2004 LeSabre Limited - Loose Steering/Service Stability

[pcmos]

I've noticed recently that my steering wheel has what seems like an excessive amount of play. Additionally I keep getting a "service stability system" warning message on the dash. The code stored on the system translates into a steering wheel angle error. To try to clarify and rule out some common causes I'd like to start by saying this is not a "center-only" issue.
My first thought was that the rack and pinion was wearing out around it's center position (143k miles). After giving the preload adjustment a little tweak I was able to clean up any slop in the rack. Here is how it behaves:

1. If I sit still in the car and gently rotate the wheel from side to side it allows about 5 degrees of rotation in either direction before there is any reaction at the wheels. With the preload on the rack adjusted you can rotate the wheel about 5 degrees plus/minus and then you feel it engage the rack which is now nice and tight.
2. While driving the car it doesn't really matter if I am turning through a curve or driving straight on the highway, I have plus/minus 5 degrees of rotation of the steering wheel before I achieve any steering response. This means I am often swinging the wheel back and forth at ridiculous angles to make tiny corrections as I am driving.
3. I pulled up the rubber cover on the intermediate shaft where it splines onto the rack and pinion input shaft. My friend jiggled the steering wheel plus/minus 5 degrees while I held my hand on the intermediate shaft. I really couldn't tell if the play was in the shaft or in the rack.
4. This past week I have been getting a "service stability system" message. When I connect to the stability system using my enhanced scantool interface I find a code that translates into a steering angle mismatch code for the steering angle sensor. In other words, the car is detecting that when I am going straight, the steering wheel is cocked at a screwy angle.

Alignment has been checked multiple times, I replaced both outter tie rod ends and all of the hardware associated with the sway bar. Tires were rotated and rebalanced. Nothing seems to resolve this issue and it is really frustrating. I don't think there is anything wrong with the stability sensor because the message it is sending to the computer is exactly what I am actually experiencing. I agree with its assessment of the problem, and if I were the computer I would be saying the same thing.

Has anyone experienced the same problem? If so, what did you do to resolve it?

[Blue Bowtie]

...if I were the computer I would be saying the same thing.

Except probably in binary.

I have not experienced this problem, but have had an intermediate shaft get loose on a vehicle with no stability control other than the "computer" in the left front seat. The worn shaft exhibited a feel much like you described, with a large null zone of input before reaction. It was like driving a 5,200 pound '71 Chrysler Imperial on 78-series bias-ply tires.

Since you were unable to determine if the intermediate shaft was the source, and seem to have systematically eliminated most of the other common possibilities, I would suggest you revisit the shaft and diagnose it further.

The only other items which should be investigated which were not mentioned are the ball joints, lower control arm bushings, rack unit mountings, and strut mountings. They are common sources of steering play, but likely would not cause the computer to react as it has.

One further question - Does this car have electrically assisted steering?

[pcmos]

... All good suggestions. Factory shop manuals provide a pretty good functional description of the Stabilitrak system. There is an accelerometer device mounted between the two front seats inside the passenger compartment (probably multiple accelerometers in one box). This device feeds information back to the computer that is then translated into three axis acceleration data. The computer then compares the results of that calculation with the feedback it receives from the steering wheel angle sensor. When your car is sliding out of control, the computer can see that the actual motion of the car (from the accelerometer) does not match your steering input (steering angle sensor.)

The computer is programmed to self-check its onboard systems. When the computer detects that the car is traveling relatively straight for a long period of time but the angle sensor is indicating some degree of steering input, it trips the code that I am receiving. My guess is that any part of the suspension or steering that results in my steering wheel being at any position other than center while my car is going straight will yield the error code.

My LeSabre does have Magnasteer variable assist steering. I really wish I could get more specific details about the construction of the rack because it sure seems as though the intermediate shaft is turning the input shaft to the rack throughout this null zone. It's very difficult to tell unless I can find a way to hold the input shaft on the rack totally fixed while attempting to rotate the steering wheel. Isolating the loose link in the system is really tricky.

I examine the upper strut mounts while rotating the wheel and I don't see any play. When I am parked with the engine off (no power steering assist) but with the key in the ignition (no steering column lock) I can really hone in on the null zone. When you rotate the steering wheel through the null zone you don't see any motion in the tie rods at all, it's not engaging the rack in the least bit, so that sort of rules out a lot of the outer suspension equipment as the source of the trouble. Based on that observation I can probably rule out tie rods as well.

I replaced the Magnasteer rack in my 2000 LeSabre for a similar feel at about 75k miles before it progressed to this point, but I never really noticed a big improvement in the null zone issue. I eventually traded the 00 for the 04 before I reached the mileage I have now on the 04. All of the improvement that I experienced when I replaced the rack probably could have been achieved by making a minor correction to the preload adjustment on the original rack. I just didn't know about the adjustment at that time.

Throughout the day today I'm going to try to experiment with different ways to isolate the play as I attempt to zero in on the offending component. I sure wish that they didn't have such an outrageous core fee for the rack because I'd like to take one apart.

[pcmos]

Ok... I seperated the lower steering shaft from the rack and clamped a pair of vice grips on the input shaft of the rack. I can feel all or most of the play originating in the rack itself. I sure wish I knew more about the way that the rack is assembled and how it could be possible for the pinion/rack interface to feel so tight while the input shaft is loose. I guess I made the silly assumption that the input shaft IS the other end of the pinion. Perhaps there is something more complicated going on in there. It's also possible that there are bushings in the rack that have worn so badly that the pinion is twisting around inside. Whatever the case it looks like it needs a new rack and I'll probably replace the lower steering shaft at the same time... next weekend's project though.

[pcmos]

I finally replaced the rack and pinion unit with a factory rebuilt that I purchased as a "new old stock" item on ebay. The advantage to go going the ebay route is that I got to keep the old rack and tear it apart. Three main observations were as follows.

1. Replacing the rack did not make a great deal of improvement in the "null zone" issue. I still have about +/- 5 degrees of play in the steering. The lower shaft did not exibit any play at all, so I didn't replace it.

2. Upon tearing the old rack apart, I found that the input shaft is part of a larger inseperable assembly that includes the spool valve and pinion. When the input shaft is turned, it twists the spool valve and directs variable pressure to one side of the double acting hydraulic cylinder or the other. The amount of twist is proportional to the torque applied to the end of the shaft and therefore the amount of assist is also proportional to the feedback from the wheel ends.

3. The service stability system error was not resolved by replacing the rack. Ultimately I determined that the problem had to be a result of the steering wheel position sensor since I had this null zone issue in the steering long before I was getting a warning message. Replacement of the steering wheel position sensor resolved the stability system warning immediately.

My guess is that the spool valve is simply a wimpy design that is intended to provide maximum assist all the time. Since it is virtually impossible to disassemble, I'll go out on a limb and say that rebuilders tend to re-use them no matter what. The longer these racks are in service, the crummier these hard parts will become. Unfortunately there isn't an easy way to address the concern. It appears as though the problem with the null zone is inherent to this pinion design. I'm going to investigate the possibility of retrofitting a rack intended for another vehicle, although it may not very practical.