WOO-HOO! Grounding mod works! (02 WRX)

Great that you got it fixed! As an electronics person however, I'd warn that if grounding the engine made a difference, there might not be a suitable path to ground while cranking, and you don't want the cranking current running through your speaker wire. It could start a fire and wreck your car. That said, if you didn't have a good ground before... how did it start?

Every car should have a fat ground - as thick as your finger - taken direct from the battery to the chassis (some Suby ones are too long), and from the chassis to the engine.

If the grounds aren't solid, the ECU cannot do its job - neither sensors nor spark&injectors can work properly.

This could happen with the *negative* battery terminal connected to the strut and the manifold??

'cranking current' - that's a new term for me.... ??

The car always started fine; it just hesitated slightly at certain revs.

Clifford - here is the thread on the forum regarding this fix. Issues discussed concern sensors being very sensitive to EM interference, and this fix addressing that *somehow*.

I am very curious about your concerns of a fire hazard. If you'd be willing to review a bit of that thread, perhaps gleaning a clearer sense of what has been done, and confirm that you are still concerned, I'd like to hear it. Not that I'm asking for a primer in electronics theory, but just that I don't see how connecting 'inert' engine components to a negative terminal could constitute a hazard.

Thanks!!

While adding additonal grounds may in fact have some benefit, you should also bear in mind that if you disconnected the battery in the process of doing this mod you also reset the ECU. This will often make the car run better for a while because it returns the timing cells to their default values and clears all the "learned" values. You may find after some driving over the next few days or weeks that as the ECU re-learns, some of the hesitation may return. Also, note that the strut tower bolts are a poor place for a ground connection because the strut mounts which those bolts go into are rubber-isolated from the frame. There may be some ground contact between the bolt and tower but there is paint underneath. The most important grounding point is from the engine manifold to the battery.

I'd read that resetting the ECU is also a 'pursued strategy' for clearing the hesitation problem, and that a recommended next step would be to drive normally for 30 minutes or so, so that the ECU resets to typical values.

Interesting point about the ground. That makes sense... I wonder if these folks removed some of that paint? Or perhaps the bold does make contact with raw body metal in the hole.

Arguably I should have ensured that the negative lead did not come off the battery at all, and that I only did one wire at a time. That would have told me considerably more.

Another reason the strut tower bolts make a poor ground is that they do not actually screw into the frame but just pass through it. The ground bolts on the inner fenders and firewall DO thread directly into the frame, and are a more suitable place to attach additional ground wires. For better contact, even with the factory ground wires, you can remove the existing bolts and make sure that enough paint is removed underneath them to ensure a good metal-to-metal contact for the bolt and lockwasher. Don't overdo it though, you do not want a lot of exposed metal beyond the area covered by the bolt and washer. After you're done you can apply touchup paint or heavy grease over the connection to protect it.

Good advice. I'll see how this pans out before reassessing additional grounding.

Kind of took the wind out of my sails, however, realizing that it may have been the ECU reset that fixed it, not the ground.

Is this a chronic problem with these cars? Seems to me that no system should settle into timings that have this kind of hesitation... How much of this owes to how the driver drives as the timings are being established? And how much owes to a 'problem' in the ECU? I've read about models that have over-sensitive knock sensors, and that a TSB has been issued in some cases.

You'd THINK the dealers would know more about this stuff than anyone else. I described the issue, and they told me I should switch to 94 gas. I'm all but convinced that the octane was _not_ the issue here... ;(

I scanned it about halfway. Lots of opinions there, mixed with a little bit of science... here's my take (not being a WRX owner).

• The radio ham mentioned that the car generates *lots* of EMI. Cars normally do, but I assume he meant much more than expected. Thing is, if there's a way for EMI to get out, there's a way for it to get in, including the car's own RF emissions mucking up other parts of the circuitry.

• The fix that seems to work is the wire to the intake manifold. This manifold is where the MAF sensor and air temp sensor reside, so if there's "ground bounce" in the grounding there, these analog signals will also bounce around. If the ECU doesn't smooth out the noise *very* carefully, it could see semi-continuous false readings either above or below the actuals. That seems the most likely cause of surging and hesitation.

• Since the main battery ground strap isn't taken to bare metal, there's a distinct possibility of poor grounding. The battery in our RX died a couple of months ago and the mobile tech that replaced it also replaced the whole grounding strap, commenting that both the strap and the fixture to the chassis were the cause of endless troubles in Subies, and he always replaced the original fittings. Even the battery terminals are crap.

• If the main battery ground to chassis is weak and you fit alternate earth wires, then when you crank the engine, at least some of that 200A cranking current will take the new route home. If the wire's not rated for 200A, it'll get very hot, hence my concern about a fire hazard. Don't do this fix unless you also ensure that the battery has a really good connection to the chassis, and the chassis has a really good connection to the engine - preferably near the starter. I don't know how this engine grounding is done in Subies, but my Alfa has a 3mmx25mm strap of braided copper bolted from the rail to the bellhousing. I found this out after an engine swap when the bozo welded the exhaust before fitting the strap, with the welder earth lead on the chassis. The weld current found its way out through the alternator, cooking it totally, all internal wiring and the diodes.

Don't forget that engine parts are fitted between insulating gaskets, and the bolts that pass through are subject to surface oxidation. Aluminium alloy protects itself with a surface layer of *very* non-conductive aluminium oxide. Don't assume that just because two things are bolted together, you have a good path for high currents or RF.

Clifford Heath.

Why not put in a tank of the 94 octane as the dealer recommended, give it a spin, and see if your hesitation is better?

What octane is recommended in your manual?

~Brian

Manual recommends 91, I've been running 94 on the recommendation of the dealer. The 94 seemed to make a difference, but the issue is still there.

Sounds like the dealer may have been going for a bit of a longshot then. If the book calls for the 91, then you're probably fine putting it in, barring anything about the condition of the motor that could possibly raise the octane rating such as carbon deposits.

I changed plug wires on my Impreza to MSG, low impedance wires, and after about 3,000+ miles I started getting aweful hesitation between 2 - 4k RPM until I changed out the plugs to an irridium resistance style. Were grossly expensive for spark plugs, but the did the trick immediately. Next time I will just use normal plug wires and not have to worry, but these silly wires better last for the life of the car. (-;

~Brian

I did. The hesitation seemed to improve, but never went away. Over time, I kind of 'got used to it', and ignored it. I'm mainly getting in a twist about it now because it is a leased vehicle, the lease is up in a month, and I am weighing whether it makes sense to keep the vehicle or let it go. So every little niggly imperfection is bothering me now.

91.

[...]

'02 WRX manual "recommends" 91, with a minimum of 89 *where 91 isn't available.* It is silent on the benefits of 94. The '02 WRX does NOT require 94, and if yours does, then the gas you're buying sucks hind tit.

To the O.P.: please continue to apprise us of your situation.

Yes, I will do... the hesitation still appears to be absent, but I do feel a *slight* difference right at 3000 rpm. Sometimes. That's about where the turbo kicks in, so I don't expect things to be liquid smooth at that point in the power band.

What I'm not clear on, and maybe someone can help here, is - some have said that the effects of this mod will be transient, and that the hesitation will come back. What I'm not clear on is how quickly. The jury is still out, but will there be a certain number of miles before I'll know for sure? Presumably this will also depend on how I drive...

Regardless - if it _does_ come back, the mod comes out, the car goes into the shop, and I'll be saying to the dealer something to the effect of '94 my @$$'.

BD