Discuss all things tuning in this section. News, products, problems and results.
My R32 now has a Bosch M4 ABS system which outputs 4 wheel speed signals along with yaw rate onto the CAN Bus (1mb/s). Because this requries changing the VR wheel speed sensors for Bosch DF11 "digital" ones, I fitted a Full Race ETS Pro controller for the centre diff, this uses TPS and accelerometers (not wheel speed) - and means you loose some of the "nanny" features the standard ATTESA gives you.
I'm starting to think about traction/stability control options.
I'm thinking I could use the rear slip rate to manage rear wheel traction in the Link. Assuming that the rears will spin up before the fronts(?)
I also thought maybe I could add a stearing angle sensor (or repurpose the OEM hicas steering angle sensor) and combine with the yaw rate from the ABS to setup some basic stability control.
Anybody here played with more complex traction control options on a Link G4+?
Can we get a full mod list on the vehicle? What is the intended use case - a particular type of racing, safety in bad weather? Is it traction limited already with the current tires?
It is a road legal track car. Only driven on the road to and from the track. Cage, brakes, suspension all the normal track car stuff.
Only used occasionally competitively for sprints but then just for fun.
AD08R tyres, never slicks.
Used year round for UK/European track days so wet/dry from 5-25degC. Occasionally snow ;)
600hp rear wheel, big single, so yes often traction limited.
G'day Alex, looking around the PCLink G4+ Traction control settings, it would seem the only option is to have the ECU provide ignition or fuel cuts, along with ignition retardation to control the amount of slip the ECU is detecting, as seen in the slip threshold table. I dont think they've implements any more advanced features yet you could use a steering angle or yaw rate sensor for as an input.
As much as I love Link ECU's, if you're wanting features like this, you'll be best off to switch to a more motorsport orientated ECU. MoTeC or Life Racing would be the first two I'd look at.
Thanks Zac, some basic slip based control on its own might still be well worth while, especially for the wet. I wonder if there are DBW traction control options in the Link? I may be going DBW in the future.
One positive of "basic" traction control is that its conceptually more straighforward and adjustments are going to be fairly intuitive. With steering angle and yaw rate as inputs you are into more complex "vehicle dynamics" maths and science stuff which could stretch the grey matter a bit.
For the cost of changing to an M150 or Syvecs/Liferacing I'm inclined to look at the stability control add-on for the M4 ABS, I think this is just a feature "unlock" in the software that requires a license. Being Bosch MS it will be very expensive no doubt, but they've done the verhicle dynamics work for you.
If I did upgrade the ECU I'd at least have the option of controlling the transfer case from the engine ECU using more than just Gs and TPS as inputs (which the Full-Race gizmo is limited to). Again danger of blowing my tiny mind with the control strategy possibilities.
Stepping back to the big picture again: what's the goal? Is the goal to try something new and learn something (science experiment, project for project's sake) or is the goal to improve lap times in a specific scenario? If you have a specific racing goal you really should step back for a second and ask if trying to do custom vehicle dynamics is the most cost effective (in terms of time and money).
As far as drive by wire traction control, well that get's complicated. Throttle closures are considered a "slow" form of torque reduction due to the time it takes to close the throttle and cut airflow to the engine. It needs to be coordinated with fuel or spark cuts which respond much faster. If you look at modern vehicle dynamics systems they have algorithms that use brake force distribution, spark/fuel cuts, throttle closures, AWD/torque vectoring to do it. It's a lot to ask a single person to come up with in what I presume is spare time.
Hmmmm, if I wanted optimum cost effective performance in 2018 I probably wouldn't start with a 30 year old car ...
Maybe you have the wrong end of the stick.
A)I find myself in the position of having four clean wheel speeds available on the CANbus and I'm asking for thoughts on using these for traction control functions available in the Link G4+ PNP specifically and any general observations on traction control implementation in 4wd circuit car scenarios.
As you will know there is a difference between what the manual says and how things play out in the real world. Link's iffy closed loop boost control being an example. If people have tried their traction control and find similar shortcomings in the implemented will l save my efforts.
This doesn't seem too fanciful a request?
B)Yes stability control is a more complex proposition and as I say above it may be that using Bosch's is the most practical way forward. Perhaps to stay productive in this thread we could leave that to one side.
Hey Alex, we use the Link traction control in our 350Z and it's quite effective. Provided you have the wheel speeds available to the ECU via CAN then there's no reason you couldn't set this up. The traction control system is relatively basic with limited flexibility, however you can set up your slip target tables as a 3D table so you could get a little creative here if you want to start messing around with parameters such as steered angle or lat g. You can configure whatever input you want to use and then use that as an axis in the slip target table.
I'd be a little mindful of the 4WD interaction though and how that pans out with your traction control strategy. Obviously when the car is RWD the traction control should be quite effective, however if you get to a point that you're transferring enough torque to the front wheels that all four are spinning then there will be wheel spin but no measurable slip. I'd definitely try it out and see how you get on.
On a separate note, I find the R32 centre diff control is pretty poor by modern 4WD standards (perhaps unsurprising since the R32 is hardly modern...). I had the opportunity to test out a car recently at the track and was a little surprised how sideways the car gets before torque is transferred forward. It all felt a little unnecessary and I can't help thinking that there would be a lot of potential gain with a properly developed centre diff controller.
Thanks Andre, sounds worth trying the traction control out. To handle the issue of all four spinning I was thinking of using slowest vs fastest wheel speed or rear axle vs mean wheel speed to calculate the slip % (the open front diff helps in a way with this).
I have done a lot of logging of the factory ATTESA R32 centre diff control pressure on track and it is pretty slow, often it takes almost half a second to bring the presure up by which point you can be pretty sideways as you say. The 33 and 34 maintain some pressure on the centre diff at all times making them quicker to react as the 32 has to come up from atmospheric pressure each time. Aside the from the ponderous response the control strategy is quite clever for 1989 though. From instance if you are standing still and apply more than 60% TPS it will pre load the centre diff anticipating a launch (oddly this isn't displayed on the factory "front torque" gauge). Like the Australian GrpA cars found, the factory ATTESA "map" isn't ideal for a track car, on the limit you need something predictable and fast enough reacting that its control inputs will not be out of phase with the driver's inputs. This is why I was willing to try the slightly crude but fast and predictably reacting Full-Race ETS controller. One of the things it lets you do is adjust the base pressure for the centre diff speeding up the reaction time greatly. Straightaway driving it you can feel the Full-Race controller adjusting the centre diff much more aggressively.
You can also use an GP PWM table on the LINK ECU to controll the Nissan Center diff.
As Andrew mentioned you can setup the axis on a 3D tables to make some speed or G-force correction.
I feel the Link traction control strategy is good for enthusiast use. There are more advanced strategies in other ECU's, but unless you spent several days to set it up there is no need for it.
It's actually quite interesting that after doing some pretty extensive testing with our 350Z on a rain soaked track and finding the traction control to be pretty damn good, I asked Link about any feedback on other gain settings. They said that despite having numerous tuners all around the world creaming at them for years about the requirement for traction control, they'd had zero feedback other than ours, which would lead you to think that it's a function many people wanted but very few are actually using.
Here's the M4 CAN traffic I have to work with (all 100Hz)
°/s Yaw Rate
g Lateral Acceleration
°/s Yaw Acceleration
g Longitudinal Acceleration
m/s Wheel Speed Front Left
m/s Wheel Speed Front Right
m/s Wheel Speed Rear Left
m/s Wheel Speed Rear Right
ABS Map Switch Position
bar Brake Pressure
Brake Light Switch
EBD Fault Lamp
ABS Active Bit
ABS Fault Lamp
g Longitudinal Acceleration (filtered value used by abs)
g Lateral Acceleration (filtered)
My initial idea is to use lateral acceleration as an additional axis on the traction control table vs slip rate.
Traction control intervention can then be increased to reduce permitted slip rate as the lateral g increases. The idea being that a tyre generates lateral force via increased slip angle, and the common behaviour of tyres is that you have a trade off between lateral and longitudinal forces (slip angle for slip rate).
In real terms: the car is past the apex coming out of aasteady state corner with neutral steer (no under/over), throttle is starting to open and steering angle is reducing. the yaw moment moves toward the rear, the rear tyres are still require slip angle to maintain lateral force but now require some slip rate to generate the longitudinal forces, promoting oversteer as now greater slip angle is need to generate the same lateral force. normally the driver manages this with careful throttle input, and with the help of the active centre diff in this case, but overshoot is easy in a slightly laggy, high power car. The traction control mitigates that overshoot.
Ideally you'd tune the traction control table for wet and for dry separately. And you could analyze logs to dial in it for maximize corner exit speed.
Andre, I think that it due to the reason, that (other than the Thunder) you need converters for the ABS sensors to make it work and a lot of folks are either unwilling to find them or simply find it too complicated. Which, if you think about it, it isn't really that big of a deal and in fact with Syvecs ECUs I've always been using the VR to hall converter - works like a charm.
Anyways, I've been only playing with RWD cars and the LinkECUs Traction control. I find it a bit more aggressive than the Syvecs TRC (that can also be the lack of my testing with LinkECUs system), but it's scary how fast you can accelerate on wet roads with it ;)
AlexJ, as Adrian mentioned you could use a G Sensor input and define the slip targets as a 3D table against speed and G forces.
Can someone upload a worked example or table for traction control for a RWD car?
Hi Everyone. Interesting conversation as I am just starting to try and figure out how to implement my Link Thunder and R33 GTR. I originally purchased a Full Race AWD controller but I have been thinking "why not just let the ECU do it all".
This is where I get stuck. I'm limited to a single PWM Aux output table with an X and Y axis. So I was firstly thinking Lateral (x) and Longitudinal (y). I have attached a picture of a dummy table.
But one of my very well educated friends was telling me that the Nissan AWD system also uses the steering angle sensor to pull out AWD duty. For example, steering hard into a corner it pulls it back to F15% / R85% to prevent under-steer but in a straight line it allows F50% and R50%.
It also uses throttle position. I'm guessing more TP = more AWD duty.
I am limited to a Link Thunder and R33 GTR at present so changing these isn't possible unfortunately.
The full Race Kit works solely from TP input, G sensor input and AWD duty. Surely I can do that too with an ECU? Maybe it's the tables that will limit me?
All ideas are welcome.
On the 32 and I'm pretty sure on the 33, steering angle is only used by the HICAS ECU.
TPS does go into the ATTESA and has a couple of different effects, one for example is to prime the transfer case when you apply more a certain amount of throttle when stationary. it also drops pressure to the transfer case when the throttle closes. The full behaviour of the ATTESA system is tricky to reverse engineer; gsensors, TPS and wheel speeds all coming into play in different modes of operation at different times.
I know of people using motec and syvecs ECUs to control the ATTESA solenoid, and the impression I get is they are using a 3D strategy (long/lat/TPS) similar to the ETSPro (which I guess you know is based closely on the Dccd pro controller developed for Subarus').