Traction control in action

[GW_Rowan]

Hi ! Just wondering if anyones got any footage of the G4X traction control in action ? 

I've got a slightly ridiculous machine (Fraser: Supercharged Type R K20a / 330hp at wheels / 670kg)  and moving to the South Island with colder road temps, and rougher surfaces, traction has become a bigger issue so just evaluating what to do about it. Running on almost the stickiest tyres I can for the road (AR-1s) 
 

Currently got a G4+ in it with an outdated Racelogic system that we just never could get working right from a sensor perspective, the project stalled for a number of years, and now revisiting it. I'm interested to see what the G4X TCS is like but footage is very very slim on the ground. Very very tempted to ditch most of the loom and go G4X + PDM. 

 

Thanks

-Rowan

[Confused]

It's something I'll be looking at closely for my Anglia very soon, so will be following, or, possibly, be the one posting up stuff if no one else has before I get round to it!

 

[Vaughan]

I don't really have any footage of it but I found it very useful in my 2GR swapped MR2 on track days, only using the ignition trim with no cuts and it would still let me slide it a certain amount when desired and wouldn't remove all the drive like OE systems (because of how I had it setup). Have also found it to be very useful driving the same car on the street when the roads are wet. My only issue with it has been that my abs sensors are wired directly to the ECU without any amplifiers and so it doesn't work below 25kph but this can be fixed with the right bit of hardware between the ABS sensors and ECU.

[GW_Rowan]

Thanks guys - @Vaughan my old ECU is down with you guys at the moment, so depending on outcome might be the push to jump to the Xtreme model. 

The Fraser is mildly terrifying on anything approaching damp, and really noticed it here as compared to up north. 

I had a RaceLogic in a TT Z32 back in the day and then (waaay back then) it was amazing. I think the approach the Racelogic had would be almost too unsettling in the Fraser as once she lets go would almost need the maximum cut level to bring it back and could be quite brutal the "on/off" nature of the aggressive fuel cuts. I then saw some footage of a Motec unit from the Motive guys in AU channel and immediately went "that's amazeballs". If the Link G4X is comparable that would definitely push me over the edge. Just surprised not many videos out there, as (good) traction control is absolutely transformative on a high HP/rwd build. I found one of a boosty boi Mini in the states but that's been largely it. The Motive guys have got an example of a high hp RWD skyline utter handful, turn TCS on - different machine. [Which I'm assuming the G4X is similar.. if so might be a good marketing thing to do!] 

Just prompted me on my lunch break to have a look, and we did get our rear speed sensor pickups working. So in theory just gotta get the front wheel speed pickups working... .

I'm assuming the G4X is quite flexible on Sensors? these are the ones I've got at the moment. 

 

[Vaughan]

1 hour ago, GW_Rowan said:

I had a RaceLogic in a TT Z32 back in the day and then (waaay back then) it was amazing. I think the approach the Racelogic had would be almost too unsettling in the Fraser as once she lets go would almost need the maximum cut level to bring it back and could be quite brutal the "on/off" nature of the aggressive fuel cuts. I then saw some footage of a Motec unit from the Motive guys in AU channel and immediately went "that's amazeballs". If the Link G4X is comparable that would definitely push me over the edge. Just surprised not many videos out there, as (good) traction control is absolutely transformative on a high HP/rwd build. I found one of a boosty boi Mini in the states but that's been largely it. The Motive guys have got an example of a high hp RWD skyline utter handful, turn TCS on - different machine. [Which I'm assuming the G4X is similar.. if so might be a good marketing thing to do!] 

We did do a lot of testing with it at the track on a BA XR6 Falcon and from memory we also did some development on our V8 RX7 which is on slicks so was definitely designed to work appropriately for circuit driving among other uses.

1 hour ago, GW_Rowan said:

I'm assuming the G4X is quite flexible on Sensors? these are the ones I've got at the moment. 

The input has to go over 1.8V to be considered On and drop below 1V to be considered off. Two wire reluctor sensors have voltages proportional to rotational speed and so there will be a minimum speed required before the ECU starts to see those signals if wired directly to the ECU. I plan to buy some amplifiers at some point to remove this issue on my MR2 , something like https://sirhclabs.com/product/vr-to-hall-sensor-converter-dual-channel/.

Your sensor appears to just be normal inductive style and so should work.

[GW_Rowan]

Awesome thanks @Vaughan  will wait and see what the story is with my current ECU and go from there. 

 

[GW_Rowan]

Out of interest I got the Racelogic system running in my Fraser this weekend (despite being 99% wired in for 8 years!). Re-confirmed "why wouldn't you run TCS if you have the option". Top of 2nd gear it really turns it into an animal..  interestingly because when you're pedalling you're drastically cutting off power, vs fuel cut is faster with throttle butterfly open, I'm hearing the BOV more

But low down (particularly cold roads) you can really feel coming into and out of fuel cuts. 

So of course I was wondering what the G4X would improve upon - as the internet is really devoid of onboard videos of it ! , and had a bit of a dig, ended up downloading PCLink. I had a look through the configuration options and I can see that for determining/configuring slip the G4X would be an improvement on the RL system - having a table of speed vs lateral G is definitely the go. [The RL system's slip %age can't be configured to drop with speed... so set well for low speed ends up being a bit hairy for high speed..] 

What was not immediately apparent is how/what controls the engine power. From what I think I've read there's configurable options to change retard and/or ignition and/or fuel cuts to cut power output. [the car doesn't run DBW so that's out of the question] Had a scan through the traction control help but couldn't see it - I read about the closed loop control but not what it's controlling .. 

So guess I'm just looking in the wrong place/being blind? 

Thx

[curly]

Torque management, at the bottom of the settings list, has to be enabled and set up for traction control to work. I too don’t have DBW, but do have a dyno and haltech rotary switch to use. I’ll hopefully be dialing it in this week on the dyno and track, and will let you know what ends up working. 

[GW_Rowan]

Thanks @curly ! I'll go check those settings out. Good luck with your setup.. 

[GW_Rowan]

@curly how did you get on with it? Very intrigued to see ... 

[curly]

Bit of a mixed bag of results for the car. As a reminder, this was the first test of a brand new build. Mk60 ABS, new engine, Link G4X, entire fuel system, Aim PDM, not much beyond the fact that it's still an E36 wasn't changed. I set position 1 on the Haltech 12 position rotary knob as a virtual aux, and set that as the disable input (don't ask me how that works) for TC. Went out for the first session in VERY slippery conditions with it disabled to verify balance of the car.  Driver reported it felt great. Had him start messing with TC for the next session on a drying track, reported it would allow slip, but then with his foot to the floor, allowed him to hold a minor slide without spinning. Difficult to get a great feel for it with the track changing as drastically as it was. 3rd session was completely dry, and we found out the flywheel bolts had all broken off, causing a major oil leak, among other issues. Parked it for the day to fix.

THREAD DRIFT 1: Always check your thread engagement. Clutch Masters provided bolts only went in 1/4" of the crank, torque spec was 80ft/lbs, they all broke. Not happy I missed this, was fixed with 1/4" longer bolts.

THREAD DRIFT 2: When you have a crankshaft issue, either at the flywheel or front damper, ALWAYS check the other end. Vibrations will travel down the crank shaft and cause issues on the other end. Knew this, and sure enough the main crank bolt had backed out. There's a good possibility it wasn't torqued properly to begin with, engine was partially built by a customer before arriving at my shop. 

Unfortunately, this still ended up being our best test session, since the 2nd time out the 6 small bolts on the harmonic damper backed out. I stupidly trusted the paint marks on the damper.

Long story short, TC without DBW worked ok, but we haven't had enough testing time to nail down PID settings and ideal slip target. I can post settings here when I get a free minute. First race is next weekend, an 8 hour enduro that looks like it'll be wet, so that'll be a good test, hopefully the engine stays together, I'll be nut and bolting for a long time.

I'll be more interested in finding out if TC, in this particular build, is contributing at all to the hardware issues I'm having. I don't think it is, but I'll only be able to explain these issues for so long before I have to come up with other theories. No idea if the engine likes this sort of constant acceleration/deceleration, so take this all with a grain of salt.

[GW_Rowan]

Hi @curly sorry I meant to reply earlier, hectic life in startup land. 

Very interesting to hear your experiences. I would be surprised if TC has any effect on the hardware side, unless your guy has 0 mechanical sympathy and just living on the limiter. I did read that certain engines may not be terrible stoked with constant ignition misfires. I'm imagining there's a difference between 'ooh there's the limiter I'll back out a smidge'. vs every corner the engines making more braps than a rotary around the corner

I've pulled the pin and going for a G4+ to G4X upgrade, for a bunch of other reasons, so the guy who's doing the work is also going to setup Traction Control. 

I've been messing with the settings on the Racelogic and it really is limiting (no pun....) to not able to adjust slip by speed. For road use it can feel good straight line coming on at 95-100 with around 5-10% slip. Gives immense amounts of extra acceleration being able to hold the throttle open, but the same slip value down low say pulling out of a roundabout it's very much wants to come on to early where you really want 15-20% slip. 

So looking forward to it ! 

[GW_Rowan]

If anyone gets this far word to the wise, in the unlikely event they follow the same path, the IE5238 sensors don't work with the G4X. Overlooked in the schematic above was the min voltage (3.6) which wasn't going below 1v. 

So now in a holding pattern waiting to get my hands (well the shop is..) on a suitable adapter to make them speak the required 1's and 0's. 

[GW_Rowan]

Well I got everything working very very briefly until the alternator pulley lunched itself (unrelated to TCS... was on the way out before)

Was hellishly addictive for that brief 10mins ! The day prior I turned on ignition timing based only, which was pretty good, but definitely needed to go to full cut based. Went with ignition cut based which sounded mega impressive and felt awesome. Definitely in the very brief 10mins it had much better engine management than the Racelogic system - the combination of ignition trim and ignition cut worked well. 1st gear is almost a waste of time in the wee Fraser and it was popping and banging away.. into 2nd and you hear the odd light popping, but it was mostly controlling it through retard alone with only slight cuts now and then. I'd had 4% target slip and it was holding it the entire way through 1st/2nd. The top end of 2nd (around 100k) was mega impressive.. you could hear retard changing as it was  controlling the slip. Felt 'almost' like it was down on power, which of course that's actually what it was doing.. 

Once I get alternator sorted I'll be tuning it and do some videos of it working. Next on the list is get the accelerometer setup correctly so we're measuring which is outside correctly, then overlay a slip table based on accelerometer/speed to reduce slip based on cornering. So say at 100 being at around 5--8% straight line, but 2-3% cornering. Whereas at say a low g 50 kay corner more like 5-7% slip. 

Annoyed, I didn't save the log with it all fully wound up and doing the business. Here's test run 1 graph, ignition retard only at around 5% target slip, with supercharger belt slipping so not reaching full power. 

 

(note the wheel speeds are out here... that's more like a 0-95 run)

[David Ferguson]

I think you'll find maximum forward acceleration with about 8-10% slip.  

[GW_Rowan]

Getting some more tuning time and setting the slip tables up how I like them, starting to trust it and had my first proper squirt through a 'chicane'. Hot roads, hot tyres (AR1s), pulling 1g laterally and getting the brave pills to lean on the TCS through it. 

Beats the pants off the Racelogic system and shows pretty graphically how there's no way I can go full throttle in this car through low speed corners on the road. 

Unfortunately my sensors are giving bunches of false slips at anything other than 5hz filter, so I need to design a new system for the rear end, once that's sorted 20hz input should get some finer grained slip control... 
 

[kiwi-racer]

This is very very cool, Im planning on doing the same thing very soon.   Interesting observation on the sensor performance impacting the responsiveness. 

[GW_Rowan]

Once I'm past my next big push at work I'm going to do some onboard videos so you can see what it's like. 

I'm assuming @Vaughan please correct me ! If I have the wheel speed filter at 20hz you'd see the a slightly finer sawtooth and therefore smoother cut application so the variance between slip and target slip (should) be a lot closer. I'll try to do some back to back runs at 10hz to see if I can feel the difference. at 20hz I was getting so many false slips it was a bit unwieldy to drive. That's with Ignition Cut. I think Fuel cut is more applicable for some engines - but we want flames

I want to tune the lateral acceleration reductions a bit more so it's a bit tighter in the corners vs looser on the straights (upto that 8% or so @David Ferguson ) the run above I was just getting a feeling for initial reductions and how quickly it would transition as you crossed +/i ve G's. Not something in my car you really want to go planting foot mid corner and getting more slip than you bargained for

It is frighteningly quick now accelerating out of a corner, you can get on the power way earlier, and have optimal traction exiting...  something that I basically other than really knowing the roads/hot tyres etc I'd have to be super careful with before. She's 400hp at the crank, 670kg at the wheels... 

I used to say my GR Yaris is point to point faster than the Fraser on unsighted roads because of that grip out of a corner, now with confidence in the Fraser I might be reversing my decision... 

As with everything the problem is just moved further along. Now you arrive at the next corner at warp speed

[Adamw]

Can you do a log with the "PID setup" set to traction control, I think the oscillation is possibly the torque recovery rate is too high (bringing torque back in too quickly when slip drops) or the proportional is too high (taking too much torque out too quickly)

[GW_Rowan]

Thanks for the offer @Adamw went for a blast this morning. I've got a pretty solid log file here. I have noticed for the first time when cold tyres it go 100% cut very quickly very much feeling like it's killing engine power, but then after it's warmed up a little less so. I can feel the oscillations so yeah if it can be tweaked that's even more fantastic (noting SAS had asked me to send the log files for him to check the settings on I just hadn't had time to ! ) 

https://drive.google.com/file/d/1lBygIEjccr6X0kv_-CD83M62gE75WHy-/view?usp=sharing

And heres' the current settings:

 

[Adamw]

11 hours ago, GW_Rowan said:

I have noticed for the first time when cold tyres it go 100% cut very quickly very much feeling like it's killing engine power, but then after it's warmed up a little less so.

This looks more the the LR wheel speed sensor is playing up to me, it is worse at the beginning of the log and gets less frequent later on.  It occasionally gives a random spike to very high speed which causes a 100% torque reduction.  

Below is one example, LR wheel speed has jumped from 60kmh to 205kmh, the engine RPM hasnt increased and the RR wheel hasnt slowed down so I think that is mechanically impossible.  Once that spike disappears the ecu quickly ramps torque back in and you can see that this upsets the car, the slip starts to increase again and causes the driver to lift off.  

So you may have to adjust the air gap or something on the LR wheel, the others seem consistent.  In most places the TC appears to be working pretty well, slip seems to be arrested quickly and controlled well.  I think I would try reducing the torque recovery rate to see if that helps reduce the slip as power is brought back in after a heavy reduction.  Try something quite drastic like halving the current value so you get an obvious change in behaviour, that will help you get a feel for what it does, then you can wind it back up if it feels too slow. 

Your slip target seems a little tight especially at low speeds compared to what I expected.  For example the 6% below at 60kmh, I would have started with something more like 12%, although I dont have much TC experience.

[GW_Rowan]

Thanks @Adamw - yes I've definitely got an issue on LR. I need to get it up on a hoist to change the gap.. it seems like it's around 1.5mm vs RR 1.2mm I actually had to adjust the wheel filter down to 5hz, as at 20hz getting a lot of false slip spikes on it, and only the occasional one at 5hz. Found it super odd that it seems to get better after warmed up.. must be so subtle the thermal expansion in the diff/mounting plate/axle (it's going off a toothed wheel mounted on the mx5 diff)

It's a very light car (670kg) with a quite square wheelbase/track, excessive slip is downright scary (it wants to oversteer very, very quickly...!) so keeping the slip targets low to start with until I get comfortable with it. 

Will have a play over the next week and report back what difference the torque recovery settings make. 

 

[Confused]

@GW_Rowan have you had an opportunity to make any more progress with this? I'm really intrigued with how you're getting on - your scenario is the closest I've found so far to my own - albeit more power and less weight!

[GW_Rowan]

6 minutes ago, Confused said:

@GW_Rowan have you had an opportunity to make any more progress with this? I'm really intrigued with how you're getting on - your scenario is the closest I've found so far to my own - albeit more power and less weight!

Yes I've gotten it largely working, but not had time really to fettle with one of the speed sensor positions so I'm getting some false spikes. 

I will say it's absolutely epic - however as I've chosen ignition based cut, err, had some mechanicals. Not sure if it's related or I'm just driving the car a *lot* harder because of the ability to get the power down and/or as I had it remapped we got an extra 40-50odd hp out of it.  I was followed (in a controlled track environment!) by someone who was trying to keep up and he's "wow - that was incredible, the noise, flames out the exhaust and you're gone!" 

It's horrendously addictive to leave it in ignition cut mode, but in my machine I know that's putting undue load on the system - given the engines' 400 at the crank and traction limited, even on warm AR-1s (and cold roads at the moment) there's a lot of high percentage ignition cuts at WOT 1st/2nd

It's absolutely night and day in my machine, as you could imagine - but instead of wheel spin that's turning into torque twisting traction. I had a bit of potential clutch slip a few weeks ago that I've never had before (again so hard to tell, is it more HP, or is it now the cutting is producing additional loads), and just had something let go in the driveline that it's waiting to get up on the hoist to see what happened. 

So does it do what it says it does? Hell yes. 
Is it tunable ? Yep 
Is it going to make you quicker ? Absolutely if tractions a problem 
 

[Arvindhmuktevi]

Thanks to you legends on this page, I have managed to get TC working on my HC92x ecu. But it feels quite violent in the way it comes on, I have seen quite a few high HP (RHM as an example) cars at Roll racing Sydney running the G4X with TC, Torque management and the cut is no where as violent.

I have tried to play around with P & D, I is missing in my PC Link software for TC, attached is my map along with few logs of wheelspin and TC doing its thing, log was completed with PID set to Traction Control.

Can anyone shed some light on what can be done to smoothen then TC cut? I have tried with Ignition cut disabled, still feels quite abrupt and almost akin to 90's JDM OEM TC's. 

Is it reasonable to expect, TC to work in the background in a seamless manner, similar to that of Motec?

Using OEM Integra Dc2 ABS wheel sensor wired into digital input + OEM Honda b18 Transmission speed sensor.

@Adamw @Vaughan

Link g4x b18 turbo ecu forum.pclx PC Datalog - 2024-09-9 8;02;02 am -derivative changes.llgx PC Datalog - 2024-09-11 8;46;06 pm TC cut 3.llgx