21 replies to this topic

[Dave Wright]

I have searched the internet but not come up with a clear explanation of how this type of limited slip differential works. Can anyone here explain? In particular, how does it compare to a Salisbury LSD? Does it only lock when one wheel begins to spin? Does it lock completely? Does it lock in the power-off condition (i.e. when braking)? Thanks in advance for any information.

[ray b]

Dave you may have more luck posting this in TNF
last time I asked about ramp angles, and clutch packs
and was told they are computer controled now!

[Bladrian]

Hi Dave

The cam-and-pawl type of ls diff is a sixties dinosaur. It basically used concentric cams with wedges (or pawls) between them that locked them when power was applied, so they were either fully locked or open - no adjustment possible. When they wore sufficiently they just stopped working entirely, leaving you with an open diff. It was basically a dead end as far as limited slip differential designs go.

[Ray Bell]

Dead end or not, its progressive action was favoured by many...

The alternatives at the time were Detroit Lockers or series production clutch-type units... or what we in Australia called "CIG Lockers" CIG being our largest supplier of welding equipment...

While on the subject of this 'sixties dinosaur' I might as well ask how far back the ZF LSD goes. I know that Stan Jones used some kind of LSD (out of a Studebaker truck) that dated back to the twenties... and it gave him a defining edge in this race...

[Dave Wright]

Thanks Bladrian and Ray for your answers.

Bladrian's description fits well with one I found on the net; as follows

In the '60s. Hewland began shipping many of their larger gearboxes with a differential they called the “Cam and Pawl" It uses inner and outer concentric fluted cam with pawls that wedge between them when power is applied. These fully lock the differential and they are non-adjustable. Since all of the parts that limit slip are made of hardened steel, the wear rate is relatively high and it takes constant maintenance to keep them working correctly. When slightly worn, they lock both wheels together, but then slip a tooth or two, causing a series of jerks. When severely worn, they act like open diffs. Many vintage cars that use Hewland gearboxes have extensively worn Cam and Pawl differentials. In other words, open diffs.

If the differential fully locks under power, perhaps the term "limited slip" does not really apply! Also, if the diff is either locked or unlocked, how does it have the progressive action described by Ray?

[Ray Bell]

I guess that's how come they wear out?

Certainly, in the time I was using one, everyone spoke of the sudden locking of the 'Detroit Locker' being a contrast to the smoothness of the ZF... I only ever had the latter, and that not for long, as the rules precluded it soon after I bought my car.

[Dave Wright]

Ray - I also found this description on the net which also says the dif is "relatively gentle in operation"

AP Racing. AP’s ‘Suretrac’ LSD is a relatively recent addition, mainly targeting the road user. The design uses cams and pawls. Consequently, despite AP’s claims that it’s a ‘novel’ and ‘unique’ concept, the design and ‘modus operandi’ is acutely similar to the established Jack Knight pawl-type, just a little more refined. The cams have a series of ‘lumps’ on them (1). The pawls have ‘lobes’ on each end (2), and are positioned horizontally between the cams, carried by fluted internal and external drums (3). When running in a straight line, no movement of the pawls takes place as there’s even traction on both wheels. Once a difference in traction occurs, the pawls slide in the flutes, engaging the cams. The amount of slip allowed is controlled by the frictional force developed between the drum flutes, pawl lobes and cam lumps, the resultant torque bias being applied to the axle with most traction. Pro’s - very ‘gentle’ in operation. Con’s - Comes as a sealed unit with ‘factory’ devised settings, so isn’t user tunable or serviceable. Won’t work with one wheel off ground - result, no drive.

Jack Knight - Their pawl diff has been around a long time, developed from the original ‘ZF’ type. Design and operation very similar to the AP Suretrac except the cams are in the drums, the pawls held radially in a cage and have rounder ‘lobes’. Pro’s - relatively gentle in operation. Serviceable by user. Con’s - Amount of slippage is a designed-in parameter so isn’t tunable. Design creates a ‘ratchetty’ feel when working. High wear rate when used in most Minis/Metros, so regular maintenance needed to maintain performance, increasing running costs. Won’t work with one wheel off ground - result, no drive. Usage - As AP Suretrac. Bill says - “Not a contender. It’s twitchy and darty, personal doubts as to whether it was designed for front wheel drive. I’ve driven rear-wheel-drive racers with ‘em fitted with no notable problems

[Ben]

The following is quoted from Carroll Smith's "The Racing Differential" supplement from Racecar Engineering circa 1994. It is also used in his book "Drive to Win";


The Cam and Pawl diff consists of concentric inner and outer cams. The inner cam is splined to the left-side drive acle, while the other floats in the housing and is splined to the right drive axle. A series of pawls is captured between the cams and located by a cage (think rollers in a roller bearing - ben). The cage is integral with the housing and therefore the ring gear and drives the cams via the pawls.

Under normal, straight-line conditions, the inner and outer cams rotate at the same speed, the pawls wedge between the inner and outer cams, and the diff is inactive. In turn conditions under low torque, the pawls slip between the cams and allow differential action - the effect is the same as an open diff. Under heavy torque loads, lateral load transfer or in conditions of reduced traction, when one wheel starts to overspeed, the cam splined to that axle increaes speed - partially wedging the pawls between the inner and outer cams, and limiting the amount of slip that can take place.

The pawls actually 'ratchet' between the cams, allowing some slip. The amount of allowed slip can be varied by the design of the cams, but only within narrow limits. The best that can be hoped for is a bias ratio of about 3:1 (75% lock) when everything is new. Note that the device acts to limit slip only after the slip has begun.

The cam and pawl has a lot going for it;

It is light
Low moment of rotating inertia
Its on and off characteristics are PROGRESSIVE and BENIGN, making it easy to drive
No instability or tyre drag understeer on turn in
Soft and progresive locking action

There are no surprises lurking in the cam and pawl - this is why F1 teams who cannot afford much testing run them (this was written in 1994).

There are disadvantages. For example, 75% lock is the maximum. The rear of the car may tend to feel a little unstable under heavy torque loads, especially on a bumpy track.

The major disadvantage lies in the suicidal nature of the device. The mechanical design ensures that the cam and pawl is inherently self-destructive. When the thing is working, the hardened steel pawls ratchet between the hardened steel cams, ensuring a high rate of wear and a constantly decreasing amount of limiting. The more the parts wear, the more inside wheelspin develops - and the more the parts wear. It is the automotive equivilent of the graveyard spiral. In the wet, the wear rate approaches infinity.

Life can be prolonged somewhat by some of the proprietary dry-film lubricating processes and the better molybdenum disulphide additive greases, but parts replacement is constant and expensive. Frequent disassembly, cleaning and repacking the moly is a must - the grease tends to centrifuge out.

Ben

[Wolf]

I could look up some pre-war limited slip diffs and post about them, if anyone's interested...

[Ben]

That would be interesting. The cam and pawl was itself fitted to the Auto Union GP cars in 1934 by Ferdinand Porsche.

Ben

[desmo]

For those of us who need a picture to get the picture, here's an exploded view of a Hewland Mk.9:

[Ray Bell]

Items 9, 10 and 11 in that diagram make up the LSD component...

However, the cage for the pawls is not show... and only one pawl is drawn in.

[Ben]

I suspect part 8 is the cage - but it is drawn from the wrong side for this to be clearly shown.

Ben

[Ray Bell]

You're probably right, the cage might be attached to the side housing.

[desmo]

Here's the whole page with the parts descriptions:

http://www.hewland-e...vga/mk9-c&p.htm

[Dave Wright]

Ben - thanks for taking the trouble to type all that information from Carroll Smith's book. It confirms the stuff I got from the net but is much clearer and has new information too. That is just what I was looking for - thanks again.

[Ray Bell]

Yes, desmo... that settles it...

The pawls are referred to as 'plungers' and that component is called the 'plunger carrier'... so it's just that it's shown from the wrong angle. The pawls (or plungers) were held so that they remained vertical to the centre of the axle, if that makes any sense?

[Wolf]

OK, my Bible, has descriptions (and cutaways) of three Selbstsperrende Ausgleichgetriebe, namely Prometheus, ZF and Rheinmetall-Borsig types...

Prometheus is quite similar to the Hewland cam and pawl described here, albeit appearing somewhat simpler in execution. ZF works on similar principle, although drive is through cage with pawls, with drive taken off from inner and outer cam track.

Now, a Rheinmetall-Borsig diff is a different beast (and apt word indeed it is) alltogether. Drive through worm engages worm gear fastened to housing, with four small worm gears freely rotating fastened to housing. Those worm gears are, on the each side of diff, engaging four 'slip-limiting' worms ("Sperr"-Scnecken), all of them (on one side) are simultaneously, in turn, engaged to worm gears taking the drive to each wheel. Sounds rather complicated (no less throgh my poor technical English), but unlike all three previously mentioned there is almost no wear, and slip-limiting effect is result of low efficiency of worm transmission... If this diff is of any relevance to the discussion, I could have drawing scanned.

[Ray Bell]

Ah yes, I've seen those complicated drawing somewhere some time...

Another strange one was the Weissman (or is it Weismann?)... I recall one day at Oran Park Phil Ward's Monaro was running away when the diff started playing up... it became very unpredictable and he lost the race.

[Ben]

Ahh, the Weismann Locker. Some more transcribing from Mr Smith's books may be in order here - unfortunately not until the Autosport International Trade Days are out the way.

Ben

[Chickenman]

Sounds like Wolf is referring to a ZexelTorsen diff or a Quaiffe style diff. If you think a Cam and Pawl diff is complicated take a look at this.:

http://www.torsen.co...rol_Article.pdf

http://www.torsen.co...ts/products.htm


Used to run a Torsen T2 diff on my 1986 Camaro Autocross and Hillclimb car. Worked fantastic, especially in the wet.

Unfortunately, for my particular application ( 420bhp400ftlbs torque through a 7.625" R&P ), durability was a limiting factor. Because of the smallish carrier the internal components ( worm gears ) simply could not be made large enough to carry the loads on them. Extreme wear was the result. In other applications this was not a problem, just with these particular vehicles.

Spent considerable time on the phone with engineers at Zexel Torsen trying to resolve matter. Ended up having to go with a Eaton style ( GM plate style posi with carbon plates ) diff. Not as forgiving as the Torsen, in fact the plate style diff is a nightmare in the rain....but it will last forever...or at least until I shear the teeth off the R&P!!!

[Wolf]

Very similar Chickenman, except this one You gave uses straight cut gears instead of worms inside... Anyways, a nice article, thanks.