True, again this is a case where the system is being used an excuse for poor driving, in damp conditions it is not safe to enter ac orner in such manner and as the manual was quoted DSC cannot defy the laws of physics. If you had no traction then the system cannot manage each wheel's speed in such a manner to point the car in the direction the sttering wheel is pointed. DSC is a computer program monitoring wheel speed; it cannot create wheel contact with the road. Again similar misconception that AWD and 4WD has created.

Again I agree the masses will likley be helped by DSC, if they do not fall ill to these misconceptions. However DSC and ASC both prevent utilization of throttle lift induced oversteer and remove understeer both of which I have used witht the systems off to avoid obstacles.

 

Having a 911 with an SM4 Autronic ECU and AIM MXL data logger I have a little experience with this stuff. In fact its far more than simple ignition timing thats measured. For every cylinder theres not only timing there's duration, both for fuel and ignition. The maps occur over three dimensions and include load measurement, throttle position, all manner of stuff including altitude, ambient temperature, cylinder head temperature and in closed loop systems there are o2 sensors of course which attempt to adjust these parameters to see the ideal burn ratio in the exhaust gases.

Its important to remember that while ecus are pretty reliable (although my first one went bad in my race car) the sensors are another matter entirely.
Even street cars are pretty hostile environments. Don't know about you, but I can live with throttle by wire, but I sure wouldnt own anything that was steer or brake by wire...err well I guess you DSC are already there with the brake by wire thing .

 

just give a concrete example of there being more redundancy built into these systems than most realize or give credit to.

Matt

 

As the others have said ...doubtful.

Just look at the quotes about following the line. If your already IN the skid, what is the car going to do?

Maybe, but still intrusive. As to how statistics are gathered, I guess you got to ask the government. Dunno

 

all accidents (that are reported to either insurance or the DMV, I'm pretty sure) have to be reported to a central data base. That's why getting statistics for the US is pretty easy. Other countries may or may not have such requirements, so getting trends out of Europe tends to be harder.

I don't remember who holds the DB... I was reading about it when I was learing about the stats on daytime running lights.

Matt

 

I was out driving in the snow today. No DSC issues as I kept speeds low and steering, brake, and throttle inputs smooth.

But, for traction control, it's not all bad. My friend's apartment has a hilly, dirt access road. It had not been plowed, so there was maybe 4 incehs of snow in it. Sure, I could have modulated the throttle to get up the hill without too much trouble. But, I also had the option, which I used, of keep the throttle steady and letting the traction control handle the power modulation to get me up the hill.

Some may say that that is making me a worse driver since I don't have to know how to get myself up a hill in low traction situations myself. Which is true. I say it's nice sometimes to have technology make my life easier.

Traction control on the MINI does work pretty well in situations like that. Coming from a dead stop, it doesn't work that well. Coming up to a snow covered hill with a reasonable amount of speed (not meaning fast, just normal driving speed) the traction control actually does work pretty well.

 

There is nothing "wrong" with having technology make your life easier. If that were true, we'd all still be making "fire" by rubbing two sticks together.

 

This is a little off-topic, but I'm new to DSC and am surprised at how slow it is to react (e.g. ~2+ seconds). If I accelerate hard either straight or during cornering, I get a fair amount of front wheel spin before DSC takes control. Is this typical DSC performance?

 

That would be the traction control. The delay is the same reason many of will get going and then go "Oh no what just broke" if we aren't expecting it to kick in. I think its more nature of the beast as it does take some time for it to recogonize the slipp then cut the power.

 

MTH is reputed to make the ASC less intrusive. Whether it goes far enough is another question.

Having driven a car with Racelogic programmable traction control, I'd estimate reprogramming the ASC to allow ~20% wheelspin would be ideal for most uses including track. My second choice would be wiring an auto-up module to momentarily disable the ASC only when the unused left middle steering wheel button (three spoke wheel) is held down.

With a Comp rear bar at full stiff and camber plates I can get the car loose just by lifting the throttle in a corner, and then countersteer + apply enough power to straighten things out before the DSC even kicks in. Unlike the ASC, the DSC is really not very intrusive at all the way it is currently programmed so I can't see why this is an issue. At least it's no issue in use for those of us with DSC rather than those who rail against it for philosophical reasons (or confuse it with the ASC). Remember it all depends on how it's programmed.

If it has to come down to equipment or software, I'd rather have LSD than traction control, or snow tires than DSC because they give you more traction rather than just making the best use of what's available. If you already have those then I see nothing wrong with adding well designed software aids to make best use of them.

Software is the biggest advance in automobiles in the last 40 years, and whether you choose to use it to parallel park your car (Lexus) or let autopilot steer you down the road (Honda, GM) is entirely up to the buyer. I know if I got a Ford GT I'd add aftermarket traction control to it, and maybe someday programmable DSC will be available too...

 

Way way way back, 2 cups of coffee ago, in post #8, Motor On mentions that the MINI was least likely to roll over.....I 've heard that mentioned somewhere before but would love to see the actual reference......

 

I'm not sure I'd think of individual cylinder control as redunancy, rather efficiency. But anyhoo, my point was that even in what one might think is a completely controlled environment with a set of well known variables, to truly extract optimal performance requires inputting a raft of info thats precisely tailored the the exact engine we're talking about.

If I think about having to brake a single wheel, I'm wondering about the effects of tire pressure and contruction, wheel weight, brake pad compound, the road surface coeffient of friction, suspension loading, camber and toe angles, etc and their impact on both the outcome and seemlessness of the system. If you have the dealer install the JCW 18:" wheels and the brake kit, is a DSC car reprogrammed? Should it be?

 

Oh yeah and by the way, for the technically inclined, I suggest you go take a look at various ECU specs, CAN bus and the like. I think you'd be quite amazed how primitive these sytems are computationally. The most powerful Autronic, for example, is powered by a 16 bit 20 Mhz (yes Megahertz!) processor. The big Motecs, and we're talking multi-K dollar units here, tout their 32 bit motorolas... which run at similarly diminished speeds. These are state of the art race units. So if you think that all these calculations are being done by incredibly powerful supercomputers... well.

 

I think the amount of milage a modern car gets, with nearly perfect ignition and fuel delivery over the life of the car, make a strong case that it's not megahertz that ensure robustness. The implication that these aren't robust systems is just wrong. For example, while CAN bus and it's derivatives are pretty simple 2 wire (sometimes 1 wire) serial implementations, they can have lots of nodes shouting down the wires at the same time, and they arbitrate just fine (by design) with all lower priority nodes shutting up when they learn that a higher priority node is asking for the bus. The ABS system is CAN based on the Mini....

While only 20 MHz sounds slow by computer standards, a single clock event is so much faster than almost anything on the car. It's a microsecond device is a millisecond world.

There is usually a trade off between complexity and robustness. I'd much rather have the robust 20 MHz CPU running my motor than the latest multi-gig CPU that has more horsepower than a car will even need, and the reliability that goes with it's design!

Matt

ps, and the systems aren't programmed from first princples, they are a reactionary feed-back based system. Is the wheel spinning? Then remove X torque. Is it still spinning, remove Y...... So change tires or wheels, and what detects wheel spin will work just the same. The levels of power removal are programmed into the ECU. As I've added more mods to my car, the amount of power removed (with the MTH mod) used to be optimal, but now isn't quite enough.... But this is with over 20 WHP of additonal power. So the envelope programmed into the current car covers a large range of minor changes.

 

True. Check out this http://www.thetruthaboutcars.com/?p=2054

If anybody things electornic nannies are going away ...

"Enthusiasts will argue that forcing drivers into automated cars is using a sledgehammer to crack a walnut. “Better driver education” is needed. ...

We— you, me, every multi-tasker scarfing a breakfast burrito, every bozo in a pickup truck convinced he’s Dale Junior, every amateur street racer driving a ZO6 with all the talent of an XBox twiddler— are the problem. For that reason alone, the fully automated car will happen.

Two thousand years, your horse was just as much a mark of wealth, virility and personal skill as a 911 Turbo or WRX is today. Millions of Saracens, Conquistadors, cavalrymen and cowboys would have told you that you were full of manure to suggest that one day, nobody but jockeys and hobbyists would ride a horse. I think it was Ferdinand Porsche who said that the last horse on earth will be a racehorse, and that the last car will be a racecar."

 

This is definately an interesting thread. What I haven't seen addressed is whether different MINI's might react differently under DSC. From my own experience, I haven't found it to be a problem at all. I don't race, so this is totally street driving. Yes, I have seen the yellow light flicker on occasion but whatever it is doing hasn't intruded on my driving at all.

Even darting out into heavy traffic (which I have impatiently done on occasion, It would be more prudent to wait for a bigger break in traffic) I have not had the engine cut on me. In fact, I'm much more worried about the cold start stumble in that regard since I'm only 200 feet from a busy 4 laner and usually need to enter the flow prior to engine warm-up. I've found a couple mild revs will get me past the stumble point.

So what gives? Are the OBC's that variable between cars or am I just a smoother starter outer than some others?

 

I've had a Mini since late 02. It was a big issue at first, and I really liked the changes to the traction control that MTH brang. But to be totally honest, I don't really kick on the traction control or the DSC nearly as much as I used to, even though I have a lot more HP to play with. The system are totally silent now, execpt for very, very rare occurances (starting from a full stop to turn onto a street with water around, or when I hammer it through tight turns with more power than the inner front can use).

I've learned to happily co-exist with the sytems. In fact, as many have said, if you are smooth, the DSC will let you get on the hairy edge of drifting!

Matt

 

Eww, the hairy edge - begets a close shave

 

That's why I like the OPTION of having it on or off.

 

My MCS only has ASC so I can't comment on the Mini's DSC performance. I always turn my ASC off because I find it totally useless and an actual danger if trying to accelerate into traffic.

I have, however, tested the systems in a Chevy Tahoe, Cad Escalade, and Cad DTS. I was on a dirt road and literally tried to make the Tahoe get sideways at 30 mph. I knew I would be able to recover if the system didn't work, but it worked flawlessly and brought the SUV right back in line. I was impressed.

My dad is a retired GM engineer who did ice testing at an airport with Corvettes. He tried as hard as he could to make a Corvette spin out on ice but could not do it with the system on. He told me he did everything wrong but it still would not spin.

I have avoided a deer in my Mini at 80 mph and I am confident that without my racing experience, I would have been backward in the median but luckily was able to keep it on the road (with snow tires!). An untrained driver would have been saved by DSC in that situation.

Perfect world: I would like the option of turning the traction control part off but leaving the "yaw" portion on to save from unexpected accident avoidance.

 

That's where I'm at also.

 

Spit,

You do realize that ASC does not have the "yaw" correct feature.

 

Yeah ASC is traction control only. What I meant was that in the deer incident I had, DSC would have saved an untrained driver while I was luckily able to avoid spinning the car even though I only have ASC.

Perfect world to me would be DSC with no traction control....then I would have the safety of accident avoidance (yaw control) but no bad effects from the traction control.

By the way, the GM high performance stuff such as Corvettes and Cad CTS-Vs have different modes for their DSC...there are normal, sport, and competitive modes...or something to that effect.

 

Ahhh... remove torque? Thats cutting power, I thought we're talking about applying the brake? I've no idea what the incremental sensitivity of the ABS pump is, but I'd imagine that if the rolling inertia changes or if the bite of the pad changes, the timeliness of the systems ability to react changes. Obviously in the end its about is the wheel spinning or not, the question is how quickly the system can react and correct for it and how a number of variables changing will effect the system.

As far as I know these systems aren't yet autonomic in nature. In fact, my understanding is that most e-throttle cars don't allow for left foot braking as there simply isn't enough computing power to square all the variables against eachother in that circumstance. The issue isn't the milliseconds that events take, its how sophisticated you intend to be in the analysis and hueristics you apply to the data. Engine mangement is all table lookup, thats fast, but its not very flexible. Change an injector flow rate and its back to the dyno to get the table right again.

I earn my living building distributed computing frameworks. Three years ago, the system I worked on (since moved onto other things) was adopted by a famous Italian automotive electronics firm to form the basis of a system to supply real time data back to a bank of computers that were analysing the telemetry which came off what went on to become the championship winning F1 cars for the past two years. When this team won the chamionship they gave a substantial piece of the credit to this system. This system was not controlling the car mind you, simply analysng the data in real time. A few members of our team consulted on the system, so I was fortunate enough to get a guarded glimpse as to what was going on. I can assure you that to simply to analyse the data in real time required astonishing amounts of computing power.

 

it's progressive, first trying to improve things by using single wheel braking to get weight transfer to help. If that's not enough, then the system cuts power. I think it does this quickly by retarding timing, and if it really has to hammer the system, then it moves the throttle body.

It's really a bummer that all we really have to go by is some superficial technical descriptions and marketing liturature. It would be nice to know actually what the current systems are doing and what compramises are built into them in the name of simplicity and cost savings. It would be a book for sure, but it would be interesting.

Matt