Weird vibration. I would like help. Rx8

Soooo....  things are getting better and better with my 3.9 v6 swapped rx8.

When I accelerate between about 65 to 80 mph I have a slight vibration.  Doesn't matter what gear (keep reading, it really doesn't matter).

So, that's 1 thing.  But wait, there's more!

If I just let off the throttle, the car vibrates MORE as it decelerates down to about 60mph.  Okay, that's kinda odd...

But wait, there's more!

If I push the clutch in and let the car decelerate from 80 to about 65, the car vibrates ALOT.  Like wtf is that about to fall off.

So, learn me.  What the he'll is going on?

I have not yet jacked up the car to see what is happening.  I will check rear diff and Trans angles to see if it's a ujoint angle issue.  I will make sure the wheels are on and tight and not wobbly.

I am honestly not sure how long this issue has existed.  My normal driving never gets above about 65 mph so I haven't really noticed it.  When I did notice it, I specifically tried to endure it to get an idea of what influenced the issue.

I do *think* my driveshaft is a bit short.  Like 1/2" to 3/4" too short so that the yoke is less engaged than is probably ideal.  

It is independent rear suspension with a ppf going from diff to Trans.  They do NOT move in relation to each other.

I don't even play an expert on public access TV, but my hunch is that it's driveshaft related. Good call on checking angles, but I wonder whether it's a worn U-joint, thing, and it certainly fed the fire when you mentioned a potentially under-inserted (less supported, more angle per amount of play) yoke...

The indifference to gear suggests output shaft or later, right?

My hunch is that the less load you have on things, the more they get to wobble. Like if you're decelerating, the engine braking holds the yoke against the splines and limits the wobble, but when you put the clutch in it's just being driven from the opposite end and is free to dance to the limits of the fit at the trans.

My two cents' worth, for which I am overcharging you by as much as two cents.

Love it.  I understood the visual you were laying down there.

The driveshaft has about 200 miles on it.  Doesn't mean it isn't E36 M3.  

My only question is:

how do I test this theory?  I don't want to pay $280 for a new driveshaft to not have that be the issue.

Definitely sounds driveshaft related. Clutch pedal releases the driveline from the additional flywheel mass and the flywheel is there to add smoothness.

Driveshaft angle, shaft straightness, balance, and u-joint clocking all play a role.

Where did the shaft come from?

Checking angle isn't too hard. A good long carpenter level can be enough for straightness.

Any driveshaft shops around that can give it a once-over?

So when torque is not being transmitted it vibrates worse at those same speeds? If the u joints in the driveshaft look good and the axles aren't bent my mind goes to a chassis issue.  Something is loose but is looser when the engine isn't actively transmitting torque to the rear end and jamming the loose thing to one side.  What do the subframe bushings and diff bushings look like?

In reply to wvumtnbkr :

That's an excellent question.

I guess if you grab it at the yoke and it doesn't have discernable play, my theory is probably dead. No slop means that if there is an imbalance wobble it shouldn't be changing with load (and again, that's my theory, not based in any actual experience with spline-play issues).

If not worn splines on yoke, then there's the opportunity for wear on the output shaft, and the opportunity for the splines just not to have been made to a good fit.

Mayb try one of those phone apps to see if you can get a sense of whether it's shaking at driveshaft RPM or wheel RPM?

I think it's probably still worth it to check angles, and sanity check that the shop phased the joints correctly (how often does a shop put a joint on 90 degrees out? Would it even matter?)... Just thinking that even if play at yoke isn't the actual cause, that's still a sliding fit (there's some clearance) and the same principle might apply to vibrations that are caused elsewhere in the shaft. So I get to keep my load-damping theory even if the vibes originate from phase or angle?

I can't think of any way to take the shaft out of the system without taking the shaft out of the system, and towing the car to 65MPH to see if the vibes are gone sounds... bad. Fingers crossed the frequency check can help verify pre or post differential... But I'm still at a loss about testing for what aspect of shaft if it's there. Hopefully someone has a handy test. Or just more practial experience with this.

Mine had a vibration above 80 with 3.45 rear end. One of the diff mounts was cracked at the subframe, eventually I ripped off a diff mount ear and replaced the subframe, and most of it went away. Not sure if it was the crack or just re clocking the driveshaft that fixed it.

RacetruckRon said:

So when torque is not being transmitted it vibrates worse at those same speeds? If the u joints in the driveshaft look good and the axles aren't bent my mind goes to a chassis issue.  Something is loose but is looser when the engine isn't actively transmitting torque to the rear end and jamming the loose thing to one side.  What do the subframe bushings and diff bushings look like?

 

Brand new diff bushings.  Aubframe bushings were still pliable and not cracked.  Just checked when I changed doff bushings (trying to solve wheelhop....  it didnt...)

This is good stuff!  Keep it coming!

P.s. wouldn't wheel speed be the same as driveshaft speed just at a factor of 4.44 (diff ratio).  Would I be able to tell the difference since vibration is an acceleration unit and they accel and decel at the same rate?

RacetruckRon said:

So when torque is not being transmitted it vibrates worse at those same speeds? If the u joints in the driveshaft look good and the axles aren't bent my mind goes to a chassis issue.  Something is loose but is looser when the engine isn't actively transmitting torque to the rear end and jamming the loose thing to one side.  What do the subframe bushings and diff bushings look like?

 

What sort of chassis thing?  The front and rear subframe are tight.  Just reinstalled both in the last month and I made absoberkeleying sure they were good.  There is a ppf that I can check for tightness.

Can a rear wheel bearing cause this?  No looseness that I can feel by shaking wheels.  No growl from a bearing either.  Just a thought.

wvumtnbkr said:P.s. wouldn't wheel speed be the same as driveshaft speed just at a factor of 4.44 (diff ratio).  Would I be able to tell the difference since vibration is an acceleration unit and they accel and decel at the same rate?

Are we talking about the same kind of app? Uses the accelerometer to tell you at what frequencies vibrations are happening?

Yes, but that factor might help you rule out wheel/tire/wheel bearing/diff... not that they seemed likely, but a fairly definitive answer to whether the vibration is occurring at a frequency that matches engine RPM in whatever gear is 1:1, or matches that frequency divided by 4.44 will tell you equally definitively whether it's "output shaft through to pinion gear" or "ring gear through to tires."

Or... and I feel like I'm being a bit of a dog with a bone, if it jumps around is that indicative of it not staying glued to shaft rotation? i.e. if it was a fixed imbalance or out of true that would shake at exactly shaft speed, but I suppose if it's jumping around across slop in a junction, it could change its relationship to shaft speed?

Moreover, I think if it's straight imbalance, that's shaft speed, but if it's phase or angle, that should be 2x shaft speed because the orientation comes around that often?

So... Let's say you're in whatever gear is 1:1, you're doing 65mph at 3200 RPM and it's shaking. You look at the frequency output from the app and you see:

1. ~6400 Hz: Phase or angle
2. ~3200 Hz: shaft imbalance
3. ~730 Hz: Tire/wheel/half-shaft etc...
4. Noisy/scattered data (but maybe centered near 3200 Hz?) slop?

I'm almost certainly getting a little carried away here from a theory != practice standpoint, and my theory should be checked.

Description screams U joint angle issue.  When it is really bad, when it is at neutral throttle and not vibrating, you will be able to hear the ring and pinion rattling against each other.  That is because the driveshaft has more inertia than the pinion gear, so the driveshaft is running at a constant speed (which it can do because it is not loaded) and the pinion is spinning up and down twice per revolution.

U joint angles must be equal (can change signs or go the same direction as long as they are equal) and for the driveshaft speeds we are talking the absolute U joint angle needs to be under 2 degrees, preferably lower than one.

Playing with rearends in an RX-7 with short gearing has made me a wizard at this sort of thing  I make adjustable links and shim mounts this way and that.

Jesse Ransom said:

wvumtnbkr said:P.s. wouldn't wheel speed be the same as driveshaft speed just at a factor of 4.44 (diff ratio).  Would I be able to tell the difference since vibration is an acceleration unit and they accel and decel at the same rate?

Are we talking about the same kind of app? Uses the accelerometer to tell you at what frequencies vibrations are happening?

Yes, but that factor might help you rule out wheel/tire/wheel bearing/diff... not that they seemed likely, but a fairly definitive answer to whether the vibration is occurring at a frequency that matches engine RPM in whatever gear is 1:1, or matches that frequency divided by 4.44 will tell you equally definitively whether it's "output shaft through to pinion gear" or "ring gear through to tires."

Or... and I feel like I'm being a bit of a dog with a bone, if it jumps around is that indicative of it not staying glued to shaft rotation? i.e. if it was a fixed imbalance or out of true that would shake at exactly shaft speed, but I suppose if it's jumping around across slop in a junction, it could change its relationship to shaft speed?

Moreover, I think if it's straight imbalance, that's shaft speed, but if it's phase or angle, that should be 2x shaft speed because the orientation comes around that often?

So... Let's say you're in whatever gear is 1:1, you're doing 65mph at 3200 RPM and it's shaking. You look at the frequency output from the app and you see:

1. ~6400 Hz: Phase or angle
2. ~3200 Hz: shaft imbalance
3. ~730 Hz: Tire/wheel/half-shaft etc...
4. Noisy/scattered data (but maybe centered near 3200 Hz?) slop?

I'm almost certainly getting a little carried away here from a theory != practice standpoint, and my theory should be checked.

I don't know.  I was just theorizing.  Do you have an app recommendation for a Droid phone?  I'm willing to try it, do the math for various driveline velocities and open it up to interpretation from the group!

Pete. (l33t FS) said:

Description screams U joint angle issue.  When it is really bad, when it is at neutral throttle and not vibrating, you will be able to hear the ring and pinion rattling against each other.  That is because the driveshaft has more inertia than the pinion gear, so the driveshaft is running at a constant speed (which it can do because it is not loaded) and the pinion is spinning up and down twice per revolution.

U joint angles must be equal (can change signs or go the same direction as long as they are equal) and for the driveshaft speeds we are talking the absolute U joint angle needs to be under 2 degrees, preferably lower than one.

Playing with rearends in an RX-7 with short gearing has made me a wizard at this sort of thing  I make adjustable links and shim mounts this way and that.

I'm familiar!  I have a 3500 v6 swapped in my 2nd gen and this in the rx8.  When I initially set it up, I had 1.2 degree of difference.  I did  change diff bushings and engine height since then.  However, neither should have changed the relationship because I am running a ppf.  I will verify in the next few days.

I'm going to guess I'm *way* more likely than Pete to be onto red herrings, and i don't have an app in mind.

 At least it's cheap and easy to peep the shaft and sanity check the shop didn't put the joints on out of phase (the yokes should mirror each other, ] and [, not ] and -, right?). Phase and angle both result in uncanceled accelerations, so I'm hoping that's not at odds with Pete's reality-based diagnostics?

wvumtnbkr said:

I did  change diff bushings and engine height since then.  However, neither SHOULD have changed...

I've been burned by this too many times on projects both in hobbies and professional life. Trust, but verify.   The PPF is a big long aluminum C channel, it can bend, but it shouldn't. You did raise the engine and remove slop from the diff mount with new bushings.

I did something similar recently on my truck. I built the rear subframe mounts with the original 1988 C4 diff bushings and a 2 years later swapped in fresh poly mounts. I had to die grinder one of the bolt holes on my batwing mount, I should probably double check my pinion angle now.

Just posting to note that I'm curious as to what you find out.

I fouled up yesterday and missed my chance to try vibration diagnosis via app in the pickup truck (if I weren't about ready to sell the truck I'd be about ready to bet a set of tires that it's cheap old tires), but I kicked off both directions on the freeway without downloading an app first.

I did not yet solve the issue.

I decided to rebuild the engine....   since they apparently like to have oil in them...

Drove it for the first time since December.  Vibration is still there.  However, I did get get a new driveshaft made that is longer.  Same thing.

I checked angles (I can check again pretty easily).  I wanna say they were with 1 degree of each other.

I read a bunch of rx8s have had bent wheels, or bad wheel bearings that lead to a vibration at the same speeds I am noticing.  I will dig into this further.  If it's a wheel bearing, I would think it would be a rear since pushing on the clutch makes such a difference.

Can't feel any play in any of the bearings, but I've had loud bearings that were obviously shot that didn't have play either.

In reply to wvumtnbkr :

I remember reading this thread about a Sirometer (VibraTach is the common/cheap brand name available):

GRM link

Maybe it'll help diagnose it.

Well....  that made me think a bit.

I do have a vibration monitoring FFT device and a couple of accelerometers.  Wonder if they would stay in place at 60mph...  wonder if I could have my wife drive the car while I watch the display.  Hmmm.

Not sure why I didn't think to use the equipment I use at work...  ya know, to track down vibrations...

I'll see if I can borrow the device on Wednesday evening.

wvumtnbkr said:

Not sure why I didn't think to use the equipment I use at work...  ya know, to track down vibrations...

Haha :)  

Are the wheels tires balanced to 0? (both of my RX8s hated the slightest imbalance with a red fury) 

Have you confirmed the balance on your driveshaft and CV shafts? 

I have not confirmed balance on the driveshaft or axles.  I'm not sure how I would do that.  I guess I could pull them out and take them somewhere.

Are you measuring driveshaft angles in both planes? Ie up/down but also side to side?

You could have a big imbalance side to side that isn't as easy to spot as up/down. 

Robbie (Forum Supporter) said:

Are you measuring driveshaft angles in both planes? Ie up/down but also side to side?

You could have a big imbalance side to side that isn't as easy to spot as up/down. 

Interesting thought.  That is gonna be way harder to figure out!

The engine is offset about an inch to the passenger side to give clearance for starter, steering, etc.  Perhaps when I raised the engine and changed the diff bushings that was enough to make the difference.

Hmmm....  two t squares should get me close I would think.