Diagnosing rack and pinion steering problems
Copyright 2015 by Morris Rosenthal All Rights Reserved
Fun, intelligent SciFi without wars or potty-mouth heroes.
- Diagnosing Ticking Noises
- Car Won't Start Flowchart
- Engine Overheating Diagnosis
- Brake Problem Flowchart
- Troubleshooting Car Steering
- Finding Short Circuits and Electrical Failures
Troubleshooting vibrations, car tire wear, steering difficulty and wheel looseness
Does the steering wheel feel loose, like it has slack in it that needs to be taken up when you turn? Even more importantly, do you feel looseness in the wheels if you shake a front tire back and forth with the car sitting on the ground? Looseness here doesn't mean that everything moves, unless your car is very heavy, shaking a front tire will cause movement. The question is whether or not the wheel moves without the whole car moving, and if you get any sort of loose clunking or clicking sounds when you shake the tire.
Return to Steering Diagnostic Chart
If shaking the tire shows that the wheel is loose, the first thing to check is that the lug nuts are all tight. If any of the lug nuts turn out to be loose, it's a good idea to remove the wheel and take a good look at the wheel studs to see if they've been damaged or cross-threaded. Studs are cheap enough to replace, every auto parts store carries them. Remount the wheel, run the nuts up snug, and follow a star pattern for the final torqueing when the car is back on the ground. Lug nuts shouldn't require loctite to stay on, so if you have a repeating problem with your lug nuts, you might want to replace them or the studs.
Return to Steering Diagnostic Chart
Does the wheel still shake after the lug nuts holding the rim to the wheel assembly are tightened? It's possible that the axle nut has loosened up, but you'll have to remove the little hub cover to check. The axle nut can't loosen up too much because it's captured by a castellated retaining washer and a cotter pin, or the axle nut itself may be castellated. But if the cotter pin was improperly installed or damaged, the axle nut might loosen up enough to allow the wheel to shake. You should never need to run the nut in several turns, if it's that loose, it's probable that the wheel bearing has been damaged or disintegrated. Ultimately, the wheel is only attached to the vehicle by the one nut which holds the axle in the wheel bearing. The lug nuts only attach the wheel to the mounting hardware, which is held in place by that axle nut. Otherwise, the wheel wouldn't be able to spin.
Return to Steering Diagnostic Chart
Since this flowchart is for MacPherson strut rack and pinion steering only, there's no upper control arm or upper ball joint in the picture. The MacPherson strut is the single point of attachment for the upper part of the steering knuckle. The spring and the plates that hold the coil spring in place, along with any rubber mountings, are pretty reliable unless there's a manufacturing defect or corrosion failure. But people do report broken coil springs from time to time, sometimes due to a noticeable loud rattle when driving down the road, though I'd expect the corner of the car to sag as well. Rubber bushings may fail with age in some climates, and it's also possible for the unibody shock tower to rot out so badly that the top strut connection is no longer secure. MacPherson strut failure is normally due to the strut damper (shock absorber) part of the assembly failing. If there's an obvious oil leak from the strut damper, it's bad, but minor leaks may be normal (at least according to my old shop manual). There shouldn't be any looseness in the mountings, if you can twist around the top of the strut (shock), the rubber bushing must be gone. The most basic test for a bad strut is just bouncing the corner of the car up and down. If the strut is bad, it will take more than two or three up and down moves for the car to settle after you stop pushing it.
Return to Steering Diagnostic Chart
If you jack up the car and get the tire off the ground, does shaking the tire with one hand on the top and the other hand on the bottom shake the whole wheel, produce little clunking or clicking sounds? That's typical for a failed lower ball joint. Ball joints look pretty much like the name, it's just a captured ball with a post sticking out the top or the bottom, which moves (very stiffly) in its housing. If you wiggle the grease fitting and you get movement, the ball joint is shot (unless the grease fitting happens to be breaking off). The lower ball joint is the main connection holding the steering knuckle (and thus the wheel) onto the car. The other steering knuckle connections are more about control and alignment. If the rubber boot on the ball joint is broken and it hasn't been greased on a regular basis, it's probably bad.
Return to Steering Diagnostic Chart
Is the tie rod loose? With the car sitting on the ground, can you shake the tire left to right, and feel looseness with a soft clunking or ticking noise? With the car jacked up and the tire off the ground, does shaking the tire the left to right produce more looseness and clinking than shaking the tire from top to bottom? The left to right looseness is a sign of a worn component in the steering linkage, which is what controls the wheel movement in the left-right direction. It could be the tie rod end, which connects the tie rod to the steering knuckle through a ball joint, it could be the inner tie rod joint, which attaches the tie rod assembly to the piston end coming out of the rack, or it could be looseness in the rack itself. I have an illustrated page with photos and videos for diagnosing tie rod problems, and replacing the tie rod end or the inner tie rod.
Replacing the inner tie rod usually requires a special tool, but it's only around $50, or you may be able to borrow or rent one from your local parts store. The real problem is that you may have to remove the rack in order to replace the inner tie rod, if it is staked or pinned to the shaft. There's generally not enough room to swing any tools around the inner tie rod joint, but you may be able to get it if you have large a large crows foot and a couple feet of extension for your ratchet. The job isn't that complicated, but any time you remove the outer tie rod, you should count the exposed threads or take a good picture first so you can reassemble it to exactly the same length - otherwise you'll change the toe and need to realign the the wheels.
Return to Steering Diagnostic Chart
I've never had a steering column completely apart, though I've had to pull the steering wheel to replace the blinker assembly. Between the steering wheel and the pinion gear in the rack, which is the gear that translates the rotation of the steering shaft to the horizontal movement in the rack, a coupling is required. My current car uses two universal joints in the steering coupling, tilt steering systems require something more complicated, a ball type joint. Couplings all wear over time, so it's possible that looseness you feel in the steering wheel, especially if it feels like there's a little slack that needs to get taken up when you go from one direction to the other direction, is a worn coupling. But a worn coupling between the steering wheel the the pinion gear shouldn't translate into a looseness in a front wheel when you shake the tire.
Return to Steering Diagnostic Chart
Of course, any problem relating to looseness or clunks in the steering can be due to the entire rack moving. There shouldn't be any looseness in the rack, not if you shake it by hand and not if you watch it while having somebody steer the wheels back and forth with the car sitting on the ground. That can happen because mounting bolts were left out following a repair, because the mounting bolts to the crossmember were improperly tightened, or because that crossmember mounting points on the unibody frame have rusted away. The latter case, the unibody rusting out so badly that the one of the crossmember bolts wasn't holding anything happened with my Omni. I've heard of people driving cars where the rack was basically just hanging in place, supported by the tie rods, and steering couplings. You'd have to be pretty whacked to drive around like that, and I can't imagine the joint to the pinion gear could last very long if it was the only mechanical point steering the front wheels.
If the rack is tight, the tie rods are good, and there's no looseness in the steering wheel or in the front wheels of the vehicle, the jerks in the steering, stuttering or surges you feel when turning may be due to a power steering pump problem. The first check is always to make sure you have enough power steering fluid, and that it isn't all frothy and full of air bubbles, or overheating and overflowing. A loose drive belt on the pump can be at fault, or the pump itself may be failing. Power steering pumps are vane type pumps, which are highly reliable, but the shaft bearings can wear out, especially if the belt is too tight, or a vane can stick, and fluid contamination can cause failure. Note that some people change power steering fluid on a regular basis, as often as they change their engine coolant, like every 30,000 or 50,000 miles. As I write this, I don't believe I've ever even made up fluid in my power steering, and that's 269,000 miles. But now that I think of it, I'll take a look later and see how the color is. New power steering fluid would either be pink or honey colored, and while you expect those colors to fade, black or brown would probably be a sign that it's overdue.
Return to Steering Diagnostic Chart
Does the steering wheel pull to one side or the other as you drive? Equally important, if you're traveling down a straight, level road, does the whole car pull to one side or another if you take your hands off the wheel? It's important to check (don't do it in traffic or on a residential street) because some power steering systems eliminate all of the road feel, so you wouldn't even know that the car was trying to pull in one direction. If you notice the car wandering back and forth, or it seems to pull one direction or the other inconsistently, it could be that toe is set incorrectly on both wheels, and an alignment is in order. Wander may also be due to loose or worn wheel bearings, or worn out bushings on the lower control arms, where they connect to the crossmember. Check for uneven tire inflation before moving on.
If the pulling only comes with hard acceleration on a front wheel drive car, you may have torque steer. Torque steer is caused by the differential sending too much power to one wheel on acceleration, normally because it senses a positive difference in the mass or tire diameter too that side. Differentials like to send more power to the side with the best traction, and can be fooled by the front tires being missmatched, or one tire being heavily weighted. The only time I ever had an issue with torque steer is when I was troubleshooting whether a vibration was caused by a bent front axle, and tried to balance it out by counterweight the axle with a custom lead weight I fashioned. Even though the weight was only a few ounces, taping and clamping it to the axle where it was bent in caused very noticeable torque steer in that direction.
Return to Steering Diagnostic Chart
Did you hit something in recent memory, like a pot hole, curb, or another car? Driving into a curb or a pothole can lead to shock force that bend on of the steering/suspension components, messing up the alignment. Sliding into a curb sideways on snow or ice is a great way to mess up the alignment. The tie rods and steering knuckle are steel parts which don't bent easily, but collision damage can also tweak the unibody in such a way that the mounting points for the McPherson struts or crossmember shift, which changes the whole geometry of the suspension and steering. Whether you can use the variability in the toe and camber adjustments to get the wheel orientation right again without having to have the "frame" straightened will depend on how badly the unibody is tweaked.
Return to Steering Diagnostic Chart
Do one or more of the tires wear unevenly? Tires should wear evenly, so that the treads at the outer edges of the tire don't wear faster on one side or the other. The treads in the center of the tire should wear at the same rate as the treads at the edges as well. Any time the tire treads aren't wearing evenly, you have problem that needs to be corrected. The first check is whether the tires are inflated properly. If the treads at the center wear out quickly, the tire is likely overinflated. If the treads on the edges wear quicker or get beat up, the tire may be underinflated, or you may drive like teenager - turning too hard.
For a car that's pulling, look for feathering on a front tire, where the treads are all worn at an angle, across the tire. Feathering is a sign of improper toe in, which can cause the car to pull in one direction. The toe in adjustment is by loosening the jam nut and turning the tie rod end. If the wear is all on one side of the tire, that indicates that the camber, which leans the tire in or out from the vertical, is bad. The camber adjustment is where the MacPherson strut attaches to the steering knuckle at the top. If you can't afford to take the car for alignment, the poor boy method is slapping a level on the tire to adjust the camber, and rolling the car back and forth in the driveway to set the toe. Strange cupping, valleys and other radical tire wear are likely caused by worn out suspension parts, like a failed strut damper (shock absorber) in the MacPherson strut, failing ball joints, or bent parts.
The problem with using a level for camber is two-fold. First, radial tires bulge out at the bottom, so if you only have a long level, you really need to use it in conjunction with a block on the wheel studs to get you distance from the rubber for a true reading. And unless you have a lift, you'll have to jack the wheel off the ground to gain room to make the camber adjustment, and the angle definitely changes when the wheel and tire are dangling on the suspension. What seems to work for me is checking the level on the ground, jacking the car up, and checking it again in the air. Then I can see that there's a half bubble or three-quarters bubble difference between the reading with the weight on the tire and the suspended level reading. So when actually changing the camber (by loosening the locking bolt and rotating the strut cam bolt) in the air, I build in that difference. When I put the car down again, I double check, and at least the last time I did this, it came out perfect.
Setting the toe is much simpler from a wrenching perspective, all it takes is two open end wrenches of the right sizes to hold the outer tie rod from moving at the steering knuckle joint while breaking loose the adjusting nut where the inner tie rod screws into the outer. To actually turn the inner tie rod in or out, there's usually grooved or finned area on the shaft to make it easy to hold with vice grips and turn. It should turn very easily, and I find it I drive up on the curve to make head room, it only takes me a couple minutes to make an adjustment. It's easy to tell when the toe-in or toe-out are the same on both sides because the car will go straight if you let go of the steering wheel on a flat road. The problem is, you could have both wheels toed in or toed out too much. Both cases will probably cause high speed vibrations on the highway, more so in turns than going straight as the tire fight against each other and are forced to skip at high speed. You'll also get a funny tire wear pattern. The only trick I know for checking toe-in or toe-out is visual inspection with a long line, so slap a 2X4 up against the tire or use a string line, and see if it's angling away or into the wheel if you move several feet in front of the car.
If your brakes are dragging on one front wheel, the car will pull. Another possibility is the steering gear valve that apportions the hydraulic force from the power assist to the proper side of the rack to push the piston out. If the steering gear valve is leaky, you may be holding the wheel straight against the force of the power steering pump. Theoretically, you could take off the power steering pump belt and drive as a test, though steering would be pretty hard. If it doesn't pull with the power assist disabled, you know that the power assist is at fault. I think that racks with power steering are designed so they can be driven without damage if the power steering fails, but don't assume, check for your model. And of course, if you have a newer car with a serpentine belt driving all your engine accessories, this wouldn't be a good test to try:-)
Return to Steering Diagnostic Chart
Is the car steering difficult, especially when the car isn't rolling? Any failure with power steering, from low pump pressure to low fluid can be at fault. Check the fluid, check the belt tension, make sure there aren't any obvious leaks. You can get a pressure gauge and check the pressure the power steering pump is generating, but you'll need you shop manual to find out what that pressure should be, along with any cautions. My Dodge manual warns against testing the pressure for more than five seconds, the pump just isn't designed to drive into a dead end load. There's a control valve right at the pump that keeps the pump from putting out too much pressure at high speeds, and there's the apportioning valve at the rack that sends the hydraulic pressure to the proper piston. If the steering is still hard when the car is rolling, low tire pressure may be the cause. Make sure that the ball joints on the tie rod ends are greased.There could be too much friction somewhere on the steering shaft, due to a bad coupling or some foreign object lodged between the shaft and the body.
Return to Steering Diagnostic Chart
Check the belt tension. A loose belt will squeal, especially in cold weather, and then quiet down when the belt heats up from friction and the rubber shrinks - remember that rubber has a negative thermal expansion coefficient. Check the power steering pump mountings, and if the noise is clunky in nature, check that the rack isn't loose (see the other side of the chart). If the noises are more like clicks and clunks, see the flowchart that includes for front end noises. But if the noise is more like what you hear from a faucet being shut off, a high frequency hiss, it may just be normal for your car. All power steering makes some noise due to the fluid getting forced through the valves, it's just a question of how well the passenger compartment is isolated. When you turn the steering wheel all the way to the stop on one side or the other, the noise will be the loudest, and you shouldn't hold the wheel cranked all the way to one side for extended periods of time.
Return to Steering Diagnostic Chart
Do you see power steering fluid dripping out of the rack, from the pump, or the hydraulic fittings? Do you need to continually make up the level for the pump to operate properly? If you're losing power steering fluid through a damaged or worn seal, it may be difficult to determine where the leak is actually coming from. You have to wipe down the rack assembly and lines with paper towels or rags, and use a flashlight to look for the leak in action. If you can't find the leak but you know it's happening from loss of fluid or spots on the floor, it may only be leaking when the wheel is turned one way or another. This means getting an assistant to turn the wheel with the car running so you can watch, so unless you own some great jack stands and are used to sticking your head under a running vehicle, you might want to give up and take it to a shop with a lift.
Return to Steering Diagnostic Chart
Does the car vibrate a high speeds? Vibrations while the car isn't moving are do to engine or automatic transmission/transaxle problems, and vibrations at low speeds should only occur if there's some gross looseness in the wheels. If you can feel the vibration in the steering wheel but not so much in the car body, ie, your feet aren't bouncing on the floor, it may be in the outer tie rod joint. If the whole car shakes, it's likely causes by a rotating part that's out of balance and amplified at a particular speed. Tires and axles are the primary culprits when the whole car starts shaking at highway speed, but drives smoot as silk around town.
Return to Steering Diagnostic Chart
Vibration better or worse with tire rotation? Vibrations at higher speeds are most often caused by improper wheel balance. This could be due to new tires that were poorly balanced, or tires where the balance weights have fallen off or been scraped off on curbs. It's also easy to pick up a highway vibration if you've bent a rim hitting a curb or a pothole. You might also have a radial tire with an manufacturing flaw, so that it will go out of balance as it fails, even though the rim weights are untouched.
Return to Steering Diagnostic Chart
If the tires show visual damage, they need to be replaced. Check the rims for new dings and dents, make sure any aesthetic features (hubs) that are bolted onto the wheel are evenly mounted and not loose. Otherwise, take the two wheels with the best tires into a garage, have them spin balanced, and but them on the front of the car when you get them back. If you still have front end vibrations, that you probably have looseness in the suspension or ball joints you haven't spotted yet, an axle problem, or a potentially serious engine problem.
Return to Steering Diagnostic Chart
Check the CV joints and the axles. If anybody else drives the car, it's possible they crashed into a curb or a pothole and didn't notice the difference or mention it. You don't want to keep driving a car with bad vibrations, the handling will be poor, it may be a sign of something about to fail radically, and it may lead to other components loosening up. If the vibration is only present when you hit the brakes at speed, the problem is likely a warped rotor (disc brakes) or a out-of-round drum (drum brakes). See the flowchart for brake problems.