Hi Simes and others.

Over 600 posts and the first time somebody has said they are pi**ed off with me and I wasn't even giving out naff advice, oh well. I don't mind any one disagreeing with me in a normal way, honest.

Mike said the drive was biased 10% front and yes that would mean a 20% diffential if it was 60/40 so I'm still a little confused with that. Every one you ask has a different opinion but it doe's work out at roughly 55/45 if you look at the final drive ratio's

The 5 door has a different front diff arrangement and maybe the same ratio as the rear diff giving a 50/50 split. I also believe you can insert a fuse on earlier 5 doors (or is it auto's) that disables the front drive for servicing etc. There is even a warning light on the dash.

Perhaps Scoobyboy can help us out with this one? Not sarcasm just a question mate.

Lee

 

I believe RA's Type R's and 22B all have mechanical rear LSD's, that's why they are more aggressive.
22B has front LSD, type R doesn't (RA might have)?

(Blanket statement now awaiting individual caveats)

Mike.

 

Mike,

Anyone who is prepared to call you either doesn't know your background or doesn't know how big you are! LOL

Thanks for the input into MickerG's simple question.

Lee

 

Actually all turbo charged imprezas have a rear lsd. Its a ramped lsd, not a plated diff. Much less complicated, but less durable too. Effective all the same.

Scoobyboy and Chris P would not have known the internals of the rear diff, as it is not a rebuildable item (by Subaru dealers, at least) and is scheduled for replacement, if it fails (at approx £1900!!).

The viscous coupling is in the centre diff and on R, RA and 22B this is progressively lockable. It is locked by a solenoid driven piston that pressurises the silicon fluid in the unit. As silicon heats up, or pressurises, its becomes more viscous and will form a solid if enough energy is there.

The 99 and 00 RA has a front plated lsd.

 

So then,

All Impreza Turbo's have LSD of differing sorts centre and rear, thanks Pete.

No problem with the Subaru guys not knowing its just they came on here abusing those of us who were correct. Some of us do have a clue after all.

Pete Croney Sir,
Final problem still to be put to bed, can you explain the torque split?

Lee

 

Thank you Professor Croney for the first decent explanation I've seen of the adjustable centre diff's operation. All other comments refer to it wearing out if not left in the 'open' position on dry roads. I always assumed it was electromechanical, relying on some sort of friction plate.

If it uses a viscous fluid like most centre diffs, why can't it be left in one of the interim settings? After all, that's how the normal viscous diff operates. Does the viscous fluid overheat, and if so why?

Sorry to be a pain, but I think this device is one of the best features of the Type R as it allows you to vary the car's grip and handling by a considerable extent on wet roads. Is it really so harmful to use it in this way? It seems strange that Subaru would fit a unit which will be knackered if used normally!

 

Thank you Professor Croney for the first decent explanation I've seen of the adjustable centre diff's operation. All other comments refer to it wearing out if not left in the 'open' position on dry roads. I always assumed it was electromechanical, relying on some sort of friction plate.

If it uses a viscous fluid like most centre diffs, why can't it be left in one of the interim settings? After all, that's how the normal viscous diff operates. Does the viscous fluid overheat, and if so why?

Sorry to be a pain, but I think this device is one of the best features of the Type R as it allows you to vary the car's grip and handling by a considerable extent on wet roads. Is it really so harmful to use it in this way? It seems strange that Subaru would fit a unit which will be knackered if used normally!

 

Hmmm...this is interesting.
About a year ago, I was considering fitting an LSD to my car, after talking to someone that had done it to his UK car with great results. This indicated to me that the car obviously didn't have it as standard.
Mike's posts backs this up, stating that UK cars have viscous thingies in the centre and rear diffs. Therefore, I presume that these and mechanical LSD's are different things, although VC's may well be a tpe of LSD?
Then Pete says that ALL Impreza Turbo's have LSD's, and that they're ramped lsd's not plated diffs!!!!
Arghhhhhhhh.....does anyone actually have any idea what the **** they're on about?

Stef.

 

Stef,

Already been said that the viscous LSD is a cheaper but effective device. The draw backs being its not instant acting and not very progresive (see earlier posts).

I wonder if the person you spoke of had a Mechanical Torsen type LSD fitted.

A far more efficient (and expensive) 4WD layout is the one involving a TorSen ( which stands for TORque SENsing) differential. This extraordinary device, invented by the American Gleasman (patent pending, 1958) and manufactured by the Gleason corporation, is based on the non-reversibility of worm gears and worm wheels (i.e. when you turn the worm wheel the worm gear turns but not vice versa). The TorSen is the only mechanism which acts like a differential and locking device at the same time. It has the advantage of being fully mechanical which guarantees its instantaneous response and progressiveness. Its main advantages therefore resume to:

Instantaneous response

The linear character of its locking to speed difference curve (smoothness)

No locking or speed difference inhibition under braking (it acts only when power is applied to it)

Integrates a "free" differential and a locking device in one part

Its compactness, the TorSen has only 8 moving parts

No wear as opposed to more traditional self-locking differentials based on friction plates



The Torsen differential
A: Differential housing
B: Out axle
C: Worm wheel
D: Worm gears
E: Synchromeshes
F: Hypoid wheel (from engine)
G: Out axle



There are mainly two drawbacks in Torsen differentials:

They are expensive devices

They tend to generate more heat under heavy use than an open differential

Due to its inherent design if one wheel is completely off the ground or completely loses traction the TorSen will act as an open differential

Is extremely difficult to assemble. Although it only holds 8 moving parts there is only one way to fit them together

The TorSen splits torque in a 50:50 proportion in no-slip conditions and can manage slips up to 20:80 ratios between the wheels it drives. The main examples of TorSen differential applications are:

The Lancia Integrale

The Toyota Celica GT4

The Audi Quattro Turbo (the earliest series used a manually locking rear differential and a TorSen center differential)

The Honda Integra Type R (front wheel drive)

The most important difference between TorSen differentials and viscous couplers is that the TorSen has a torque sensing characteristic while the VC has a rotation sensing characteristic. That's why TorSen differentials only lock when power is applied to them whereas viscous couplers lock both when power is applied and while braking.

Lee

 

A limited slip differential is a limited slip differential, whether it is a mechanical or viscous-coupling device. They all do the same basic thing, transfer torque from the axle that is slipping to the axle that isn't, hence they limit the slip. They don't <I>have</I> to lock solid, usually this is undesirable except in more extreme racing cars.

The best place to start on differentials is at

 

Great link (howstuffworks) DavidRB, thanks for that !

 

Phew !! thats was enlightening (and entertaining ) stuff.

Thats more like it...I was beginning to think the BBS had gone soft, what with posts like 'Anyone going skiing in Scotland' that appeared in the General Questions a few days ago

Den.

 

Hi Guys,

Glad to see you've sorted thing out...

I'm not sure that all 22b's have front LSDs, I thought that was an option you could go for when you ordered the beast, is it was with STI's, R's and RA's IIRC.

Ages ago I decided that the main advantage those EVO 5 and 6's have over scoobies in the handling stakes was their helical style mechanical front LSDs (similar to TorSen or Quaife ATB). I fitted a Quaife ATB LSD to the front of my STI Type RA VLimited and am happy to report that I was pretty much right on the ball.

Not everyone's cup of tea, but it was amusing to be told that it was a dangerous thing to fit to a car by someone who had one in his EVO and didn't know it... ho hum!

Moray

PS: Scoobyboy, Can you confirm you work for a subaru main dealer? I'd like to know, because I want to keep my dangerous grey import serviced safely...

 

MorayMackenzie,

The Scooby mechanics have been a little quite.
Asked him before which dealer he works at, no reply.

Lee

[This message has been edited by logiclee (edited 20 February 2001).]

 

Hi

we still haven't got to the bottom of what the Torque split is on UK cars.Don't people want to put this issue to bed once and for all.There has been disagreement on this for a couple of years now.

Here are some points that don't make sense.

Mike Wood said that the front and rear final drives have a different ratio of 3.9 and 3.54, and that 3.9 was the effective ratio.
If you fitted a 3.9 and a 3.54 to a car equipped with a centre diff then the effective ratio would be in the middle which is 3.72:1.
It would also mean the center diff and viscous coupling would have to absorb this difference in rotational speed and would turn once for every 3 turns of the road wheels.
This would surely bugger the viscous coupling in no time.Can anybody explain this?

Also if a car has a torque split other than 50:50 achieved through for example an epicyclic gear train then it would show up on the test I did on my car and it did not(see earlier post).

I could sort this out once and for all if I had the following information.

1)What is the design of the centre diff,is it a bevel gear diff or an epicyclic type?If its a bevel gear diff are the gears the same size for example?

2)how many teeth are on the crownwheel and pinion on the front final drive located in the front gearbox?.

3) in my earlier post I decided there must be a pair of transfer gears that transfer the drive from the centre diff to the propshaft.Scoobyboy has sort of confirmed they do exist.Can anybody else? If they do exist how many teeth are on each of them?.(I think these gears account for the different final drive ratios)

And finally does anybody have any diagrams/pictures etc they could send me.

Logiclee you seem to have a very good knowledge can you help?

Pete C and Mike T you must have seen loads of boxes in pieces can you help?

Mike Wood is there anybody at Prodrive that could give us a better answer?

I am going to get to the bottom of this one way or another.

Andy

 

Lee.
A bit more than I needed to know, but thanx!

Stef.

 

Stef,

No Problem LOL

AndyMc,

I too can't see the viscous centre diff being used to absorb the rotational differences of the different ratio front and rear diff.
It has already been said that an intermediate gear is used to output drive from the centre diff to the front wheels, this could be used to bring the gearing back into synch but would mean a 50/50 split.

I can't think of any other way than some sort of mechanical type LSD splitting torque which we know the Scoob doesn't have, or an epycyclic gear set which would have shown up in your test.

PLEASE CAN SOMEONE CLEAR THIS UP?

[This message has been edited by logiclee (edited 21 February 2001).]

 

in answer to morays question i do work at a dealer ship but i'm not going to say what one because we will be inundated with time wasters asking silly questions for no reasons .
p.s we do not service grey imports because we only deal with people who don't lower the quota for cars shipped over from japan.

 

Be afraid.....be very afraid

Seems there are Subaru dealers whose technicians don't know all that they should

Oh, and I may be wrong, but I wouldn't have thought that import quotas were affected by "grey" or personal imports

D

 

I'll try again.

The reason the 'turning the wheel and another one turns at the same speed' test doesn't work is that the gearing is effectively identical to each wheel. If it wasn't there would always be slip through the centre diff which would create enough heat to lock it solid.

Some other 4wd cars with vc diffs are sensitive enough (Calibra Turbo is a shining example) that you are advised to change all 4 tyres at once if you need to replace 1 as the transmission cannot cope with the difference in diameter between new and part worn tyres.
The Subaru transmission is much stronger and the centre diff is able to cope with this situation without too much of a problem.

As the UK Turbo drives the front pinion at a different speed to the rear by using a 1 to 1.1 step off, the final drive ratios are also different by the same ratio ie 3.54:3.9
This has the effect of giving more torque to the front than the rear which balances the static weight bias to the front, whilst retaining identical wheel speeds.

With regard to differentials, all UK Turbos have vcs centre and rear, a front lsd is an option on some Jap spec cars and is fitted to our Group N cars. The Type R has a plate rear diff which can be quite agressive in it's operation, hence the dramtic sideways when it runs out of traction.

Regards

Mike

 

Thanks Mike

 

right then diablo uk quotas are affected by imports because the uk can only import so many cars from japan per annum this is why subaru decided to bring in the legacy from the states where there is no restrictions on how many can be bought in. this then gave subaru the chance to bring in more turbos to reduce the waiting list.
so every grey import is 1 less uk spec car that can be bought over

questioned answered.

 

Hi Mike

Thanks very much for that.I think your post has confirmed some working out I did the other day.I think I might know the number of teeth on the various gears.
Yesterday I put my car on the ramp at my mate's works and we confirmed that the rear final drive is 3.54.

I did a bit of working out and came up with the following numbers

If the front final drive has 39 teeth on the crown wheel and 10 teeth on the pinion this would give the already accepted ratio of 3.9:1

If the rear final drive also had 39 teeth on the crownwheel (the same as the front,I suspected it might be the same to reduce production costs etc ) but 11 teeth on the pinion this would give the other accepted ratio of 3.5454:1.Very convenient

We had already established that the drive to the rear passes through a pair of transfer gears and as I have said previously, I think it's these which convert the rear ratio to the same as the front ie 3.9:1.

When I worked out what the ratio of these gears must be I came up with 1.1:1,which is the exact same number you stated in your post.

It also happens that two gears,one with 22 teeth and the other with 20 teeth have the perfect ratio of 1.1:1.

Bingo it all fits.How very convenient

If this is correct it means

-When driving in a straight line the center diff has no motion to absorb other than that caused by uneven tyre wear, which will prevent unnecessary wear and drag.

-The effective final drive ratio would be what we already knew it was, 3.9:1

-It also means that my test would produce the results that it did.

If any of this is not correct I think you'll agree it's one hell of a coincidence that it works out so perfectly.

Now its time for me to be a pain in the a**e again.
You have said there is normally no motion in the center diff due to front and rear final drives having the same effective ratio ie front 3.9:1.Rear 3.5454 * 1.1 =3.9:1

This means if the center diff is not turning it will be having no effect on the car and it may as well be a solid shaft!.It must also mean the torque will be split according to the effective final drive ratios which you have also said are both the same.Given all this I don't see how the split can be anything other than 50:50.
Do you see what I mean?

I think we are very close to putting this issue to bed once and for all.Thank F**k

Andy



[This message has been edited by AndyMc (edited 22 February 2001).]

 

Hhmmmm, confused again now

Andy, I see where your coming from.

Mike? Help

Lee

 

Ah but...

torque is related to rpm. Two shafts, driven by the same motor, but rotating at different speeds will give a different torque reading.

This is why a rolling road has to be callibrated for each car. The operator will hold the car at, say, 3000 rpm and then initialise the RR computer. It then works out the gearing, by measuring the wheel speed for this rpm, and calculates a constant to apply to the torque measured.

The higher the speed of rotation, the higher the torque.

Now what interests me is this... on a rolling road, are the two rollers attached? If not, we have established that the cars will show different torque at either end and different models of car will vary this split, so how do we get a reading?

If they are connected, it would negate torque split and show a single figure, which is an average.

Does anyone know?

ps I remember Phil Gardner taking his car to a rolling road and being horified that the rear wheels were a blur and the fronts were stationary.

 

quote (from Mike Wood)
The Type R has a plate rear diff which can be quite agressive in it's operation, hence the dramtic sideways when it runs out of traction.
unquote

Got anything to say about this then Mr Croney nudge nudge

 

The capital of France is Paris

 

Pete,

Agreed that two shafts driven by the same motor with different rpm will not have the same torque, but Andy is saying that the effective gearing on the front and rear wheels is the same with no rotational difference in the centre diff.

This would mean the same torque is applied to the front and rear wheels, hence a 50/50 torque split.

Help, I can't sleep at night!!

Lee

 

Dear Scoobyboy, your right! Sarcasm is 'the lowest form of wit.' Unfortunately for you, it's also 'the highest form of intelligence.'

Shut up.

 

grow up this has been laid to rest