My car tried to kill me...

[johnl]

A few nights ago I was negotiating a fastish sweeping left hand bend over a crest, to be suddenly confronted with another car heading toward me at excessively high speed in the middle of the (narrow) road. I backed off and the cars line tightened (as wanted and expected) so that my two left wheels were just off the edge of the dampish tarmac, which caused the start of a very minor oversteer. A touch of opposite lock and the car immediately started to straighten out as expected, no dramas, but then, the cars' stability control system (VDC) decided to apply one (or more?) of the brakes, which then flicked the car in the opposite direction, into the path of the oncoming car. There was no accident, but it was a scary moment.

The problem seems to be some latency in the system, it doesn't respond instantly. So, by the time it does, the driver is already doing something about the problem. At this point the driver is already doing X to straighten the car out, and then the car starts doing Y. These two different actions may conflict (or are unexpectedly cumulative?), and the car becomes unpredictable, and thus difficult to control.

This is the first time this has happened to me on a tarmac road (not counting the time – on day of purchase still with near bald rear tyres – when it went stomach churningly weird at low speed on a very slippery roundabout). But, I live on a dirt road where it is very very easy to slide the car enough for the VDC to kick in, and the results are always similar, the car becomes unpredictable and very unresponsive to driver input. When the car does react to driver input, the reaction is unpredictable, and its' response to driver correction makes it feel as if the steering has become substantially disconnected from the front wheels...

This 'VDC' system may well work OK, if we assume that a given driver is insensitive and takes no immediate and effective corrective action to control a slide. However, if the driver is in fact already taking the correct corrective action, then when the VDC system kicks in it seems to create a dangerous problem. Too many cooks in this particular SPAM... (why does this site automatically and insistently replace the word 'k i t c h e n' with the word "SPAM" )

Can the VDC be turned off? I know that Alfa doesn't give us the option (a convenient on / off switch), but is it possible to do so by altering the computer programming in some way? I suspect the system could be disabled by pulling the ABS fuse, but this could cause other issues?

Regards,
John.

[poohbah]

Only (stupid) solution I can think of is to trade down to a slightly older model - neither of my 156s were fitted with VDC, which I believe became standard equipment after 2002.

By the sounds of it, it really isn't set up to deal with gravel - which is somewhat of a problem in country Australia.

I am regularly impressed with the handling of my '99 V6 in emergency situations. I was travelling at 110kph (WA limit) on country road the other day, when I spotted at the last minute a sheet of guttering had come off a trailer and was lying directly across the lane I was in. I was probably only 30-40m from it when it became visible (it was painted gray and was almost completely invisible). I barely had any time to apply any braking but managed to swerve right (at speed) into the vacant RH lane - only to see a second piece of guttering lying completely across that lane too about 20m ahead. Not wanting to slash my tyres on the sharp metal edges, I managed to swerve left and get around it without making contact or losing any control.

Felt like I was in the high speed chicane at Monaco - and the car handled it with aplomb. My son merely looked up from his phone and said what was that? Obviously, I never left the tarmac, but it was impressive nonetheless.

Its a pity the driver who dropped the guttering (and who had parked in the scrub beside the road and was waiting to collect it when the road was clear) didn't think to wave at the oncoming traffic to warn them the guttering was there.

[Neil Choi]

Is the ASR in a post 2002 JTS same as VDC as you mentioned (I think it is), if so, there is a button to turn off ASR, that is the case in a 2002 JTS.

Also I would suggest NOT to pull the ABS fuse to alter the ASR or VDC or ABS, as it upsets the balance of the car greatly, especially the braking bias. 

How I know?  Without ABS, the shift of weight to the front of car and the subsequent lightening of the rear causes the rear to wag and the rear brakes to lock up very easily, which results in swapping ends. 
I found this out a few times at Winton (just locking without swapping ends) with the last attempt spinning and sliding 180 degrees towards the outside wall, stopping 6 inches short luckily.  Then I puddled around and came in.

[Neil Choi]

Also I believe the ASR retards the accelerator if your steering angle is too great.  Found that out at Broadford, trying to accelerate out of a corner and the JTS wouldn't until it straighten.  Then I turned off the ASR and a little more happy.

[johnl]

Poobah,
I agree the basic handling is very good (especially with my 20mm rear ARB), despite the stock spec TRW front dampers being pretty awful (and they are new ones all round, dreaming of Bilstein...). It's when the electronics try to take over that it all falls in a hopeless and dangerous sloppy heap. I don't need to be saved from myself, I can do a better job, at least than this particular system seems able to do...

While I'm having a whinge, I don't like the ABS either. I can cadence brake better than it can. It gets really bad when I'm (instinctively) cadence braking and then the car tries to do it too, adds a lot of braking distance and feels bloody awful. Of course on a dirt road you actually do want to be able (if you choose / need) to lock up the front wheels, as this gives the shortest stopping distance on gravel roads (not tarmac though).

On the other hand I don't really mind the ASR system as it seems rather benign with no Mr Hyde to its' Mr Jekyll, but I do tend to turn it off, if I remember to.

Regards,
John.

[johnl]

Also I believe the ASR retards the accelerator if your steering angle is too great.  Found that out at Broadford, trying to accelerate out of a corner and the JTS wouldn't until it straighten.  Then I turned off the ASR and a little more happy.

Neil,
Even with the ASR switched off I haven't found wheelspin to be a significant issue with this car, it seems to put the power down fairly well (its not a GTA...). Having said that, on a wet tarmac road I have encountered significant wheelspin even with the ASR left on. It does eventually step up to the plate, but like the VDC also doesn't seem to react particularly quickly. Unlike the ABS which is a bit overly enthusiastic IMO.

Where the ASR is noticeable is accelerating hard on gravel. Switch it off and I do get increased loss of grip, but on gravel the car will accelerate significantly harder with a bit of wheelspin. Personally I'd rather switch it off, I can modulate the pedal well enough...

Regards,
John.

[poohbah]

John, if you really want to disengage the VDC, I found this on youtube.

https://www.youtube.com/watch?v=wE-3wNEs_zY

[johnl]

Poobah,
Thanks. I have come across references to the VDC system being 'disengaeable' by that or a similar method, but my understanding from what I've read is that it only works with the GTA version...? Anyway, I'll try it and see what happens, let you know...

Assuming it does work as per the video, I suspect the car will nag me to death with "VDC failure" warnings and chimes...

Regards,
John.

[poohbah]

Best of luck John.

[Colin Edwards]

The traction control / stability control system can react abnormally if the transducers are telling it the car is doing something it in reality is not doing.  Things like significantly odd tyre pressures or tyre diameter / circumference can stooge the chassis computer into thinking the wheels are spinning at a different rate than they are in reality.  Also the steering angle sensor can inform the computer the wheel is turned to the left when in fact it is turned to the right!  Imagine you have a bit of left opposite lock on and the wheel sensors confirm this however the steering shaft sensor tells the computer the wheel is turned to the right!  In this instance the ABS is going to try and correct an opposite lock drift that is physically impossible.

I'd be checking tyre pressures and tyre circumference.  Then have the steering shaft / angle transducer checked or calibrated.  Has the car recently had a wheel alignment and the steering wheel is a tad "off centre"?

[johnl]

Poobah,
Tried the 'pedal trick', and nothing. Not surprised.

Colin,
I have aligned the car personally, with weight equivalent to my weight placed in the driver seat (75kg). Front camber has been corrected so that right and left are within 0.1° of each other. Rear camber is equal to about the same amount, maybe 0.2°. Camber adjustment is achieved by 'slotting' any holes that determine camber, e.g. the four holes that locate the casting to which the upper 'wishbone' is attached (in this case the right side upper wishbone has been moved inward by about 4mm, to correct the 'stock' right side neg camber being 0.4° less than the left side camber). Caster has also been equalised in a similar manner by moving the left side wishbone mount backward by 3mm (more on how I determine caster if you ask).

This is likely within the degree of error with which I can actually measure it with the equipment I have (basically an electronic inclinometer, string lines, two different length 'straight edges', and a steel ruler). I suppose I should expand a little on my method.

This includes lightly machining the outer edges of the rims (my special technique, which allows it to be done 'on car', ask if you want to know) to ensure that there is no lateral 'wobble' in that part of each wheel from which camber (and toe) measurement is taken. Yes, each wheel on this car did have some 'wobble' of up to about 1mm, which has been 'corrected' (a wheel that is about 1mm off square will result in an inaccuracy of a bit less than 0.1° when measuring camber, or toe). As a result of this each wheel has a 'squared off' lip at the outer edge (parallel with the wheel face) where some metal has been removed. This squared off 'ring' area ranges from about 0.5mm to about 4mm in width, depending on how 'out of square' that part of the wheel was (i.e. innermost or outermost part of the 'wobble', and degree of wobble), and how much metal needed to be removed to correct it. If you look closely at the edge of the rims this isn't all that pretty, but then if you look closely the rim edges also have some curb rash marks (previous owner), so visually it's no worse really, and the wheels are silver in colour, so the 'raw' aluminium doesn't really stand out.

For camber measurement the car was placed on three shim stacks (one per wheel, other than the 'highest' one) corrected for height to within 1mm or so (using the inclinometer and a long straight edge). Camber was directly measured with the inclinometer and a shorter straight edge (placed across the wheel from machined edge to machined edge, top to bottom). Toe was measured from parallel string lines directly to the machined rims. This was all done very carefully and has been done a number of times on this car, every time I make any adjustment that might affect alignment. I've aligned a number of cars using this basic method, the results are good so long as you think and proceed carefully.

Now, the front toe is set with the steering wheel held at the straight ahead position. This doesn't mean that the steering wheel will be straight when driving in a straight line on a 'flat' road. This is affected by tyres as well as alignment, and why so many cars do not drive with the steering wheel dead straight even when all alignment parameters have been adjusted (with wheel at the 'straight ahead' position) to be near perfectly equal side to side, being why it is often necessary (even when new tyres are fitted all round and the suspension is in good condition) to adjust the centering of the steering wheel post alignment (by shortening one tie rod by X and lengthening the other an equal amount). This can be caused by tyre 'conicity' which can be worn into a tread, or the same term seems used for tyres where the belts are laid a bit off centre (i.e. a tyre fault in manufacture), creating an 'efective' conicity even with new tyres (because the tyre will be a bit stiffer on one side than the other side due to the belt offset). It can also be caused by the 'PRAT' (plysteer residual aligning torque) that is inherent in all tyres to greater and lesser degree.

PRAT is not the same thing as tyre 'conicity', is independent of tyre wear, and is not a fault per se but inherent in the way tyres are made (in varying degree). Its directionality cannot be altered by swapping the tyres side to side or 'flipping a tyre on it's rim (PRAT can and often does also cause issues with steering pulling slightly to one side). It just has to be lived with, hopefully it isn't bad enough to cause an issue, and it's affect with a given tyre can vary on different cars with different alignment settings. PRAT is created by the inherent and unavoidable angle of the radial belts (which is diagonal across the tread face under the rubber), especially the outer belt, and to a lesser degree by tread pattern if the pattern has a 'diagonality' to it.

Sorry, I'm rambling on, and on...

My car does require some correction of the tie rod lengths to achieve a perfectly 'straight' steering wheel, even when camber, caster, toe etc are very very close to being equal side to side (each front equal to each other, each rear equal to each other). This changes a bit if I swap front tyres to back and vice versa, with the current front tyres being Pirelli P1 and the rears being Maxxis Victra. This is not at all unusual, i.e. it's unwanted but common when different tyres are fitted to an axle line, i.e. the steering wheel angle can change all else still being equal (or it may not, depending on a lot of things. i.e complex interactions beyond my understanding). My tyres are different brands, but the same nominal size, there is likely to be some small difference in circumference, but I expect it would be minimal, and less than likely with four tyres of the same brand / model where one axle pair were near new and the other axle pair quite worn. My tyres are all near new.

Still, it could be an issue for the electronics, but I tend to doubt it because the steering wheel is straight when travelling on a dead flat surface (i.e. not sloping off one way or the other), and I suspect that would be all the computer would likely be concerned with other than detected differences in wheel rotation and direction of lateral acceleration, rate of yaw etc. (?).

At any rate I don't think the VDC is not working as intended, it seems to do the same things whether the car slides to the left or to the right. It just seems a bit dull witted and slow to react.

Regards,
John.

[Colin Edwards]

Yep!

[bazzbazz]

May I suggest recalibrating the steering sensor. For problems like this it does do wonders at times.

As all these sensors and systems are interwoven, it just being out a mm can upset things somewhat.

Baz

[johnl]

Thanks Baz.

As I said before, but can't be certain, I don't really think the VDS is malfunctioning. But, out of curiosity, how is the steering sensor recalibrated?

Regards,
John.

[bazzbazz]

Its done via MulltiEcuScan or AlfaOBD