Steering Issues

[dod1695]

The RX-8 has electric power steering, unlike many cars that have hydraulic, and when this is working well gives tremendous driver feedback as it is a major component in the overall handling package.

However, like any system it can have its issues.

On the electrical side the system demands a high current from the battery, so if the battery is tired or the connections are dirty/loose, you will experience issues with the steering.

There are also a pair of connectors under the airbox (just where the coolant tank overflows) that can also get corroded or loosened if someone's been working nearby.

In extreme cases this may cause the steering warning light to appear, but intermittent connections may just cause a sudden loss of power assistance - the steering still works but is heavy.

If this happens to you, check the battery and its connections first.
If these are okay then it might be the main body earth point - follow the black lead to here

battery ground.jpg (97.78 KiB) Viewed 10670 times

and check that this is clean and tight.

The other, more common, issue is that the universal joint at the bottom of the steering column gets dried-out by the heat from the exhaust manifold and this can make the steering stiff, either constantly, or intermittently, sometimes in one direction more than the other, something just off straight-ahead in either direction.

Even if you're not yet experiencing steering issues, it's a good job to give this joint some lubrication.

Note that this issue doesn't appear on the US Mazda forum, as their steering is on the other side, well away from the exhaust

To lubricate the joint...

• - jack the front of the car up so both wheels are off the ground
  - remove the driver's-side wheel to give clear access
  - unclip the rubber flap (equivalent to the one on the nearside for the plugs)
  - locate the steering column and follow it down to the UJ
  - put the key in the ignition to release the steering lock
  - turn the steering wheel from lock-to-lock and feel if there's any resistance
  - it should spin freely and a quick nudge should send it all the way. (it probably won't)
  - get a can of motorcycle chain lubricant (any brand as long as it's suitable for O-ring chains)
  - spray a generous coating over the UJ, turning it around by push-pull on the front disc.
  - repeat the 'turning the steering wheel' check and you should find that the action is much freer.
  - replace the rubber flap and front wheel.

When you test drive it, it might feel a little strange, especially if there was a marked difference between the before and after steering test, as you would have compensated in your driving.

You might have noticed an increase in "tramlining" or be finding it difficult to hold a smooth line on tightening or opening bends.

This is because, when the UJ dries out, there's a major increase in stiction (static friction) which then causes you to over-correct minor steering inputs as you overcome the initial resistance.

This process is also good for fault finding as it shows that the mechanical side of the system is sound, assuming there are no clicks or tight-spots afterwards, and any further issues are likely to be electrical.

[warpc0il]

See Vic's DIY item on lubing the universal joint

[warpc0il]

Ray recently had an incident with his steering, that may be a one-off or could be an indication of things to come, as our cars get older...

Car was fine last Sunday for the meet at Mazda HQ on Friday went to collect the car to go to the Essex meet only to find I had no power steering and an OBD reader borrowed from Shane at Sun Motors, adjacent to my garage, revealed a DTC B2278 'Steering Torque sensor malfunction which would not clear.

A call to Aston at Essex Rotary and he provided me with not one, but two power steering modules, to possibly sort the problem, which I picked up at the Essex meet.

So today was spent removing the ECU to access the steering module and replacing mine with each of the borrowed moules to see if this removed the problem. No joy with either of these, so replaced my original and set about removing the battery tray and air box and tray to gain access to the connectors at the torque sensor and the wiring loom. A circuit test gave a 13ohm reading which is within the 12-15ohm spec so all seemed to be in order and pointed to the torque sensor itself. Sprayed the connectors and reseated them and put everything back and prepared to park the car up until I could speak again to Aston to try and find a resolve. Lo and behold power steering back again, which is all good but does not reveal why it happened in the first place .

Hey Ho RX8 foible strikes again. A vote of thanks to Aston again for being at the end of the phone over a weekend to offer help and advice.

[warpc0il]

This was posted by PeteH and adds to the related knowledge base.

More people seem to be experiencing stiff steering, often only in one direction. This can be due to poor electrical connections, or a dry UJ, but it can also be traced to the steering rack.

I thought it may be useful to show where the steering torque sensor is, and suggest a possible reason for why it might go wrong.

The steering UJ connects to the rack via the pinion. On this pinion sits a dust cover:

1.jpg (241.78 KiB) Viewed 6681 times

This dust cover just slides off the pinion to reveal a plastic housing. This housing has the electrical connection for the torque sensor, a pinion seal, and behind the seal is the pinion bearing. This rack is from a 2006 car, and the amount of rust and muck in the area is quite significant.

2.jpg (261.79 KiB) Viewed 6681 times

The plastic housing is attached with three bolts. After removing the bolts the housing slides off the pinion to reveal this odd looking device:

3.jpg (249.88 KiB) Viewed 6681 times

There are a couple of pins which slide on a ramp arrangement:

4.jpg (80 KiB) Viewed 6681 times

If you apply torque to the steering wheel you twist a torsion bar at the bottom of the pinion. This causes the pins in the pinion to slide up the ramps, and causes the sleeve to slide up the pinion. If you apply torque in the other direction the sleeve will slide the other way. So the amount of torque you apply is proportional to the movement of the sleeve. This can be seen in the following video:


The amount that the sleeve moves is critical for the weight of the steering. If the sleeve doesn't move very far then the steering will be very heavy, and if it moves a long way then the steering will be very light. Mazda tuned the system very carefully, and you can see one of the important tuning methods on the pictures; there is a spring under the sleeve. The stiffer this spring then the heavier the steering. It should (in theory) be possible to change the feel of the steering by changing this spring. If I can work out how to remove this spring then I may have a fiddle with this.

The final part of the mechanism is the sensor itself. This is actually in the black plastic housing, and is a position sensor.

5.jpg (348.72 KiB) Viewed 6681 times

The metal ring on the sleeve slides up and down, and this movement is picked up by the position sensor in the housing. This generates the electrical signal which is interpreted by the steering ecu.

So that's where it is and how it works. What can go wrong? Well, there are two important seals that protect the workings; the pinion seal to protect the pinion bearing, and the o-ring that seals the housing to the rack. Looking at the one I took apart, both of these were only just managing to function, and corrosion was clearly starting to creep inside the housing. I suspect that if crud gets onto the pinion shaft then the movement of the sleeve could become unpredictable. If the pinion shaft is starting to rust then the sleeve could move downwards but not upwards (giving asymmetric steer weight). If the spring started to rust then the sleeve could move one way much more easily than the other way (again giving asymmetric steering weight). And if the pins were causing wear of the ramps then all sorts of odd steering weighting could happen.

If you have odd steering weight, and you are fairly confident mechanically, then it may not be necessary to replace the whole rack. It could be worth removing the black plastic housing from the pinion and just having a look. Perhaps a simple clean and lubrication of these parts could fix the issue.

Thanks Pete

[warpc0il]

There have been issues reported over the years of steering being "stiff in one direction", which hasn't been a mechanical issue with the uj, proved by checking with the wheels off the ground, but relates to the loss of power steering assistance in that direction, and is often related to the steering not correctly self-centering.

In most cases this has occurred after the car has been left for a long time with either the battery disconnected or it has been allowed to go flat. This in turn has resulted in a corruption/deletion of the Power Steering control unit memory.

Until recently it could only be cured by getting the car to a Mazda dealer who could use their factory diagnostic equipment to reset the memory, though many don't seem to know how, and the response is often "you need a new steering rack".

A new steering rack should cure the issue, as it forces a reset of the system, but at a fitted cost approaching (or in some cases exceeding, the book value of the car!

OC Members have determined that it is now possible to perform this reset yourself, using the Forscan program and appropriate connection hardware.

Details of this solution are provided here, for OC club members
viewtopic.php?f=58&t=83896

[warpc0il]

If you do find yourself having to replace the steering rack then this How-to-Guide has been recommended by a member.

http://www.jefflewis.net/rx8-steering_rack.html#1

It seems quite comprehensive and I like the style, which I would encourage to any of you considering creating for your next How-To feature.

However, my recommendation for this guide comes with three caveats, from me:
- I've not had to do this job on mine or anyone else's car, so can't compare with first-hand experience
- Jeff's car is LHD and there are some obvious, and less obvious differences, compared to RHD
- I've not read through the article in detail, and anyway I'm dyslexic...

plus the standard caveat from the OC applies, in that we (members, committee & officers) can't be held responsible for any outcome due to following this or any other guide.

[warpc0il]

Copied here from another tread, to make it easier to find...

I shall describe the most common fault in the RX8 steering system in the most "non technical" way i can.
First issue is that after many years of usage the torque sensor goes faulty and sends very wrong numbers to the P/S ecu.
The ecu having a very rudimentary 8 bit system thinks "hang on a minute,this can't be right,i must be broken"
What happens then is that you get a B1342 Ecu Internal Fault code and power steering stops working.
Now,because of that the ecu goes into a boot loop and even if you replace the rack or the torque sensor with a good or new one it will still not work unless you replace the ecu.Because of the boot loop the ecu will not entertain any torque resets as well.
What most people do is only replace the power steering ecu which will make the steering work but the joy will be for a limited time until that gets thrown overboard by the old steering issue.
The correct existing solution on the market is to replace the sensor or the rack with a new or known working part and replace the ecu as well.
I have not long developed a little software that can read and write the flash memory on the power steering ecu thru a pass thru device.
What that does is save owners from the hassle of having to replace power steering ecu when it goes into boot loop by reflashing the "broken" ecu with a good flash that has torque calculations set to 0 which basically turn that into a healthy working ecu.
I have succesfully tested this solution on three cars and upto now none of them had issues.
The solution might be worthless now as a power steering ecus can be widely located for a few quid but i am sure that as soon as the supply dries up there will be a few people interested.
Hope it helps