Noise on turning (video)

[cdscivic]

Need some assistance and hoping you fine folks can help. Recently had some service done to the nsx, and it drove fine post service. A month or so afterwards it started to make a horrible noise only on turning (either way). I also noticed the car rear end wobbled back and forth with the noise when I pushed it into my garage in neutral. What are your thoughts on what it could be?

Attached is a video, forgive my talking but there are some good examples of the noise throughout particularly at the end. Nothing is rubbing, it's all stock suspension and wheels.

1996 Targa 5spd.

Noise video

 

[Old Guy]

Have you put it up on jack stands (or otherwise raised the rear end) to check for damage to the tripod joints on the driveshafts? Seriously damaged tripod joints will make a lot of noise; but, it is not clear how they would make the back end wobble from side to side. A major wheel bearing failure can allow the wheel to wobble and shake the car; but, usually you get some bearing noise before you get a major failure. With the back end of the car up in the air grabbing a hold of the wheels and shaking the wheels side to side should indicate whether you have a bearing that is so far gone that it is allowing the wheel to wobble. While you are in there poking around with the wheels off check for play in the ball joints and suspension pivots if the driveshafts and the wheel bearings are OK.

If the problem is a failed wheel bearing, given the amount of noise it is making I definitely would not be driving the car anywhere.

 

[cdscivic]

Yes sir, everything seems tight underneath. Recently had a large service done and everything was torqued to spec etc.

Some questions if you don't mind:

Could the axle be bad but still look and feel OK on inspection?

Are the axles inherently different sizes/lengths (like a b series cv axle pair)?

Could this be differential related? Mechanic said fluid was "more sparkly than it probably should be"

Thank you so much sir.

 

[cdscivic]

Have you put it up on jack stands (or otherwise raised the rear end) to check for damage to the tripod joints on the driveshafts? Seriously damaged tripod joints will make a lot of noise; but, it is not clear how they would make the back end wobble from side to side. A major wheel bearing failure can allow the wheel to wobble and shake the car; but, usually you get some bearing noise before you get a major failure. With the back end of the car up in the air grabbing a hold of the wheels and shaking the wheels side to side should indicate whether you have a bearing that is so far gone that it is allowing the wheel to wobble. While you are in there poking around with the wheels off check for play in the ball joints and suspension pivots if the driveshafts and the wheel bearings are OK.

If the problem is a failed wheel bearing, given the amount of noise it is making I definitely would not be driving the car anywhere.

Ok some additional information.

Car recently had an engine out service where the car drove great after for a month or so (3-5 drives).

During the service, the intermediate shaft bearing and both rear wheel bearings were replaced.

With the car on the lift and no weight on the wheels, the noise is not present and neither is the wobble. The rear wheels turn at different speeds when the steering wheel is turned/wheels rotated.

 

[e9coupe]

Definitely sounds wheel speed related and not engine or gearbox speed related. Hard to believe there's no noise without a load on it. You used a stethoscope or screwdriver in the ear to listen? Obvious, but wheels are torqued?

 

[cdscivic]

Definitely sounds wheel speed related and not engine or gearbox speed related. Hard to believe there's no noise without a load on it. You used a stethoscope or screwdriver in the ear to listen? Obvious, but wheels are torqued?

I have not, I had it towed to a shop but even they are a little unsure. When I inspected here at my house in the air, wheels/axle nuts/axels all looked ok to me. I shook everything pretty well too. Nothing seemed inherently loose or bad to my eyes.

The mechanic thinks diff is likely the culprit because everything underneath looks tight and clean. That's why above I was asking if a cv axle can be bad but look ok.

Sad guy is sad asf.

 

[e9coupe]

Get it up off the ground with the wheels spinning and you'll probably figure it out

 

[cdscivic]

Here's what he sent me

"I did take one wheel off, lowered the car so the other side was sitting on a brick, holding that side wheel still with the wheel on. On the opposite side I took a long torque wrench and 36mm socket and turned it with the axle nut and it broke away at around 100 ft pounds. I think the nsx-t breakaway spec was 40 foot pounds, and the R is around 100. However, at a certain point it would get sticky and kind of stop unless I applied even more pressure."

 

[Old Guy]

The left and right axles are different. You can't diagnose a failed trip pod joint by visual inspection of the outside (unless the boot is obviously damaged). With both wheels up in the air and parking brakes off rotating the wheels back and forth should make severe tri pod damage apparent.

I also noticed the car rear end wobbled back and forth with the noise when I pushed it into my garage in neutral.

It is not apparent to me how a differential problem would cause the rear end of the car to wobble back and forth, particularly when you are just pushing the car forward when there should be no speed difference between the left and right rear wheels. The LSD can make noise going around corners where you have a speed difference between left and right wheels. There have been cases where individuals have replaced the transmission fluid with Red Line MTL which has friction modifiers which caused early lock up of the differential going around corners and a lot of noise. Flushing the transmission a couple of times resolved this problem. A fill with straight GM synchromesh modified would also likely cause the same problem.

According to the 1991 service manual, the pre-set torque for the test your mechanic described (right rear wheel up in the air and the left rear wheel blocked / restrained) is anywhere between 43 - 101 lb-ft. It seems like your differential pre-set torque is in the correct range. The differential changed in later cars; but, I thought that was after 1996. I am confused by the following statement.

However, at a certain point it would get sticky and kind of stop unless I applied even more pressure."

Does that mean sometimes the pre set torque is exceeding 101 lb-ft? If so, has anybody ever fiddled with the differential or perhaps it has been filled with something with a friction modifier?

I definitely do not understand the following statement

With the car on the lift and no weight on the wheels, the noise is not present and neither is the wobble. The rear wheels turn at different speeds when the steering wheel is turned/wheels rotated.

You need to clarify that. Is the second part unrelated to the car being up on the lift?

 

[Honcho]

The only things that could cause the wobble like that are:

1. Sticking/failed LSD clutch pack
2. Failed CV joint
3. Bent/warped hub and bearing
4. Blown ball joints

If you are not using OEM (or Torco) fluid in the trans, the diff could be the culprit. Does the wobble go away when driving, e.g., increasing speed?

 

[cdscivic]

I will attempt to address all these amazing replies as best I can.

It is not apparent to me how a differential problem would cause the rear end of the car to wobble back and forth, particularly when you are just pushing the car forward when there should be no speed difference between the left and right rear wheels.

Sorry for the confusion, wobble was only present during turning when pushing the car in neutral. It was accompanied by the noise in the video.

There have been cases where individuals have replaced the transmission fluid with Red Line MTL which has friction modifiers which caused early lock up of the differential going around corners and a lot of noise.

Confirmed with my mechanic Honda MTF was used.

Does that mean sometimes the pre set torque is exceeding 101 lb-ft? If so, has anybody ever fiddled with the differential or perhaps it has been filled with something with a friction modifier?

Yes, he is saying the wheel would stick a little and require additional force to move. I will confirm with him if that force exceeded the 101lb-ft but pretty sure it did. To my knowledge the differential has never been fiddled with and this issue was not present with the previous owner or during my ownership until very recently. As far as I know nothing with friction modifier has been used, most of the service records i have been given are from Acura dealerships prior to my ownership, which could indicate the use of proper fluids (but maybe not?).

With the car on the lift and no weight on the wheels, the noise is not present and neither is the wobble. The rear wheels turn at different speeds when the steering wheel is turned/wheels rotated.

This is definitely confusing, sorry about that. Essentially, this was test #1 on the lift. Raise car, rotate/move wheels and see if CV Axles appear bad while moving. The noise did not present itself while this test was being performed and visually i am told the axels looked OK during this test. My apologies for the confusing wording.

 

[cdscivic]

The only things that could cause the wobble like that are:

1. Sticking/failed LSD clutch pack
2. Failed CV joint
3. Bent/warped hub and bearing
4. Blown ball joints

If you are not using OEM (or Torco) fluid in the trans, the diff could be the culprit. Does the wobble go away when driving, e.g., increasing speed?

1. My greatest fear - $$$$
2. My hope, but this is diminishing as time passes.
3. New acura oem hubs on both rear wheels were installed upon recent engine out service
4. These passed inspection on service

Mechanic used Honda MTF fluid for refill, the GOAT of Trans Fluid.

I only noticed the wobble while pushing it but logic tells me it was there while driving. (Un)fortunately once the noise started I parked the car immediately so there wasn't a lot of time for me to really notice much more. I didn't really get up to speed past 15+20 mph on the last drive.

I will say, now thinking about it more and more the noise was present on the second to last drive but was not nearly as loud or bad from what I remember. I recall barely hearing it and thinking I had run over something that suck to the tire or something, but I was 2 blocks from home so I parked it, inspected and didn't see anything. Next drive was the video which made it one street over and I turned around and parked it for good until fixed.

 

[Old Guy]

I am curious about the 'wobble'. Is the rear of the car actually moving from side to side or is it just shaking / vibrating? I don't see how a differential problem could cause the back end of the car to actually move back and forth sideways. If the differential is locking up (or has locked up) in turns the inner rear wheel might hop (more like skip) a little bit which could create vibration and noise.

Does this noise get worse with tighter turns? The speed difference between the inside and outside wheel increases as the radius gets tighter which increases the tendency for the inside tire to skip. If this describes what is happening then it likely is a differential problem. The fact that your mechanic reports pre set torque values that might be over the max setting suggests that the differential might have transitioned from being an LSD to something closer to a closed differential (as in no differential). I have no experience with the NSX differential so I am not exactly sure how that happens without somebody getting in their and fiddling with it.

You said the mechanic reported that the transmission oil appeared 'silvery'. That is never a good thing. The root problem might not be the differential; but, some other failure in the gearbox. If that failure has created 'stuff' which has found its way into the LSD clutches it may effectively be changing the coefficient of friction (just like Red Line MTL does) resulting in lock up of the differential. The manual transmission has an oil pump strainer that is accessible from the outside. If they didn't do this the first time, have your mechanic remove and examine the strainer for debris. This will effectively require another transmission flush unless they are quick like a bunny. Depending on what they find on the strainer and what the drained oil looks like you could have two options

1. lots of bad stuff suggests that the transmission needs to be dropped and the transmission / differential disassembled so that it can be examined to find out what is going on
2. not a lot of bad stuff - perhaps try flushing the transmission fluid again to see if you can wash whatever might be contaminating the LSD clutches out of the clutches. As I recall, it took the unhappy owner of the transmission full of straight Red Line MTL more than just one flush to restore normal differential operation. I would only try this option if the transmission has no other symptoms such as bad shifting or howling noises that might indicate a failed bearing.

The always safe answer would have been #1; but, these days not all mechanics can rebuild a manual transmission successfully. That makes the flush and see what happens option (#2) attractive. However, if something such as a bearing going bad is taking place that is not going to get better by itself and could ultimately lead to transmission shaft damage which would be very expensive (and problematic if the parts are not available). That is why you need to make that assessment of whether something has changed with the transmission (new noises or hard to shift). Tough choices.

The up side to dropping the transmission is that in the process of removal you will be dealing with suspension parts and the half shafts which will allow for close inspection and conclusively ruling out those components as the cause of the noise. If by happy circumstance you discover something like a bad half shaft then you can stop the disassembly process.

 

[e9coupe]

If it were me...

My background... I buy salvage vehicles (and private party) for the company I work for and a lot of them are "Repairable" vehicles. We sell them as-is as projects but a good number of them have suspension damage so we repair them to run and drive to help customers evaluate them better. We see everything in the way of damage, weird issues, etc. When we have an issue like yours the first thing we do is put it on the lift with someone inside the car and run the drive train while the techs listen. (I mean I've heard people do this, we would never! ;-)) 9 times out of 10 we get a good read on the problem and have a direction for repairs. You need to recreate the noise while listening.

Just sayin'.

 

[Honcho]

The fact that your mechanic reports pre set torque values that might be over the max setting suggests that the differential might have transitioned from being an LSD to something closer to a closed differential (as in no differential).

The NSX differential is actually fairly simple. It's a clutch pack that is released by a fixed spring pressure. At rest, the clutch pack is engaged, providing torque to both wheels. On cornering, when the torque split between the outside and inside wheel exceeds the spring clamping pressure, it will compress the spring plate and release the clutch discs, turning it into an open differential. On the "regular" NSX, this is about 50 lb/ft. On the NSX-R, it's about 100 lb/ft. It's not a coincidence that this is 2X the standard value- the NSX-R just uses two stacked spring plates instead of one, effectively doubling the spring rate.



If the spring plate were to become jammed somehow, it would have the effect of permanently locking the differential as you suggest. I've never heard of that happening in a NSX, but it's possible.

 

[cdscivic]

You guys are the best for continuing to reply, hondatech would not be as amazing as this. (Source: owned many hondas and was on HT for a long time, if you know you know!). I will again do my best here.

I am curious about the 'wobble'. Is the rear of the car actually moving from side to side or is it just shaking / vibrating? I don't see how a differential problem could cause the back end of the car to actually move back and forth sideways. If the differential is locking up (or has locked up) in turns the inner rear wheel might hop (more like skip) a little bit which could create vibration and noise.

Yes, it looked like it was dancing. A legit wobble probably a few inches each way (visual estimate) back and forth. It was shocking to witness.

Does this noise get worse with tighter turns?

I only drove it 2 times but to my panicked ears it seemed like it did.

You said the mechanic reported that the transmission oil appeared 'silvery'. That is never a good thing. The root problem might not be the differential; but, some other failure in the gearbox. If that failure has created 'stuff' which has found its way into the LSD clutches it may effectively be changing the coefficient of friction (just like Red Line MTL does) resulting in lock up of the differential. The manual transmission has an oil pump strainer that is accessible from the outside. If they didn't do this the first time, have your mechanic remove and examine the strainer for debris.

Interesting development. I have poured over the maintenance records and a placed called Mechanics Direct in Mass back in 2016 DID in fact use Redline Trans Fluid. I see no other records of transmission fluid change until my known service 2 months ago where Honda MTF was put in.

The mechanic is checking the strainer today, he drained the Honda MTF they filled up 2 mos ago yesterday (that was when he referenced the sparkly description). Will report back on this when he lets me know.


now on to @e9coupe :

put it on the lift with someone inside the car and run the drive train while the techs listen.

I have offered to come perform the driving, unsure if they will accept.

and finally @Honcho
Interesting, i wonder if the redline being in there for so long (allegedly) could cause a lock up once it was drained and MTF put in? Fixed with a flush or new diff?


Again, thank you all so very much for the replies. Seeing my dream car be a paperweight is basically killing me inside.

 

[e9coupe]

Yes, it looked like it was dancing. A legit wobble probably a few inches each way (visual estimate) back and forth. It was shocking to witness.

Hmm. This reminds me of a BMW circle track race car I had with a welded diff. When we'd push it in the shop and turn it would wobble like crazy since the wheels were trying to turn at different speeds but couldn't.

 

[Honcho]

Interesting development. I have poured over the maintenance records and a placed called Mechanics Direct in Mass back in 2016 DID in fact use Redline Trans Fluid.

I can't believe Mitch would be dumb enough to use Redline in there- he's one of the top NSX techs in the US. Perhaps the customer insisted?

Interesting, i wonder if the redline being in there for so long (allegedly) could cause a lock up once it was drained and MTF put in? Fixed with a flush or new diff?


Again, thank you all so very much for the replies. Seeing my dream car be a paperweight is basically killing me inside.

All Redline and the other non-wet clutch fluids do is alter the desired lockup friction between the clutch plates. I don't think that fluid could cause the drastic behavior you're describing. At worst, you might get weird behavior while cornering hard or excessive clutch wear. But, the shimmy you're describing sure seems like what would happen with a fully locked diff, e.g., full-time posi-traction.

The only other thing I could think of is that the diff bearing is somehow loose in the carrier. This would permit left-to-right movement of the differential in the transmission case and transmit that same movement to the axles. But, such a failure would be VERY noticeable, the transmission would be noisy, and you may not be able to engage any gears. Here's a picture of that bearing on the differential- there's one on each side. Maybe whoever opened the trans forgot to reinstall the bearing race in the trans case? They tend to fall out during disassembly...

 

[cdscivic]

This is the receipt I have, date is cropped but it's 2016.

I am not sure the trans was ever opened (definitely not during my short ownership this far).

The more we discuss, the more it seems like there is a differential failure as uncommon as it may be.

 

[e9coupe]

Not to change the subject but how do you like your SOS 275 clutch? I just put mine in last night but have never driven with one.

 

[cdscivic]

Not to change the subject but how do you like your SOS 275 clutch? I just put mine in last night but have never driven with one.

There was a different clutch put in in 2021, so I am unsure.

 

[e9coupe]

There was a different clutch put in in 2021, so I am unsure.

Well that's not very reassuring. LOL

 

[cdscivic]

Well that's not very reassuring. LOL

Lol it was still in good shape with the service I got a few months back. Can confirm this clutch feels nice but I am unsure what clutch it is other than clutch masters.

 

[Old Guy]

The NSX differential is actually fairly simple. It's a clutch pack that is released by a fixed spring pressure. At rest, the clutch pack is engaged, providing torque to both wheels. On cornering, when the torque split between the outside and inside wheel exceeds the spring clamping pressure, it will compress the spring plate and release the clutch discs, turning it into an open differential. On the "regular" NSX, this is about 50 lb/ft. On the NSX-R, it's about 100 lb/ft. It's not a coincidence that this is 2X the standard value- the NSX-R just uses two stacked spring plates instead of one, effectively doubling the spring rate.

View attachment 186083

If the spring plate were to become jammed somehow, it would have the effect of permanently locking the differential as you suggest. I've never heard of that happening in a NSX, but it's possible.

Which differential are you talking about? The OP has a 1996 model which should have the later differential design where the central gear and carrier assembly use helical cut instead of straight cut gears like the early cars. Helical gears create a side thrust when they are in motion which I thought engaged the clutch to stop the inside wheel from spinning. This is Honda's description

1995
A new torque reactive limited slip differential was developed to minimize spinning the inside wheel on NSX models equipped with the manual transmission. This unit uses a multi-plate clutch and new, helical-type planetary gears. When traveling in a straight, the amount of slip between the rear wheels is controlled by the force of a preset spring-loaded disc imparting a force on the multi-plate clutch. In a tight corner, however, the force of the spring-loaded disc is overridden by the thrust force of the new helical-type planetary gears, thus preventing the inside wheel from spinning and enhancing stability. In testing, this unit improved acceleration time out of a corner by 10%.

My interpretation of the operation of the '95+ differential is that it goes from allowing limited wheel speed difference / wheel slip (looks more like an open differential) to much less limited slip (looks more like a closed LSD) when rotation of the helical gears causes side thrust which compresses the clutch plates. The is more like the Dana clutch type diffs that I am am sort of familiar with - when the pinion gears go into motion thrust on the pinion shafts forces the pinion carrier halves apart which compresses the clutches which tends to transition to more like a closed differential. Like the pinions in a Dana LSD the helical gears have no relative motion until you get a speed difference between the left and right wheels. You want to allow for a limited amount of speed difference (open differential) during normal turns; but, once the inside rear wheel lifts and starts to spin the helical gears go into big time operation which then compresses the clutches transitioning to more of a closed differential.

I will admit that the role of the single clutch disc with the spring plate on one side of the carrier and the multi disc clutch on the other side of the carrier is a little opaque. I find planetary gear LSD operation to be slightly less intuitive than the old school Dana 'pumpkin in the back' differentials.

The pre set torque values of 40 - 100 lb-ft that I quoted were for the 1991 differential. The 1997 + service manual sets the pre-set torque test at 87 - 203 lb-ft so if anything cdscivic 's pre-set values are towards the lower end of the design values. The service limit is 43 lb-ft. I do not have the service manual that applies to the 1996 model year. It would be good to confirm that the values for 1996 are the same as for 1997.

 

[Honcho]

Which differential are you talking about? The OP has a 1996 model which should have the later differential design where the central gear and carrier assembly use helical cut instead of straight cut gears like the early cars. Helical gears create a side thrust when they are in motion which I thought engaged the clutch to stop the inside wheel from spinning. This is Honda's description

Ahhh, I missed that it was a 96. That LSD works in a similar way- the wheels start out locked and the helical cut gears gradually apply a side load during cornering that eventually compresses the spring plate to release the clutch pack, allowing the wheels to turn independently like in an open diff. Where the 91-94 is ON/OFF, the 96+ is gradual. Both diffs use a spring-loaded clutch pack however- the only difference is the helical gears. In fact, both diffs use the 41696-PR8-F00 spring plate to load the clutch pack. On the 96+, it uses a slipping clutch disc instead of the steel spring plate to interface with the spring. It's a neat design and definitely better/easier to modulate than the 91-94 style. I always wondered why they didn't use it on the R. An answer could be that the 91-94 diff stays 100% locked until breakaway torque, where the 96+ starts losing lock right away until breakaway. It seems they set the breakaway higher too.