How does the AWD work?

[ericem]

Just wondering, in 1st gear with a auto tranny it does 50/50 split right? Because my front wheels spin sometimes for a bit, then it kicks in the rear, whats wrong? Is it the tcu? SHould i reset it?

[Manarius]

There's still a split second of reaction time for the TCU. Completely normal. Consider that you have a car nearly 15 years old. Consider also that you have a tranny, while electronically controlled, is still bound by mechanics. You can't expect 50/50 split as soon as you put it in first, you have to cause the TCU to want to throw the split.

[nipper]

Reverse and first, along with WOT force a 50/50 split. It takes a split second for the hydraulics to react. Your car is fine.

 

nipper

[RallyKeith]

If you have an automatic then there really isn't a center diff. It's actually a hydraulic clutch pack that acts like an open diff that is biased towards the front under most driving. Only once the TCU detects wheel spin will it begin to lock up, and then it's got a 50/50 split. That is the most you can ever get in an auto. There is no way to put more than 50% power to the rear wheels. So, what your car is doing is completely normal. At first 90% of the power is to the front wheels. You accelerate and the weight shifts towards the rear and the fronts lose traction and begin to spin. The TCU senses that and proceeds to lock up the "center diff" spliting the power and stopping the wheel spin.

 

If you had a manual it still acts like an open diff initially, but there is an actual diff in there. In a manual you can actually get more than 50% power to the rear before the viscious coupling begins to lock up the diff.

 

Subaru did a great job with the ad campaign "from the wheels that slip to the wheels that grip" but it doesn't really work that way. A subaru can never take all the power away from any one wheel, or axle. The best it can do it lock everything up so that each wheel is seeing 25% of the power.

 

Keith

[All_talk]

 

Subaru did a great job with the ad campaign "from the wheels that slip to the wheels that grip" but it doesn't really work that way. A subaru can never take all the power away from any one wheel, or axle. The best it can do it lock everything up so that each wheel is seeing 25% of the power.

 

Keith

Not exactly true, open diffs are torque balancing devices, the two axels are always outputting the same torque (not the same speed or power). As long as no wheels are slipping both axels get half the available torque, and in the case of the Ej 5MT with the center diff, all 4 wheels get the same torque, 25% at each wheel… until a wheel starts to slip. When slip occurs in an open diff the output is still balanced, but is equal to the wheel with the least grip. So worst case, one wheel gets no traction (think one wheel off the ground), its outputting no torque and so is the other one that’s stopped on the ground. But in a LSD diff the VC (or clutch packs) use the speed difference from axel to axel to apply torque across the diff. The free spinning wheel still applies no torque, but the wheel with traction does (equal to the torque created across the VC) and starts to turn. And since “power” is a function of speed (RPM) and torque, the Subaru AWD system in the 5MT with its center VLSD does in fact “transfer power from the wheels that slip to the wheels that grip”.

 

Gary

[jamal]

Also, the STi and VDC or VTD or whatever Legacies can do a little better. The Legacy uses brakes on individual wheels to keep one side from spinning, and the STi has two LSDs and a center diff that can lock up and send nearly 100% of the power to the front or rear. they, along with the AT WRX, also have rearward bias.

 

On pavement with the FWD fuse I was getting front wheel spin all over the place on pavement. With it out I don't notice unless I have it in D and it's pretty slippery.

[RallyKeith]

and the STi has two LSDs and a center diff that can lock up and send nearly 100% of the power to the front or rear.

 

Note the flaw in your logic. If a diff is locked, everything sees the same power, therfore you can't have 100% anywhere but the input. The power is applied equally on the output.

[RallyKeith]

Not exactly true, open diffs are torque balancing devices, the two axels are always outputting the same torque (not the same speed or power).....

.... When slip occurs in an open diff the output is still balanced, but is equal to the wheel with the least grip. So worst case, one wheel gets no traction (think one wheel off the ground), its outputting no torque and so is the other one that’s stopped on the ground. Gary

 

Good point!

[jamal]

Note the flaw in your logic. If a diff is locked, everything sees the same power, therfore you can't have 100% anywhere but the input. The power is applied equally on the output.

 

Imagine the front wheels are on rollers, the rear wheels are on the ground, and the center diff is locked. Are the front wheels going to do any work to move the car forward? no. Is the car going to move forward from the full power of the motor going to the rear wheels? yes. A locked diff can transfer all the power to one end whereas an open or lsd cannot.

[cannonball]

Imagine the front wheels are on rollers, the rear wheels are on the ground, and the center diff is locked. Are the front wheels going to do any work to move the car forward? no. Is the car going to move forward from the full power of the motor going to the rear wheels? yes. A locked diff can transfer all the power to one end whereas an open or lsd cannot.

 

But a locked diff is a locked diff. Depending on gear ratios there would be no transfer from front to back at the point the diff locks. In other words once the diff is locked it is what it is.

[jamal]

Okay, back to the front wheels being on rollers. There is no resistance from the rollers, and therefore no force exerted by the rollers to the wheels. Newton's second law anyone? The force exerted by the front wheels is zero, therefore all the power goes to the back.

 

Edit: 2nd law.

[ericem]

There's still a split second of reaction time for the TCU. Completely normal. Consider that you have a car nearly 15 years old. Consider also that you have a tranny, while electronically controlled, is still bound by mechanics. You can't expect 50/50 split as soon as you put it in first, you have to cause the TCU to want to throw the split.

 

ok, well my car slips llike real noticeably, like on dry ashspalt the wheels smoke, but its definitly AWD, because with the FWD fuse in place, cetainly does spin the wheels for a long time, but i have changed the tranny with a rebuilt one, and it has replaced solenoid, and clutch, and the TCU was changed, why is there such a delay? My dad told me when the car was new it never had a problem engadinjg, as soon as he floored it, it was impossible to notice any wheel spin, at ALL, even in the winter, so how do i force 50/50 split, just put it in first with manual mode in? I tryed that spins the front wheels still for bit, then u can feel the rear kick in, think i should change the tranny fluid?

[johnceggleston]

clarification please:

 

any time the a/trans is in first or when you manually put the selector in first do you have 50/50 split. i just asumed it was only when you placed in first gear or reverse that it was 50/50, all other times it was computer controled. but even when it's is 50/50 you still have computer input be cause you don't have torque bind, right? i guess there's a difference between 50/50 split and locked.

 

john

[Manarius]

Changing the tranny fluid isn't going to make the solenoid engage any faster. If you're getting wheel spin, first I'd like to know how since I know my 130hp EJ22 is not capable of wheel spin even in FWD mode. Second, I think that you're over exaggerating the problem. I'd like to know how you can get your car to break rubber on dry pavement.

 

Assuming there is something wrong, it would require removing the tranny to either replace it or some parts inside it. Why are you so concerned about 50-50 split? You're not going to need it on a daily basis anyways. And if you're trying to race people, I'd suggest against it - that's how young people like us get killed.

 

Clarification: When the selector is in the 1st selector, the split will be 50-50 if the computer detects slipping. In other gears, the computer adjusts as necessary, using 50-50 in a worst case scenario.

[cannonball]

Okay, back to the front wheels being on rollers. There is no resistance from the rollers, and therefore no force exerted by the rollers to the wheels. Newton's first law anyone? The force exerted by the front wheels is zero, so all the power goes to the back.

 

I agree with this statement.

 

I love AWD discussions.

[johnceggleston]

 

Clarification: When the selector is in the 1st selector, the split will be 50-50 if the computer detects slipping. In other gears, the computer adjusts as necessary, using 50-50 in a worst case scenario.

 

got it.

 

thanks

[wallaceg]

clarification please:

 

any time the a/trans is in first or when you manually put the selector in first do you have 50/50 split. i just asumed it was only when you placed in first gear or reverse that it was 50/50, all other times it was computer controled. but even when it's is 50/50 you still have computer input be cause you don't have torque bind, right? i guess there's a difference between 50/50 split and locked.

 

john

 

When in 1/2/R, the 4EAT transmission is NOT a 50/50 split. A 50/50 split would cause severe torque bind, which does not happen when you downshift. What actually happens is that the duty-c solenoid is set to a 50% duty cycle, which means that the rear driveshaft is engaged 50% of the time. Similarly, the split is NOT 90/10 when in D. The duty cycle of the duty-c is set to 90%, which means that the rear driveshaft is engaged 10% of the time. If the TCU senses that the front wheels are slipping, then it will engage the rear driveshaft up to 100% of the time automatically.

 

In order to get a 50/50 split between the front and rear wheels, the duty-c must be set to a 0% duty cycle, which engages the rear driveshaft 100% of the time. You can install a simple "kill switch" to the duty-c in order to manually set the split to 50/50, but this causes severe torque bind and will ruin your transmission if done frequently on dry pavement.

[johnceggleston]

When in 1/2/R, the 4EAT transmission is NOT a 50/50 split. A 50/50 split would cause severe torque bind, which does not happen when you downshift. What actually happens is that the duty-c solenoid is set to a 50% duty cycle, which means that the rear driveshaft is engaged 50% of the time. Similarly, the split is NOT 90/10 when in D. The duty cycle of the duty-c is set to 90%, which means that the rear driveshaft is engaged 10% of the time. If the TCU senses that the front wheels are slipping, then it will engage the rear driveshaft up to 100% of the time automatically.

 

In order to get a 50/50 split between the front and rear wheels, the duty-c must be set to a 0% duty cycle, which engages the rear driveshaft 100% of the time. You can install a simple "kill switch" to the duty-c in order to manually set the split to 50/50, but this causes severe torque bind and will ruin your transmission if done frequently on dry pavement.

 

ok so if you are in a situation and know you need more power to the rear wheels, put the selector in 1/2/r and you start off with more rear power. not 50/50 power but more than the standard in d. but the computer can't / won't vary the power because your manual selection has over ridden it.

 

did i get it right. and i think it would take an engineer and a 4wd dyno to determine just what the power split was, and i don't really care, but if i'm in something deep i know enough to get it right the first time because by the time you try again, you may worse off.

 

some one posted here once before that his dad explained off-roading in the old days, as driving in 2wd till you get stuck and then egaging 4wd, turning around and going home. this doesn't work for awd.

 

john

[cannonball]

If there is a 50/50 split for the AT in 1/2/R would it have the ability to shift power back to the front if the rear wheels slipped or is it truely locked?

[nipper]

If there is a 50/50 split for the AT in 1/2/R would it have the ability to shift power back to the front if the rear wheels slipped or is it truely locked?

 

thats not how it works. There is no shifting of power, just sharing it. You get between 90/10 to 50/50 split. The 50/50 is in low and reverse.

The newer versions plays with the brakes to transfer more power to the non slipping wheel, basically you can do the same thing in an older one by slightly applying the brake (or on a manual by applying the hand brake).

 

It doesnt decide to dissconnect power fropm the front and send it to the rear, or visa versa.

 

The front wheels basically propel the car. If the front wheels spin faster then rear, then power is applied to the rear. There is almost no way for the rear wheels to spin since the car is biased toward the front.

Now if if the car was biased towards the rear, then if the rears spun, power would be applied to the fron untill everything was spinning the same speed.

 

nipper

[johnceggleston]

If there is a 50/50 split for the AT in 1/2/R would it have the ability to shift power back to the front if the rear wheels slipped or is it truely locked?

it sounds to me like that the computer measures the front vs. rear speed and increases the power to rear when its slower. it would then stand to reason the if the rear speed was greater it would reduce power to rear. this would give the front more pull. correction, it may be that it can't turn the rear wheels faster. the rears will only turn as fast as the front, slower sometimes - but not faster. the rear wheels can spin but not faster than the front.

 

but i think we have to start thinking about this in the way it measures the difference, speed not power. as long as the wheels front and rear are turning the same rpms, the computer does nothing. when the front spins faster, more jucie to the rear until they catch up, or get within spec. as the speeds match again, less to the rear. the duty c apparently has a minimum operating cycle of 50%, standard is 90%. 100% operating cycle is FWD and 0% is locked (torque bind).

 

ok , my eyes are glazing over.

 

edit: i started writing befroe nipper and ended after him. obviously he knew the answer, i had to think aboutit.

[cannonball]

thats not how it works. There is no shifting of power, just sharing it. You get between 90/10 to 50/50 split. The 50/50 is in low and reverse.

The newer versions plays with the brakes to transfer more power to the non slipping wheel, basically you can do the same thing in an older one by slightly applying the brake (or on a manual by applying the hand brake).

 

It doesnt decide to dissconnect power fropm the front and send it to the rear, or visa versa.

 

The front wheels basically propel the car. If the front wheels spin faster then rear, then power is applied to the rear. There is almost no way for the rear wheels to spin since the car is biased toward the front.

Now if if the car was biased towards the rear, then if the rears spun, power would be applied to the fron untill everything was spinning the same speed.

 

nipper

 

OK it is becoming clearer. I now understand that the active system is one or the other and not between. My thinking was that it could go any ratio between 90/10 and 50/50. The ratio for the MT is infinite which gives it the "continuous" nomenclature. I have been enlightened. Thanx.

[ericem]

Look all i want to know is why is there a delay? There should not be a delay, all everyone says is the car is 13years old, i dont CARE, it should drive like when it was new, the tranny is pretty much new, and the TCU is new. Should i reset the tcu, and see if it fixes the problem?

[Manarius]

Look all i want to know is why is there a delay? There should not be a delay, all everyone says is the car is 13years old, i dont CARE, it should drive like when it was new, the tranny is pretty much new, and the TCU is new. Should i reset the tcu, and see if it fixes the problem?

I think you're expecting Deus ex Machina - not the case for this car. I'm still interested in how you're spinning out on dry pavement.

[firstwagon]

I get a slight bit of front wheel spin if I nail it on a slippery surface. Hard to notice really.

 

I can burn rubber no problem on dry pavement if I put the fuse in and stand on it from a stop.

 

(91 Legacy LS Wagon)