Skip to main content

Limited Slip Differential Types Compared

BMW M2 equipped with an eLSD - BMW ©

A few weeks ago, I posted about traditional clutch-type limited slip diffs (LSD's) and how they work. You can read about those in the previous post: How Limited Slip Diffs Make You Faster on Track. But as you might know or have learned from reading the article, they aren't without their faults, which means engineers are always working to get around those limitations.

You may not be surprised to learn that something like the Ferrari 488 GTB doesn't use a traditional limited slip diff, but it's not limited to super cars, far from it. Cars like the Golf GTI, the Civic Type R, various Mustangs, Corvettes, and BMW M cars, and even the Lexus RC F and GS F, all avoid a traditional limited slip diff in favour of one of these technologies. To keep things simple, I'll focus on two wheel drive vehicles. The vast (vast) majority of principles apply to all and 4 wheel drive vehicles, but there are some subtle differences that I'll cover in a future post.

Gear-type Limited Slip Diffs

Exploded view of Torsen T2-type differential - JTEKT ©

The name is a bit misleading because the vast majority of differentials use gears (yes, some differentials don't. More on that in a future post on AWD/4WD diffs). But gear-type limited slip diffs refer to a differential that uses a mechanism of gears to provide lock-up instead of clutches or viscous fluids as discussed in the previous article. If you've ever heard of Torsen or Quaife diffs, that's what they do. They are probably the two most popular manufacturers of these types of differentials, but there are others.

I won't get too much into how they internally work (stay tuned for a future post on that!) but what I'll focus on instead is how they affect the car differently compared to traditional LSD's. The main difference is how they lock. Instead of using a set of clutch packs for instance to lock or partially lock the differential, they use binding. Most of those differentials utilize helical gears which - inherent to their design - transfer part of the torque as a thrust force that pushes the gears outwards. By doing so, the gears progressively bind with the diff case and each other, in effect locking the differential.

Since the force generated is proportional to the force (torque) being transferred by the gears, lock up is proportional to input power (i.e. how much power you're applying). That's why they are called Torsen (Tor for Torque and Sen for sensing) because they "sense" torque and lock up in response. If you're off the power, it's basically an open differential because there are no forces to bind the gears and lock the diff. As you roll into the power, it progressively and smoothly locks up.

Quaife ATB (Automatic Torque Biasing) front LSD - Quaife ©

The benefit to that is, because the diff is more open off power, you can typically get away with higher torque bias ratios than non gear LSD's without seeing as much of the side effects of higher locking (resistance to turning and understeer as discussed in the previous post). The higher TBR allows better traction performance as discussed in the last post.

Moreover, the fact that lock up is proportional to input power means the diff locks up as you send more power, without the need for slip. Non-gear LSD's need slip to work as intended. If you are going around a turn with no inside wheel slip, a traditional LSD diff is waiting for slip. The same is true for driving in low traction conditions like snow or ice.

Only after one wheel begins to slip does a traditional LSD begin to do something you want. This is also the cause of the potential for "fish-tailing" discussed with traditional LSD's (and absent modern stability control systems) because it responds to slip and can shuffle power back and forth between two sides.

Mustang Shelby GT350R at Atlantic Motorsport Park - Graham MacNeil ©

With a gear type differential, both wheels are limited before they slip. The more power you apply, the more the diff (progressively) locks, preventing excessive slip from happening in the first place because both wheels are locked to a degree. And because they are more forgiving of a higher TBR, cars can better put power down, accelerate faster out of turns, and generally perform better on track. Examples of cars that use this type of diff include the Civic Type R, Toyota 86/Subaru BRZ, various Mustangs (Boss 302, Perf. Pack, Shelby GT350/GT350R, and the upcoming Shelby GT500), Focus RS (front differential, final edition), WRX STI's (front and rear diffs), and some Audi Quattro models.

Brake-based Differential Lock

A car that uses brake-based limit slip action utilizes an open differential just like described in the first diff post Why an Open Diff Doesn't Work on Track. But the car attempts to solve the shortcomings by applying the brakes to individual wheels. If you roll into the power and one wheel spins, the car's computer realizes that and applies the brakes at the spinning wheel.

In high performance driving, this solves the two shortcomings of an open differential: loss of grip due to a spinning wheel and under utilizing good grip at the loaded, outside tire. As we've established, an open diff transfers equal torque to both wheels. When you're coming out of a corner, there is plenty of weight transfer, meaning the outside wheel is loaded and has higher grip. It can transfer more power. In order to distribute torque where you want it (unevenly; more torque going to the outside loaded wheel), the brakes are engaged at the inside wheel. That slows down the wheel that doesn't have as much grip which would be spinning excessively because the power you are sending exceeds the available grip.

3rd gen Ford Focus ST at ASCC autocross - Kevin Doubleday ©

In other words, the car is artificially creating resistance at the low traction wheel. The diff is still transferring equal torque to both wheels, but the brakes increase the torque holding capacity of the inside wheel, and the diff as a whole as a result, thereby allowing the diff to transfer more torque overall. Think of it as brakes "gripping" the wheel and tire instead of the tire gripping the road. Half of that torque goes to the outside wheel where it can all be used. The downside is wasting some power simply turning the low traction wheel against the brakes. The second problem is, as a result of trying to power the low traction wheel against the brakes, the brakes can overheat and prematurely wear.

You'll hear a lot of owners and reviewers complain about their effectiveness, or lack thereof. The problem is the application, not the tech. In a Focus ST or a Golf GTI (non PP), you don't have liberally sized brakes, brake cooling ducts, brake system (thermal) capacity, etc. In reality, an optimized brake based setup can work very well. McLaren uses them, for example. You won't hear too many people complain about their performance.

McLaren 720S - Graham MacNeil ©

In a future post, I'll explain in more detail why they don't work very well most of the time (in everyday cars). But even for humble cars, the benefit is that the tech is extremely flexible because it provides complete uncoupling and independence between the two wheels when no lock up is needed and infinitely variable and adjustable bias ratio when you do. You have non of the shortcomings of mechanical LSD's and all the benefits of limiting slip and sending power where it can be used. If you can get around the heat and overuse of brakes, then it's a very capable technology. Examples of cars using this technology are the Focus or Fiesta ST, Golf GTI (non PP), Subaru WRX's, Jeep's, and, well, McLaren's...

Electronic LSD's (eLSD's)

eLSD is often used to refer to electronic limited slip differentials. They attempt to combine the benefits of all the above without any of the downsides (what a concept!). It is essentially a clutch-type limited slip differential as the one discussed in the previous post. Only here, the clutch packs are controlled and engaged electronically. You only get lockup when you want. This solves most of the shortcomings.

Eaton UltraPosi Electronic Limited Slip Differential - Eaton ©

It doesn't have the same resistance to speed differentiation unless activated so it can have a higher lock-up with no downside (understeer). It is electronically controlled so it can selectively lock and unlock as needed based on real time calculations and inputs from various sensors. It doesn't have to worry about brakes overheating. It doesn't get confused by road conditions like a traditional LSD.

It could also preemptively lock prior to slip happening to prevent slip based on the car anticipating the need for lock, such as rolling into the power in low traction conditions or exiting a corner. It can give you high diff lock when you need, say, exiting a slow narrow turn, and low or no lock in corner entry, reducing the understeer you'd get with a high lock/bias ratio.

2016 BMW M4 at Atlantic Motorsport Park - Graham MacNeil ©

In summary, it gives you higher traction performance while eliminating most of the downsides of higher traction performance since it selectively locks only as needed. Its only downside, really, is complexity due to the electronic controls and actuators, resulting in having more parts to go wrong. Examples of cars using this technology are the Mk7 Golf GTI PP, Camaro SS 1LE, Corvette Z06/ZR1, Porsche GT3 and GT2 RS, Ferrari 488 GTB, and BMW M cars.

Torque Vectoring Differential

A torque vectoring differential is very similar to an eLSD. The main difference is that eLSD's only transfer torque from slower to faster. An eLSD is a basically a typical LSD, say a clutch-type, where the clutches, and therefore amount of lockup, are electronically controlled instead of passively and mechanically activated based on the difference in speed between the two wheels on an axle. That means that they can control when and how much to lock up, but after that, the same principles apply and torque is biased from the faster spinning wheel to the slower wheel. On the other hand, a torque vectoring diff can transfer torque in either direction and can do so independently of a speed differential.

Lexus Torque Vectoring Differential (TVD) - Lexus ©

The benefit is better control (from the car) and wider range of adjustability, allowing the car to correct and/or improve more frequently and in more situations. Moreover, a torque vectoring differential doesn't transfer torque by locking up. Instead, most utilize actuators (clutches and/or motors) and planetary gear sets to overdrive or under-drive each wheel independently. The best way to simplify them is to think of them as having two tiny gear boxes on each half-shaft/wheel. Changing gearing in each gear box changes how much torque is sent to each wheel, thereby independently controlling how much torque (and therefore power) each wheel gets.

What does this mean? It means you have no resistance to speed differentiation (i.e. understeer) but as much torque transfer to either wheel as you want (can be up to 100% of torque sent to the diff). The downside of lockup (understeer) is eliminated, a huge plus to start with. Then you get the best torque bias you could get (basically infinite if designed for 100% bias), allowing you to utilize every last bit of traction available, and to top it off, you can have individual wheel torque control to help the car handle.

Lexus RC F - Lexus ©

For example, if the back end is starting to slide, you could vector torque to the inside to bring it in. This has the same effect as a stability control system applying individual braking to a wheel or more to bring the car back in shape but braking means scrubbing off speed. Torque vectoring doesn't. You could also transfer torque to the outside wheel to help the car rotate if the car is understeering.

It's a technically wonderful system. The range of adjustments is huge. Opportunities to improve the handling are plentiful. It can make a huge difference in how the car handles. The only downsides are added complexity, weight, and a less "mechanical" feel. Examples of cars using this type of diff are the Lexus RC F, Lexus GS F, late-model Mitsubishi Evo's (rear diff), and a number of Audi S and RS models (rear diffs).

Which one is best?

Well, brake-based systems are hugely flexible and simple, but because of the heat issue and power wasted, I'd only take one if I had to choose between one of those and an open diff. Otherwise, my preference is the gear type LSDs, even though they aren't as capable as electronically controlled ones.

My 2012 Mustang Boss 302 at Atlantic Motorsport Park - Jenn Earle ©

Their only downside, really, is that they don't work if you lift a wheel completely off the ground, but some of them have preload springs to overcome that short coming, which is very rarely a problem to start with. Otherwise, gear type LSD's provide the best performance without going to an electronically controlled system, in my opinion. They understeer less compared to a clutch-type with similar lock up. They also don't have clutches to replace so they are also lower maintenance than clutch types. Compared to electronically controlled LSDs, they're more natural feeling, simpler, and cheaper to maintain. They demand more of you, expect less of the car.

But the same can be said for better tires, better dampers, stiffer chassis, etc. that could help make up for skill level, so I can see people arguing both ways. I would imagine that someone learning on and sticking to torque vectoring technology on track would see it as natural and consider it a baseline; anything else is a compromise. That wasn't the case for me. I didn't drive cars with torque vectoring technology until a few years ago, so they will always seem like the deviation from norm - the cool tech that manipulates the car to your advantage, not the bare essentials.

With that said, in a FWD or FWD-based AWD car, I would absolutely and unquestionably give in and take torque vectoring or eLSD's. In a heart beat. The drivetrain layout is working very hard against you to prevent the car from doing what you want it to do. The tech does its best to mitigate the limitations. The way I see it: it almost undoes the crime that is FWD-based, transverse engine layout as opposed to improve the text book longitudinal front engine, RWD layout.

Follow Rams Eye The Track Guy on Facebook and Instagram!


  1. Very thorough article. I think clutch type LSD has one big advantage, that it could change the lock up rate by adjusting the angle and tightness of the clutch packs, which is crucial for finetuning that last bit for a specific car + track application. Torsen LSDs are set and done, no adjustment possible.


Post a Comment

Does An Aftermarket Grille Really Increase Airflow?
I put a Saleen S281 grille to the test to answer that question.

Stock Suspension S197 Mustang With Square 305/30/19's
What you need to fit a proper size square tire setup.

How Limited Slip Diffs Make You Faster on Track
What you need to know about how they put power down and pros and cons.

Can Telemetry Explain Schumacher's Talent?
A comparison between Schumacher's and then team mate Herbert's data.

Cayman GT4 Track Review
The first Cayman with proper (911-challenging) power.

Is an EcoBoost Mustang any good on Track?
Two days at the track in a Mustang short 4 cylinders.

2016 BMW M4 DCT Track Review
It's quick (properly quick). But is it fun?

Can a stock Golf Diesel handle a Track Day?
Not your every day track beater.

🔥 Most Visited This Week

Falken Azenis RT615k+ Street and Track Review

Last year, I picked up a 2009 Lancer Ralliart to do a long term test with it as a dual duty track/daily. One of the first things I knew I was going to do was put a decent set of tires on it. The car came without OEM wheels which was actually good because I didn't have to hesitate about getting a good set of aftermarket wheels to support going wider. Thankfully, my friends at YST Auto Halifax  set me up with a great set of Superspeed RF03RR wheels. The Wheels I had never even heard of Superspeed but I trusted the good folk at YST Auto who mentioned some customer cars running on track with them. These wheels are rotary forged which is basically a prerequisite to be taken seriously in this market populated by companies like TSW and Fast Wheels. The wheels looked like a high quality, well finished wheel and each had a "QC" check sticker on. Just for appearances? Maybe, but I found no defects. The wheels seemed easy to balance (didn't need many weights) and at 18.1 lb. f

Michelin Pilot Super Sports vs Firestone Firehawk Indy 500 - Street Review

I've been a huge fan of Michelin PSS tires and exclusively bought them for the Mustang over the last four years. So how did I end up here? This year, I was hugely interested in trying an "R-comp" tire. I had my eyes set on Bridgestone Potenza RE-71R's for two simple reasons: price and reputation. Although not a true "R-comp" tire on paper, it performs like one by the account of every single test and review I've read (down to wear rates...). They seem like they're easily the most affordable (from a big brand) R-comp tire and combine that with a reputation for having tons of grip, it was an easy top contender. I had my concerns, though. For one, I'm told and have read that they are an autox tire, not really designed for high speed, pressure, and temps associated with open track. For another, the Mustang is a heavy car (as far as track cars are concerned) being roughly 3,800 lb. (including driver), which will amplify the unwanted open track load

2004 Audi TT 3.2 Quattro DSG Track Review

Before getting into this, I have to confess something... I had never driven an Audi TT before. Not until this one, anyway. But that hasn't stopped me from forming an opinion about it from the comforts of my own couch while reading and watching reviews online. After all, if you've never done that, do you even know what the point of the internet is? Now, we all interpret reviews differently. Call it confirmation bias if you will, but if you like a car, you'll read a review and look at the positives as what makes the car great and the negatives are but a few quibbles you have to live with. If you don't like a car, the positives are a few things the manufacturer got right while screwing up everything else. It's a bit harsh to put the TT in the latter category, but that's where it ended up for me... I never took the TT seriously. The problem with the TT for me isn't that it's a Golf underneath, per se. There is nothing wrong with a performance car sharing a

Michelin Pilot Sport Cup 2's vs Bridgestone Potenza RE-71R's

I never thought I'd ever run Michelin Pilot Sport Cup 2's on my 2012 Boss 302. The cost is astronomical and they are supposed to last the least of anything comparable. So how did I end up with (nearly) fresh Sport Cup 2's? A complete fluke. I came across a lightly used set with only a few hundred miles and no track time; 305/30/19 takeoffs from a GT Performance Pack Level 2 (GT PPL2). I knew my 71R's were getting very worn before the season started and likely wouldn't last the whole season, even this short one. The price was far better than a new set of RE-71R's, a little more than half, and local Time Attack rules (Canadian Automobile Sport Clubs) recently made 180 and 200 TW tires equivalent, meaning no PAX or PIP point penalty for going with 180 TW tire like the Pilot Sport Cup 2's. I have been very curious about how PSC2's compare to RE 71R's but I stayed away due to their being painfully expensive and, up to last year, their 180 TW rating would