Ted - Bend, OR

Drove over the Willamette Pass after a couple of days of hard snow. Plows, graders and a screw were out in force together with skiers anxious to get at the fresh powder. Every form and flavor of sutes, 4x4 trucks and AWDs were no match for my TT. Very pleased.

180Q, Blizzak MZ-01, 205/55/R16 89Q, steel rims, plactic Audi hubcaps.

Marius B

Finally we got really slippery conditions here too, so I was able to try a few silly things:

1) When at low speed (about 30 mph), I eased it into 1st and then released the clutch pedal abruptly, so the front wheels broke traction and lost some speed. The rear wheels maintained their speed and the Haldex DID NOT engage, and I could clearly hear the Haldex hydraulic pump working (mid-frequency gnarling noise), because of the difference in speed front to rear. This indicates that we should not have to worry about rpm mismatch leading to rears breaking traction when downshifting in fast corners — there doesn’t seem to be any torque to the rear in these situations, if we take it that the same applies to higher gears.
2) When accelerating uphill in 1st, with some wheel-spin on all wheels, I applied the hand-brake: The Haldex clutch released immediately, maybe even before the brakes started working, so there was no "binding" or struggling between power and brakes. With the rear wheels arrested there was the same, clearly audible sound of the hydraulic pump, but no clutch engagement, of course, again due to computer override. Seems there’s no mechanical reason not to use the handbrake to bring the car around if needed.
3) True on-spot donuts don’t seem to work with Haldex in the snow! The best I could get was tight four-wheel drifts within a 30-40’ circle, but no real pirouettes with the axis inside the perimeter of the car. Maybe the Haldex software somehow detects what is going on and just rules against it, cutting rear power? If so, how can it detect the difference from a prolonged accelerating spin in an open curve? Anyway, this hardly contitutes a basis for complaints to Audi.
4) No attempt at braking in low-speed, very slippery curves could make the rear step out on its own. Good to know for everyday driving in extreme conditions.

I’ve had this 225 for 6 months now without opportunities to check these things before, so I’m sure there must still be others here who found this interesting. Yes? ;-))


Marius

Ted Princehorn

Upon reading your post it became self-evident that I need to study-up on the Haldex system. Thanks. (nt)

Neo, Barcelona

Marius, thanks for sharing your testings! (nt)

Graham

I have good from a respectable contact in Audi that a directive came from 'The Top' that the Haldex in the TT must allow a handbrake turn at upto 50kmh!!!

Which of cource raises two thoughts (at least)

1) has anyone tried it - other than Audi test drivers ?

2) what happens over 50kmh?

In the case of the TTq (and most likely the S3 too) it seems that whilst the Haldex is still operative under normal braking, as soon as the ABS or handbrake are applied, handbrake possibly only up to a cetain speed, it disengages.

Regards, Graham
No snow here, may try it on the beach )

Matt Harris

Marius, thanks for your very interesting post. What happens under heavy throttle on mixed traction surfaces such as snow and ice? Is there "snatching" of the drive line fore and aft or side to side as traction varies wheel to wheel? And is the Haldex hydraulic noise present?

Thanks Marius!

John S

The most the Haldex could do is lock up giving 50% torque to the rears. This system allots the rears with a proportion of the front's torque. The only way I could see doing tight donuts would be equipping the car with Saab type front wheel handbrakes.

AkRascal

can't explain the physics, but having driven on snow most of my life, lemme clue yu, this car does donuts... with wild abandon, I might add.

Marius B

No perceived "snatching" in any situation. The only sideways intelligent distribution is done by the computer applying ONE front brake if that wheel is spinning faster than the other.

No Haldex noise in any "unprovoked" driving state — the system won't let fore-and-aft speeds be sufficiently different.

Marius B

In most lock-up situations torque distribution would be around 50% — oscillating around that figure because of resonnances in driveline and body movement — but if one axle got good grip while the other was a slick patch, the one with the grip would get up to 100% of it.

For tight donuts — has been doable with Golf Syncro-like visco coupler on the rears, without handbrake at all.