Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Toyota 4WD systems explained



  • cliffy1cliffy1 Posts: 3,581
    Don't those shop manuals cost about $120? You spent that kind of money so you could verify my statements on an Internet bulletin board? Wow. I guess I feel honored.

    As to your story of the 5% incline, I too am curious. It is my experience that many drivers haven't a clue as to how to properly operate their vehicles in the snow. Momentum is everything when dealing with extremely slick surfaces as you described. It is quite possible that you were the only one to hit it at the proper speed to make it up.
  • wwestwwest Posts: 10,706
    I was able to sit there and observe many drivers, and you're right, some of them clearly didn't have a clue. The approach to the incline wasn't such that you could reach it with very much momentum, it was a relatively narrow "cut" in the snowbank between the lower and the upper parking areas.

    Over on CL GS thread there is lots of discussion about the need for a mechanical LSD in the GS430 vs the standard TRAC. On the GS series I have expressed my thoughts that the TRAC will always outperform a mechanical LSD. I have PSM on my AWD 911 and while I don't yet have very many miles on it I am pleased with the PSM so far.

    But in listenting to some of the counter arguments I have begun to wonder if the electronic TRAC is really appropreate for an LSD replacement on an SUV.

    I don't think I have EVER known of an SUV with an LSD on the front. It seems to me that the feedback to the steering wheel would be so severe as to maybe yank it from your grip (like a Jeep in part-time 4WD on dry pavemant). Lexus' has told me that the TRAC implementation on the RX is much more "moderate" than my 92 LS.

    Maybe that's how they get around severe steering wheel feedback but I suspect that would also be pretty detrimenatl to AWD perfromance, especially if the RX always remains predominantly FWD biased, as I also suspect.
  • wwestwwest Posts: 10,706
    Ever since my very first car I have always bought the shop manuals, not just so I could DIY,M, but os I could understand the designs and overall workings. When I bought my first Jeep, an 85, it turned out to be such a mish-mash of parts from various sources that I didn't pay much attention to the manual and I didn't buy one for the 92 Jeep.

    So for all that time, 87 to 99, I never knew why the full-time mode of the 4WD Jeep system was virtually worthless. I knew that often in adverse conditions I always had to go to part-time (locked transfer case, center differential{?}).

    I probably wouldn't go to the trouble of buying the Sequoia manuals if you hadn't said it uses the same TRAC system for LSD substitution as the RX. Obviously that manual might help me understand my RX.
  • cliffy1cliffy1 Posts: 3,581
    Well, the TRAC system is the same, but the effect is different due to the different types of center differentials. You would probably be better off with an RX shop manual. It would behave the same if you are talking about its effect only at the front or rear wheels, but the Sequoia manual will not enlighten you on the effect on the center differential.
  • wwestwwest Posts: 10,706
    I have the shop manauls for my first RX, a 2000, and I have them on order for my 2001. The problem is that there is absolutely no information about how TRAC works or is supposed to work.

    I have personally concluded, at least for the RX, that TRAC is only used right to left or vice versa. And maybe more moderatly in the front vs the rear due to the steering wheel feedback that would likely otherwise result.

    Additionally, if TRAC were used to pro-actively allocate torque distribution front to rear then the longer time constant VC would be useless. However briefly, it needs a sustained disparate turning rate F/R in order to heat the viscous fluid and thus increase the coupling coefficient across the VC, itself across the center diff.

    Both the RX and the HL have an open, free, center differential, actually all three diffs are of this type. Across the center diff is a Viscous clutch, coupling, which is supposed to limit the "freedom" of the center diff if its two outputs develop a disparate turning rate.

    Becuase the RX never acts like an AWD, always "acts" more like a FWD than anything else, I have come to suspect the VC is there for marketing purposes only, it doesn't have enough "pre-load" to actually have any affect on drivability on LTS, Low Traction Surfaces.

    I came to this conclusion after reading the implementation details of the Chrysler T&C AWD system.

    Chrysler says that since the front wheels do 80% of the braking there cannot be a VC, at least not a "tight" one, between the front and rear during severe braking when the front wheels are likely to turn at a substantially slower rate than the rear. "Telegraphing" the slow front roll rate to the rear through a tight VC coupling, or a VC coupling that quickly tightens with disparate F/R roll rates, might brake the rear wheels to so great a level that the driver loses control of the vehicle.

    So Chrysler's VC is bypassed by an over-running clutch so the rear wheels can over-run the front during severe braking.

    But that's not all.

    So that the T&C can remain an AWD in reverse gear when otherwise the over-running clutch would open the driveline to the rear, Chrysler adds a "dog-clutch" to lock the front and rear drivelines in reverse.

    Now, this is Chrysler, CHEAP, inexpensive, shoddy designs, that has gone the extra mile to assure their customer's safety and provide a truly multi-mode AWD.

    Now do you see why I suspect the RX's VC isn't really functional?
  • drmpdrmp Posts: 187
    You mentioned ...."Becuase the RX never acts like an AWD, always "acts" more like a FWD than anything else, I have come to suspect the VC is there for marketing purposes only, it doesn't have enough "pre-load" to actually have any affect on drivability on LTS, Low Traction Surfaces."

    I am with the impression that HL and RX (also RAV) has a center differential with 50/50 torque split and supplemented with VC (mild version?) to limit front to rear wheel speed difference. Torque is always distributed to all wheels at all times.

    I believe Chrysler (and Volvo?)has no center differential, just VC to connect the front to rear axles. Torque here is fed to the front and if the front slips, the VC catches up and send some torque to the rear. Others has a VC substitute such as that of GM versatrac, MDX and CRV.

    Just my $0.02.
  • wwestwwest Posts: 10,706
    Can't speak for them.

    But the T&C and the RX AWD setup is VERY simuliar.

    Both have three open differentials, front, center, and rear. Absent the VC, what that means is that as long as all four wheels maintain something approximating equal traction the engine torque will be distributed accordingly. But if one wheel, or wheels, begin to slip, then ALL available traction will be routed to the wheel or wheels with LEAST traction.

    The T&C shop manual is very explicit and detailed even including a good theory of operation and some cause and effect, and even some engineering background regarding implementation decisions.

    The RX shop manuals, on the other hand, seem to be written as secretively as is Lexus "HELPS" about the overall operation of their system.

    So the Chrysler first.

    The Chrysler's VC is placed in series with the rear driveline. Chrysler says that its initial pre-load, which determines the VC's coupling coefficient in latent (no wheelspin) mode, is high enough that during severe or even moderate braking it might "telegraph" the slower turning heavier braking duty front wheel's rotational rate to the rear wheels, defeating the ABS at the rear which might lead to rear wheel lockup and loss of control.

    Obviously things would quickly get out of hand with sustained severe braking, say on a ice covered slippery roadbed, when the VC reacts by further increasing the coupling coefficient.

    So Chrysler has placed an over-running clutch around the VC so that the rear wheels can freely over-run the fronts in these situations.

    Given the dearth of detailed information about the RX AWD setup some of what I say about it is something more of a guess than actual knowledge.

    What I finally did was enlarge the pictorial from the RX shop manual of the "transfer" gearing in the area where torque is split front to rear. I'm sure I can state as factual the issue of an open differential for basic F/R torque split.

    What remains somewhat fuzzy to me and others is the placement of the VC in this mix. Several mechanical (but with no automotive experience or background)engineers believe the VC is mechanically coupled ACROSS the center differential and others think it is more like the T&C. It is clear to me that the answer cannot be readily determined From the enlarged drawings and Lexus refuses, so far, to clarify the issue.

    Regardless it is clear to all that there is no over-running clutch. The only conclusion one can derive from this is that the RX's VC never reaches a high enough coupling coefficient to affect the need for ABS at the rear. Either that or Lexus doesn't care about our safety and I can't, will not, given the evidenceto the contrary below, accept that as an explanation.

    Shade-tree testing. Done on both a 2000 AWD RX and a 2001 AWD RX with VSC and TRAC.

    With the rear wheels off the ground on a roller jack and the front wheels chocked against forward movement and the transmission in drive and the engine RPM increasing slowly and carefully the rear wheels would spin freely and the front wheels would rise higher on the chocks as engine RPM rose. I was too scared (timid?) to get much above engine idle, didn't even reach 1000 RPM in this testing mode.

    In the opposite configuration, rear wheels chocked, etc, I could raise the engine RPM all the way to 2000 with no perceptible effort by the rear wheels to climb the chocks, and the front sspinning rapidly.


    With all four wheels elevated.

    First I blocked the rear wheels from turning with 1x2" strips of light duty lumber, allowing the fronts to turn freely. In drive mode with RPM as high as 2000 you could hear the wood being stressed but no breakage.

    In the opposite configuration the 1X2" strip blocking the left front broke just as soon as I raised the throttle above idle.

    As far as I could tell TRAC was never activated and certainly there was no indication of same. I could cause VSC to come alive by turning the steering wheel off-center as I raised the engine RPM.

    Then New Year's day I watched an AWD RX with VSC fail to climb a 25 yard 5% incline of packed snow and ice that my AWD Aerostar had no trouble at all getting up. I also watched as a new 4runner failed to get up this slope.

    To be fair most of the vehicles that tried didn't make the grade, maybe only about 10% did. When I offered to help the 4runner driver by "pushing" (manually) he seemed to get fairly peeved. He even tried it in low range with no success.

    Come to your own conclusions.
  • cliffy1cliffy1 Posts: 3,581
    A viscous coupling is by definition a limited slip. Go to the following link for a complete explanation of how a VC works.

  • wwestwwest Posts: 10,706
    Has two design, manufacturing parameters, that are important to us.

    Initial coupling coefficient is determined by latent, non-heated, viscosity and apparently this parameter is most often "set" by controlling the "size" of an air bubble within the hermetically sealed VC assembly. Until this air bubble is completely compressed, virtually non-existant, the increasing VC fluid volume within the chamber due to increasing temperature will not result in an increase in the coupling coefficient.

    The second parameter has to do with the formulation of the fluid, how fast will the fluid's volume increase with rising temperature. The faster the fluid "expands" with rising temperature the higher is the "rate of attack" of the VC.

    It appears to me that the RX's VC chamber may be filled with air and not much else.
  • pschreckpschreck Posts: 524
    Are you sure the fluid doesn't simply thicken? Why do you say it expands?
  • drmpdrmp Posts: 187
    You mentioned this about RX: "With the rear wheels off the ground on a roller jack and the front wheels chocked against forward movement and the transmission in drive and the engine RPM increasing slowly and carefully the rear wheels would spin freely and the front wheels would rise higher on the chocks as engine RPM rose. I was too scared (timid?) to get much above engine idle, didn't even reach 1000 RPM in this testing mode."

    With the rear wheels spinning freely and the front not slipping, it indicates that it has a center differential. Otherwise, if it is just plain VC it would require the front to slip first before the rear starts spinning.

    Your subsequent experiment would indicate that the VC on the RX's center diff biassed to one direction.

    Here's an interesting and really informative article but a bit dated.

    Happy reading.

  • wwestwwest Posts: 10,706
    In the beginning we were all taught that fluids become more thin, viscosity lowers, with increases in the fluid temperature, and vice versa.

    So when the VC was announced some years ago that seemingly relied on fluids acting inverse to this law I was taken by surprise. But in the end I figured that some scientist had figured out a formulation to create this affect, increasing viscosity with increasing temperature.

    Now I can't say for sure that the above revelation was wrong, but after recently reading a fair amount about VC implementation and design I have come to believe that VCs actually operate in the following way.

    The two sets of clutch plates and the viscous fluid are placed in a fixed volume sealed chamber. The fluid, silicon normally, is specially formulated to dramatically increase its volume (all{?}fluids expand with temperature) with just slight increases in temperature.

    So, what we have is a VC wherein the viscous fluid "effectively" increases it viscosity but what is actually happening is the fluid "pressure" within the chamber is what is actually doing the work, increasing the coupling coefficient between the two sets of clutch plates.

    Really only a guess based on the way manufacturers control the actions of a VC, air bubble, etc.
  • wwestwwest Posts: 10,706

    Not at all, I need all the help, input, I can get.

    As I said the RX has an open, free, center differential, just like the front and the rear. The center differential seems to be configured very much like the pictorial you linked.

    Except some of us think the VC is mechnanically mounted ACROSS this center differential, coupling its two outputs, thus limiting its "freedom" if the wheels if either output begin to slip. Others of us think the VC is simply in series with the center differential's output to the rear driveline, just like the T&C.

    Even enlarged, the RX's pictorial representation can be readily interpreted, argued, either way.

    In any case it is pretty clear to me, based on my testing, that the RX AWD, even with the VC fully "activated", remains primarily FWD torque biased. But none of us have been able to determine, from the information available, how this torque bias to the front is accomplished.
  • drmpdrmp Posts: 187
    I remember it was mentioned before that there are the cheap kind of VC and the more expensive kind. I think the cheap one was used by Talon/eclipse to supplement the center diff. It has a TRANSIENT "hump" on it's torque curve corresponding to fluid heat expansion that pushes the discs to touch one another. This was reported to have inconsistent locking effect.

    The more expensive one have more numerous discs and rely primarily on "viscous drag" for locking effect.
  • tidestertidester Posts: 10,059

    I think the increased viscosity with temperature doesn't have much to do with the pressure, though it's a good guess!

    It was discovered that the behavior of a viscous fluid can be the reverse of what is expected (newtonian fluids) by the addition of certain polymers to the fluid. I think it has to do with the unraveling or stretching of the polymer molecules as the fluid is warmed.

  • pschreckpschreck Posts: 524
    We should be getting snow today. Can't wait to try out the Sequoia 4wd system. Hoping the stock tires are up to it. I don't think they'll be any real problem as all the other snow perfomance reports I've read have been glowing.
  • wwestwwest Posts: 10,706

    Just what I believed until I ran across the actual VC manufacturing process. They actually do control its performance "on-set" by injecting an "air-bubble" of a specific size depending on the level of coupling coefficient they want vs increases in temperature.

    Perhaps there is more than one way to skin a cat.
  • spdmtr5spdmtr5 Posts: 111
    Is there a site that explains what the fluid really does when it heats up;thickens or expands?Who makes it? Does it need to be replaced due to breakdown?Can the injectd air bubble be changed to "tune" the VC?I guess the question really is what does this fluid really do when heated.My knowledge of chemistry is basic but I think that I could understand a paper on the subject.
  • wwestwwest Posts: 10,706
    Yes, it can fail but the only thing I have seen on the subject is in the 911 AWD owner's manual, it says as little as twenty seconds of uncorrectable wheelspin and its VC is toast.

    I have considered finding someone to "re-build" the VC in my RX so it does a better job of reacting to wheelspin and more quickly and firmly distributes the torque. But in the interests of safety, my own, that would likely require that I add an over-running clutch like the Chrysler T&C.

    So I guess I will, for the moment, live with an RX300 AWD that isn't.
  • wwestwwest Posts: 10,706
    Torsen vs VC

    Most importantly, the torsen (center differential) does not lock or inhibit speed differences during braking, thus allowing all 4 wheels to rotate independently at their own speeds when no power is applied. The torsen diff'l only locks in a power application situation while the VC locks during acceleration OR BRAKING. The torsen has a torque sensing characteristic while the VC has a rotation sensing characteristic.

    The VC's rotational sensing characteristic initially caused lots of problems for the engineers. Electronic anti-lock braking systems, ABS, rely almost entirely on speed differences between the four wheels to detect a locking wheel. Thus, when the VC tries to force all 4 wheels to turn at the same rate, during moderately heavy, or severe braking, it creates serious difficulties for an electronically implemented ABS system, and since it would telegraph the front wheels' (slower) rolling rate to the rear wheels it might cause loss of control of the vehicle.

    The engineers had to use a variety of HACKS to get around this problem. Mitsubishi delayed ABS for awhile for its first generation GSX, then finally decided to make ABS and rear VC limited slip mutually exclusive options. The VW syncro system simply disconnected 4WD the moment the brake pedal was depressed via a secondary clutch. Most other vehicles using this implementation of VC have a very simular disengage feature. The very successful World Rally Championship Lancia Delta Integrale even went so far as to apply a bit of power (via the engine ECU) to reduce the drag of the VC when the brakes were applied! Some very crude (sophisticated?) VC systems used an overrun device (Chrysler T&C) conceptually simular to a bicycle crank. This meant that while 4WD was disengaged during braking it was also inoperative when reverse was engaged! (The T&C overcomes this VC implementation flaw by adding a dog-clutch to lock the over-running clutch in reverse).

    The easiest VC "HACK" was to


    of the fluid in the VC coupling, so that it never (NEVER{?}) "TELEGRAPHED" the front wheels' rolling rate to the rear wheels, possibly causing them to lock up and cause loss of vehicle control, and ABS could still perform its task, rather than have the more heavily braking front wheels "dictate" their rolling rate.

    This also meant that the VC's "normal, and overall locking effectiveness was reduced, in some cases significantly so, but that might be quite acceptable for a vehicle used primarily below the snowline and never off-road.

    The VC's "attraction" is its simplicity and cheapness, not it's sophistication.

    Lexus RX300 AWD.. CHEAP ??!! NON-SOPHISTICATED ??!!

    None of this tells us why the RX300, with an open center diff'l but with a FLACCID VC, still remains predominantly FWD.


    Any questions class?

    Original document prepared by Eliot Lim.
    Editing by WWEST
  • wwestwwest Posts: 10,706
    I have seen a site explaining the need for a viscous fluid for these applications, VCs) that more dramatically increases its volume with increasing temperatures, including graphs of the rate of volume increase of some fluids vs others. I think the conclusion was that silicon based fluids of various formulations had the most dramatic increases in volume.

    Sorry, I don't remember the specific site.
  • cliffy1cliffy1 Posts: 3,581
    We're about to get our first snow as well down here in VA. Let me know how it works out for you. I own a Tundra and am still hopeful the 2003 model will include the Active-Trac system. My lease expires in November and the timing would be perfect.
  • spdmtr5spdmtr5 Posts: 111
    I would like to have some of the viscous fluid,place it in a pan, heat it,then determine if it really does thicken.Fudge does but then when it cools it sure doesnt thin down.And we are glad of that.
  • wwestwwest Posts: 10,706
    The viscous fluid will thicken with heat in the same way does the fudge, boiling off some of the liquid content.

    Do you like yours with walnuts or pecans. I vastly prefer pecans.
  • spdmtr5spdmtr5 Posts: 111
    Pecans!The only way.Could it be that the system in my wife's Trailblazer is really more reactive than a VC? I've not driven on ice-she has- and said it worked great.She used the 4awd position on the selector.The TB also has a LSD.
  • drmpdrmp Posts: 187
    "None of this tells us why the RX300, with an open center diff'l but with a FLACCID VC, still remains predominantly FWD."

    This is just a guess, I pressume the RX's traction control (part of VSC) is more agressive on applying the brakes to the rear thus quickly rerouting torque to the front. It is less agressive on braking the front since it affects steering and thus routing less torque to the rear on slippery situations.
  • spdmtr5spdmtr5 Posts: 111
    It may be that an increase in traction in the rear of a FWD car may cause more handling problems than increased traction to the front on a RWD;i.e. on the RWD if it's sideways additional traction to the front would tend to pull it straight,while adding traction to a front driver might skew the rear sideways.Pure FWD can be a handful on slick downhills.They want to rotate around the weight mass up front and you can't add any power with your right foot to control the slide.
  • wwestwwest Posts: 10,706
    In the past I have proposed that TRAC is only used right to left or vice versa, never front to back, etc. Lexus "HELPS" has so far not been willing to confirm either way.

    My reasons:

    TRAC is an electronically implemented system with the only mechanical "time delay" being the time needed to pressurize the brake calipers. Lexus says that TRAC will activate on detection of "impending" wheelspin. The VC on the other hand works on a fairly long time constant, it typically takes many milliseconds to heat the viscous fluid through the friction of the disparately rotationing VC clutch plates. Additionally the on-set of the VC activity must be designed for a fairly long delay to prevent its activation during turning manuvers and to prevent undue interference with the ABS.

    So the VC's incremental activation is totally dependent on sustained disparate rotational rates of the two sets of clutch plates and this would never happen if the instantly acting TRAC were to be used to allocate torque front to rear or rear to front.


    Anyone ever hear of a front axle LSD?

    Both of my Jeeps had rear LSD and a part-time 4WD mode with a locked center diff'l. The thought of having a front LSD had never enterd my mind and I don't remember any sales document offering it as an option.

    I doubt that TRAC can be designed to function very well as LSD on the front axle. It seems to me that any left to right, or VV, TRAC activity on the front axle would provide such a horrid level of jerking feedback to the steering wheel that it would be extremely discomforting to many drivers.

    If not forewarned might they react to a suddenly jerking steering wheel in basically the same way people who first encountered pulsing brake pedals reacted?
  • drmpdrmp Posts: 187
    Front LSD. Yes, they are used by nissan cars and infiniti FWD cars (SE and touring versions). These however are not the mechanical clutch type but (drums roll please!) the VC type. If you wonder how a VC could be attached in parallel with the diff, it is still similar to the link I showed earlier. The VC's one end is attached to the differential casing and the VC's other end is attached to one of the axles. VC is ideal for FWD since it has negligible initial binding and has gradual increase in coupling torque in relation to differential speed. In contrast, the mechanical LSD has significant initial bind (60 lb-ft) and non-linear torque increase which can easily be felt in the steering wheel.

    Traction control. This can be fine tuned to equalize wheel speed on both sides thus acting as electronic LSD. I believe this feature is used by Audis, VW's and Volos. I haven't heard complaints about unfavorable steering effect/feedback due to traction control activation but I theorize that it can.
Sign In or Register to comment.