Haldex Sport controller = increased wear & tear?

[ZPrime]

Guys, don't forget about brake bias! Engine braking with the rears engaged is like increasing the braking power in the rear, which makes your braking less effective! You want the front and the rear to have the same braking potential, which means weight transfer, rotor diameter, pad composition, tire size and power distribution have to be considered.

http://stoptech.com/technical-suppor...alance-matters

Engine braking with the rears engaged just means that some of the available stopping friction is already being used by the engine's drag. It means that the brakes + rotors should actually have less work to do since the drag from no throttle is being evenly distributed to all wheels vs. just the front. Remember that the fronts also have to give you friction to turn, so if you can take some of the load off of them and put it in the back that will help.

Also, note that our cars have automatic brake proportioning in the ABS unit (or so several sources have said on this site). The bias should remain relatively the same even if you do change brake components.

[Mr_Peach]

And apparently not just me, but allot of folks who have the gen 2 controllers have been smoking something when we perceive the car pushing more from the rear as we accelerate through / out of corners, or thinking we can do better donuts in a snowy parking lot, etc etc???????

I still believe, Jake (cough cough hack hack you-take-it!)

I feel it, but I feel it when I'm pushing hard enough that the fronts are breaking traction.

A lot of folks associate 'breaking traction' with 'burning rubber' or 'sliding sideways', but in terms of AWD onset (or ESP onset) it happens at a lower, more subtle, level.

The 'engine braking' thing (or not) is harder for me. I thought the drivetrain felt more solid on 'lift'. Less lash and all that.

[ZPrime]

The 'engine braking' thing (or not) is harder for me. I thought the drivetrain felt more solid on 'lift'. Less lash and all that.

It's entirely possible that the rear may still stay engaged for some period after throttle LIFT (i.e. coasting). What Keir said above is that it immediately disengages upon brake application, regardless of switch setting. Many (most?) of us were of the understanding that our controller with the switch in race was identical to the Gen1 Competition ("orange") controller, which keeps the haldex clutch engaged even when the brakes are pressed.

Now, what the behavior of the rear is when coasting in race vs. sport may be different - won't know until (if?) it is logged.

I would love to go mess around with it today but it's freaking 90F out (in spring, in Cleveland!?) and I know my little vagcom PC won't be happy with it... I may log something on the ride home anyway though. The hard part is going to be logging everything I want - there's no way to log data from multiple controllers simultaneously, which I'd need if I want throttle position and brake status I believe.

There's also no way to check the status of the switch (stock/race/sport) so I need some way I can note that in the data, unless I start a video recording and try to match the timecodes... Pain in the ass.

[motrrrpsycho]

I still believe, Jake (cough cough hack hack you-take-it!)

LOL!! Woooww maaan, this haldex thing is faaaaaar ouuuuuut

[Fast1one]

Engine braking with the rears engaged just means that some of the available stopping friction is already being used by the engine's drag. It means that the brakes + rotors should actually have less work to do since the drag from no throttle is being evenly distributed to all wheels vs. just the front. Remember that the fronts also have to give you friction to turn, so if you can take some of the load off of them and put it in the back that will help.

Also, note that our cars have automatic brake proportioning in the ABS unit (or so several sources have said on this site). The bias should remain relatively the same even if you do change brake components.

I don't think that's how ABS works. Doesn't abs remove braking power from a slipping wheel?

For example, imagine that the car is heavily biased towards the rear. Under heavy braking, the rear wheels naturally would begin to slip and the ABS system would reduce braking torque in the rear to stop it from slipping, but you could still have a lot of potential up front that is not being utilized! Let's put down some numbers down. Ill use the numbers in the linked article.

Under 1G of decceleration, a vehicle weighs in at 1800 pounds on the front axle and 600 pounds on the rear axle due to the weight transfer. As a result, If we assume that all the tires have the same coefficient of friction, the rear will require less braking torque to lock the wheels when compared to the front. How much? 50 percent less. So of it takes 700 foot pounds to lock up the front under 1g of decceleration, then it only takes 350 foot pounds to lock up the rear.

Now let's assume that the brake bias is 50/50 split, highly exagerated yes. The rears lock up first, so you are only applying 350 foot pounds of torque to the front, which has the potential of 700 foot pounds before the wheels lock up.

What If you just SLAM on your brakes until all four corners are applying 700 foot pounds of torque, the rear begins to slip and the ABS takes care of it right? Maybe, butI wouldn't want to be riding on an ABS system for any length of time plus the response time of the ABS system could hurt the braking performance anyway. Any bit helps.

What about front bias? Most cars from the factory are set with just a little bit of front brake bias. It makes the car more stable and typically makes ABS work less compared to modulating four wheels instead of two. In general, relying on ABS is bad practice under racing situations. Simply changing the pad composition can help bring everything back into equilibrium when a variable is changed. Tire pressures can also help, as well as clever static weight distribution and many other factors. But that also affects suspension tuning, which is another topic all together

[LWNY]

I think there is an misunderstanding of haldex being engaged so that you can let engine braking slow down all 4 wheels vs haldex unit still engaged when the brakes are applied. Using engine braking is useful to prevent an independent wheel slip when you want to do some deceleration, but once you hit the brakes, you don't want haldex unit to stay engaged. If one wheel locks up, you will end up with a spinning opposite wheel, thus when your inside wheel locks, your rear inside wheel will be acting like a torque vectoring differential, putting all its power on the outside wheel and spinning you out. And if you throw ABS into the picture, every time one wheel locks up, ABS will modulate that brake, making the opposite wheel go up and down in speed, making the car extremely unstable.

So haldex stay engaged while throttle release is good for engine braking

Haldex stay engaged when brakes are applied is asking for trouble.

[Fast1one]

I think there is an misunderstanding of haldex being engaged so that you can let engine braking slow down all 4 wheels vs haldex unit still engaged when the brakes are applied. Using engine braking is useful to prevent an independent wheel slip when you want to do some deceleration, but once you hit the brakes, you don't want haldex unit to stay engaged. If one wheel locks up, you will end up with a spinning opposite wheel, thus when your inside wheel locks, your rear inside wheel will be acting like a torque vectoring differential, putting all its power on the outside wheel and spinning you out. And if you throw ABS into the picture, every time one wheel locks up, ABS will modulate that brake, making the opposite wheel go up and down in speed, making the car extremely unstable.

So haldex stay engaged while throttle release is good for engine braking

Haldex stay engaged when brakes are applied is asking for trouble.

Those are my thoughts as well. Under very heavy braking conditions, beyond the limitations of the tires, the ABS system and Haldex system will fight each other.

On top of that, the ABS system is a control system and as such will have a response time and settling time. If the system is under damped, the braking modulation will overshoot and undershoot trying to find the happy medium where the wheel is not locking up or under performing. If its critically damped, you still have the response time to consider. Over damped and the car will not slow down as fast as it possibly could.

Saying that the ABS system will take care of brake bias is a poor band aid for a system that will not perform to its full potential. Just because it's "close" doesn't make it right. Let the brake design dictate the brake bias, not the all wheel drive system or the ABS system

That being said, I would never take the ABS off of any car. It's a good safety net, even for race cars

[ZPrime]

The ABS and Haldex won't "fight." Haldex has communication from ABS and traction control, it's not blind and stupid like a Torsen mechanical diff.

Regarding brake bias, I'm just parroting what I've seen from Stoptech and other brake manufacturers. They all claim it's not a concern to upsize rotors and calipers on the front of the car without doing anything to the rear, and they attribute this to the braking system automatically proportioning as needed. I am not sure how they think this happens though.

As far as leaving the rear engaged under braking, HPA says it all for me:

Developed by racers for racers; the Competition Controller, now in a stealthly look, identified with an engraved "Competition" marking, transforms the Haldex equipped 4-Motion platform into a weapon on the track. Similar in principle to the standard performance Haldex upgrade, the Competition Controller features one MAJOR difference.

Back in the ’80s the Audi Quattros dominated the race track with their unbelievable ability to brake late when entering the corners. This ability was a result of the rear axle staying engaged under full braking allowing for a combined mechanical and frictional stopping force to be applied across all 4 wheels. The Competition Haldex Controller offered from HPA mimics this by keeping the rear axle fully engaged while under braking. Where the stock Haldex and HPP units release the rear axle into a free wheel situation as the brakes are applied, the Competition unit keeps the rear axle engaged, generating additional mechanical force to assist in slow down and maintain dynamic chassis balance.

The Competition Controller is a must for modified vehicles with high torque outputs. The experienced driver will benefit from the additional chassis control afforded by the equal deceleration rates of the front and rear axle. This allows the driver’s input to dictate the vehicle dynamics under braking, corner entry and exit rather than falling victim to the abrupt disengagement of the rear axle. This added stability and control translates directly to faster lap times; making the Competition Controller a must for the competitive driver.

Originally it was implied (maybe even stated explicitly?) that our Gen2 unit + switch in "race" was equivalent to the G1 Competition ("orange") controller, including the engaged rear whilst braking.

The default behavior for the Haldex (in "sport") would still disengage the rear, which is great for the street or people who don't know car control.

The whole point of the switch + Race was to give us further control over the rear. I agree with HPA - I'd rather have everything stay engaged (for balance reasons) than have the Haldex deciding that it needs to open the rear clutch. I also autocross and run the occasional track "Fun day" though, so I am not everyone.

[Fast1one]

The ABS and Haldex won't "fight." Haldex has communication from ABS and traction control, it's not blind and stupid like a Torsen mechanical diff.

Regarding brake bias, I'm just parroting what I've seen from Stoptech and other brake manufacturers. They all claim it's not a concern to upsize rotors and calipers on the front of the car without doing anything to the rear, and they attribute this to the braking system automatically proportioning as needed. I am not sure how they think this happens though.

Fair enough. I didn't know the Haldex had an ABS signal. Link to information? Sounds neat!

In regards to brake bias, smells like marketing bull to me. Of course it doesn't affect brake bias, they wouldn't sell BBKs otherwise!

As far as leaving the rear engaged under braking, HPA says it all for me:


Originally it was implied (maybe even stated explicitly?) that our Gen2 unit + switch in "race" was equivalent to the G1 Competition ("orange") controller, including the engaged rear whilst braking.

The default behavior for the Haldex (in "sport") would still disengage the rear, which is great for the street or people who don't know car control.

The whole point of the switch + Race was to give us further control over the rear. I agree with HPA - I'd rather have everything stay engaged (for balance reasons) than have the Haldex deciding that it needs to open the rear clutch. I also autocross and run the occasional track "Fun day" though, so I am not everyone.

Not really sure if this is a fair comparison, again sounds like marketing bull to me. If the car is not very fast to begin with (R32) AND you stick R comps on it, it's very likely that you can stand on the brakes and not have it lock up. If this is the case, you want all of the drag you can get and as evenly distributed as possible assuming the braking system is sized accordingly to perform in a balanced manner.

For regular folks like us running street tires, I'm not sure if adding rear bias with then AWD system would help. In either case, if the distribution remains constant under heavy braking, it definitely wouldn't hurt things. But I still think the brake bias should be adjusted to compensate for the change in the torque distribution. Or maybe it could turn out to be perfectly biased, considering the R is front biased from the factory.

Wasn't this about wear?

[motrrrpsycho]

Originally it was implied (maybe even stated explicitly?) that our Gen2 unit + switch in "race" was equivalent to the G1 Competition ("orange") controller, including the engaged rear whilst braking

Yes, I can't speak for everyone / other sources of info, but this was explicitly stated to me when I called HPA and asked very specific questions about this.

The whole point of the switch + Race was to give us further control over the rear. I agree with HPA - I'd rather have everything stay engaged (for balance reasons) than have the Haldex deciding that it needs to open the rear clutch. I also autocross and run the occasional track "Fun day" though, so I am not everyone.

Exactly - that was the major selling point for me, and it seems I'm not the only one who was under this impression, so I'm hoping HPA comes back to this thread and says "oops - our bad, gen2 race mode IS actually equivalent to gen 1 / gen4 competition mode (with regard to rear axel engagement under braking) - we retract the earlier statement in this thread" -OR- "we regret to inform the community that we have, in the past, incorrectly stated that the gen2 race mode is equivalent to gen 1 / gen4 competition mode (with regard to rear axel engagement under braking), when in fact, this is not the case".

HPA - please set the record strait

[LWNY]

The ABS and Haldex won't "fight." Haldex has communication from ABS and traction control, it's not blind and stupid like a Torsen mechanical diff.

This is likely not the case. Haldex might know when braking is applied and could be programmed to then disengage, but the don't talk to each other, braking into a turn, they can't work in conjunction with each other to modulate haldex clutch and ABS.

[Mr_Peach]

About the dyno, is the front and rear rollers connected to each other? Like with a belt or something? Never dyno a haldexcar on a 4wd dyno where the front and rear rollers are separate units, it will kill your haldex.
.

This is something I did not know, that there are 4wd dynos with belt linkage. Makes sense.

[motrrrpsycho]

^ ditto - now I'm wondering what HPA used to produce the afore mentioned plots. If one where they are not linked will kill Haldex, then presumably they ran on one that was linked. If so, and the front wheels actually do NOT need to be at max slip (relative to rears) to send max power to rear axles, then maybe....... maybe we DO feel more power pushing from the rear?????? (TWSS ) for realz tho - would be nice to get the strait facts here and not have to depend on butt dynos and speculation!

[JRutter]

From HPA's site:

Superflow’s Autodyn 30 is a dual purpose dynamometer. Featuring an industry norm inertia mode, wide open throttle HP pulls are generated with unmatched repeatability. For furthering HPA’s DSG and turbo development, the Autodyn 30’s Eddy Current Brake assembly will facilitate long steady-state durability runs allowing our engineers and product team to test any user condition at the flick of a switch. Superflow stands at the top of its industry with both rollers connected through a drive shaft to avoid any ESP and ABS sync issues commonly found when testing today’s technical AWD platforms.

From Superflow's site:

SuperFlow's new AutoDyn 30 all-wheel drive chassis dyno model lets you test four-wheel drive, all-wheel drive and models equipped with modern anti-lock braking and traction-control systems that insist that all four wheels are turning at similar speeds. The standard, single-power-absorber model is perfect for most passenger vehicles including modified, high performance vehicles.

SuperFlow's design includes trunnion-mounted differentials that allow SuperFlow to measure torque at each axle, individually, so you can evaluate center differential torque bias.

The AutoDyn 30 AWD chassis dyno standard configuration accommodates wheelbases from 92.5" (235 cm) to 130" (330 cm), which covers most cars and light-duty trucks. If needed a longer wheel base is available.

All SuperFlow chassis dynamometers are calibrated at the factory to correct for internal windage and parasitic losses using an AC motor and sophisticated electronics for precise measurement and superior repeatability.

[Sales@HPAmotorsports]

Thanks John.

Not sure where the speculation lies, we've answered this question many times. We provide dyno graphs showing you exactly why you would want to get the controller.

In terms of wear and tear, we've never had any reason to believe it would increase wear on the Haldex unit itself, or the rest of the drivetrain.

Haldex racing does not make units for rear ends that would have an issue if such a unit was installed.

[motrrrpsycho]

JRutter - thanks for the dyno info! Sounds like a very cool dyno.

HPA - thanks answering on the OP's question / thread topic! Regarding my speculation comment - what I meant is that the topic of how Haldex transfers torque is till confusing (at least to me) so would be nice to understand it better and not have to speculate as to how to interpret the dyno plot.

When I first posted the dyno plot, I was assuming that it basically "proved" that under normal acceleration load (not just in loss of front wheel traction, but "normal" meaning all 4 wheels have equal traction) that X% power is ALWAYS going to the rear axles via Haldex, and that the Gen 2 controller allowed you to shift the amount of torque that is ALWAYS going to the rear so that more goes to the rear. When comments were posted about dyno plots not necessarily reflecting real driving conditions, I then assumed that the front and rear drums were not connected and that this would make Haldex think the fronts are slipping and would cause max power to be routed to the rear, thus the controller only changing what the max power transfer was during front slip conditions.

So... there seems to be allot of speculation around what Haldex does / does not do, what the term "slip" actually means / how it applies, how representative (or not) the dyno plots are of how torque is applied to the pavement and so forth. Or to ask using a hypothetical example scenario:

In a 500 yard/meter strait line acceleration test starting from 0 MPH with all 4 wheels having equal traction (dry, warm, clean asphalt, good tires on the car so really good traction), using 80% throttle will the % of torque sent to the rear be roughly equal to / represented by the dyno plots through the entire distance / duration of the acceleration?

[petef]

When I first posted the dyno plot, I was assuming that it basically "proved" that under normal acceleration load (not just in loss of front wheel traction, but "normal" meaning all 4 wheels have equal traction) that X% power is ALWAYS going to the rear axles via Haldex, and that the Gen 2 controller allowed you to shift the amount of torque that is ALWAYS going to the rear so that more goes to the rear. When comments were posted about dyno plots not necessarily reflecting real driving conditions, I then assumed that the front and rear drums were not connected and that this would make Haldex think the fronts are slipping and would cause max power to be routed to the rear, thus the controller only changing what the max power transfer was during front slip conditions.

You may be confused - no worries.

First possible confusion - under acceleration all four wheels do not have equal traction. During acceleration weight shifts to the rear which unloads the fronts and causes them to lose traction. They begin to slip waaay before all the available engine torque is transferred to the pavement. Meanwhile the rears gain traction because of the rearward weight transfer.

Second confusion - the Haldex only sends power to the rear tires when the fronts start to slip/spin. Otherwise the split is 90/10 by default. As long as you drive gently so that you don't cause the fronts to slip you will never engage the Haldex whether its in Race or Normal.

Third confusion - during a dyno pull (esp with an FI motor) the front tires will be overloaded (slipping) during most of the pull so there will be rear power transfer during most of the pull. If the engine is strong enough you'll easily hit "maximum" rear transfer as shown in the dyno traces. No dyno tricks needed.

Pulling it all together - having the Haldex in Race allows more power to go to the rear wheels when the fronts lose traction. This is beneficial in some situations but not all. Race mode can put enough power to the rears in corners that they break loose and cause power oversteer (generally a good thing if you know what you're doing).

VW knows most 'Murican drivers can't deal with oversteer so they kept us safe with an 80/20 max split for our own good. Or more likely to avoid lawsuits.

[LWNY]

You may be confused - no worries.

First possible confusion - under acceleration all four wheels do not have equal traction. During acceleration weight shifts to the rear which unloads the fronts and causes them to lose traction. They begin to slip waaay before all the available engine torque is transferred to the pavement. Meanwhile the rears gain traction because of the rearward weight transfer.

Second confusion - the Haldex only sends power to the rear tires when the fronts start to slip/spin. Otherwise the split is 90/10 by default. As long as you drive gently so that you don't cause the fronts to slip you will never engage the Haldex whether its in Race or Normal.

Third confusion - during a dyno pull (esp with an FI motor) the front tires will be overloaded (slipping) during most of the pull so there will be rear power transfer during most of the pull. If the engine is strong enough you'll easily hit "maximum" rear transfer as shown in the dyno traces. No dyno tricks needed.

1. The OP seems to imply that he wants equal traction meaning all on dry surface, not under loss of traction condition.

2. Haldex can send power to the rear axle w/o the front wheels spinning. The controller receives information from steering wheel angle, accelerator depression rate to determine whether the likelyhood of potential front wheel traction loss and pre-engages the clutch before there is actual slippage. That is why the preferred way of driving the haldex AWD unit is not to get off the brake and slam the accelerator right away, since doing that does not give the haldex unit enough time anticipate the potential wheel slip situation. Getting on the throttle moderately and then going full throttle a short while later will usually result in power laid out in all wheels.

3. From the dyno graph, One can assume the max clutch pressure determines the how much power goes to the rear axle, with its peak power transfer of 170 power units, which is indicates some slippage given the front axle still delivers more power. This will also be different depending on the gear the dyno run was on since the torque output from the transmission would be different in each gear.

In a 500 yard/meter strait line acceleration test starting from 0 MPH with all 4 wheels having equal traction (dry, warm, clean asphalt, good tires on the car so really good traction), using 80% throttle will the % of torque sent to the rear be roughly equal to / represented by the dyno plots through the entire distance / duration of the acceleration?

This should be calculatable, they could just take the reading off the oil pressure reading off the haldex unit to determine how much of the torque is transferred to the rear axle, integrate it over the true torque output based on RPM/gear to determine the torque transfer cutoff point to the rear axle. But with a straight line acceleration where you don't loose traction won't be that interesting. If those data are collected over a road course, you can determine when AWD power are laid down and when it is not.

[ZPrime]

2. Haldex can send power to the rear axle w/o the front wheels spinning. The controller receives information from steering wheel angle, accelerator depression rate to determine whether the likelyhood of potential front wheel traction loss and pre-engages the clutch before there is actual slippage.

Not on Gen 1 and Gen2. The hydraulic pressure to engage the clutch comes from the speed differential between the front and rear axles. Granted, the Haldex clutch can engage in "as little as 1/8th of a tire rotation" according to the VW documentation, but there is absolutely no way to "pre-engage" the Haldex on a Mk4 or Mk5 car.

The GoRf (Mk6) is on Haldex v4, which uses an electric pump for hydraulics and it can pre-engage.

That is why the preferred way of driving the haldex AWD unit is not to get off the brake and slam the accelerator right away, since doing that does not give the haldex unit enough time anticipate the potential wheel slip situation. Getting on the throttle moderately and then going full throttle a short while later will usually result in power laid out in all wheels.

This is generally the best way to go fast on any car. Smooth = fast. You can drive an Evo and a GT-R like a pillock because the computers are more sophisticated (they have far more information than the Haldex does) and because the couplings are more advanced (electronically engaged clutches, etc).

For whoever it was that asked for more info on the signals Haldex can see - I've found two different documents. VW SSP 206 is Mk4 specific, but the ideas are similar. VW SSP 333 is for a "2004" Mk5 chassis (it's from Germany and they got the Mk5 around then), so I'm assuming this is the Gen2 coupling. It's also covers the Transporter; interesting to see that the Transporter's version is similar but different from ours.

[LWNY]

Not on Gen 1 and Gen2. The hydraulic pressure to engage the clutch comes from the speed differential between the front and rear axles. Granted, the Haldex clutch can engage in "as little as 1/8th of a tire rotation" according to the VW documentation, but there is absolutely no way to "pre-engage" the Haldex on a Mk4 or Mk5 car.

The GoRf (Mk6) is on Haldex v4, which uses an electric pump for hydraulics and it can pre-engage.

There is always a speed differential between the front and rear wheels, given most of the weight are in the front and it compresses the tires more thus making its rolling diameter slightly smaller. Plus, one is never driving in a straight line forever, there are always turns, however small, where the rear wheels don't follow the front exactly, causing more rotational differences. Thus, when in a turn, it is where the greatest rotational difference between the front and rear without any wheel losing grip, the haldex pump has generated more than enough pressure to pre-engage the clutch if it deems required (based on accelerator position, amount of G force encountered, steerinng wheel position, etc).

This is generally the best way to go fast on any car. Smooth = fast. You can drive an Evo and a GT-R like a pillock because the computers are more sophisticated (they have far more information than the Haldex does) and because the couplings are more advanced (electronically engaged clutches, etc).

On a torsen based car, you can go from full brake to full throttle since there will be no slippage during that transition, but on the haldex based unit, if you go from full trail baking to full throttle as soon as you hit the apex, you will go into serious understeer, especially in low grip condition, and even if the haldex engages, it usually can't undo the understeer because the coefficient of friction on a sliding tire is lower than a tire in grip, thus the front tires will keep on sliding. But if you just give that slight throttle for a split second before jamming the throttle (no need for smooth), the haldex unit already determined you are in a situation where you power should be spread to both axles and you will not encounter wheelslip at the front, and if there is wheelslip, it should be on all 4 wheels and you will be going sideways.

[petef]

There is always a speed differential between the front and rear wheels, given most of the weight are in the front and it compresses the tires more thus making its rolling diameter slightly smaller.

Can you point us at info that supports this?

Plus, one is never driving in a straight line forever, there are always turns, however small, where the rear wheels don't follow the front exactly, causing more rotational differences. Thus, when in a turn, it is where the greatest rotational difference between the front and rear without any wheel losing grip, the haldex pump has generated more than enough pressure to pre-engage the clutch if it deems required (based on accelerator position, amount of G force encountered, steerinng wheel position, etc).

This doesn't change anything Zprime or myself said. The Gen 2 Haldex cannot pre-engage, mate. It has to encounter front wheel slip before anything happens. No slip = no Haldex action.

[Sales@HPAmotorsports]

Can you point us at info that supports this?



This doesn't change anything Zprime or myself said. The Gen 2 Haldex cannot pre-engage, mate. It has to encounter front wheel slip before anything happens. No slip = no Haldex action.

Stock, with no wheel slip you still get power to the rear wheels. Its simply biased heavily to the front wheels providing the power.

As can be seen by this dyno graph ( the wheels do not slip on the dyno, and its linked it does not think its slipping )

[AmRando]

1. The OP seems to imply that he wants equal traction meaning all on dry surface, not under loss of traction condition.

Yes, this is exactly what I was asking. However, it appears from all the debate and Keir's commentary that this is NOT the case. The 'Race' Haldex does not turn the .:R into a *RWD-biased car*in everyday conditions (well, 50/50 anyways), which is specifically the answer I was hoping for and was also the wear & tear issue I would expect if this were the case, ie: 40k down the road having run with that heavy rear-torque bias.

As per Keir's and others' answers, all the Haldex change does is increase the speed, frequency and 'volume' of torque that is applied in conditions when warranted. This is obviously better than stock, but not quite what I had hoped for.

[petef]

Stock, with no wheel slip you still get power to the rear wheels. Its simply biased heavily to the front wheels providing the power.

As can be seen by this dyno graph ( the wheels do not slip on the dyno, and its linked it does not think its slipping )

I stand corrected (don't you hate it when that happens?) Two missing data points (no wheel slip on dyno and linked rollers) make all the difference, yes?

This also answers the OP's question pretty clearly. If you run in Race you're putting more torque through the rear diff on a continuous basis. This will increase the wear on the rear diff and possibly the Haldex (not sayin' its a bad thing, just trying to be accurate).

Part 2 - F/R slip is required to move the Haldex off the default splits shown on the dyno, ie, more to the rear than the nominal split (90/10 or whatever). Without slip there's no change in the nominal split, correct?

Part 3 - what is max transfer possible front to rear - 50/50? Or could we momentarily see 20/80 in some situations?

Much appreciated - thanks.

[AmRando]

This also answers the OP's question pretty clearly. If you run in Race you're putting more torque through the rear diff on a continuous basis. This will increase the wear on the rear diff and possibly the Haldex (not sayin' its a bad thing, just trying to be accurate).

Part 2 - F/R slip is required to move the Haldex off the default splits shown on the dyno, ie, more to the rear than the nominal split (90/10 or whatever). Without slip there's no change in the nominal split, correct?

Isn't this contradictory? If you're running in Race mode you're NOT 'putting more torque through the rear diff on a continuous basis', I'm pretty sure that's what everyone has said? The drivetrain will continue to use the default bias UNTIL it encounters wheelslip, and then it will pump up to a 50/50 bias. So in everyday, mundane driving it won't be creating additional wear on the drivetrain components constantly while driving- that was really my concern. If in wheelslip conditions it's doing so, well so be it- that's only a small percentage of average use anyways.


Really what I wanted to know was two-fold:

Whether the controller changes the handling of the car in average real-world use (possible oversteer at the limits, obviously) and whether this change would induce significantly more wear on the rear drivetrain. And if I read the thread correctly, it sounds like no on both counts.

[Sales@HPAmotorsports]

Isn't this contradictory? If you're running in Race mode you're NOT 'putting more torque through the rear diff on a continuous basis', I'm pretty sure that's what everyone has said? The drivetrain will continue to use the default bias UNTIL it encounters wheelslip, and then it will pump up to a 50/50 bias. So in everyday, mundane driving it won't be creating additional wear on the drivetrain components constantly while driving- that was really my concern. If in wheelslip conditions it's doing so, well so be it- that's only a small percentage of average use anyways.


Really what I wanted to know was two-fold:

Whether the controller changes the handling of the car in average real-world use (possible oversteer at the limits, obviously) and whether this change would induce significantly more wear on the rear drivetrain. And if I read the thread correctly, it sounds like no on both counts.

Yes this changes your day to day impact with the Haldex. As quoted from our site as well:

" The advantages of this HPP upgrade are it’s ability to predict the onset of torque. This upgrade takes into account the TPS signal more so than the OEM software and therefore can proactively begin applying power to the RWD clutch packs before wheel spin begins."

Either controller in Race ( Competition or switchable ) will apply more torque to the rear axles in all situations, dry, wet, snowy, balmy, windy, even in a tornado you've got more to the rear axle. If your foot is to the floor, the most amount of torque available ( IE whatever is not being used by the front axles ) will be sent.

Should be no perceivable drivetrain wear. Its hard to say this as we'd have to have someone drive their car stock, while driving a completely separate clone with the Race controller to give a back to back comparison. However Hadlex does refuse to make controllers for rear ends that cannot take any more power then the OEM has allowed for in other makes/models.

[petef]

Isn't this contradictory? If you're running in Race mode you're NOT 'putting more torque through the rear diff on a continuous basis', I'm pretty sure that's what everyone has said? The drivetrain will continue to use the default bias UNTIL it encounters wheelslip, and then it will pump up to a 50/50 bias. So in everyday, mundane driving it won't be creating additional wear on the drivetrain components constantly while driving- that was really my concern. If in wheelslip conditions it's doing so, well so be it- that's only a small percentage of average use anyways.

What may trip people up (including me ) is there are two parts to the Haldex behavior and we tend to mix them up or only focus on one:

1) What happens when all wheels have full traction vs. 2) what happens once the fronts lose traction.

Kier is saying that the 'full traction state' always sends power to the rear - the switch changes the F/R ratio (and some dynamic behavior). The dyno charts show this behavior, not the dynamic aspect.

With front tire slip a dynamic transfer occurs - stock might roll at 90/10 around town but if you hit a slick spot you might get as much as 50/50 transferred briefly and then once the fronts grip it goes back to 90/10. This is what I was focused on.

So - daily wear and tear in the 'full traction state' is increased by the Sport and Race settings over stock because the Haldex is always putting more torque to the rear no matter what.

Dynamic F/R transfer is caused by front wheel slip so the switch settings don't affect it (much - Kier says the transfer is a bit more aggressive).

Not contradictory so much as two halves of the same coin?

Looking back at this thread I see half of us were talking 'full traction' behavior while the other half were talking 'dynamic transfer'. Dang.

[Saabstory02]

Petef: To answer your "Part 3" question above:

Yes/no. Haldex will send up to 100% of the available torque to the rear; but you have to remember we are talking torque here. As was described beautifully earlier in the thread: if you have 0 traction on a wheel then you have 0 torque. So; if both front wheels are slipping heavily and have no traction, all available torque will be routed to the rear to keep forward momentum possible. However; in routine, even track driving, I don't think you could "realistically" see much beyond a 50/50 power scenario.

I suppose you could get momentary points where the rear is getting more like 75% of available torque; but that would be because the front wheels have near 0 traction and you are likely sliding in an understeer nosedive anyway

[AmRando]

What may trip people up (including me ) is there are two parts to the Haldex behavior and we tend to mix them up or only focus on one:

1) What happens when all wheels have full traction vs. 2) what happens once the fronts lose traction.

Kier is saying that the 'full traction state' always sends power to the rear - the switch changes the F/R ratio (and some dynamic behavior). The dyno charts show this behavior, not the dynamic aspect.

But, no- he isn't.

As quoted from our site as well:

" The advantages of this HPP upgrade are it’s ability to predict the onset of torque. This upgrade takes into account the TPS signal more so than the OEM software and therefore can proactively begin applying power to the RWD clutch packs before wheel spin begins."

Either controller in Race ( Competition or switchable ) will apply more torque to the rear axles in all situations, dry, wet, snowy, balmy, windy, even in a tornado you've got more to the rear axle. If your foot is to the floor, the most amount of torque available ( IE whatever is not being used by the front axles ) will be sent.

These conditions appear to be all based on the TPS signal, not the traction state.

He's not talking about 'full traction state', because as most here seem to have agreed real-world conditions are always changing. Keir's contention that 'Yes this changes your day to day impact with the Haldex' is a Marketing-speak-bias -50/50? because YES, the controller will kick in more aggressively more often, but it isn't ON BY DEFAULT at a 50/50 bias at a steady speed on dry pavement (.. or is it? Noone seems to have literally answered this). During the dyno pull, this is an acceleration state generating uneven traction- OF COURSE the race controller is active. But if you ran that dyno test with the cruise control on, what happens when the system reaches a steady speed with stable traction conditions? Does it continue to distribute 50% power(my correction) to the rear indefinitely, or does it reapply this to the front wheels once that state is reached?

I think the misunderestimation here is what factors the race controller accounts for and whether it really has an 'off' state at all - all FWD torque under ideal conditions as per the stock controller.

-TPS signal - throttle state (the onset of torque)
-wheel slip
-???

[LWNY]

Yes, to Saabstory02's quote. Too many ppl get torque and power mixed up with each other. 50:50 power distribution is not the same as 50:50 torque distribution. I try not to mix the two up, but when jumping back and forth between the two on discussion of torque transfer and power transfer, it is easy to assume it is the same discussion.

Can you point us at info that supports this?

It is quoted in the PDF doc above. It is also a common knowledge of how wheel rotation works on a car.

This doesn't change anything Zprime or myself said. The Gen 2 Haldex cannot pre-engage, mate. It has to encounter front wheel slip before anything happens. No slip = no Haldex action.

It is quoted in the HPA post:


" The advantages of this HPP upgrade are it’s ability to predict the onset of torque. This upgrade takes into account the TPS signal more so than the OEM software and therefore can proactively begin applying power to the RWD clutch packs before wheel spin begins."

That means, before torque is applied..thus before the onset of slip. One thing of issue is that, this quote is stated for the HPC controller. It actually is a fuction of Haldex. You do not need the HPC controller to have preemptive AWD

I think the misunderestimation here is what factors the race controller accounts for and whether it really has an 'off' state at all - all FWD torque under ideal conditions as per the stock controller.

-TPS signal - throttle state (the onset of torque)
-wheel slip
-???

As the doc indicates, haldex takes info from Accelerator pedal, steering wheel angle sensor, accelerometer sensors, wheel slip sensors and factor all of them in to determine when clutch engagement is necessarily.

And as the doc state, the pressure valve allows full pressure on the clutchpack even when pressure is not generated anymore by the differential pump. Thus if you are still driving on the twisties after recoving from a low grip situation, it can stay at full AWD mode until deemed unnecessarily. That "unnecessarily" consists of many factors, such as speed, acceleration, etc. Scenarios are all described in the docs from above.


As to the wear and tear of the haldex. The oil pressure of 1500 psi allows it to transfer around 1500 ft/lb of torque transfer. Anything beyond that, the clutch will slip and not transfer that to the rear axle and indeed will cause wear to the clutchpack. That would more likely occur on the very low gears where the torque out to the driveline is very high.

As for the drivability, it will allow for for a more tail out feeling, especially if driven hard. If the haldex clutchpack locks, the rear axle will be forced to turn at the same rate as the front axle. This will cause it, instead of taking the shorter path that the rear axle usually takes, a longer path dictated by the equal amount of rotation required, thus a wider path, hence a feeling of the rear coming out in turns...not unlike the torque vectoring differentials.

[petef]

It is quoted in the PDF doc above. It is also a common knowledge of how wheel rotation works on a car.

Thanks - will read the doc with great interest.

[petef]

But, no- he isn't.

Is it too late to apologize for stinking up your thread? I really thought I understood the Haldex but clearly...

I'll go quietly you honor...

[AmRando]

Is it too late to apologize for stinking up your thread? I really thought I understood the Haldex but clearly...

I'll go quietly you honor...

[AmRando]

That means, before torque is applied..thus before the onset of slip. One thing of issue is that, this quote is stated for the HPC controller. It actually is a fuction of Haldex. You do not need the HPC controller to have preemptive AWD

This part I understood from the start, the race controller is simply more aggressive in doing so.

As the doc indicates, haldex takes info from Accelerator pedal, steering wheel angle sensor, accelerometer sensors, wheel slip sensors and factor all of them in to determine when clutch engagement is necessarily.
And as the doc state, the pressure valve allows full pressure on the clutchpack even when pressure is not generated anymore by the differential pump. Thus if you are still driving on the twisties after recoving from a low grip situation, it can stay at full AWD mode until deemed unnecessarily. That "unnecessarily" consists of many factors, such as speed, acceleration, etc. Scenarios are all described in the docs from above.

So it will not stay in this state indefinitely.

As to the wear and tear of the haldex. The oil pressure of 1500 psi allows it to transfer around 1500 ft/lb of torque transfer. Anything beyond that, the clutch will slip and not transfer that to the rear axle and indeed will cause wear to the clutchpack. That would more likely occur on the very low gears where the torque out to the driveline is very high.

And if the controller is not in the engaged-state indefinitely, that allays my concern about wear and tear - I was afraid a full-time engagement over 30.. 50, 70,000km would cause excessive wear on the system.

As for the drivability, it will allow for for a more tail out feeling, especially if driven hard. If the haldex clutchpack locks, the rear axle will be forced to turn at the same rate as the front axle. This will cause it, instead of taking the shorter path that the rear axle usually takes, a longer path dictated by the equal amount of rotation required, thus a wider path, hence a feeling of the rear coming out in turns...not unlike the torque vectoring differentials.

And this addresses Keir's contention that 'Yes this changes your day to day impact with the Haldex.' The car will feel different, but not to the extent I had been led to believe. Thank you one and all.

[oidoglr]

It's too bad that the stock controller can't be reprogrammed to do a demo mode like APR chip tuning for 6 hours or whatever, so owners could get a fair idea of the difference in operation of this upgrade.

I, too, have been curious if the Haldex upgrade makes the R drive perceptibly different even in normal day-to-day driving, or if it is something that really needs to be appreciated at limits.