Topic: OEM FD3S Differential Limitations and Options

The OEM Torsen rear end, found in the FD is regarded as the best choice for road coarse or autocross FDs due to it's efficient engagement and power transmission; however, it is not well suited for high shock applications such as drag racing. As noted in a Node's thread, shock introduced by hard launches and wheel hop can begin to chip gear teeth in the torsen unit at near stock levels of power.
For drag racing applications there are several plug and play options. Cusco and KAZZ offer aftermarket LSD that offer power handling capabilities which are beyond the structural capacity of the diff housing itself, but these units cost around 800 dollars and do generate a bit of clanking noise during sharp turns and deceleration. Another option that was realized over time is a hybrid transplantation of an older clutch type diff out of an S4 T2 directly into the FD housing in place of the Torsen LSD. A T2 conversion will cost somewhere in the neighborhood of 500 dollars after labor and will again be able to handle more torque than the stock housing can handle. It's also worth noting that the axles will need to be upgraded for drag cars making power in the early-to-mid 400RWT range. 300M axles are the popular choice for addressing this problem. These axles require the tranplantation of the OEM outer joints onto stronger 300M shafts. A step-by-step axle conversion tutorial can be found HERE.

Knowing which T2 Clutch to use and why:

As per this thread. The S4 Turbo 87-88 clutch type LSD is the thread of choice for an FD diff upgrade, and will bolt directly into the FD pumpkin. Other S4 non-turbo models utilized a clutch-type LSD as well, but with only a 7" rear, while the S4 T2 utilized an 8" rear. Later S5 turbo cars utilized viscous differentials.

Thoughts

For most cars a T2/Kaaz/Cusco + 300M setup is sufficient to hold give or take 500-600RWHP, depending on factors such as casting quality, road surface, tire selection etc etc., But as I mentioned previously the diff casing is a weak point, as are the OEM inner stubs. Jim Lab covered this topic fairly thoroughly in one of our driveline discussions linked below so hopefully he won't mind me posting this tidbit:

And this is a classic example of housing failure (ignore the red circle).


More Thoughts/Questions

So why does the housing break? There are various factors at play here and we will by no means address all of them:

Considering the effects of a stiffened PPF and inadequate motor mount and diff bushing durometers:

But this is not the entire story, because their are internal forces at play as well which are also testing the structural integrity of our OEM diff and tranny casings in combination with external torsional forces and driveline shock. These factors combine make for an effective means of grenading the stock diff casing.

I maintain that the reduction of wheel hop and driveline rotation are the best way to reduce the amount of shock transmitted through the driveline and is therefore the most effective means of keeping the diff from greanading as power will be applied in more smooth and uniform manner and as a result decrease the likelyhood of cracks and breakage occurring. An IRS setup in conjuction with a transmission that is mounted to the chassis via appropriate durometer bushings is IMO the best way to deal with this issue.

There are two IRS setups currently available for the FD:

Paul Imbrogno has developed an IRS setup (GB Link HERE)designed to utilize the OEM diff and axles. The cost is approximately $1,000US and requires the modification of the stock subframe.


Jim Lab has developed a 8.8 Cobra IRS setup utilizing Ford axles. The Cobra IRS setup is believed by Jim to be the stronger of the two as it eliminated the achiles heal Mazda diff casing and Mazda axle stubs. The Cobra IRS is not currently available for purchase to my knowledge, though the jig has been purchased by a forum member so the kit may go back into production at one point.



The open question with this kit remains, how much power can this 8.8 Cobra IRS setup actually hold? To my knowledge is not particularly bullet proof. Perhaps Jim can provide some background. Cobra 8.8 IRS strength limitations are briefly addressed HERE.

Conclusion

The conclusion that I have drawn is that we do not really have a 100% bullet proof, bolt-on option, so choose the one that best suits your power goals and budget. I personally have already invested in 300M axles and a custom DS so I'm going with Paul's IRS.

Afterword

I hope this thread will be of some assistance, it is my small effort to bring tech threads back in style ;o).

Robert


Some More Background Reading:

Greg's T56 conversion thread:

• https://www.rx7club.com/3rd-generation-specific-1993-2002-16/t56-tranny-fd-609461/page2/

Node's Torsen thread:

• https://www.rx7club.com/showthread.p...highlight=diff

Rebuilding info:

• https://www.rx7club.com/showthread.p...highlight=diff

T2 Conversion Tutorial

• https://www.rx7club.com/showthread.p...highlight=diff

T2 diff specifications:

• https://www.rx7club.com/showthread.p...highlight=diff

 

Good info and thanks for taking the time to summarize all the disparate information spread across multiple threads.

Unless one is here everyday and is a long time member, its hard to fully understand the multiple factors affecting tech items (such as this) and searching only allows you to read certain pieces of information valid at a certain point in time.

Bravo!

 

TY, The whole point was to create a substantive and comprehensive thread for a commonly searched subject. I was searching for old threads from guys like Vosko and realized that a great deal of the information that I originally used back in the day is no longer yielded in my searches. This is nothing new, but it should help answer a lot of queastions that I've been getting over PM. People constantly advice new members to use the search function, but to be honest most of the threads that you'll find when doing a search are sparcely relavant and shabby. This thread was created for searchers so no need to bump guys.

ltr

 

Nice post. I also am sending my subframe up to Paul for the same reasoning you are. I already have a KAAZ unit and 300M axles. While I do think the Cobra IRS is a bit stronger, and the extra gear options would be a plus, I also think a major contributor to the housing breakage isnt directly related to the strength of the housing, but yet its mouting locaitons. A majority of cars using an IRS rear suspension with a "performance" base are running the sub-frame mounted diff, and not using a PPF/torque arm. On our factory setup, engine/trans vibration and any rotational force is transmitted through the PPF directly into the "snout" of the diff. This coupled with wheelhop and any twisting motion in the differential itself just cant be a good combination.

 

How to ID a TII diff:
https://www.rx7club.com/automotive-news-lounge-22/s4-clutch-typle-lsd-vs-s5-viscous-lsd-460891/

Pictures of broken stock torsens:
https://www.rx7club.com/3rd-generation-specific-1993-2002-16/need-some-diff-help-607965/


I had some slop in my diff ever since I bought my car. It was only a matter of time until my diff gave out, as seen in the above thread.

 

Excellent additions. Does anyone have comprehensive information the PNs for the Mazdaspeed clutch packs/rebuild kit and their OEM contemporaries. I'm assuming that rebuild kits are still available from Mazda?

I forgot to mention that one large advantage of the T2 setup is its rebuildability.

 

The oversized part # for the clutch disc is: P020-27-257

(which isn't a mazdaspeed specific part.)

 

From Greg:
There is a solution to the Mazda stock rear diff case breaking. The craddle that Peter Farrell, Ray wilson, Adam Saruwatari, Ari Yallon, and Dan S. that they fabricated and used/uses. Like I had stated in another thread, each is a little different in design, but do essentially the same thing.
When you wheel hop or come under Hard load, the pinion wants to push back and up...well since the connecting point is the nose of the case where the PPF attaches, its just breaks the case there or near there as it wants to push up.
Adding one of those general diff braces that everyone sells adds to the breaking factor as it gives less deflection for the diff to move.
The craddle encases the whole diff and the PPF frame attaches to the outer part of the craddle and transfers all the torque/load from the front to rear and essentially lifting the whole entire diff up instead of it trying to lift it up at the front where the PPF attaches and cracking the diff case in the middle. Now the weak link would be the PPF breaking, but if you could fit the general style diff brace in addition to the craddle then essentially you are eliminating all of the weak points that move. I would rather break a PPF frame then have to go through and replace the diff case, ring and pinion and have to go through and set everything up again.
Peter would cut 1.3xx 60 ft times with the stock rear diff case and had no issues.

 

Why an OEM IRS setup makes sense (assuming that the current axle weaknesses will be resolved, or that the current axles are suitable for the power goals of the majority of FD users). -

FYI DSS is currently working on 300M CVs for the FD and FC.

These are ofcourse just my opinions, but my intuition tells me that the diff becomes significantly more susceptible to breakage when it is in a flexed state. Flick a relaxed sheet of aluminum foil and it will not break. Flick a stretched piece of foil and it will. Same concept

There are those who think that simply mounting solid bushings on the rear end of the diff has the same effect as frontal bracing, but this is an oversimplification it seems and is fundamentally wrong in my respectful opinion. By supporting the front end of the differential you are eliminating the flexing by transmitting the rotational forces to the chassis. Furthermore, there is no longer a PPF or TA to transmit such forces from the power plant to the rear end, and you are creating a steel wall that the pinion must push against to escape.
Solid bushings will decrease wheel-hop, certainly, but at a cost. When used in conjunction with a PPF they will increase the amount of strain experienced at the front center of the differential as they will no longer act as shock absorbers for the differential as the PPF attempt to role the differential over. Visualize the rear end of the diff is held completely still by its carrier while the free-hanging front end is twisted by the PPF. Again, this is not enough to break the diff casing on its own, but I have a hunch that this is creating enough strain on the diff to make it susceptible to the leverage which is caused by the pinion which is attempting to escape.
The incorporation of solid or high durometer bushings alone, results in a trade-off between wheel hop and torsional casing flex. Yes, you will decrease wheelhop, but you are now requiring the diff casing to absorb more torsional strain and leverage than before because these damaging forces are no longer partially absorbed by the diff carrier bushings.
The pinion is already trying hard enough to escape without the casing in a stressed state. It is a mistake to solely focus on the pinion and ring gear and ignore the casing.
The incorporation of a front mount on the diff also absorbs upward, rearward and sideways forces from the pinion, and thus decreases the likelihood of a breakage occurring at the front end of the diff. Paul's IRS (shown below), for example, incorporates a steel mounting plate bolted directly over the nose of the diff which is machined flat to accommodate this. When the pinion try to climb it is climbing directly against this brace and is therefore climbing directly against the chassis via two moderately stiff bushings. Take a careful look and see that this setup is fundamentally different than the stock setup. The forces that were once being applied directly to the front and center of the differential casing are now being transmitted almost directly to the chassis.






Now examine the photograph below and notice where the breakage is occurring. The differential is not breaking at its nose is breaking at its center where the casing is the most thin and where the differential is stressed internally by the lever that pinion/front of the diff, and externally by the rotational forces exerted by the PPF. And there is your aluminum foil. I have yet to see one photo of diff cracked at the actual nose end; this is undoubtedly due to the thick casting at this portion of the diff.
Now closely look at the photos above and follow with your mind the path that the forces are following. The pinion is pushing against a wall which is the front brace, and the front and rear ends of the diff are necessarily held in sync. It should be clear that this setup is not the OEM crap-box PPF setup. It is fundamentally different. It is almost equilaterally supported and reinforced by the rear sub-frame so that the rear sub-frame becomes its brace. Also considering that no torsional forces are coming from the tail end of the trans because there is no longer a PPF.

 

How the Cobra IRS failure is different than the OE Mazda unit: When an FD differential comes apart at the snout**, the pinion gear and the ring gear tend to chew eachother up somewhat, so you can pretty much plan on replacing the entire case. The driveline and PPF are also no longer attached to the differential, so if you're traveling at any serious rate of speed, you have vastly increased potential for further damage to the car. The only "good" thing is that the transmission tunnel braces should hold the PPF and driveshaft up off the pavement.

When the rear cover on a Cobra diff breaks, you get fluid all over your exhaust. It will probably smell. Then you get a new rear cover and some more differential fluid and clean up your exhaust.

Evolution of the Cobra axles:
2001 Cobra axle vs. FD axle. Note the difference in diameter in the axle shafts.[img]file:///C:/DOCUME%7E1/ROBERT%7E1/LOCALS%7E1/Temp/msohtml1/01/clip_image001.gif[/img]

2003 Cobra axle vs. 2001 Cobra axle. Note the difference in beef here.

[img]file:///C:/DOCUME%7E1/ROBERT%7E1/LOCALS%7E1/Temp/msohtml1/01/clip_image001.gif[/img]


Jim's Axles: Now consider that we're replacing the center shaft on the 2003 axle with a 300M replacement. The Driveshaft Shop rates the completed axle for 900+ horsepower. I know for a fact from talking extensively with Ari Yallon that the OEM axles won't hold anywhere near that kind of power, even with 300M center shafts.

Other Considerations:
The C4 IRS is much weaker than the Ford 8.8 and uses axles that adapt well to FCs, but not FDs.

A C5/C6 differential would require extensive work to be mounted in the car (it has no front mounting provisions) and to cap the front of the differential (which normally bolts to a transmission). Then a pinion flange would have to be fabricated, because the Corvette has none. After that, you'd have to figure out how to adapt the inner stub of the Corvette axles to the outer joints of the RX-7. The Cobra uses axles that are easily adapted to the RX-7 outer hardware, which is one of the reasons I made the decision to use it. The C5/C6 doesn't.

After all that, you'd still be stuck with an 8.25" ring gear (C5) vs. 8.8" ring gear (Cobra). I shouldn't have to tell you that the larger the ring gear (the FD's is a puny 8.0"), the more mass and strength it has. Supposedly the C6 differential has a 9.0" ring and pinion, but you'd still have all the other fabrication problems to deal with, and at the time I made the decision to use the Cobra 8.8" IRS, obviously the C6 differential wasn't available yet.

Why Jim sees gearing options as beneficial:
The 4.10s and 3.90s work just fine with stock or nearly stock engines, although they aren't exactly conducive to improved gas mileage. However, 4.10s and even 3.90s make 1st and 2nd gear basically useless for a V8 car with 400+ RWHP wearing street tires. I also think you'll find that anyone building a turbo car would benefit from the availability a much lower (numerically) gear ratio also, since boost is load dependent.



The goal of the 8.8" IRS project: To provide access to a wide range of gear ratios for various applications. The fringe benefits were that the assembly is much stronger than the Mazda OEM parts (whether you want to believe that or not), and it doesn't add any weight to the car. I don't see how you can argue with that.


Update:
34 people now have that option, and others seem to be very interested because my kits resell very quickly and I've sent many people to Alex Hagedorn (TT_Rex_7), who now has my welding jig, templates, and blessing to produce more of them.

It appears that you have a solid design, my only concern would be NVH since the [rear] of the diff is solidly mounted to the subframe. So far, the only one who has complained has been Chris Heiser, who bought Moroso gears that whined excessively. Of course, my cradle got the blame for everything, as usual. LS1-FD panicked and changed his rear mount based on Chris' posts. Still, it's not a bad idea to add bushings to the rear mount as long as they're rigid enough not to allow excessive movement.

 

wanklin,

nice job. a welcome summary of an issue that an increasing amount of fd owners will face as they tune up there motors.

if you make more than 450 and you push it hard occasionally you will be dealing w drivetrain problems. the thread is about solutions.

either tune up your drivetrain or be prepared for lots of rear end lube and broken parts under your car.

i am leaning towards keeping the rearend ( i have a T2 LSD) and doing some supportive brace for the nose. i would like to use more than the two bolt holes on the driver's side of the case and the one small top threaded bolthole.

i will probably pull my rear crossmember next week.

howard coleman

 

There's so much info spattered about, it only made sense to consolidate it, especially concerning the gravity of the drive line issue as the power levels are increasing (the Rotary vs. V8 race). I wonder if the 20Bs will be able to keep up ;o) One thing's for sure, these cars are going to start breaking the 1000hp barrier sooner than later and in order to do so a lot of things behind the engine are going to need to change....

You said it Howard. Push it too hard with the stock equipment and you can end up like this guy.

 

Rob, great reading, thanks for taking the time to do this.

Just a note, in your most recent post above, we can't see this picture:

2003 Cobra axle vs. 2001 Cobra axle. Note the difference in beef here.

[img]file:///C:/DOCUME%7E1/ROBERT%7E1/LOCALS%7E1/Temp/msohtml1/01/clip_image001.gif[/img]

 

The cobra diffs are pretty beefy, but wheel hop kills them. There is a girdle that aleviates the problem. I run one on my car. The next week point is the axles. The driveshaft shop has an answer but its pricey. I don't think the rotary is going to produce the torque that you need to worry about breaking the cobra axles.

 

How to differentiate between an S4 T2 and S5 diff at the junkyard: https://www.rx7club.com/showthread.p...=1#post6670731

 

and what about cars that are build for high speed track racing?

We have the Pettit race car over here and it had an old school DANA 44 bolted in there. with no slip at all. it's like driving a cart through the turns. Inside wheel starts hopping up and the car gets unbalance like you can't imagine.

I suggested to the team to put the OEM diff back in. What could be other options? OEM diff with aftermarket internals and/or new diff?

 

Ofcourse this is only my opinion, but a car built for that type of racing would do fine with a Kazz/T2/Cusco setup in the OEM housing, especially since the drivetrain isn't subjected to hard launches. Ofcourse I would suggest mounting the nose end of the differential to the rear subframe and ditching the PPF. What type of powerplant and output levels are we talking about? Viscous LSDs are the preferred choice for roadracers, but they don't hack it once you introduce shock or large amounts of torque into the equation.

 

Rob, i should have said 'in post #10, we can't see this picture'.

 

Yeah I know, I wish I had mod capability so I could just quickly fix these things. ;o/ I'm just going to let that one go as I don't feel that it's really critical to this thread. Thanks for the heads-up.

 

Thanks, Power output is between 400 and 500hp. Strictly track racing but with short and long turns.

Any recommendations about gears and diffs is welcome; I'm not too familiar with that.

The Dana is like a fixed axle and no way of good use for trck racing

 

speedworks,

i don't know the nature of the track surfaces you race on but i raced competitively for 22 seasons on fairly smooth surfaced circuits and always felt a solid axle worked very well.

we bent the axle tubes for about 1.5 degrees neg camber and ran a Detroit Locker. i recommend it as an excellent LSD.

good luck,

howard coleman

 

I just found this thread, good info.

 

GreGreat write up, Appreciate all the work you’ve been doing Wanklin, with regards to the drive line and transmission solutions. Wish I could help in some wayat write up, Appricate all the work youve been doing wanklin with regards to the drive line and transmision solutions. Wish i could help in some way

 

One of the cars here is running an aluminium cased Harrop IRS rear end with power north of 500 and has had no issues with standing starts which you'd be doing over there - definately not a spool. Believe they are 9" based so gear ratios are never a problem.

http://www.harrop.com.au/index.html

Probably be prepared for a bit of coinage though!

 

FWIW re the OEM diff.

a friend has an LS1 FC and i have ridden in it and driven it lots over the last 3 years. he has absolutely no mercy as to drag starts.... all the time. he runs it at the strip, 11.8 at 119.

i can't believe how he bang shifts it and how he launches. he has had no diff problems at all. granted he is making just a hair under 400 rwhp but it has alot of torque off the line and changing gears.

it is brutal on the rear end.

my guess is if it were an FD, the rear end would be in pieces but the FC has a very well supported pinion.

i am definitely losing my PPF and going to build a really neat brace using more than just the 3 common pickup points. it will be complete in March.

during the summer i will acquire the parts i need to go T56.

hc