Ford Focus RS Forum banner
1,141 - 1,160 of 1,178 Posts
Hi All, my RS still trying to kill me with torque veering and it really takes fun away. Can't trust it on uneven cambered straights or sometimes on weaving dirt roads. I had it do it at 200km/hr on a tarmac uneven straight on a closed backcountry road. It moved sideways about half a meter steering all by itself. Can't trust it, on edge. Ok on race track. Totally unlike torque steer. Like dancing with an axe murderer.

Anyway, I've been watching the RDU tuning stuff for a long time. What I want is to disable the vectoring and have constant torque so it behaves as much like a LSD as possible. I don't care much about changing the torque ratios. At least if it can be set like a standard LSD and not vector then I'll know if this is the root cause of the problem.

Can it be done? What do you recommend? Looking forward to your comments re RDU config to stop torque veering. Desperately hope anyone has a solution or advice. I have the old daftracing handbrake etc but will buy the new stuff from Nutron snd hopefuly get @TomekRST to help set it up.

I'm also working on a custom suspension to get about 6.5" travel with more compliance. One of the thoughts is if I can keep the wheels more in contact with the ground maybe it will be less nuts. I'll put this up in the lifting the RS thread once I am more progressed.
Cheers
James
You don't want to disable Torque vectoring.
Pull drift compensation (PDC) is what's trying to kill you, disable that.
 
Hi All, my RS still trying to kill me with torque veering and it really takes fun away. Can't trust it on uneven cambered straights or sometimes on weaving dirt roads. I had it do it at 200km/hr on a tarmac uneven straight on a closed backcountry road. It moved sideways about half a meter steering all by itself. Can't trust it, on edge. Ok on race track. Totally unlike torque steer. Like dancing with an axe murderer.

Anyway, I've been watching the RDU tuning stuff for a long time. What I want is to disable the vectoring and have constant torque so it behaves as much like a LSD as possible. I don't care much about changing the torque ratios. At least if it can be set like a standard LSD and not vector then I'll know if this is the root cause of the problem.

Can it be done? What do you recommend? Looking forward to your comments re RDU config to stop torque veering. Desperately hope anyone has a solution or advice. I have the old daftracing handbrake etc but will buy the new stuff from Nutron snd hopefuly get @TomekRST to help set it up.

I'm also working on a custom suspension to get about 6.5" travel with more compliance. One of the thoughts is if I can keep the wheels more in contact with the ground maybe it will be less nuts. I'll put this up in the lifting the RS thread once I am more progressed.
Cheers
James
Second the turning off of PDC.

As you've been messing with the suspension, check the strut top mount (top hat) bolts. I couldn't get a good torque on them because it had a bunch of old pre-installed Loctite on them. You don't have to remove the cowl or wipers, just the fasteners. Cleaning the bolts and reinstalling with new Loctite helped out a lot.

You could also need an alignment, are you getting Drive Mode Unavailable errors as well?

Addressing those 3 things helped out a lot with the random pulling to the side on crowned roads.
 
Thanks. Yep, the PDC was turned off long ago. I do get the odd drive mode unavailable error but not often. What's causing that? Suspension is all good so I don't think its that. PDC works on the front wheels and this effect is hugely rear dominant.

BTW, a while ago now I got a suspension shop to install some hardrace parts and lower control arms - btw I live in a unit so no easy workshop. Weirdly, the lower control arm to the subframe bolt backed out dramatically while on track (not lock tighted maybe?). Super dangerous. Anyway, it moved from one position limit to another in a fraction of a second, and then back again. Timing random. Boy did it start steering from the rear - the ultimate in torque vectoring. The torque veer effect feels identical, and not much less severe, than that. it certainly doesn't feel like front torque steer.
 
Putting this up as a refresher. Is this the latest? I grabbed this table from Hanks web site. Great site by the way. Hanks RS web site

I am reading through this whole thread to get as up to speed as possible. In the meantime, does anyone have any comments on what it does with SWVEC off? What parameters is it still using to calculate? Thanks all.

Any driving experiences on twisty tarmac or dirt in any of these modes would be appreciated too.

Image
 
Sorry for bombing the thread with a lot of stuff. I think I have a partial solution. I never fully looked into the difference between sport and track. Track puts RDU in aggressive mode. Sport and Normal leave it in Normal. An unspoken assumption is that aggressive mode in Track assumes smooth race tracks. The unspoken part is that it exacerbates torque veering when road surfaces are rough - hence track use only. Sport leaves RDU in normal, which I assume reduces torque veering.

I use Track a lot on the road (mainly to get relaxed ESC cause I couldn't be bothered turning it all the way off). Over the next week or so, I'll do some experiments on a known bad bit of road. Test conditions are:
  1. Track mode
  2. Sport mode
The hypothesis is that torque veer will be less in sport mode. Thats OK/ I can just turn ESC off completely and I have tractive suspension so I can dial up what I want. So now Sport = Rally/Fast road, Track = track, Normal = boring.

It seems obvious now, but has anyone looked into this or has any comments?


Image
 
I'm busy writing a design document on this whole thing with a detailed analysis of torque veering. This will cover my take on how to remediate it, how to get a car which is supple, sporty, flat-rising, low jerk, multi-purpose and effective. Very different to the hard lowered springs and uprated sway bar approach. I'll put it up shortly when ready. As a taster, this is the direction I think I will be going in. It requires me to get custom long travel tractive suspension made, be able to remap the RDU maps, and then a lot of testing. I may be on the wrong track here and certainly welcome any comments or discussion.

Image
 
HELP NEEDED!!! Does anyone know how to make custom RDU maps and to load them into the RDU itself. I know people have done it in this thread before but the knowledge seems a bit lost and I can't figure out how to do it by myself. Prepared to pay for help if required. I've tried contacting @TomekRST on a few channels but no luck so far. Also, if anyone knows how to contact @Kacper that would be great. @Hank if you can chip in that would be great :)
I'll already have Tomeks old setup for the drift brake and will upgrade to his new setup once I get this capability sorted. Thaks all.
 
Get yourself up to speed on Python and firmware programming
 
Still waiting for that. Right now you can only do it via Carloop or ESP32
I've got this through Tomeks Nutron kit for the handbrake. however, does anyone know how to turn off ice mode. I'm using a dummy ABS and changing wiring for dirt khanacross events to stop spearing through the finish garage and into the pits. Ford is dumb, ice at 30C, hah.
 
Get yourself up to speed on Python and firmware programming
I'm a software systems engineer, so know the last thing I want to do is go down individual development with all the risks in bricking the thing and huge time requirements. It's not what I'm trying to do. I am more a customer of this not a developer in that all i want is to configure the RDU not develop it or modify strategy (though that would be nice). All I want to achieve is:
  1. Build my own map files in the Nutron map editor and then convert them to loadable files. I can't find a converter yet but will look trough Github again. Does anyone know.
  2. Load the desired map file onto the RDU. Lots of people have done this. I'll look at Hanks web site, I think he has instructions.
I'm still trying to contact Tomek to buy the device, I have a very old one for my handbrake but I'm sure its outdated and would like to install the new version permanently. Also trying to contact both Tomek and Kacper re how to use mapping editor on the Promoto site.

@Pezzar, please let me know how your work on the RDU went. I'm sure it will help me, thanks.
 
Thank you for this. I’m going to print and add it to my obsessive collection of RS info.
My first 3rd gear fast pass was a gripping experience.
Yep, anyone who has experienced torque veering, as in a 3rd gear fast pass as you say, will never forget it.
Happy to send you the powerpoint pack. not sure how to get it over to you. In the mean time here is my best guess at what happens when we lose traction completely on one wheel. The distinction being that a normal LSD detects the wheel spinning and then starts transferring torque to the other, outside wheel, in a gradually ramping up fashion and typically not fully locking whereas torque veering just sticks with the clutch setting if it is behaving naively.

I am firstly setting up to log the RDU input and output messages to see if I can identify the circumstances when it goes nuts - the edge cases. I am working on a man-in-the-middle adapter which goes between the RDU and vehicle Canbuses - splits it. If I can identify edge case conditions I can then spoof or scale the input parameters to move it to a better position in the map, for example halving the torque output by halving accelerator pedal setting, when in a veer condition, or some such treatment. I am also doing a lot of suspension work so I'll have two sets of rear springs. One with the standard native understeer ratio (it's about 0.93 F/R) and the other with flat ride more towards oversteer (about 1.15). With the flat ride setup, I'll be scaling back the torque vectoring to some lesser value, say 50%, so the RS feels more normal. I'll be using a multi-position switch so I can vary the scaling and also enable or disable the torque veer nannie - if needed (it may limit track times slightly for example - only time will tell.

This whole thing is a bit of a compromise using a man-in-the-middle attack. Nutron and Tomek are working on a much more sophisticated system which effectively directly drives the clutch pack settings and has direct mapping and can be much more sophisticated in behaviour. However, this is due later in 2025 so I'm working on the MIM approach because its within my capabilities and I know I can get something working which will let me try different torque vectoring weightings fairly easily. Next step after that is dynamic control of veering. Mine is sort of like trying to predistort inputs to get close to the result I want - a bit of a bodge but should get me close. It helps being semi-retired so I have some time ;-) Once Tomek gets his solution out and it covers edge cases and the things I need I'll most likely move to it - but I see this being at least a year away and I need to get suspension working sooner.

Image
 

Attachments

1,141 - 1,160 of 1,178 Posts