different diff types. It seems at least somewhat likely that this could end up being a pointless endeavor if the game’s diff choices are brought into line with the real world.

In real life i think there'd be no obstacle in changing the diffs of your car to what you like most, on most racing cars it would be prohibited by rules, but the oprion can then be locked in the setup screen for those car classes.
A typical example could be my favourite car: the lancer Evo IX, depending on version and year, may have open/geared/clutch diff on front wheels and also some difference on how the Active Yaw Control controls torque splitting.
From what i read, in real life lots of tuners chenge all 3 of the differentials, and it's up to them to choose what kind to use.
Also in real life super/hypercars a lot of times you get "particular" behaviours from electronics, that a lot of times mostly act as something that can be approximated using a combination of viscous, geared and clutch in parallel on that axle, for example: center differential on audi/lambo/R34 are electronic but very similar to a viscous in how they act, but probably on a rain setting you'd want it to also act a bit like a clutched one (and so be faster to give torque to front wheels).

 

[BrunoBæ]

However in general a car will use one kind of LSD, though it won't necessarily be a clutch pack LSD,. Which is why I find Bruno's opinion puzzling. Cars are not restricted to clutch pack limited slip diffs and sims shouldn't be either.

OK I understand what you are saying - and I have honestly no hard or very educated oppinion about this.
But taken from all(?) the other sims I have used (except pCars2/AMS2) I have only seen the clutched diff.
So I just thought that this was because this kind of diff was the most common in RL.

And I have also to admit that because its often extremely difficult to correlate a theoretical correct change in as example the power/coast settings with the resulting outcome in car behaviour - so my personal conclusion have been: why the hell make it even more complicated (using 3x different diffs ) when the sim devs cannot even make the clutched diff to function correct.

CatsAreTheWorstDogs: This postulate is not a personal vendetta against the diff devs - but something I have concluded after reading peoples advices about as example how you should set up the coast in rF2 or iRacing to get the corner entrance more oversteering(rotating).
Because I have forund advices saying the complete opposite of each other.

 

[hotak]

CatsAreTheWorstDogs: This postulate is not a personal vendetta against the diff devs - but something I have concluded after reading peoples advices about as example how you should set up the coast in rF2 or iRacing to get the corner entrance more oversteering(rotating).
Because I have forund advices saying the complete opposite of each other.

That's probably because the real answer to every setup question is "depends on how the rest of the car is set up and on its characteristics".
I have to say that if you consider the "physical effect" of diff setup (not letting inside wheel spin for acc, not letting less grip wheel lock for coast, locking wheels together mid-corner for preload) is pretty easy to understand when you need to change it and in what direction.
For my driving style for example i tend to drop preload to have it rotate mid corner, find the right coast that enable me to have a bit of turning during braking without the car wanting to spin, then increase acc untill i get an understeery exit that also allows me to power oversteer if i need so (not letting inside wheel spin means understeer, but also means easier to spin external one)

 

[BrunoBæ]

is pretty easy to understand when you need to change it and in what direction.

Just so much to your "the physical effect of diff setup is pretty easy to understand":
Here is an example of how different sims often use the lock settings either rather confusing or as I mentioned above completely opposite.

You say increase Differential to achieve less over steer but in my experience it is the opposite.
I guess the question is; does the percentage slider mean percentage of lock or percentage of slip?
---
answer:
Yes in my experience more diff lock on power side decreases (exit) over-steer. (makes the car go straighter).
The higher the number the greater the lock. (opposite in some sims/games)
So we have different experiences.

RFactor 2 Advanced Car Set up Guide

[ATTACH] Why There have been some very good and some exceptional set up guides written for ISI Motor (and other) sims in the past including the...

forum.studio-397.com

 

[BillGuy]

What cars would you guys consider the most “textbook” behaving for trying things with the clutch LSD? Similarly, have you found any cars that are not giving you expected behavior with the clutch LSD compared to what you’re getting with other cars?

I’ve spent the last couple of days messing around with the clutch LSD’s power and coast settings on the f309. To be honest, I’m not really feeling anywhere near the magnitude of change that I grew used to with those settings in AMS 1 (generally speaking for most any car).

In AMS 1 if you cranked the power or coast to their maximum “effectual values” the cars were often close to undrivable. That’s just not what I was getting with the f309 in AMS 2. Again, the overall magnitude of the effect from the diff settings changes just felt way less than what I grew accustomed to in AMS 1. So, did I just pick an “oddball” car when I chose the f309 to experiment with or is this how all the cars behave with the LSD?

Side Question:
Anyone want to hazard a guess as to what the deal is with the differential for the F-Trainer cars where the diff section is completely empty and grayed out? I think I asked this back in one of the general release threads but I never got any takers.

 

[Shadak]

What cars would you guys consider the most “textbook” behaving for trying things with the clutch LSD? Similarly, have you found any cars that are not giving you expected behavior with the clutch LSD compared to what you’re getting with other cars?

I’ve spent the last couple of days messing around with the clutch LSD’s power and coast settings on the f309. To be honest, I’m not really feeling anywhere near the magnitude of change that I grew used to with those settings in AMS 1 (generally speaking for most any car).

In AMS 1 if you cranked the power or coast to their maximum “effectual values” the cars were often close to undrivable. That’s just not what I was getting with the f309 in AMS 2. Again, the overall magnitude of the effect from the diff settings changes just felt way less than what I grew accustomed to in AMS 1. So, did I just pick an “oddball” car when I chose the f309 to experiment with or is this how all the cars behave with the LSD?

Side Question:
Anyone want to hazard a guess as to what the deal is with the differential for the F-Trainer cars where the diff section is completely empty and grayed out? I think I asked this back in one of the general release threads but I never got any takers.

It seems like some cars dont really react to any LSD. I tried the F3 and it just doesnt change behaviour at all, even with Geared which is usually pretty crazy.

I usually tested this with Ultima Race. Try that one, the angles alone are fairly noticeable. Using an open Geared LSD on coast nearly rotates you along your own axis at turn in

actually im gonna do a quick bug report on the F3 thats a weird one and there could be more.

 

[David Wright]

What cars would you guys consider the most “textbook” behaving for trying things with the clutch LSD? Similarly, have you found any cars that are not giving you expected behavior with the clutch LSD compared to what you’re getting with other cars?

I’ve spent the last couple of days messing around with the clutch LSD’s power and coast settings on the f309. To be honest, I’m not really feeling anywhere near the magnitude of change that I grew used to with those settings in AMS 1 (generally speaking for most any car).

In AMS 1 if you cranked the power or coast to their maximum “effectual values” the cars were often close to undrivable. That’s just not what I was getting with the f309 in AMS 2. Again, the overall magnitude of the effect from the diff settings changes just felt way less than what I grew accustomed to in AMS 1. So, did I just pick an “oddball” car when I chose the f309 to experiment with or is this how all the cars behave with the LSD?

Side Question:
Anyone want to hazard a guess as to what the deal is with the differential for the F-Trainer cars where the diff section is completely empty and grayed out? I think I asked this back in one of the general release threads but I never got any takers.

Answer to side question: presumably the cars do not have a limited slip diff.

When you compare AMS1 and AMS2 are you using a "conversion" table to convert ramp angle and clutch numbers to % locking? Also are you removing the 300Nm viscous lock these cars have?