Paul, excellent article you wrote about the new Subaru STi. I'm guessing the STi will be similar to the one we'll receive here in the states. If there is anyone who hasn't read the article..check it out here: http://www.apexjapan.com/ "Look under Sti..much more than a pretty face"

Since you have driven this fine car and many others too, my question is: How useful is this DCCD to somebody who uses the STi as a daily driver on normal (non raceway) roads?
Personally, I think it would be fun to have a switch that would select the power "bias" front and rear nad hope it comes standard (or at least an option) on the USA STi. But, is it useful?? When would you want it on something other than 50/50? What setting for dry conditions, or wet or snow? What is the true adjustable settings. 10/90 (front rear)? 20/80, 30/70, is 100 rear and 0 front even possible. Could you elaborate on the settings.

Thank you very much for your insights and I look forward to future articles....

BTW, what is your comparison of the EVO 8 vs. STi? :)

Truly,

Davishttp://www.apexjapan.com/news/sti8/f_34s.jpg

P.S. of course if there is anyone else that has experience of the DCCD feature..fire away!

BMW system is 40/60 and I understand that the EVO tarmac level puts more power to the rear as well, though I don't know the percentages. More to the rear gives it more of a RWD feel with more safety by putting something to the front. Wouldn't it be nice to have a system like the Skyline where it is RWD until power in front is needed?

USDM EVO VIII power distribution is 50/50.

Generally speaking you'd leave the switch in the Auto position, which allows the torque to shift as needed.

Switching it fully forward is normal for gravel, snow, wet pavement, wet grass.... (John Felstead are you reading?! :)) - torque 50% front/rear

Switching it fully rearward for "spirited driving" or the track (dry). - torque ~35/65 front/rear

I have not driven a car equipped with the "auto" position, but it should behave like the current WRX we have (the torque split is not "locked" at 50:50, it "defaults" to that absent different load conditions, and it can shift as required via the center VC).

Switching it fully rearward doesn't make the car an oversteering powersliding monster, but it does alter the "feel".

Paul or John, feel free to jump in here.

Glenn

DISCLAIMER for those with interesting imaginations: These general observations assume 1) Relatively sticky dry tarmac is the substrate, and 2) these apply to a steady-state condition. These do not apply to 1) Racing on ice across a lake, or drag racing on ice, 2) Going into a 40 MPH corner at 100, hard braking, followed by lift and steer, etc etc...

The center diff f/r torque distribution will alter the oversteer/understeer balance of the car.

FWD - Full power understeer

RWD - Full power oversteer

AWD 50/50 - same as FWD, full power understeer

AWD 35/65 - or more RWD bias is critical to provide less understeer and more neutral full power balance

Combine AWD 50/50 with a limited slip rear diff and you will always end up in an understeer situation under full throttle.

Does it really matter for commuting? No.

- Steve

AH! thanks guys for the info. I posted in the transmission forum and got very little info.
http://forums.nasioc.com/forums/showthread.php?threadid=294792

Seems you get a better response if you post in news and rumors. STI's got everyones attention here:)

I read in some thread that the DCCD tranny can't take high horsepower applications, so if you are gonna mod a car to high power levels w/ the subaru 6MT, opt for the non DCCD one.

zzyzx, I can tell you haven't drive a wide variety of cars.

I've gotten power oversteer in FWD cars (hard but very very possible)

Power understeer is the default mode for lots of RWD vehicles.

Oh and a Rear LSD on a RWD/AWD car can make for a very very tail happy car. My bloody subaru wagon on ice swings the tail out to a 35-40 degree angle when launching hard in a STRAIGHT LINE. In the wet power oversteer is all you get. More to the point My old AWD (part time) Topaz with it's std. rear LSD could get oversteer due to the diff, even with the rear driveline NOT GETTING ANY ENGINE POWER! It would "sense" too much slip between the rear wheels in a tight corner and cause the inside to break grip causing the rear to step out a good distance.


Never make geral statements about cars, as they will allways have a notible exception. (in most cases)

Ok, then I'd better not comment unless I'm prepared to write a 50 page treatise on chassis dynamics... :rolleyes: :lol:

If you didn't get the hint from the terse description I gave above, I was generalizing... my bad. My general statements above are probably only good 85-90%.

Of course, I could have written up a 500 word "disclaimer" notice, which includes an "if on ice" section. :lol:

For those of you who want to learn this stuff beyond superficial generalizations, then buy a good book like Carroll Smith's "Tune to Win"...

BTW, as to the "default mode" of RWD cars, I've yet to find that toggle switch that says "Give me oversteer". ;)

- Steve

The "under power" balance of the car is so dependant on tires, the suspension setup and what you did right before you got on the gas the discussion of it as it relates torque split is pretty moot.

The DCCD is just another variable which can alter the balance of the car.

I expect with the DCCD set forward that (as the split is locked when you set the DCCD out of auto) the car performs better on snow and gravel (if you're a moderately experienced driver). With the stock center VC you can get a "see-sawing" action due to the VC response time (the automatics work faster, both versions).

If you were in a lower grip situation (with the wrong tires) I'd be tempted to use the auto mode though.

I can recall driving RSquire's 22B at Donington that with the DCCD set to fully rear the tail would come out a little under power through the jog into the main straight but the front wheels were still pulling enough to keep it going the right way.

(of course if you lift at the wrong time then no diff setting is going to stop you taking a little spin)

Another thing that will change how the car behaves under power is the front diff (e.g. open, Suretrack), etc.

There is no information on that released yet - interesting the EVO US has an open front diff.

Glenn

I don't know if you saw this, but in another thread Paul talked about DCCD and how good the system is:

2004 Subaru STi (http://forums.nasioc.com/forums/showthread.php?s=&threadid=289775&highlight=DCCD+auto+paul+hansen)

If you pan down to near the bottom there is a detailed description of the feel of DCCD. He mentions that "[he] would gladly pay up to 5k extra on nearly any turbo Subaru to have that system. It works wonders that [he] thought [he'd] have to buy an RWD for. "

It sounds to me that it would be very useful on normal roads-- RWD feel with electronic AWD drive to keep things under control.

Eric

This is kind of off-topic, but I always thought the US WRX would be better off with a front LSD than a rear LSD. On pavement, I never spin the rear tires, it's only the front inside tire that ever spins. As for the rear, either it's sliding or both wheels are gripping.

I can see why the rear LSD would help on gravel or wet roads, but seems a front LSD would be better for most real world (and autocross) situations.

Hopefully we will be seeing the front diff as well, but somehow I doubt it. We shall know tommorow.

Sorry - you are experiencing the twin effects of time zone difference (+16 hours depending on your location!) and recovering from the flu.

The DCCD in automatic setting gave the most transparent feel of AWD I've experienced yet, but I must hasten to add that my entire experience with it was on a track in rather extreme circumstances. Extrapolating that to road driving, I could probably say that for commuting, DCCD-A is unnecessary, but:

It did feel more responsive and, oddly enough, lighter than the standard AWD system. This may have been due to the large difference in engine power and delivery of such between new and older Subaru's.

The main thing to remember is that I'm heavily performance oriented towards mountain-road style driving. For that, the new STi/w DCCD-A is so far the best car I've encountered short of a heavily massaged Miata, and I have to say it felt more balanced than even that Miata, which is saying a hell of a lot.

The GT-R doesn't feel as good. While it theoretically is in RWD mode, the truth is that it's electronics interfere quite quickly, and in a rather obvious way. Then there's it's weight...

Originally posted by pio!pio!
I read in some thread that the DCCD tranny can't take high horsepower applications, so if you are gonna mod a car to high power levels w/ the subaru 6MT, opt for the non DCCD one.

Now, it comes in a car that's making 300+hp and 285ft/lb's of torque - that's a rather high-power application right there. As far as power levels far in excess of that, I have no comment -the car has only been out for a short period of time in Japan, I don't see how the tuners have tested it out thoroughly already.

Cheers,

Paul Hansen

Thanks everybody for defining some aspects of the DCCD.
Paul, hope you feel better..the flu just plain sucks! Thankstoo for your hands on details of the DCCD and the STi. Very well written article:) Its a pleasure to have you posting here.
I suppose we'll know tomorrow if we will see the DCCD on the STi.
Most of my driving is on tight twisty (50-60mph avg.) hilly/mountain roads so I was especially interested in the DCCD. Of course, the weather here in Ohio changes from sun/rain/sleet/snow all the time so AWD is so much fun and hopefully the DCCD will just add to that.

If its an option, I'll take it..thank you very much.

Cheers,

Davis

Originally posted by Turboy
USDM EVO VIII power distribution is 50/50.

Actually its more liek 60%/40%

Bzzzzt: 50/50.

http://mitsubishicars.com/MMSA/jsp/evo/tech.jsp

I've had a lot of questions about the DCCD, like how to use it, when to use what, what happens when you use it, etc. This thread from our friends in the UK helps a lot (http://www.scoobynet.co.uk/bbs/thread.asp?ThreadID=143750&Page=1), but still leaves me with some questions. Exactly what will happen (not mechanically, I understand that :p) if I switch it to full lock and drive around. Sure, I'll get shudder, noise, and some scruffing at low speeds, but what happens if I try a hairpin at a good clip? When would you all use the different degrees of lock (bearing in mind that I think it's infinitely variable, using the 6 position display as a reference)? There's a diff oil temp light, so under what condition could I make that warning light go on? Full lock and a day of driving, or what?

I'm just anxiously awaiting summer so I can get my grubby little hands on my STi :D

Until subaru gives out some sort of "The Complete Idiots Guide to DCCD" I'm completely lost. I know its an Active center diff and from what apex japan said it does have automatic settings. I'm only assuming it has a feature to operate as an open differential when the handbrake is pulled... otherwise they wouldn't have the handbrake suggestions in the short manual.

I can't wait until someone who really knows how to drive these cars sits down and starts describing the system in detail. Maybe a Karl Schielble or Ramman Lackeman (ok I just murdered that dude's name but they're both scca pro rally subaru drivers).

I have a feeling that after break-in its going to take me a month just to figure out what to do and when for performance driving on this thing.

-Tom

I'm sure johnfelstead can answer this if he runs across it.

Yeah, John wrote a lot of good informational stuff in the scoobynet post, but he didn't really get into situational driving and the what if's that I'm really curious about :)

The DCCD system on the new car is different from the DCCD on the older gray market STi's in the UK right now.

If set to say, 40% front and 60% rear, will it still transfer power between the front/rear wheels if they slip?
Will it transfer itself to 10/90% for example?

Thanks

Some of you may be helped by researching the Torquemada, it was a DCCD grown from the i-club members. Still in development.

The way I have read it in my research is that in diff free mode it's 35% front, 65% rear as it's normal open bias. Open diffs still can be designed to have a bias, and that's this one's natural bias. Then, by actuating the DCCD you are basically making it a tighter and tighter LSD, all the way up to full lock which would be 50% front, 50% rear mechanical coupling.

The diff never actually mechanically transfers power to ther rear wheels, that's just the diffs natural bias. That means that even in diff free mode, the car will behave more like a RWD car than a FWD one, and you can manually make it tighter.

At half the adjustment range, for example, you will be at around 50% lock for the LSD portion of the diff... This means that the fronts would have to unload WAY more than normal to make them lose traction. So it's more apt to stay with it's natural 35f/65r bias.

With the diff free, if the front tires were on ice and the rears on concrete it would spin the front tires. But all load being equal, it will put more power to the rear tires. And, by adjusting the DCCD, you will reduce this tendency...

It's a bit of a vague concept, but think of the above situation with the front tires on ice....
With diff free, the front wheels will spin...
With DCCD at let's say 10%, you will slowly pull forward off the ice patch, and as soon as the tires are off the ice, the power would go back to 35/65.
With DCCD at 75%, you would quickly move off the ice and resume natural bias.
With DCCD at full lock, you would shoot off the ice.

So the system basically controls how quickly the car can compensate for dissimilar grip or loads.

At least that's my understanding so far...
:)

-Jason Saini
www.over6racing.com www.kingmotorsports.com

your analogy re ice is wrong i am afraid.

when i get time i will write a users guide. We have a few months yet. :lol:

The DCCD is the same as the older versions except you have the option to run in auto mode, once you switch to manual its exactly the same. So all the auto is doing is choosing the lock ratio for you.

Full Lock: On an older STi, I actually preferred this over the 35/front 65/rear setting, because the system didn't try to route power on me. Power-on drifting of the edge-of-traction sort was more comfortable, and I could balance the car on the throttle easier.

35/65: the car has a more RWD bias, which the owner of the car preferred because he could kick the tail out more. Depending on your driving style, it can be a bit faster around the corners, or it could make you slower. Still fun either way.

New Sti:
DCCD-A: Less oversteer than the 35/65 setting, but rather adequete amounts. The slight understeer that is dialed in to the system actually made for faster speeds through the course, for there was higher levels of driver-error correction available. Rather stable for a very aggresively RWD-biased system.

Normal driving: Just leave it in automatic. Despite your initial tendencies to play with the new toy, the manual settings are best left for when you are doing some serious driving and have a style that works with a specific setting. Even in automatic mode, RWD-style drifting is only a throttle stab away.

Cheers,

Paul "I do know how to drive these cars!" Hansen

John, if you have a moment could you explain what would happen in each of the ice analogies? That's how I interpreted how things would work from reading the Torqmada thread over on i-club, among other things...

It's all a vague concept, it would be good to hear from someone with experience! :)

-Jason Saini
www.over6racing.com www.kingmotorsports.com

John F- WE look forward to that DCCD manual. (don't forget to copyright it:)) Us Yanks need it explained in normal non-engineered terms sometimes.
Basically, I think we are looking for an example like this:

In setting A: (whatever A is) the car has ??? power going to the front wheels and ??? power going to the rears. And what would be a good use for keeping the DDCD in setting A. PLease add when to use setting ??? on snow and raing and gravel and dry.

In setting B: (whatever B is) the car has ??? power going to the front........etc...etc.

Explained in each setting plus the real world usage of why you would use it in this setting. That would be great!!! Thanks!!

P. Hansen- Thanks for exaplaing the auto mode. I'm sure many of us will keep it in that mode. However, if we were to do some "serious driving" on the road (under the speed limit of course;)) we would like to understand why we may want to change it out of auto and into manual. Like I mentioned above, real world environments for these settings. Which one to use?

Many thanks to you guys.

Cheers,

Davis

Would Auto mode be the best, since the car can always move power to the most efficient wheels?
"From the wheels that slip to the wheels that grip"

We cant change the DCCD manually as quickly as the computer can, so it seems best to leave it to the computer, especially on loose surfaces.

For some reason I always forget about Paul Hansen on here... Apex Japan has given way too much good information over the last few months and thrown me into fits of jealousy over the Nissan Skyline GTR-N reviews... A car that I think is just my style... but I'll never drive one.

Thanks for some of the clarifications on the DCCD-A. But the question I have is what setting would be used more for rally style driving or autocross? Would locking down the diffs give you more stability under super hard hairpin manuvers or would less lock or even open be better to redistribute more power as traction neccessitates.

Also, which way allows for faster rotation of the car? What about the handbrake turns for those serious 15 foot radius ultra tight hairpins?

-Tom

PS - Paul, you da man

The auto mode is much more like the current car (wheels that slip to the wheels that grip).

The DCCD out of auto mode locks the torque split whereever you set it.

Glenn

Originally posted by DistantTea
FThanks for some of the clarifications on the DCCD-A. But the question I have is what setting would be used more for rally style driving or autocross? Would locking down the diffs give you more stability under super hard hairpin manuvers or would less lock or even open be better to redistribute more power as traction neccessitates.

Also, which way allows for faster rotation of the car? What about the handbrake turns for those serious 15 foot radius ultra tight hairpins?

You see, that's where theory starts getting filtered out, and reality seeps in. There is no perfect mode for conditions, because 5 drivers would be fastest in 5 different modes. In theory, you would want as much power to the front as possible for rallying, because understeer is encouraged in most rally setups. OTOH, autocross is about power-on drifts by most people, which would point to a setting locking as much power to the rear down as possible. Yet there are some people that are so smooth and have enough on-the-limit control that they'd never slide it like that.

Personally, I liked it best in automatic mode, at least partially because I didn't know the track that well (Fuji Speedway is a mongrel of a track) and because I quickly realized I didn't know this particular Subaru that well anymore. It has changed a lot since even the STi bugeye. If I was far more familiar with the car, I might prefer to switch to manual mode.

However, and this again is a personal preference, once I go to manual mode, I would find the setting that I'm more comfortable with and stick with it. It may be fun to play with when you are, well, playing around, but when you are at 10/10's, the last thing you want to do is change the way the car behaves to your reflexes. That's probably why I prefered my mates ver. 4 coupe with the DCCD set to lock - when cornering under power, I could literally feel out the state of grip at all 4 corners, which made it a snap for me to steer purely by throttle. The facelift STi's new DCCD gets quite close to that even in automatic mode. With manual set to the rearmost setting, it had ferocious amounts of throttle-on oversteer (especially for an AWD) that, once I got used to it, was very progressive and had none of the vagueness about it that, say, a WRX or Legacy B4 would have as the center diff starts moving power around.

And I'll admit that I'll probably never, ever use the handbrake. Even the tightest turns down out of Nikko (1st gear all the way down the 47 curves), I'm more than certain that the throttle will rotate it as much as needed. At present, I can only imagine what the 2.5-litre turbo will be like, but if it is like that, then rotation isn't going to be much of a problem. STi will probably have to dial in a bit more understeer to keep even the good drivers out of the weeds...

Cheers,

Paul Hansen

Originally posted by shirokuma

Now, it comes in a car that's making 300+hp and 285ft/lb's of torque - that's a rather high-power application right there. As far as power levels far in excess of that, I have no comment -the car has only been out for a short period of time in Japan, I don't see how the tuners have tested it out thoroughly already.

Cheers,

Paul Hansen

What is the rumor about DCCD being not suitable for very high-power applications based on? The only weakness that I can think of would be overheating of the rear differential *IF* torque is biased to the rear *AND* the car is raced. In that case the dash would light up telling you that you cooked the R180. A rear diff cooler would solve that problem. Prova uses one, and mounts the heat exchanger where the license plate should go.:)

I think the rumors are based on the "old" system. I need to find the description I have of the "new" system.

Originally posted by shirokuma


With manual set to the rearmost setting, it had ferocious amounts of throttle-on oversteer (especially for an AWD) that, once I got used to it, was very progressive and had none of the vagueness about it that, say, a WRX or Legacy B4 would have as the center diff starts moving power around.


So in this setting the car was not shifting torque anymore?
Is AUTO mode the only one where shifting occurs?
And again: What can cause a differential to overheat?

Good stuff :D

But what I'm still curious about is what happens when you lock the diff? Am I going to be driving a mad oversteer/understeer monster? If I throw it into a tight turn with full lock, will the car turn like it's on rails, powerslide like a stunt driver, or will it flip over and explode ejecting me from the underside with a cowboy hat on my head? I want to know how the car will react under different locks. I can imagine that the car will be somewhat like a mix between my RS and my Mustang GT in full open. It'll have better grip than the GT, but still have more power to the rear wheels, and I could probably kick it out if I tried. And if I set it to full lock, would I be able to kick out the rear at all, or would I be driving constantly at maximum traction the tires could provide (making me handle on rails), but always keeping me that much closer to the edge of powerslide territory?

(Ejection from underside of car with cowboy hat is not typical. Your results may vary)

Originally posted by Glenn Wallace
I need to find the description I have of the "new" system.
So where is it!
Find it now now now!!!! :alien:

I have read presumptions, in other threads, that the US STi DCCD relies on a hydraulically controlled multi-plate clutch. This may not be true. It's possible that US Patent Application US 2002/0042320 A1 (http://l2.espacenet.com/espacenet/viewer?PN=US2002042320&CY=ep&LG=en&DB=EPD), filed Sept 27, 2001, is a result of DCCD research and develpment. It represents an alternative to the multi-plate clutch that is less bulky and much simpler. I guess we won't know for sure until the US STi service manuals are available in 6 months.

Originally posted by Fido
But what I'm still curious about is what happens when you lock the diff? Am I going to be driving a mad oversteer/understeer monster? If I throw it into a tight turn with full lock, will the car turn like it's on rails, powerslide like a stunt driver, or will it flip over and explode ejecting me from the underside with a cowboy hat on my head? I want to know how the car will react under different locks.

None of the settings will change the handling in such a major way. They will effect it's handling at the limits in differing ways. Nobody is going to be able to claim that because it was in such-and-such setting, it became an uncontrollable beast.

You see, the problem here is I don't know how you drive -do you turn in early, right on time, late? How much throttle do you apply, how much braking where and when? All of those will affect the handling far more than the settings on the DCCD system will.

Oh, and a reply to something that John Felsted mentioned earlier. This DCCD system is not the same as the previous DCCD. Not only the automatic mode is different, but the way it behaves is much different also. It's much better. It's very quick to respond, and I would say that it was rather transparent in operation compared to the old DCCD system. Cost included, I would actually choose the new STi over a older type 4 or 5 GC8 at half the price, literally for the new DCCD-A. Well, ok, for the new motor and suspension setup also, but the DCCD was very impressive.

Cheers,

Paul Hansen

US STi DCCD relies on a hydraulically controlled multi-plate clutch

This is the 'old' system. Really I swear I had something on the new DCCD system vis a vis the VTD system in the auto.

I believe DCCD is just current VTD plus manual locking. DCCD-A = VTD in Auto WRX. The VTD system uses multi-plate clutch to control the locking of the planetery differential. DCCD adds the user control as how much you want to lock. Handling wise, a fully locked diff. makes car turn wider not tighter but handles loose surface better. On a tighter turn, you want diff. to open so that front and rear axles can travel at different rate but still be able to lock it when one end lose traction. So, it's really hard to say what setting on the DCCD makes car handle better. VTD or DCCD-A does a quite good job adjusting the center diff. automatically, IMHO. Maunal control in DCCD offers better control just like AT vs MT. Hence you still need change the locking contantly based on the driving considtion (road condition, turn or straight line etc.) So, asking which position in DCCD offers better handling is just like asking which gear in MT is better.

Does the car have any G sensors in it?

If so, could the DCCD-A (while in automatic mode) proactively (sp?) shift torque to the rear under acceleration? IIRC this is what the Porsche 959 did. Also, if G-sensors ARE present, then could yaw control be the next EVOlution of the Impreza? The EBD uses lateral G sensors, right?

Originally posted by shirokuma

None of the settings will change the handling in such a major way. They will effect it's handling at the limits in differing ways. Nobody is going to be able to claim that because it was in such-and-such setting, it became an uncontrollable beast.


Hi Paul and John Felstead,

Just wondering... i think what lots of the guys are asking is this: when set to a non-auto setting, does this just affects the diff's natural torque split before slippage is detected? i.e. will the DCCD, even in non-auto mode, re-route torque when slippage (front or rear shafts) is detected? Or will it keep on pouring out the torque in the driver selected split?

Sleepy "Huge fan of apexJapan and John's 'ring videos" Scoobie

Shirokuma - Thanks, you have a point. I was just curious as to what might I expect to feel under different settings, but I guess what I'll feel depends on how I drive :D

I guess the best way to answer these questions is to wait until I drive it and find a nice out of the way place and test the settings out. Mostly I'm sure I'll leave it at auto, but when I want to play I'd like to be able to change the settings and know a little of what to expect. Only time will tell :)

Originally posted by sleepy-scoobie


Just wondering... i think what lots of the guys are asking is this: when set to a non-auto setting, does this just affects the diff's natural torque split before slippage is detected? i.e. will the DCCD, even in non-auto mode, re-route torque when slippage (front or rear shafts) is detected? Or will it keep on pouring out the torque in the driver selected split?


Even a fully locked mechanical diff. can transfer torque when one axle is slipping. You don't need computer to do this. I think DCCD does no let directly control how torque is distributed. Rather it lets you select how much locking you want on the center diff. which in turn will change torque split indirectly.

OK - sorry to bring a reality check to the party, but DCCD is generally binnned for the very high power tuner cars.

<info>the Prova racing car does not have DCCD. The cooler in the rear bumper is for the rear diff, btw. There is a seperate tranny cooler up front. </info>

The Zero Sports car that does amazing laptimes at Tsukuba runs a centre viscous lsd. In fact most of the really hi-po tuner cars here are running the same'ole centre diff we see in the WRX. Some also use the STi centre diffs (info is available in many places on the net about them, so I won't bore you).

There is a point where the DCCD is over powered by whats going on around it, simple fact. You can also overheat the rear diff (hence the sensor on it) - and you can add a cooler pretty easily coz the diff cover has the requisite connector holes in place IIRC.

Now, I also know of places that can upgrade the DCCD for you power maniacs :D Like AVO Turboworld and their people in Japan... a bit far, but you could probably convince them to open in the US if lotsa people ask politely :)

B4. the holes to use for a rear diff cooler are there.... They are the fill and drain holes

OK - sorry to bring a reality check to the party, but DCCD is generally binnned for the very high power tuner cars.

As discussed, this is a new generation of DCCD. I don't think the tuners have had access to it yet.

Glenn

I was wondering whether DCCD-A uses a steering angle sensor as one of its inputs to decide when to do what. This is how ACD works on the latest EVOs and it a piece of beauty! The more you turn, the more loose car you get! Anyone has any insights about this and the DCCD-A?

Thanks

Fedja
__________________
95 AWD #34 SM
"Let boost be with you...
...and bring some timing too!"
http://www.geocities.com/esp_er

Originally posted by MrAWD
I was wondering whether DCCD-A uses a steering angle sensor as one of its inputs to decide when to do what. This is how ACD works on the latest EVOs and it a piece of beauty! The more you turn, the more loose car you get! Anyone has any insights about this and the DCCD-A?


The VTD in AT WRX already does this. There is no reason the more advanced DCCD doesn't.

Originally posted by Foxbat


The VTD in AT WRX already does this. There is no reason the more advanced DCCD doesn't.

Where would a steering angle sensor be located on the car?

Originally posted by Foxbat
The VTD in AT WRX already does this. There is no reason the more advanced DCCD doesn't.
So, is there a sensor or not? :confused:
It sounds kind of silly if ECU wouldn't know which way and how much the car is turning and adjust the settings in the DCCD at the same time, doesn't it?

Fedja
__________________
95 AWD #34 SM
"Let boost be with you...
...and bring some timing too!"
http://www.geocities.com/esp_er

Isn't the steering angle sensor one of the building blocks needed to have yaw control?

Originally posted by aspera
Isn't the steering angle sensor one of the building blocks needed to have yaw control?
I believe that is used by both: AYC and ACD


Fedja
__________________
95 AWD #34 SM
"Let boost be with you...
...and bring some timing too!"
http://www.geocities.com/esp_er

The VTD in AT WRX already does this. There is no reason the more advanced DCCD doesn't.

You are wrong, the VTD only does the torque distribution, you are thinking of the VDC that is not in the AT WRX, but in the Legacy VDC. You don't automatically get VDC with VTD.

Glenn

Nope. I was talking about steering sensor affects VTD locking behaviour not VDC. I'm glad Subaru didn't put VDC in WRX.

Holy crap I'm lost. There are waaay too many abbreviations for me to follow what's going on.

VTD equiped AT WRX uses 4EAT's TCU to collect data from G-sensor, ABS sensor, steering sensor as well as throttle input to determine how to best adjust the VTD torque split for best traction and handling. 5MT uses simple VC LSD and hence don't need/use all these sensors.

Originally posted by Foxbat
VTD equiped AT WRX uses 4EAT's TCU to collect data from G-sensor, ABS sensor, steering sensor as well as throttle input to determine how to best adjust the VTD torque split for best traction and handling. 5MT uses simple VC LSD and hence don't need/use all these sensors. Sorry, but there is no G-sensor and there's steering sensor on any USDM WRX. And the ABS wheel sensors are used only by the ABS. The TCM does recognize when the ABS is active and acts to eliminate the influence of engine braking.

Sorry. What you said is true for 5mt but ABS sensors are used by TCU to determine wheel sliping because that's the only way to determine wheel slipping in a 4EAT Subaru. And you can read the rest from Subaru's own web site for how VTD works.

Foxbat vs. Jon *and* Glenn:)

Sorry, but there is no G-sensor and there's steering sensor on any USDM WRX. And the ABS wheel sensors are used only by the ABS. The TCM does recognize when the ABS is active and acts to eliminate the influence of engine braking.

Jon, you are half right. There is a longitudinal G sensor on all WRX's as part of the ABS system. (book 4, page ABS-2)

Book 4 page AT-2 lists all the inputs to the TCM, no steering angle sensor.

Foxbat, please quote the manual page or URL which says the AT WRX (or VTD Legacy) has a steering angle sensor.

Again, I believe you are confused with VDC.

Glenn

What does the longi, longitu, ack!, longitudinal G sensor do exactly? Could it be getting confused when the WRX hits a sharp bump? Where is it and how is it mounted? This might be a mod in the making.

The longitudinal G sensor is there so that the ABS system won't get confused when all 4 wheels are on ice, i.e. no traction on any wheels.

The reason I say there is a steering 'sensor' (you don't really need a sensor) is because the TCU/TCM have to differentiate between wheel slip and turns so that it can quickly lock the diff. when wheel slips but not when car is turning. On a paper released by Subaru (linked on this site couple months ago), it explains in detail how VTD helps enhance car's turning performance by changing the degree of diff lock in different stages of turn.

Here is a snip:


Yaw velocity feedback control

The transfer torque according to the estimated is
modified according to a deviation between actual yaw
velocity and target yaw velocity. The target yaw velocity
is calculated from steering angle and vehicle speed. When
actual yaw velocity is smaller than the target yaw velocity
(understeer behavior is excessive), the transfer torque is
decreased to improve steering response. The other way,
the transfer torque is increase when actual yaw velocity is
large (oversteering behavior), so that stability of vehicle is
improved. It insure a driver against unexpected spin
motion of vehicle.

Control according to steering angle

To get smooth differential function of the center
differential, in order to reduce a running drag and to
improve a steer feeling in tight turn, the transfer torque is
decreased according to steering angle when the vehicle
speed is very low. However, the transfer torque is not
decreased below a value setting from estimated . It
keep the vehicle stable when the vehicle makes a standing
start with steer on slippery road.

Control for throttle-off in cornering

If the throttle is rapidly released in hard cornering or on
slippery road, the vehicle may show oversteering
behavior. In the new VTD control, when hard cornering is
observed from yaw velocity and vehicle speed, and an
throttle-off situation is detected, the transfer torque is
increased to retain sufficient stability.

Control during braking or ABS operates

When the center differential is locked, wheel speed of
each wheels can not be controlled independently. The
other hand, especially on icy road, vertical load of rear
wheel may be insufficient to recover it's wheel speed
quickly from nearly wheel lock situation. A optimum
distribution of a engine braking torque and inertia of
engine and transmission between front and rear wheels has
a effect of improving recovery of rear wheel speed. In the
new VTD control, when hard braking with ABS operates
or forecasted of the operation of ABS, the transfer torque
is set to suitable value to improve controllability of wheel speed for ABS.

Sorry I lost the original link. Only have a PDF file on my computer.

I'll say it again, there is no steering angle sensor on the WRX AT.

Send me the PDF if you can. I believe even SOA marketing had some confusion originally between VTD and VDC.

Glenn

Glen, I bet he has a pdf origionaly written FOR the Outback VDC as it uses the same tranny that ended up in the Auto WRX, just w/o the fancy stability and traction control junk on it.

Kha0S posted the original link to the VTD doc. Why don't you pm him see if he still keep the link around.

sorry, had a very busy week so not been able to keep up with this debate.

Can someone tell me, forgetting all the electronics, whether the physical diff unit inside the USDM STi is the same as that used in the previous generation of DCCD transmitions? Ie. an elctomagnetically modulated locking system.

If it is then it's going to behave in exactly the same way. The controller in the manual DCCD just varies the magnetic field based on potentiometer position, nothing special at all. If the diff unit is the same in the DCCD-A then the only diference will be an inteligent control of that eltromagnetic field, the mechanics will be the same.

If the DCCD-A has a physically diferent locking mechanism, then we are talking about two entirely diferent systems.

As i understand it, the DCCD and DCCD-A are the same physical diff units, but i could be wrong.

I dont know of anyone taking the DCCD out to run high power engines in their TypeR/RA's. No idea where that idea comes from?

Glenn, where are the pics of me being a loon in that field? The tyres werent exactly suitable. ;)

I'll try and come up with something a bit more how to drive DCCD in a while, but i could do with knowing for sure if the electromagnetic centre diff unit is still being used.

Originally posted by johnfelstead
Can someone tell me, forgetting all the electronics, whether the physical diff unit inside the USDM STi is the same as that used in the previous generation of DCCD transmitions? Ie. an elctomagnetically modulated locking system.I cannot say that the US STi's center differential is operated "the same as that used in the previous generation of DCCD transmitions." However, here in the US everyone does presume it relies on "an elctomagnetically modulated locking system." But few believe it's the same (or even similar to the) system you have on your old STi.

And that belief is based on what exactly?

The A bit of DCCD-A is just the computer system that is giving the input to the electromagnet, replacing the drivers brain, left index finger (right for you guys) moving the potentiometer.

If you can show me the two diff units being physically diferent, then yes, they may cause a diference in the cars handling. If you cant show me a physical diference, then the cars with react in exactly the same way.

I would be very surprised if your DCCD-A uses diferent hardware to the non Auto DCCD, because the basic information is identical as regards 65/35 torque split with a 0-100% locking ratio available. It's a well proven system. The Japanses spec STi8 has a DCCD-A too, the A is just the computerised enhancement to make the system easier to use.

Does anyone know for sure or is this belief just someone speculating?

So, I am still failing to see the answer to the question that started latest bunch of posts. Someone said that there might be the steering angle sensor in there, but used only by the ABS controller. Others are still saying that such a sensor doesn't exist...

You keep me confused, that is for sure!!! :)


Fedja
__________________
95 AWD #34 SM
"Let boost be with you...
...and bring some timing too!"
http://www.geocities.com/esp_er

John,

The DCCD-A diff is essentially the same as the good'ole DCCD in your RA V limited, with a tricky controller on the side.

DCCD-A AWD controller sensors/inputs: wheel speed sensors, G sensor, throttle sensor, ABS controller, brake switch, handbrake switch, and mode switch/manual dial.

As many know already, the AWD controller adjusts the diff's lock and torque split on the fly. Also, due to the extra stress placed on the drivetrain, the rear diff is a mechanical LSD (front remains Suretrac).

Originally posted by MrAWD
So, I am still failing to see the answer to the question that started latest bunch of posts. Someone said that there might be the steering angle sensor in there, but used only by the ABS controller. Others are still saying that such a sensor doesn't exist...

According the press information, which isn't going to be terribly technical, the 4wd controller takes in information from the brake switch, ABS controller, side brake switch, DCCD Mode switch, DCCD Manual settings, G sensor, throttle sensor and sensors on all 4 brakes. No mention of a steering angle sensor.

Cheers,

Paul Hansen

well in that case it's going to handle exactly the same way when you put the car in manual DCCD mode.

I would be interested in how this Auto mode is going to work, because i use the diff diferently depending on whether its wet or dry, the speed i am doing, the type of driving i want to do etc.

come on SOA, do yourself a favour, fly me over and give me a USDM STi to play with for a week on road and track. You know you want to. :D:lol:

John,

I think you'll recognise the little display and the dial ;)

http://www.subaru.co.jp/impreza/sedan/05/imgs/photo_506.jpg

Perhaps a more informative explanation:

In manual mode, the DCCD-A does theoretically perform the same as previous DCCD systems. However, the different rear mech. LSD and the rear suspension geometry changes provide more aggressive power delivery in the rear in conjunction with better toe control and tyre contact patch management. Less squat and reduced push are also features of the new car's setup. Personally, the system seemed to react faster while in manual mode, but that may be new car vs. old car syndrome.

Now, the DCCD automatic mode is certainly different, and quite good. It's pretty hard to control torque split and lock-up by hand while doing a corner at 10/10's. It will stop wheels spin at the front, helping keep a better line through the corners.

This is offered without the benefits of long-term testing (soon, maybe), but other than in snow and gymkhana, I don't think there is any benefits to using the manual mode on the DCCD system. In all likelyhood, you will be faster with it in automatic mode while enjoying a better feel to boot.

In my talks with STi, they've mentioned adjusting the setting changes to the DCCD system, and not specifically in regards to the automatic setting map, so it's likely that there are some adjustments to the system from the older versions. If I was a bit more technically glib, I would be able to help better, but my base mechanical knowledge is still around the weber carburator rebuild level! :)

Cheers,

Paul Hansen

There are lots of things i would love to do with my DCCD settings whilst driving. I looked into an active system a while ago for it, but there were issues with the number of inputs avaliable to make it really worthwhile. I think the DCCD-A will be a great peice of kit.

What would be even better would be an option to write your own diff maps. I can see that coming pretty quickly if someone manages to crack the controller coding.

One of the bigest improvements i could see from an active controler was under braking. For track use having the DCCD set to open gives a fantastic balance on high speed circuits in particular, but under very heavy braking you arent maximising the rear brakes. One of the things i would do is increase the diff lock to around 50% under heavy braking, that would give a better eficiency and more stability and then as the brake is released i would have the diff go open again.

I am sure the DCCD-A map will be doing this kind of thing, plus a lot more besides, it should be interesting. Sometimes with limit handling you dont want a computer controlling what the car is doing. It may have inbuilt max slip angle limits that stop the car being FUN, which is where the manual settings would come into their own.

pity it's still not white, it stands out more as something special that way. ;)
http://www.btinternet.com/~john.felstead/scoobynet/DCP03222.JPG

Originally posted by johnfelstead
One of the bigest improvements i could see from an active controler was under braking. For track use having the DCCD set to open gives a fantastic balance on high speed circuits in particular, but under very heavy braking you arent maximising the rear brakes. One of the things i would do is increase the diff lock to around 50% under heavy braking, that would give a better eficiency and more stability and then as the brake is released i would have the diff go open again.John, I believe your car has a mechanical brake proportioning valve with a fixed F/R ratio. Do you think that the US STi's EBD (Electonic Braking Distribution) might alter the list of things you want?

Not entirely sure what you mean? My car has no ABS and doesnt have an adjustable brake bias. If you mean has it got a valve in the rear system that reduces brake force under dive, i am not sure, but my car doesnt dive much anyway with it's uprated suspension. Using the diff lock under brakes helps make the rears share the braking loads more, as the front/rear axles are more coupled. This is one of the reasons why i drive with some diff lock in the wet on Motorways, it makes emergency braking much more eficient and helps prevent front wheel lock up. With an open diff its very easy to lock the fronts up in the wet with my 330mm AP 4 pots.

I dont like ABS on a car used on track, you lose a lot of brake feel with electronic systems getting in the way. I have no idea how the EBD system will feel so cant really comment on that yet.

Originally posted by johnfelstead
I am sure the DCCD-A map will be doing this kind of thing, plus a lot more besides, it should be interesting. Sometimes with limit handling you dont want a computer controlling what the car is doing. It may have inbuilt max slip angle limits that stop the car being FUN, which is where the manual settings would come into their own.

Don't worry - at least on the JDM version, they didn't dial any fun out of the system. I can say with a somewhat straight face that it will certainly let you go past the car's limits without interfering in the least. :D

It does calm the rather exuberent RWD tendencies of the Type C in comparison to the manual setting however, so drivers with a greater hoon index than I may enjoy it a bit more on the manual settings.

Cheers,

Paul Hansen

Are we gonna bring up just HOW sideways it can go now Paul? :eek:

hehe :devil:

Originally posted by anotherB4
Are we gonna bring up just HOW sideways it can go now Paul? :eek:

hehe :devil:

Oi!

BTW, you didn't get to see the hooligan in front of me at the chicane go completely sideways (farther than your pic, John!) and hold it there while "adjusting" his way through the 90 and then swinging it around for the next 90. The fun index is rather high on the new STi!

Cheers,

Paul Hansen

Originally posted by shirokuma
According the press information, which isn't going to be terribly technical, the 4wd controller takes in information from the brake switch, ABS controller, side brake switch, DCCD Mode switch, DCCD Manual settings, G sensor, throttle sensor and sensors on all 4 brakes. No mention of a steering angle sensor.
Thanks Paul!

So, after reading a lot of posts on this subject, it looks like that DCCD is basically a selector for the amount of the lock in the center differential. It can go from the full lock (which makes it 50:50 between the front and the rear wheels) to full open (which is 36:64 split between the front and the rear wheels).

There is no such a thing as the steering sensing opening/closing of the center diff and you actually get whatever you have set the dial to. If all of the above is correct, than EVOs ACD diff is more advanced system and helps driving a bit further. I guess that is why Prodrive is bringing in that new ATD system which was mentioned on the ScoobyNet (http://www.scoobynet.co.uk/bbs/thread.asp?threadid=167283). Of course we have a couple of threads in this forum too, so check the first one (http://forums.nasioc.com/forums/showthread.php?s=&threadid=303357) or the second one (http://forums.nasioc.com/forums/showthread.php?s=&threadid=303114) as you wish (this is really just to make US2JDM happy :))

Things are definitely getting better for the US car enthusiasts!!! Keep coming!!



Fedja
__________________
95 AWD #34 SM
"Let boost be with you...
...and bring some timing too!"
http://www.geocities.com/esp_er

Originally posted by MrAWD
So, after reading a lot of posts on this subject, it looks like that DCCD is basically a selector for the amount of the lock in the center differential. It can go from the full lock (which makes it 50:50 between the front and the rear wheels) to full open (which is 36:64 split between the front and the rear wheels).

I think you have that slightly wrong. With the centre diff fully open you have a 65/35 torque split with no direct limited slip coupling between front and rear axles. What this means is if you get the rear wheels spinning, all torque gets transfered to the rear making it RWD. This is exactly the same effect as having an open diff in a FWD car where one wheel is spinning on ice, all drive goes to the spinning wheel and your stuck there until you slow the spinning wheel enough to allow some torque back to the stationary wheel.

Now what you get if you run the diff locked is a solid coupling between front and rear axles. This means that you still have a 65/35 torque split, but you wont ever get the situation of the front and rear axles spinning at diferent rates. Locking the centre diff doesnt give you 50/50 split, it gives you the mechanical ratio as set by the front/rear axle coupling.

Having the centre diff set somewhere between 0% lock and 100% lock gives you a certain amount of available slip between front and rear axles, this % will alter the handling balance.

This is why the japanese manual shows a torque split of 65/35 and a sliding scale of 0-100% lock and 50/50 is not mentioned in there at all. This takes a bit of thinking about but thats the way it works i believe.

Aha... so I had it backwards in my earlier ice analogy. :) I think I'm finally starting to understand DCCD. Thanks John! :D

So can you run an STi on a rear-wheel-dyno then without damaging anything ala the Nissan GTR?

-Jason Saini
www.over6racing.com www.kingmotorsports.com

John:

THX for a nice explanation. But to make my picture whole can you tell me how torque is shifted when slippage occurs in this case? Cause it looks like the bias is set firm.

John F- can you expand more on the questions I have below. I think it will may answer many of the questions from us DCCD noobs:)
Cheers,
Davis


Originally posted by Davis K Powers


Basically, I think we are looking for an example like this:

In setting A: (on the DCCD HUD) (whatever A is) the car has ??? power going to the front wheels and ??? power going to the rears. And what would be a good use for keeping the DDCD in setting A. PLease add when to use setting ??? on snow and raing and gravel and dry.

In setting B: (whatever B is) the car has ??? power going to the front........etc...etc.

Explained in each setting plus the real world usage of why you would use it in this setting. That would be great!!! Thanks!!

Many thanks to you guys.

Cheers,

Davis

Excuse me John, I'll just give an example of the way the system works (right off the DCCD-A sample map supplied in the PR materials).

Please check this link, the bracketed "(letters)" match the points on the visual about a third of the way down the page. http://www.subaru.co.jp/impreza/sedan/05/index.html

On the way into a corner (A), full braking and ABS doing its thing, the torque split is 65r:35f with the diff open (to permit the ABS to do it's thing). During turn-in (B) the sensors tell the diff what it should do and it gradually locks the diff until at the clipping point (C) (where max G is achieved, lock-up is about 50%) - when it senses the accelerator being opened and G's dropping, it steeply increases lock-up until the exit where lock-up is full and the torque split is set to 50:50 (D).

It sounds wierd, but I can promise that the result is slingshot performance out of corners without the push that is usually associated with full throttle corner exits in std. 50:50 centre diff cars. And yes, as Paul mentioned, tire shredding drifts are eminently possible.

edit to add: there is a visual on the Japanese site that ties in, so added points to match it.

Thanks Jon! It is getting a bit more clear now! :)

But, I do have another question...

You where saying that even with the full lock ratio is still 35/65 between the front and the rear. If things are at the full lock and car is moving straight, some of the wheels would have to be spinning at the different speeds (or spinning) or something would have to brake. Help me here, please?

Thanks


Fedja
__________________
95 AWD #34 SM
"Let boost be with you...
...and bring some timing too!"
http://www.geocities.com/esp_er

I think in full lock there is no way to have any other distribution then 50:50. This is how this whole thing work. When you change the amount of lock at the DCCD it transfers extra torque to front wheels from 35/65 to 50/50.
I do not see it possible any other way .

A fully locked center diff can have torque split vary from 0/100 to 100/0 depends on which end is slipping.

Originally posted by RafalW
I think in full lock there is no way to have any other distribution then 50:50. This is how this whole thing work. When you change the amount of lock at the DCCD it transfers extra torque to front wheels from 35/65 to 50/50.
I do not see it possible any other way .
Same here, but John motioned up there that you still get 35:65. Here is the clip that is talking about that.
Originally posted by johnfelstead
Now what you get if you run the diff locked is a solid coupling between front and rear axles. This means that you still have a 65/35 torque split, but you wont ever get the situation of the front and rear axles spinning at diferent rates. Locking the centre diff doesnt give you 50/50 split, it gives you the mechanical ratio as set by the front/rear axle coupling.
That is why I was asking the extra question! :)

Originally posted by Foxbat
A fully locked center diff can have torque split vary from 0/100 to 100/0 depends on which end is slipping.
Well, that the part that I have issue with! If something is fully locked, how it can go differently on two different ends? Are you saying that at the full lock, rear wheels will be spinning faster then the front ones then?

Sorry, but still confused! :)


Fedja
__________________
95 AWD #34 SM
"Let boost be with you...
...and bring some timing too!"
http://www.geocities.com/esp_er

I like the clip. It shows how DCCD in auto mode might be beneficial in corners.
So my guess is:
Auto mode - dry pavement use
manual lock - slippery and loose surfaces
Is it right?

No: auto mode - varying conditions
manual mode - constant conditions (e.g. dry pavement, deep snow, etc)

Some of you guys are still talking about the functionality of the old system.

With the new system, when you go to manual mode, the torque split is fixed at whatever the thumbwheel is set to.

Glenn

oh gawd, people are getting confused about this big time. I dont have time for this now, off to Monte Carlo in a few hours, but when i get back i will write a better explanation.

Glenn, you are getting mixed up too. The centre diff LSD doesnt alter the torque split, it alters the locking % ratio of the diff, two diferent things entirely. Torque split is based on the ratio of the front and rear output drives, not the amount of LSD lock you have. The LSD alters the amount of slip that is allowed front/rear but the ratio of torque distribution wont change unless you get wheels spinning, when the wheels start to spin the torque naturally goes to the point of least resistance and goes outside your 65/35 split anyway, such as 100% RWD in a full power slide with an open centre diff.

Just think about the escort cosworth, that has a 65/35 torque split on the road car. In order to get a 50/50 torque split such as we have on the Group A rally car i run, you have to change the gears/chain in the transfer casing to change the torque split to allow a 50/50 split. You then put a higher rated NM centre diff in to alter the LSD effect (you can do both independantly), just the same as increasing the centre diff lock.

Torque split is all to do with front/rear transfer ratios, it has nothing to do with how stiff you make the centre diff lsd. You can have a 65/35, 50/50 or whatever split you like with no LSD at all. Think about it.

I know this will be tough stuff for some to understand, and it needs more time to explain, but you should get this after thinking about the cossie reference glenn?

OK. col de turini is calling. :D

so basically most of us are thinking that when hearing the terms 50:50 and 35:65 in manual mode, that we are locking the dif at these ratios.

instead, the normal torque split is 35:65 and can increase to 50:50 as the dif starts to lock and couple the two ends of the car. what the dccd controls is a set locking in a sense.

here's my question. if the setting on the dccd is on its lowest, does the dif start to lock automatically and the dccd setting being a PRE-lock so to speak, or does it stay at that setting between open and locked? i'm guessing the latter as otherwise it would be controlling it much like the automatic mode would.

The centre diff LSD doesnt alter the torque split, it alters the locking % ratio of the diff, two diferent things entirely. Torque split is based on the ratio of the front and rear output drives, not the amount of LSD lock you have.

I understand what you are saying, and I believe you are accurately describing the "old" system.

I think new system works differently somehow, perhaps because of the VTD components in it.

Originally posted by johnfelstead
The centre diff LSD doesnt alter the torque split, it alters the locking % ratio of the diff, two diferent things entirely. Torque split is based on the ratio of the front and rear output drives, not the amount of LSD lock you have. The LSD alters the amount of slip that is allowed front/rear but the ratio of torque distribution wont change unless you get wheels spinning, when the wheels start to spin the torque naturally goes to the point of least resistance and goes outside your 65/35 split anyway, such as 100% RWD in a full power slide with an open centre diff.
So, it looks like that torque split is a result of the transfer ratio between the front and the rear wheels. Different kind of gearing will change this from 50:50 to what ever the ratio is.

So when the power is applied, more torque will be applied to the lower gearing side and spin those a bit sooner than the other one.

What is confuses me here is that different transfer ratios will create difference in the speed between the two ends. In order to make wheels to go at the same pace, there will be some slip in the differential. If you have a regular viscous coupler LSD in there, it would have to slip all the time and possible wear out eventually. I guess, type of the coupling that is in DCCD is not affected by this and it can last long enough to be useful.

Now, regardless of the amount of locking in the center diff, torque will be transferred the same way. The only thing that I am not so sure about is when diff is fully locked than one set of wheels will be spinning at the different speed from the other one, which will create some strange feel on the steering wheel and whole drivetrain. Kind of like driving in the corner with the locked diff, and inside wheel is jumping over due to improper speed forced by the locked diff.

Fedja
__________________
95 AWD #34 SM
"Let boost be with you...
...and bring some timing too!"
http://www.geocities.com/esp_er

One more thing about this torque bias part. If it is true that torque bias is coming from the different transfer ratios (which makes sense), then the speed of the front wheels will be different from the speed for the rear wheels.

So, in order to send more torque to the rear wheels, they would have to have lower ratio, which will make them to spin slower than they would at 50:50. But, when the car is in the turn (especially turning that is typical for the AWD vehicles, when the rear end is hanging further out) rear wheels will be crossing larger radius than the front ones. Since they are already spinning slower than the front ones, now the different between the speeds that car is trying to make on each corner is even bigger...

Well, I know that this is confusing and that I am confused my self, but I am just trying to find the light at the end of the tunnel! :)

Later

Fedja
__________________
95 AWD #34 SM
"Let boost be with you...
...and bring some timing too!"
http://www.geocities.com/esp_er

Yep!!!... it's offical.......IM LOST!:eek:
I was understanding ( at least I thought I was ). Then you all started talking in squares.:confused:

Glenn, show me the mechanical diferences in the diff units and i'll concede it will work diferently to the "old" system. If you cant show me that the hardware inside the gearbox has changed, then it will behave exactly the same.

The car never gets a 50/50 torque split, it is always 65R/35F but the allowed slip between front and rear varies depending on how locked up the centre diff is.

As an example, i can jack my front wheels off the ground and have the rears on the ground, say changing a wheel. With the diff open ie 0% lock i can spin the front wheels as though the car was RWD. If i then go and stick the thing into 100% diff lock, the fronts wont spin. That is because to spin the front the rears have to move too because they are locked together. The ratio is controlled by the front/rear transfer coupling ratio, not how locked the diff is though.

I can see why people will get lost on this subject, it's not easy to get your head round and to be perfectly honest, it's all acedemic bollox anyway, understanding this can help a little in the driving, but feel matters more than understanding the mathematics behind how it works. Hopefully i can write something that explains the feel and real world effects, rather than blinding people with science that is hard to explain and hard to understand. I'll give it a go anyway, watch this space. ;)

I understand exactly what you are saying; the diff guts are indeed different. I am not 100% sure, but I believe the combination of the new computer and the new diff guts is closer to what I describe than the old system (automatic/manual controlled "torque split" rather than variable "lockup" on the same split).

I've sent you a pm. :)

John F- yes we look forward to an "real world" explanation. Thanks again for sharing your thoughts thoughout this enire post.

PLease give us some examples on why you change the DCCD on the road, track, weather conditions, etc..

Cheers,

Davis

Originally posted by johnfelstead
Glenn, show me the mechanical diferences in the diff units and i'll concede it will work diferently to the "old" system. If you cant show me that the hardware inside the gearbox has changed, then it will behave exactly the same.

The car never gets a 50/50 torque split, it is always 65R/35F but the allowed slip between front and rear varies depending on how locked up the centre diff is.

As an example, i can jack my front wheels off the ground and have the rears on the ground, say changing a wheel. With the diff open ie 0% lock i can spin the front wheels as though the car was RWD. If i then go and stick the thing into 100% diff lock, the fronts wont spin. That is because to spin the front the rears have to move too because they are locked together. The ratio is controlled by the front/rear transfer coupling ratio, not how locked the diff is though.

I can see why people will get lost on this subject, it's not easy to get your head round and to be perfectly honest, it's all acedemic bollox anyway, understanding this can help a little in the driving, but feel matters more than understanding the mathematics behind how it works. Hopefully i can write something that explains the feel and real world effects, rather than blinding people with science that is hard to explain and hard to understand. I'll give it a go anyway, watch this space. ;)

OK, I'm bringing up the DCCD thread again. (shoot me if necessary:))

John F. if you would, please writeup a real-world explanation, it would be great. (your scientific reasoning sounded fine but I think it went over our heads..at least mine.)

I suppose what I am asking is this:

In what circumstances would you change the DCCD from auto setting? What effect does it give in setting a, b, c etc..?
What setting is best for each road condition? dry tarmac? gravel? rain? snow?

Many thanks!

Davis

Originally posted by Davis K Powers


OK, I'm bringing up the DCCD thread again. (shoot me if necessary:))

John F. if you would, please writeup a real-world explanation, it would be great. (your scientific reasoning sounded fine but I think it went over our heads..at least mine.)

I suppose what I am asking is this:

In what circumstances would you change the DCCD from auto setting? What effect does it give in setting a, b, c etc..?
What setting is best for each road condition? dry tarmac? gravel? rain? snow?

Many thanks!

Davis

In my never ending quest to fully understand DCCD I've come across this thread.....let's finish it now that we've had our STi for a while and come up with some real-world examples of proper DCCD manual mode use.

moveit!

Originally posted by MrAWD
One more thing about this torque bias part. If it is true that torque bias is coming from the different transfer ratios (which makes sense), then the speed of the front wheels will be different from the speed for the rear wheels.

So, in order to send more torque to the rear wheels, they would have to have lower ratio, which will make them to spin slower than they would at 50:50. But, when the car is in the turn (especially turning that is typical for the AWD vehicles, when the rear end is hanging further out) rear wheels will be crossing larger radius than the front ones. Since they are already spinning slower than the front ones, now the different between the speeds that car is trying to make on each corner is even bigger...

Well, I know that this is confusing and that I am confused my self, but I am just trying to find the light at the end of the tunnel! :)

Later

Fedja
__________________
95 AWD #34 SM
"Let boost be with you...
...and bring some timing too!"
http://www.geocities.com/esp_er

think of it this way:
In a rear wheel drive car, the car is being pushed from the back, the front wheels get no power. but going in a straight line, all 4 wheels are still spinning at the same speed.

the difference in speed between right and left wheels is due to the front and rear diffs, not the center.

So, is the conclusion 35f/65r the fixed torque split?
John? Glenn?

Do a search for "DCCD newbie question" or something like that. It will tell you everything you need to know, isn't that long and is easy to understand.

I thought this is *the* thread. :confused: (Search is how I found this thread to begin with).

does anyone know any info on a jdm sti ra v2? whats the most to the rear...