Part-time AWD transmission????

[johnceggleston]

If I understand the system correctly, you could make the computer think the rear is slipping relative to the front, or to put it more correctly, that the front was slipping so more power needs to be diverted to the rear. Under those cases, the rear would indeed be going slower than the front.

 

You can never feed the rear with more power (torque) than the front, though.

 

 

Does anyone know what the rear or front speed sensor is, and where its located?

 

 

Dave

 

the front speed sensor also controls the speedometer. i don't know much about how it works, but to over simplify it , it counts the revolutions. it's mounted in the passenger side of the front differential near the fire wall. it has a wire coming off of it. it is driven by a geared shaft off of the differential carier which rotates the same as the front drive axles.

 

the rear speed sensor is simplier, the sensor, a magnet?, notices every time a 'tab' onm the output shaft passes by ( i think there are 4 tabs). the sensor is before the rear differential. nipper tells me it's a square wave, i think. this sensor is the rear most wire on the passenger side mounted into the rear extention housing ( 95 - 99 auto trans).

 

any conversions that need to be done are covered by the ECU or TCU.

[nipper]

You dont want to use the speed sensors. Not only can you affect your spedo, but if the ECU sees that the drivelines are not catching up to each other, it may stop trying to correct the speed differnce between front and rear to keep from overheating.

 

I know it will do this at 50 mph (ie broken front CV joint), but even in that scenario you still have spinning output shafts front and rear. This threshold is unknown. You may aslo affect transmission shiftpoints. this is really an unknown area.

 

It would be easier to just directly control the solenoid.

 

 

nipper

[Gloyale]

 

It would be easier to just directly control the solenoid.

 

 

+1

[98obster]

+1

 

+infinity

Reading all the posts about the solenoid switch has become as nostalgic as watching It's a Wonderful Life. I installed the switch last year but I don't think I am fooled into thinking I can now offroad-the rear diff is still open. Anyways, I want to make sure my clutch pack will last as long as possible with the occasional use of the switch?. 1) make sure the transmission has had time to warm up? may not matter but I flip the switch (open circuit) only when I'm stopped and close it as soon as I don't need it. Winter tires are required before I use the switch. And I should probably flush the AT fluid every 16-18 months or so.

 

from what I've read in past threads, without the switch, if I just put it in 1st gear I am driving with 50/50 split?

 

a hard question: does the switch guarantee a 50/50 split or is that still determined by the condition of the clutchpack. I've heard AWD does diminish somewhat but would this still force 50/50.

 

jon

[nipper]

+infinity

 

a hard question: does the switch guarantee a 50/50 split or is that still determined by the condition of the clutchpack. I've heard AWD does diminish somewhat but would this still force 50/50.

 

jon

 

Think about this question. Any clutch's ability to transmit torque is limited by the condition of the clutch. If the clutch is tired, glazed, slipping, it's not going to transfer all the power it can.

 

Same goes for an engine, if its tired it wont make all the power its designed for.

 

If your clutchpack is dead, or tired, no matter wht you do to the solenoid, its still going to have any problem you had before installing the switch.

 

nipper

[98obster]

Think about this question. Any clutch's ability to transmit torque is limited by the condition of the clutch. If the clutch is tired, glazed, slipping, it's not going to transfer all the power it can.

 

Same goes for an engine, if its tired it wont make all the power its designed for.

 

If your clutchpack is dead, or tired, no matter wht you do to the solenoid, its still going to have any problem you had before installing the switch.

 

nipper

 

Yeah, I probably knew that. thanks. I do think aout the continuum between 90/10 and 50/50 and where the mean would be and how often the TCU would actually ask for 50/50 and for how long. The Duty C works a lot harder when it's draining pressure than when it's not-no?

 

It would appear it may be time to add my 98 to the old generation (at least older) forum when I read about the new Impreza and VDC-not a lot of switches to be installed on the '08 ya think. And in the last 3-4 years how many manufactures have added AWD versions-Subaru still best imho.

[nipper]

Yeah, I probably knew that. thanks. I do think aout the continuum between 90/10 and 50/50 and where the mean would be and how often the TCU would actually ask for 50/50 and for how long. The Duty C works a lot harder when it's draining pressure than when it's not-no?

 

It would appear it may be time to add my 98 to the old generation (at least older) forum when I read about the new Impreza and VDC-not a lot of switches to be installed on the '08 ya think. And in the last 3-4 years how many manufactures have added AWD versions-Subaru still best imho.

 

You get a 50/50 split at full throttle, reverse, and low gear. You get a 10-50/90-50 split depending on line pressure (internal tranny pressure) and the difference in speed between the front and rear speed sensors. Line pressure is basically the pressure from the front pump, which is directly connected to the engine. This is how the older system determens engine RPM. This is a mechanical spool valve, which works in conjunction with the Duty C.

 

There is a mysterious sensor in the center of the car, the G sensor. I can not find anywhere, nor can determine (ie dont want to screw with it) to see if it operates with the TCU, ABS, SRS. That may be another input.

 

Newer systems use individual wheel speed sensors, steering angle sensor, G sensor, brake input, Throttle, engine torque, moon and tide charts to determine where to send power when.

 

 

hope that helps.

 

nipper

[98obster]

You get a 50/50 split at full throttle, reverse, and low gear. You get a 10-50/90-50 split depending on line pressure (internal tranny pressure) and the difference in speed between the front and rear speed sensors. Line pressure is basically the pressure from the front pump, which is directly connected to the engine. This is how the older system determens engine RPM. This is a mechanical spool valve, which works in conjunction with the Duty C.

 

There is a mysterious sensor in the center of the car, the G sensor. I can not find anywhere, nor can determine (ie dont want to screw with it) to see if it operates with the TCU, ABS, SRS. That may be another input.

 

Newer systems use individual wheel speed sensors, steering angle sensor, G sensor, brake input, Throttle, engine torque, moon and tide charts to determine where to send power when.

 

 

hope that helps.

 

nipper

 

yes it does. thank you wise sifu.

[bulwnkl]

This is an area where I become confused as to the exact function of the 'newer' Subie center distribution systems. Straight mechanical open diffs like on my XT-6 I can comprehend: They're always 50/50 torque split until and unless you mechanically lock them. Then they're infinitely variable torque split up to essentially 100/0 - 0/100 depending upon traction.

 

But, the over-simplified diagrams and/or English-as-a-second-language descriptions I see on various web pages make the 'new' systems look like nothing more than either a straight open differential with a clutch pack mounted aft of them or no diff but a solid mechanical link fore and aft with a clutch pack to decouple the rear driveshaft. The first scenario makes no sense and so I dismiss it. Is the second what they're doing? That wouldn't be 50/50 torque split (though I guess it would be more like 50/50 power split) but at least that way I could see how they can claim what they claim for splits and how the cars are basically FWD unless the system 'sends' power backwards. ?

[nipper]

This is an area where I become confused as to the exact function of the 'newer' Subie center distribution systems. Straight mechanical open diffs like on my XT-6 I can comprehend: They're always 50/50 torque split until and unless you mechanically lock them. Then they're infinitely variable torque split up to essentially 100/0 - 0/100 depending upon traction.

 

But, the over-simplified diagrams and/or English-as-a-second-language descriptions I see on various web pages make the 'new' systems look like nothing more than either a straight open differential with a clutch pack mounted aft of them or no diff but a solid mechanical link fore and aft with a clutch pack to decouple the rear driveshaft. The first scenario makes no sense and so I dismiss it. Is the second what they're doing? That wouldn't be 50/50 torque split (though I guess it would be more like 50/50 power split) but at least that way I could see how they can claim what they claim for splits and how the cars are basically FWD unless the system 'sends' power backwards. ?

 

Ah yes grasshopper, but try to take the stone from my hand....

 

See this is the old thinking.

 

Come to new thinking. WHy go through the expense and complication of mechanical systems, when everything else is already existing on the car.

 

You have Traction control, ABS, and more importantly ABS with Electronic Brake-force Distribution (EBD) . This is where the magic happens.

 

The brake system is the LSD. When the system sensses a wheel spinning, and no forward motion, it can apply the brakes to the spinning wheel. Th brake application is usually light engough to stop the spinning wheel, but not lock it up (unless it deems it needed). This is how the car can actually stop three wheels and have all the torque go to just one.

 

This is not a new idea, as Baja Beetles had individual parking brake handles to do the same thing. It uses the simplicity and design of an open diff to its advantage.

 

nipper

[bulwnkl]

Indeed, I am familiar with using the brakes to send power where it's needed most. I just meant the center 'differential' itself.

[nipper]

The new systems are no differential, but a center clutch pack, or an electronically controlled viscous coupling and center diff, or just a viscous coupling with a center diff.

 

nipper

[bulwnkl]

The new systems are no differential, but a center clutch pack, or an electronically controlled viscous coupling and center diff, or just a viscous coupling with a center diff.

 

nipper

Alright, so:

 

- The center clutch pack with no diff is equivalent to a 'straight-through' or 'solid' driveshaft with a clutch pack to couple/de-couple the rear output shaft,

- The center diff with a viscous coupling is... is... here's where I'm losing something. If there's a 'regular' differential with a coupling downstream to the rear (i.e. on the output shaft that goes to the rear), then that coupling would always have to be engaged in order for the car to move. If it was not, the 50% of the torque that is always coming rearward from the diff would simply spin that side of the diff's output and the car would not move (until that coupling's fluid heated up and locked the clutches) just as a pickup with one rear wheel off the ground or in mud will not move, it will just spin the one wheel with no traction.

 

So, is the coupling somehow integrated into the center power divider? Is that center divider truly a regular differential (i.e. constant 50/50 torque split by definition), or is the term 'differential' just used out of convenience? Which brings me to my next question: How does a planetary system work in dividing power fore and aft? I only know how to use planetaries as gear reduction mechanisms.

[nipper]

Subarus are FWD by default.

 

Yes the rear clutch pack can be considered a straight through connection for arguments sake. The clutchs are designed to tolerate some slippage, in conjunction with the Duty c operation and the pressure variation of the spool valve.

 

Any Duty solenoid is called that because it cycles on and off very quickly. It is not designed to stay constitnly on, though some tolerate it better then others.

 

The center diff, lets go to a rear diff. Power is split 50/50 between both wheels as long as they both have traction. Once one looses traction, all the power goes to that wheel. A LSD type mechanisim (generally speaking) will transmit torque to the non spinning wheel through a set of clutches.

Now with the same set up as the senter differential, the same thing happens front and rear.

 

ANY AWD or 4WD vehical is in essence 2wd (one front one rear) with open diffs. If it is a really cheap AWD system (no center viscous coupling) it can easily be one wheel drive (with the one wheel getting all the torque if its slipping).

 

THis may or may not help

http://www.autozine.org/technical_school/traction/tech_traction_4wd_2.htm

 

nipper

[bulwnkl]

Subarus are FWD by default.

 

Yes the rear clutch pack can be considered a straight through connection for arguments sake...

 

...The center diff, lets go to a rear diff. Power is split 50/50 between both wheels as long as they both have traction. Once one looses traction, all the power goes to that wheel. A LSD type mechanisim (generally speaking) will transmit torque to the non spinning wheel through a set of clutches.

Now with the same set up as the senter differential, the same thing happens front and rear.

 

ANY AWD or 4WD vehical is in essence 2wd (one front one rear) with open diffs. If it is a really cheap AWD system (no center viscous coupling) it can easily be one wheel drive (with the one wheel getting all the torque if its slipping).

 

THis may or may not help

http://www.autozine.org/technical_school/traction/tech_traction_4wd_2.htm

 

nipper

Yes, that website is one of the sites I was talking about above.

 

Let's use engineering definitions for power and torque for a moment. In those terms, a 'regular' or open differential always distributes torque (but not power) 50/50 by design and definition, agreed? Because of this, and again in engineering terms, the power distribution can be from 100/0 - 0/100, though torque distribution remains 50/50 all the time, agreed? So, as you say, with an open diff when one wheel loses traction all the power goes to that wheel, however it still only receives 50% of the torque.

 

What I'm trying to decide whether I have clarity about is the differential/viscous coupling combination. It appears to me that, even though the pictures I've seen make it look like the viscous coupling is present on the rearward, downstream side of the diff, that clearly cannot be the case. If that was the case, the open diff would transmit 100% of the power (engineering def.) to the rear output shaft and the car would not move. Then, the viscous coupling would heat up and start to 'bite,' and the car would move but it would, in effect, have a simple open center differential with extra, unnecessary complexity.

 

So, it appears to me that the viscous coupling system must be integral to the differential itself and act as the limited slip mechanism would in a clutch-type LSD on a pickup. Is that the case, or am I still missing something?

 

Also, I'm starting to think that the way a planetary system is incorporated here is as a very, very slight gear reduction system aft (meaning on the rear output shaft side) of the center diff as a means of putting some heat in the viscous coupling system to keep it 'tight' or in other words ready to lock tight with any additional wheelspin at all. This appears to me to be an effort to alter torque distribution so as to try to keep power distribution (engineering def.) closer to 50/50 but still allow the vehicle to turn corners and such without stressing the drivetrain or scrubbing tires too much. Is that what they're doing?

 

If all these assumptions/guesses are correct, that would mean that only the Subies that don't have an actual differential in the center are the ones with the simple clutchpack (non-viscous) behind a 'solid' link (logically speaking) between forward and rearward output shafts. Is this correct?

[nipper]

Now your confusing me, and i'm the automotive engineer . you may be over thinking and confusing terms.

 

A differnential is a mechanical computer. the input shaft turns twice, the two output shafts turn twice. Make a left turn, left will turn once, right will turn three times. It basically averages the rotation of the outputs to match the drive shaft. Avg of three plus 1 divded by two (axles) is two.

 

There is no gear reduction in the clutch pack, this would self desturct the viscous coupling, same as driving on a flat. It can not handle a constant speed differential.

 

Manual tranny. All 4 wheels get torque (lets not use the term power anymore since people take it to mean HP, torque or both equally) on dry ground thorugh three differntials. The front axle turns at a differnt speed then the rear axle due to curves, sterring corrections, etc. This is why you need a center diff in AWD. When there is a specific differnce in speed between front and rear for a specific amount of time (hence the lag for a vc to hook up) the fluid heats up from internal friction. Usually this friction is caused by altenating plates with holes in them. they sheer the fluid, it hetas up, expands, and engages a clutch. Now you have a mechanical 50/50 split since one axle is spinning much faster then the other one. On dry ground for the most part the front rear speed differential is not so great as to kick in the VC in normal driving (normal is not power slides or autocross).

 

It is no differnt then an LSD differntial

 

http://auto.howstuffworks.com/differential.htm

 

nipper

[Gloyale]

You get a 50/50 split at full throttle, reverse, and low gear. You get a 10-50/90-50 split depending on line pressure (internal tranny pressure) and the difference in speed between the front and rear speed sensors. Line pressure is basically the pressure from the front pump, which is directly connected to the engine.

 

Technically, you dont; get a 50/50. You get Whatever amount is generated by the increasing of Transfer pressure to it's highest premapped value. In some cases, espescially with half worn clutches. The highest premapped value is not enough to fully engage the clutches and transfer 50/50

 

This is how the older system determens engine RPM. This is a mechanical spool valve, which works in conjunction with the Duty C.

 

Even the first 4EATs used a tach signal to the TCU to determine RPMs. Not fluid pressure.

 

 

 

If your clutchpack is dead, or tired, no matter wht you do to the solenoid, its still going to have any problem you had before installing the switch.

 

Like I just said. If you install a "lock switch" you force maximum possible pressure to the transfer clutches. Even old, worn clutches are going to grab pretty hard under that pressure. It really helps.

[nipper]

Technically, you dont; get a 50/50. You get Whatever amount is generated by the increasing of Transfer pressure to it's highest premapped value. In some cases, espescially with half worn clutches. The highest premapped value is not enough to fully engage the clutches and transfer 50/50

 

I was just going to argue with you, but am i the only one to notice all the 4eat and tranny stuff is off the endwrench site?

 

nipper

[nipper]

Technically, you dont; get a 50/50. You get Whatever amount is generated by the increasing of Transfer pressure to it's highest premapped value. In some cases, espescially with half worn clutches. The highest premapped value is not enough to fully engage the clutches and transfer 50/50

 

 

 

Even the first 4EATs used a tach signal to the TCU to determine RPMs. Not fluid pressure.

 

 

 

 

 

Like I just said. If you install a "lock switch" you force maximum possible pressure to the transfer clutches. Even old, worn clutches are going to grab pretty hard under that pressure. It really helps.

 

The pressure to the clutc p[pack is affected by the duty C solenoid and the mechanical spool valve. The spool valve is directly affected by line pressure, and the operation of the Duty C bleeds off some pressure from the spool valve. I dont consider the spoo;l valve directly affected by the TCU. The spool valve can operate independetly of the Duty C in case of duty C failure.

 

I found a back door to the endwrench drivetrain archives.

http://www.endwrench.com/archive/sysdrive.html

 

nipper

[nipper]

Not what i was looking for but its close

 

http://www.subaru.ca/WebPage.aspx?ArticleID=460&WebPageID=4998&WebSiteID=282

 

 

nipper

 

 

PS hehehe while i was looking i found a forester site discussing rear diffs and LSD, and finally they said "lest just make it simpler and learn to use the brakes"

 

hehehe

 

some truth to that

[bulwnkl]

I'm pretty sure we're either talking past each other or one or the other of us is talking way below or way above the other.

 

A differnential is a mechanical computer. the input shaft turns twice, the two output shafts turn twice. Make a left turn, left will turn once, right will turn three times. It basically averages the rotation of the outputs to match the drive shaft. Avg of three plus 1 divded by two (axles) is two.

Yes, I agree. Simultaneously there is always a 50/50 torque split between the 2 output shafts of the diff in consequence of its design.

 

There is no gear reduction in the clutch pack, this would self desturct the viscous coupling, same as driving on a flat. It can not handle a constant speed differential.

I agree with that but some of the information in or linked from this thread seems to suggest it's being done on purpose. I ask yet again: How does one incorporate a planetary system into the center differential? The planetary system is referenced at least here on USMB routinely as a component of the VTD system.

 

Manual tranny. All 4 wheels get torque (lets not use the term power anymore since people take it to mean HP, torque or both equally) on dry ground thorugh three differntials.

OK, so then in contrast to an earlier post in this thread, M/T Subies remain not basically FWD cars, but basically AWD cars with a limited slip center diff which uses a viscous instead of a more 'traditional' friction system? There is a decent explanation on the Canadian site you posted that answers my question about where and how the viscous system is located in the output stream.

 

I can't find one of the Endwrench articles that I was looking for on that backup site. Hopefully it'll come back or something. In any case, at least one thing that remains unclear for me given the present lack of Endwrench articles is how the VTD system operates. Is there a conventional center differential in place as there is in the M/Ts? Is it the 4EAT (phase II) that has no center differential but rather just a 'solid link' with a clutch pack (either friction or viscous) to decouple the rear end? Last I thought I read, the VTD system uses a planetary system somehow and a viscous (not conventional friction) pack somewhere in the center. What's the deal with the planetary system? Where is it and what is its function?

[keltik]

Well i made it through the last 3 pages without getting too confused so im going to celebrate by posting something hardly relevant.

 

Subaru got it right using a computer controlled clutchpack. Honda use a similar clutchpack with its own internal pump with a valve system that increases fluid pressure to the clutches as the speed differential between front and rear wheels increases.

 

This means you get nasty FWD understeer as you plough into a corner, which gradually builds to AWD oversteer - and if you try to correct the slide it will do all kinds of gay things.

 

Nissan and Volvo use EBD systems. A nice locked center diff and then a big computer that makes all the brakes do the work. Brilliant! No need for LSD's with this system. But once again the computers cant adjust very smoothly so you can get some wacky handling on loose surfaces.

 

No why the hell didnt everyone just use a viscous coupling and call it a day?

[nipper]

No why the hell didnt everyone just use a viscous coupling and call it a day?

 

Computers are cheaper in the long run and dont care about mismatched tires or flats.

 

nipper

[CNY_Dave]

Ah yes grasshopper, but try to take the stone from my hand....

 

See this is the old thinking.

 

Come to new thinking. WHy go through the expense and complication of mechanical systems, when everything else is already existing on the car.

 

You have Traction control, ABS, and more importantly ABS with Electronic Brake-force Distribution (EBD) . This is where the magic happens.

 

The brake system is the LSD. When the system sensses a wheel spinning, and no forward motion, it can apply the brakes to the spinning wheel. Th brake application is usually light engough to stop the spinning wheel, but not lock it up (unless it deems it needed). This is how the car can actually stop three wheels and have all the torque go to just one.

 

This is not a new idea, as Baja Beetles had individual parking brake handles to do the same thing. It uses the simplicity and design of an open diff to its advantage.

 

nipper

 

 

I don't know if they do this, but they could use the ABS in reverse (so to speak) to allow the braked wheels to spin exactly as fast as the unbraked wheel(s) by pulsing the brakes juuuuust right. That'd be a lot more effective than just clamping on the braked wheels a specified amount, and would be the only way to be sure the braked wheels weren't holding the car back.

 

 

Dave

[nipper]

I don't know if they do this, but they could use the ABS in reverse (so to speak) to allow the braked wheels to spin exactly as fast as the unbraked wheel(s) by pulsing the brakes juuuuust right. That'd be a lot more effective than just clamping on the braked wheels a specified amount, and would be the only way to be sure the braked wheels weren't holding the car back.

 

 

Dave

 

They could but....

 

Its impossible to modulate a wheel with no traction. The wheel with traction has the mass of the car behind it. Once you loose that traction, you just have a spinning mass of 100 lbs, where as the brake system normally (for arguments sake) modulates a 800 lb mass per wheel.

 

nipper