MX-5 OC Forum

Rich M:

The only thing I would disagree with is that the posts here are worth nothing. The world has changed with the internet and manufacturers must keep up and improve their communication. If I were a manufacturer I would not like to log on and see these posts.

Manufacturers don't like to see these posts, but really - would I spend millions on a recall based on posts on the internet? A rival company or even random people may have set this up as a hoax, or people are misidentifying parts, or including data from other models, or double posting, or posting on behalf of their mates who's brothers wife met a man at a wedding who had heard that a bit of metal in a diff had once stolen a Snickers bar from a half blind shop keeper, or something. Then when you look at the volume of internet traffic - I could employ a twenty man team of experts and still not correct every mistaken post on Pistonheads. Plus would I really care that a handful of people who aren't ever going to be direct customers are occasionally unhappy at the thought that maybe something might be wrong? As long as I'm legally covered and they aren't legally wrong there isn't any need to do anything.

When you already emply a network of trained* people in every major town in every country you sell cars in its better to trust the data from them.

* I know some people object to the thought of people at main dealers actually be trained in anything, but compared to the average person on the internet they are very reliable. Plus it's their job.

Rich M:

Clutch Judder. That line of thought caused me to find this which coincides with the same years of production. Severe clutch judder really does put a snatch on the transmission and would certainly have the effect of engaging and disengaging the LSD hard and fast when it occured. Do you know if your car has had a replacement clutch plate and is it smooth these days?

http://www.mx5oc.co.uk/forum/forums/t/276.aspx

Edit; This was a newly introduced LSD so its durability margins would have not been known and it would also not have been tested with a juddering clutch in the tests Captain Muppet explains above.

Interesting! Good point.

Rich

I don't know if my car has had a replacement clutch.  There is no evidence of a change in any of the documentation that came with the car.

The clutch is fairly smooth, no worse than any previous car I've had, although I have to say it's not quite as smooth as my present Ford Focus.

Phil S:

Rich

I don't know if my car has had a replacement clutch.  There is no evidence of a change in any of the documentation that came with the car.

The clutch is fairly smooth, no worse than any previous car I've had, although I have to say it's not quite as smooth as my present Ford Focus.

Hi Phil,

The link, http://www.mx5oc.co.uk/forum/forums/t/276.aspx  says if it's been done under warranty there should be a record of it. A phone call to your local dealer with the VIN should bring it up. Otherwise it may have been done outside the network or maybe you're just fortunate to have a smooth one anyway. Sounds to me like you will be fine with a smooth clutch and no bits.

I find this very re-assuring. I've had a BMW 3 series, a Passatt, and presently have a Mazda 6 which all juddered. On damp days I slip it gently to warm it up, it seems that condensation may be a factor. If you've never experienced it, it can be very bad - to the point sometimes with the VW in partic where I had to dip the clutch and start pulling off again as it was hammering so much. If some Mk2.5s were so bad they were being done under warranty then that must have been very bad indeed.

Might be worth anyone who now has a failed LSD checking with Mazda to see if their clutch has been replaced under warranty as it would most certainly indicate that it would have been hammering the LSD as well.

My Mk3 clutch is very smooth but I don't know if they all are. I checked the bite point and gave it a bit of work before buying it just to be sure it wasn't about to expire as it felt so smooth (you know the "am I buying a torque converter here?" thoughts). Perhaps Mazda did some development there and possibly on the LSD too, I can hope...

Clutch judder or abuse. I think this is over and out for me on this.  

I think I'm right in saying (someone will put me right if not) that the engine in a Mk3 MX5 like yours is very similar to that in a 2006 Ford Focus like mine. Wasn't it a MZR Duratec engine or something like that?

As I said that clutch is supremely smooth in the Focus and therefore would probably be the same in the MX5 Mk3.  My 2.5 is not quite as smooth but still reasonably good, no clutch change AFAIK.  As I say perhaps I've just been lucky.

Hmm.

http://forum.miata.net/vb/showthread.php?t=175842

http://forum.miata.net/vb/showthread.php?t=314048

http://forum.miata.net/vb/showthread.php?t=248422

While many US States have annual vehicular safety tests, which might involve some sort of brake test, quite a few don't. In the first thread above, poster "Darmie" is listed as being from South Carolina, which is one of the States that doesn't have an annual inspection (the law was repealed), so its unlikely this car has ever had any kind of rolling road brake test. The third thread gives the OP as being from California. California only carries out safety inspections on new cars. So it seems the same breakage can occur under other circumstances besides that described in this thread.

 Hi there all,

Just to clarify for all reading this,and as one of the first people involved,as a qualified insurance engineer i have a lot higher level of experience and qualifications than technicians in the dealer networks,and used to be a warranty engineer on behalf of manufacturer's.

This particular differential will not fail through clutch judder,as the torque is not applied in the same way through the tabs,is doubtfull it would fail from standing start abuse in a standard vehicle,which is why it would have complied the specifications Mazda would have applied,and they same would apply to drifting,as both wheels would lock up under power when wheels slip is detected,if the vehicle is over tired(larger wheels and tyres hence more grip)or excessive power applied(turbo,supercharged,or tuned)the only way this diff is failing is due to torque being applied in a way it was not designed for.

Of the differentials seen,and inspected,all the tabs have sheared in the same way,the only way this can happen is due to a latent defect,which in my opinion is through poor manufacture,and the slots the tabs sit in not being radiused,and applying high pressure to a 90 degree angle in the slot against the tab,causing fatigue,then sheer stress on the tab causing failure.

The only thing that could cause this is a high level of torque being applied with one wheel locked,the other rotating,applying pressure to the tab causing it to shear,whether this occurs due to the rbt or not,i cannot prove,but as i cannot think of anything driving a wheel from the wheel side back to the diff,it would seem likely.

To cover the issue of people not being worried,or thinking it may not cause an accident,i have now seen two incident's on this forum and another ,where the tab has caught between the crown and pinion,and snapped the teeth off,one of them locked,the other did not,but punched a hole in the diff casing.

If you have found tabs,your diff may still work,but it will be half as strong and more prone to failure due to twice the pressure being exerted through the two remaining tabs.

As for it not being in Mazda's interest,any owner of a Mazda may be a prospective owner of a new Mx5,for them not to take action is very nieve..this topic does not affect me,as i dont own a vehicle fitted with one,the reason i did get involved was due to concern,and to help the people on the forum involved,do i think there is a risk involved?YES!!!do i think its dangerous?YES!!!do i think Mazda should deal with it?YES!!! under the sale of goods act,an item should be fit for purpose,and under EU Legislation this can apply up to 6 years...is it "reasonable to expect this diff to last for 6 years?yes..it should last the lifetime of the vehicle...when is the last time you replaced a diff on any other car you have ever owned????think about it...this is in no way a normal situation.

If you would like to know what to do,please read the start of the thread and read mine and others comments.

regards

Tim

AT:

Hmm.

http://forum.miata.net/vb/showthread.php?t=175842

http://forum.miata.net/vb/showthread.php?t=314048

http://forum.miata.net/vb/showthread.php?t=248422

While many US States have annual vehicular safety tests, which might involve some sort of brake test, quite a few don't. In the first thread above, poster "Darmie" is listed as being from South Carolina, which is one of the States that doesn't have an annual inspection (the law was repealed), so its unlikely this car has ever had any kind of rolling road brake test. The third thread gives the OP as being from California. California only carries out safety inspections on new cars. So it seems the same breakage can occur under other circumstances besides that described in this thread.

Yep, well spotted AT. All those 3 are 2003 cars and it won't be an RBT causing it as you say. So a design/manufacturing fault or provocation from clutch judder or hard use.

Anyone found any Mk3 cars with the problem yet?

sidewayssam:

 Hi there all,

Just to clarify for all reading this,and as one of the first people involved,as a qualified insurance engineer i have a lot higher level of experience and qualifications than technicians in the dealer networks,and used to be a warranty engineer on behalf of manufacturer's.

This particular differential will not fail through clutch judder,as the torque is not applied in the same way through the tabs,is doubtfull it would fail from standing start abuse in a standard vehicle,which is why it would have complied the specifications Mazda would have applied,and they same would apply to drifting,as both wheels would lock up under power when wheels slip is detected,if the vehicle is over tired(larger wheels and tyres hence more grip)or excessive power applied(turbo,supercharged,or tuned)the only way this diff is failing is due to torque being applied in a way it was not designed for.

Of the differentials seen,and inspected,all the tabs have sheared in the same way,the only way this can happen is due to a latent defect,which in my opinion is through poor manufacture,and the slots the tabs sit in not being radiused,and applying high pressure to a 90 degree angle in the slot against the tab,causing fatigue,then sheer stress on the tab causing failure.

The only thing that could cause this is a high level of torque being applied with one wheel locked,the other rotating,applying pressure to the tab causing it to shear,whether this occurs due to the rbt or not,i cannot prove,but as i cannot think of anything driving a wheel from the wheel side back to the diff,it would seem likely.

To cover the issue of people not being worried,or thinking it may not cause an accident,i have now seen two incident's on this forum and another ,where the tab has caught between the crown and pinion,and snapped the teeth off,one of them locked,the other did not,but punched a hole in the diff casing.

If you have found tabs,your diff may still work,but it will be half as strong and more prone to failure due to twice the pressure being exerted through the two remaining tabs.

As for it not being in Mazda's interest,any owner of a Mazda may be a prospective owner of a new Mx5,for them not to take action is very nieve..this topic does not affect me,as i dont own a vehicle fitted with one,the reason i did get involved was due to concern,and to help the people on the forum involved,do i think there is a risk involved?YES!!!do i think its dangerous?YES!!!do i think Mazda should deal with it?YES!!! under the sale of goods act,an item should be fit for purpose,and under EU Legislation this can apply up to 6 years...is it "reasonable to expect this diff to last for 6 years?yes..it should last the lifetime of the vehicle...when is the last time you replaced a diff on any other car you have ever owned????think about it...this is in no way a normal situation.

If you would like to know what to do,please read the start of the thread and read mine and others comments.

regards

Tim

Hi Tim,

Sorry I don't follow what you're saying. The cones are the friction surfaces (their rough side contacts the back side of the diff gears) and so I'm assuming the tabs fit into recesses in the cage/case of the diff.to stop the cones rotating. If that is the case then the tabs will take the hammer under any use or clutch judder.

You can see that the friction surface on them is only on one side, the plain side must be located (connected with and prevented from rotating separately from) the cage/case for the LSD to function.

Please explain the function of the tabs if you understand differently.

Perhaps Taff can explain how the cones and tabs locate in the case (if they do) as he may have both the case and broken cones.

Okay, I think it is time for me to give my two pence worth.

This picture from taff eunos is one that interests me.

taff eunos:

 

In my simple world that to me suggests that it is this crack that is happening first and the loss of the tabs is consequential to that.

This still does not get us any nearer the root cause of the problem, but I would be looking for mechanisms that would cause the crack to form where it has and run in that direction.

Rich M:

Perhaps Taff can explain how the cones and tabs locate in the case (if they do) as he may have both the case and broken cones.

My understanding of the purpose of the dish (with tabs) in this differential design is that the tabs fit into recesses in the diff cage and are effectively a `stator` with the outer shoulders of the side gears being the effective `rotor` although in practice they are all `rotors` but under normal conditions such as cornering, having more traction on one or other wheel etc. the two are encouraged to equalise their respective differing rotational speeds by the action of the differential pinion carrier gear and the side (output) gear producing a thrust force outwards and causing a `clamping-force` between the dish and outer shoulder friction action (similar action to an oil-immersed friction brake) to equalise the speeds of the differential cage and the slower rotating output flange.

They (plates) appear to be `floating` and able to centralise to the shoulder in operation, obviously there is a film of oil at all times which is able to be squeezed out (via the segmented surface of both the shoulder and the dish) as the clamping force is applied by the pinion carrier gear and side gear.

If Im wrong about how it operates then Im all ears as to how the infernal pile of rubbish is supposed to work, all I know is its an awful design and, at least on the years from which the failures are cropping up, bound to go wrong thanks to seriously piddle-poor manufacture

Dr. Eunos

 Hi Mark,you are correct in what you have said,and taffy has given a good explanation,what happens the area that grips on the inners surface clamps up,applies the torque to the outer casing when it locks,and the whole carrier then rotates as a whole allowing both wheels to rotate at the same speed.

But when you apply torque the opposite direction by rotating the wheel under force,the carrier rotates,and applies the torque from the slots in the carrier to the tabs ,which then try to transmit the force back through the ring above,it would appear from what i can see that when this happens,the tabs are not a perfectly tight fit in the carrier,and are being pushed from the top edge of the tab (like a lever)creating a fulcrum at the bottom in the ring,because it has been machined/pressed with a 90 degree edge,and as it not radiused ,is causing stress tears.

In normal operation the force is transmitted from the bottom of the tab,to the carrier,in reverse the force appears to be from the top.Really it should be a interference fit with no movement at all...

This appears to be from what i can see ,the cause of the tabs tearing off..hope this helps....prior to seeing these diffs,i had never encountered one with this design,so all new to me as well,

hope this helps!clear as mud i know....under normal operation,as you said,it should be fine,its just when the force is applied the opposite direction.

and just as a p.s, not sure on this ,as it is the chicken and the egg....but if two tabs snap off,you would have force applied to one side of the ring ,and not the other,would it be this that has caused the cracked ring on the other images???

regards

Tim

taff eunos:

Rich M:

Perhaps Taff can explain how the cones and tabs locate in the case (if they do) as he may have both the case and broken cones.

My understanding of the purpose of the dish (with tabs) in this differential design is that the tabs fit into recesses in the diff cage and are effectively a `stator` with the outer shoulders of the side gears being the effective `rotor` although in practice they are all `rotors` but under normal conditions such as cornering, having more traction on one or other wheel etc. the two are encouraged to equalise their respective differing rotational speeds by the action of the differential pinion carrier gear and the side (output) gear producing a thrust force outwards and causing a `clamping-force` between the dish and outer shoulder friction action (similar action to an oil-immersed friction brake) to equalise the speeds of the differential cage and the slower rotating output flange.

They (plates) appear to be `floating` and able to centralise to the shoulder in operation, obviously there is a film of oil at all times which is able to be squeezed out (via the segmented surface of both the shoulder and the dish) as the clamping force is applied by the pinion carrier gear and side gear.

If Im wrong about how it operates then Im all ears as to how the infernal pile of rubbish is supposed to work, all I know is its an awful design and, at least on the years from which the failures are cropping up, bound to go wrong thanks to seriously piddle-poor manufacture

Dr. Eunos

 

Hi Taff,

Yes, completely agreed. I hope that's also how my description of it came across too, and that which Robbie posted from Mazda also.

What I couldn't see from here(!) is how the tabs are located. Your description sounds exactly as I'd imagined and I'm intrigued that the cone plates also seem to be floating to some extent. Makes sense but also means they start with some clearance which may not help if they are hammered.

Mark raises a good point with your pic. I've posted it again. Looking at the tabs carefully we can see the burrs on the edges from rotational forces in both directions (this LSD may be 2 way). The crack near your pen is on the edge of the tab showing most burring. Applying force in the direction of the red arrow would cause the fracture of the cone plate at the root of the tab.

The green arrows point to this burring which is on all tabs. The tabs clearly take the force when the LSD is active. If all tabs break off the diff will still work but just as a regular diff with no LSD as the cone plates will be free to rotate. If one tab goes the others will surely follow soon due to the increased work load. This may explain the case here where one plate has lost all tabs and the other none - but looks just about to start.

The yellow arrow points to where the tabs have broken away on the other plate but have remained in place for some time. There is considerable burring and polishing which wouldn't be present if the tab had broken off and dropped away immediately. That made me think the tabs were in a recess in the cage/case as you say they are.

sidewayssam:

 Hi Mark,you are correct in what you have said,and taffy has given a good explanation,what happens the area that grips on the inners surface clamps up,applies the torque to the outer casing when it locks,and the whole carrier then rotates as a whole allowing both wheels to rotate at the same speed.

But when you apply torque the opposite direction by rotating the wheel under force,the carrier rotates,and applies the torque from the slots in the carrier to the tabs ,which then try to transmit the force back through the ring above,it would appear from what i can see that when this happens,the tabs are not a perfectly tight fit in the carrier,and are being pushed from the top edge of the tab (like a lever)creating a fulcrum at the bottom in the ring,because it has been machined/pressed with a 90 degree edge,and as it not radiused ,is causing stress tears.

In normal operation the force is transmitted from the bottom of the tab,to the carrier,in reverse the force appears to be from the top.Really it should be a interference fit with no movement at all...

This appears to be from what i can see ,the cause of the tabs tearing off..hope this helps....prior to seeing these diffs,i had never encountered one with this design,so all new to me as well,

hope this helps!clear as mud i know....under normal operation,as you said,it should be fine,its just when the force is applied the opposite direction.

and just as a p.s, not sure on this ,as it is the chicken and the egg....but if two tabs snap off,you would have force applied to one side of the ring ,and not the other,would it be this that has caused the cracked ring on the other images???

regards

Tim

Hi Tim,

Yes you have it now :) What I I think you'd missed in your last post (where you still blamed the RBT) was that the diff will not be operating as a LSD if there is no load on the prop.

Rich M:

Yes, completely agreed. I hope that's also how my description of it came across too, and that which Robbie posted from Mazda also.

TBH I hadnt read yours or Robbie`s description of how it works (not for long though...) I had my own idea about it going back some 9 months ago, once one tab breaks the loading on each other tab on that side is increased, its guaranteed failure

denzibaby sent me the bits from his own failed diff, he would better know how the tabs on the dishes are located in the diff carrier as he has had first hand experience of dismantling one.

On another thread Robbie was advising an affected owner to `carry on driving until it fails altogether` or words to that effect see: http://www.mx5oc.co.uk/forum/forums/p/14186/83333.aspx#83333

this is why I became so very incensed at the situation in the first place. Robbie also works for a Mazda dealership