[torsen] Torsen and Haldex type differentials - UPDATE HAC

QSHIPQ at aol.com QSHIPQ at aol.com
Mon Nov 18 11:29:39 EST 2002

Dave et. al:
Let's get a baseline of where the torsen/haldex center differential 
technology has come to in the last couple years.  Haldex type came on the 
scene late in the game and has done amazingly well in a short couple years, 
in racing and production car/truck application.  Zexel Torsen has meanwhile 
come to the states in a big way (Buffalo NY based) and hasn't sat on it's 
laurals in any stretch of the imagination.  They aren't going away, just 
getting smarter with teaching old spiders new tricks.....

Comments inserted
In a message dated 11/17/02 3:07:02 AM Central Standard Time, 
Dave.Eaton at clear.net.nz writes:

>the audi torsen can't handle high torque applications?  it handles the rs6,
>the largest torque application audi has, ditto the w12 in the old a8...

The T2 torsen is different than the T1 torsen found in Jim A's 1991 200tq.  
The T1 (referred to as the  "old university special" = M021-DHU) is no longer 
produced by Zexel Torsen.  The new T2 was adopted to the audi line starting 
with the A4 in 1995, Audi <all> by 1997.  The T1-M021 was pretty well known 
(especially by the RX7t boys) to be good for 400hp, no more.  There is no 
listing on the torque limits of the T2 (I believe it's actually lower), but 
let's assume it to be the same as the T1.     The changes Zexel put in the 
basket design shows that the limit of torque to the new T1 University Special 
(012000) has been raised to 5000nm (*.737 =) 3685lb/ft of torque.  Wow you 
say?  Hold on for the fine print, that's AFTER the final drive, so you have 
the gear ratio and the final drive to factor.  

For instance, let's take the RS6 450hp * 3.57 (1st) * 3.197 (FD)  = 
5135lb/ft.  What might be more concerning is those darn tuners.  Say Herr 
Meyer puts 550hp (stage 1?) engine package:  550 * 2.2 (2nd) * 3.197 = 
3868lb/ft (6277 in 1st)

>haldex?  this is the name of a company, not a differential.  haldex makes a
>range of (mainly truck-based) equipment - brakes, abs systems, hydraulic
>systems etc.

I prefer to call the Haldex for it's design, not it's marketing.  It IS an 
electronic LSC by defintion.  Acquiesing: I'll refer to "haldex type" (see 
Keith's listing of specifics to the manufacturers marketing terms).  To be 
fair, I should also refer to Quaifes as 'Torsen type' differentials, which 
I've pretty much done all along.  

>haldex also makes an awd differential, called the lsc.  interested to
>understand how audi could have installed it in "full-time awd" mode?

Well, actually it is now, by definition.  95% of the torque is supported by 
the front wheels, electronically.  Electronically, audi could have chosen any 
torque ratio f/r up to lockup (50f/50r).  This is strictly a choice by audi, 
the engineers of the device claim it could  have been something other than 

>certainly the jeep application you mention is the same lsc unit, in the same
>application as audi use.  nvg started using their haldex lsc-based unit for
>jeep at about the same time as it debuted in vwag applications.  you can't
>call it full-time awd in one, and not the other - because it isn't.

See Keith's comments here, no sense in going thru them again.  To me, 
"synchro" isn't full time awd in any stretch of the imagination. 

>there is no doubt that an electronically controlled lock-up clutch (which is
>what the haldex lsc is), it the future of awd systems - cheap, and fully
>controllable.  "torque levels virtually limitless" makes no sense.  the
>haldex lsc has an overload protection valve - but is not operational at that

At which point it allows slip, that's different than Trg = 0.  

Now, let's take a look where everything has gone in the last 6 years, it sure 
appears the marketing boys have seized some opportunities, and have changed 
the engineering boys perspectives some.

Torsen T1/T2 technology allows more torque, which means they were able to 
keep audi in line for 1 more car.  This T2 technology isn't without some 
sacrifice to the BR.  The really wide T1 BR (I always suspected as part of 
it's inherent design problem) has been modified to a much lower BR in the 
later audis (1995>).  In fact, torsen recommends that the T1 be used above 
2.5BR.  Since we know audi is using the T2, I'd suspect strongly that the T2 
torsens don't have the wide BR of the old T1 in the pre 95 cars.

What's really more exciting in my opinion is the T3 technology.  This is 
Torsen's attempt at a Torsen limited slip differential lock.  Actually 2 
differentials, 1 a fixed split (say 65r/35f), then torsen intervention when 
slip occurs up to 30% shift.  This technology has the best chance of 
competing with haldex type couplings.  But, given our car engineers (Keith 
too:) directives/propensity to have electronic control, Haldex type will be a 
tough nut to crack.

WRT Brett's comments:  A gear wears, period.  Shims wear too.  The longer 
it's in service the more it wears.  Even Zexel will only guarantee a TBR as 
it is delivered from the factory.  They TELL you it reduces from what's been 
delivered over time.  Yes, I have pulled dozens of torsens, no I don't pull 
them apart and check them for wear, I don't know what to check for, tho I do 
know what fails.  That said, I did try shimming one for S&G, not an easy job, 
and when you look at bronze bearings (just like in a turbo), my mind  doesn't 
at all think "lifetime". Now virtually impossible, since you can no longer 
order old T1(MO21) service parts readily, tho you can get some of the thrust 
washers from FAG.   A k24 "can" last a real long time if stock, so can a 
torsen, is it operating the same as it did with 5kmiles on it?  I don't think 
so at all.   I doubt Zexel would even tell you that.

If you look at the new T1/T2 technology, it addresses some of the T1 
identified shortcomings, including noise, backlash, basket structure and 
torque handling limits.  I don't read "bulletproof" into any of those 
concepts wrt the original T1 pre95 audi applications.  I know the originals 
break, I've seen broken ones.  My interest in them is casual, more 
specifically, causal, as in chassis handling outcome.   I believe there is 
better than either the old T1, the new T1/T2, and haven't seen the T3 yet.  
The Haldex hasn't really shown it's potential in any audi... Yet.  

What's happening in center box applications is exciting, even Zexel is moving 
forward.  My own thinking is that the increases in torque handling by all 
these devices is going to dictate that lower BR will be more commonplace in 
centerbox awd applications and, if this restricted design parameter 
incidentally keeps a understeering quattro understeering all the time, I'm 
happy for it.  Until then I remain...

hopeful and well scarred.

Back into my basket, er, web.

Scott Justusson
Locker dude


More information about the Torsen mailing list