Skywalker

Car is a 2004 w/ 67K miles. Stock suspension.

After getting new tires installed (Continental Extreme Contact DW 245/40-18 - old tires were Hankook RS-2 235/40-18), I have been noticing a subtle pull to the left upon accelerating and a subtle pull to the right on decelerating. No noticable pull, other than due to road camber, ruts, etc., on braking. Searching the forums has turned up posts with similar symptoms, but in each of those cases where they posted their solution, they found a mismatched tire - either size or brand. I've checked and double checked and all my tires are the same brand, size and are at the same pressure to a fraction of a psi.

To the best of my ability, I have visually inspected the F/R suspensions and the only abnormality I've found was torn boots on all 4 sway bar links. New links have been ordered, but I doubt this will fix the issue (fingers crossed, though).

Once I get the links replaced, I will get the alignment checked. In the meantime, I'm looking for other things to check so that I don't waste my time on the alignment if there's something else wrong. One poster on one of the posts I found hinted at some more obscure issues that could have the same symptoms, but they didn't elaborate. I'm looking for those obscure issues so that I can check them out and fix if necessary before taking the time and spending the money to get an alignment.

Wingznut

iTrader: (1)

Try rotating the tires (front to back) and see if the situation changes, better or worse.

fuztupnz

iTrader: (4)

If rotating tires front to back doesn't change anything, check that you don't have a hanging brake caliper and that they all move easily on the slides.

Skywalker

I had thought about rotating the tires, but never got around to it. I'll get it done this weekend, if not sooner. (Probably this weekend, though.)

I can kill two birds with one stone - I can inspect the brakes while I'm rotating the tries.

The brakes were recently redone a few months back - pads, rotors, greased slider pins, new boots for the pins, brake hardware, etc., so I hope nothing's amiss there. Plus, I could see how a dragging caliper could cause a consistent pull to one side, but I can't reason how it would cause a pull one way on accel. and another way on decel.

Thanks for the advice. Please keep it coming!

Brettus

iTrader: (2)

You have bought one of the few FWD RX8s ever built .
Be happy that you own such a rarity .

RIWWP

iTrader: (2)

Rear tire pressures too. I had this issue shortly after getting my Miata. The tire pressure difference was changing the rolling diameter of the tire enough that it was actually engaging the LSD and sending more power to the higher tire pressure wheel.

godesshunter

What he said... sorta... You may have a torque push from the rear. (commonly known as torque steer in FWD) I suggest taking a real good look at all your rear suspension bushings. Specifically the ones on the right rear. Opposite your pull.

During torque push, the right rear may be getting pushed back a bit more than the left side. This will throw out your thrust angle forcing the car to "steer" to the left from the rear much like a fork lift.

When you let off the power, the suspension relaxes and goes back to a neutral state giving you the feeling that the pull has changed directions.

It may be difficult to determine with the car jacked up and suspension hanging. You may have to have a shop with a 4 post rack check it out from underneath. Any shop you got to for an alignment will have this type of lift so be sure to mention it to them before the alignment is done.

You can do that whenever. Sway bar links do not effect alignment so there is no need to wait for them to be in. Good luck. Let us know what you come up with in the end.

Skywalker

Thank you!

I was talking with a friend at work who races karts and he had a couple of interesting suggestions from his karting and other automotive/racing experience. He suggested measuring the actual circumference of the tires and seeing if there were any differences. If there are, put the two tires closest in circumference on the rear. He felt there could be differences in manufacturing lot, etc. that might be at work here. Also suggested that if a bead wasn't seated properly, it could affect the tire such that the circumference could be off.

All easy enough to check once the car is in the air.

Given my work schedule and the need to make an appointment with my preferred alignment shop, Friday a week from now will be the earliest I could have it checked/aligned, so (weather permitting) I can have all this done at that point.

I will post an update if/when I get this figured out.

Bladecutter

Way back, I had similar behavior on a Porsche Boxster when the right rear strut had failed, and leaked out all its goodness.

Once the rear struts were replaced, the car tracked straight on accel and decel.

BC.

godesshunter

True but you would feel the high spot bouncing the car as it hits the road during revolutions.

Did you ever find the culprit?

Skywalker

OK, just came in from finally doing a more thorough inspection with the wheels off (and replacing the sway bar links). Used a tailor's tape measure to check the circumference of the tires and found the following:

LF 81-3/8" Date code: DOT AF2P 3VX 1909

RF 81-2/8" Date code: DOT P52P 3VX 1711

LR 81-2/8" Date code: DOT P52P 3VX 1711

RR 81-6/8" Date code: DOT AF2P 3VX 2009

(I left everything in eighths to make the differences obvious and noted the mfg codes to see if there was a pattern.)

The fact that the RR tire was a half inch bigger in circumference amazed me. It fits the symptoms as well. (Imagine a cup lying on a table with the larger, open end on the right. If you push it, it will "turn" left.)

Notice also that the two tires with the same circumference were made the same month.

Anyway . . . I swapped the right side tires. Now the tires with the same measurement and date code are both on the rear. I'll test drive later and report back. If this fixes it, I guess I'll have to go back to the tire store or Continental and see what my options are. My concern is that when the time comes to rotate the tires, I'll be back in the same boat.

Also checked the shocks the best I could and saw no obvious signs of leaking. Fronts were easy, just lift the boot. Rears weren't as easy as the top of the shock body is hidden by the car body even with the suspension at full droop.

Skywalker

Well, I'm pretty sure I've got the problem licked for now. Drove the car yesterday afternoon and couldn't detect any pulling. Flat road, 3rd gear, hands off the wheel, foot to the floor - no pulling. I'll drive it again tomorrow on my usual route to work and that will be the final test to see if I notice the pulling.

Now, to see if I can get that tire exchanged for one that is the proper circumference.

Goes to show how sensitive and precise the RX-8 is that a ~6% difference in circumference would be so noticeable and annoying. It is comforting to know that there isn't anything noticeably wrong with the suspension, although I would have almost welcomed an excuse to replace the shocks, because you might as well replace the springs while you're in there, right?? :-)

RIWWP

iTrader: (2)

6% more torque to one wheel over the other is plenty noticeable ...

Skywalker

Some info on DOT code decoding I found. Turns out I have tires made in two different plants according to this:

First locate the letters "DOT" on the sidewall of the tire. Nearby will be the DOT code. DOT codes are 10 to 12 digits long. BTW the digits can be numbers or letters.

The first 2 digits are a code for the manufacturing plant.

The next 2 digits are a code for the tire size.

The next 3 or 4 digits are a code for the type of tire.

The last 3 or 4 digits are the date code. The format is week/week/year/year or week/week/year. These are always numbers.

Starting in the year 2000, the date coding used was 4 digits. That means the largest number you should see for the year is 09. Before 1999 the format was 3 digits. 1999 and 2000 are transition years, so you will find both 3 and 4 digits.

The date code only has to be on one side – and it is permissible for there to be a partial DOT code, so long as one side has the complete code.

mambotaino

I had the same pulling to the right problem on acceleration. I had the rear right with less than 20 PSI and the remaining three with 33 PSI. corrected the rear right to 33 PSI and test drove it and the problem is gone. It was that simple of a fix for me. Thanks to everyone on the forum and I hope this helps.