Recently did a vacuum hose replacement job on our (nearly) stock '94. The idle had been missing some, but the hose job did not cure it. The plugs were 12 K mile-old NGK BUR7EQP (Leading) and BUR9EQP (Trailing). After chasing the idle miss elsewhere I threw up my hands and replaced the plugs. I found that the #1 rotor plugs had a brown sticky substance over almost the whole length of the insulator, from the center electrode all the way down to the hex of the plug body. Most of it was glommed around the hex. This was mostly on the leading plug but some also on the trailing. The #2 rotor plugs were clean. Replacing the plugs seems to have cured the idle miss problem.

The business ends of all the old plugs looked normal, with a tan coating over about 3/4 of the surface and the remainder black carbon.

Does anyone know where this gooey stuff might have come from? It was underneath the blue plug wire ends, i.e. inside them. I installed the plugs and they were clean when they went in. I did use Permatex anti-seize on the plug threads, but this stuff is more like the adhesive used on vinyl electrical tape, except thick, like a soft epoxy or some kind of non-hardening sealant. And it's not on the threads.

Thanks for any comments!

 

pictures?

 

It comes from the stock knock sensor, which is located above the front rotor housing spark plugs.

 

Yup, the stock knock sensor is located above the front plugs and it is filled with dark gooey stuff. Over time it starts to melt and drip down on the plugs.

 

What is the solution to this?

-s-

 

replacement

tom

 

is it a problem though??? And how easy is it to replace?

Another question... you mentioned putting anti-sieze on the plugs. Is it always recommended to put anti-sieze on the plugs prior to putting them in? Is there a negative effect to it of too much/little is added?

thanks.

 

always put it on the plugs. Just don't goop it on there, just dab it around and thn use your fingers and smooth it out. If you ever have a plug stick on you and you gaul out the thread taking it out, your engine's shot if the threads are ruined.

Tim

 

FYI - 12k miles is actually quite a while for plugs on a rotary. I tend to change mine around 10k miles or so.

 

Wow, that was a shock. So the knock sensor is full of goo? What is the effect on it of losing it over time; does it screw up the sensitivity? Is it necessary to replace it now because it has lost function? I appreciate the responses!

FYI, here is that requested picture.

 

Dealer quote for OEM knock sensor: $235.50

Known alternative to Mazda part? And is it really necessary to replace the leaky one, or is the knock sensor still working?

 

The knock sensor is basically just a microphone that listens for knock in the engine and as I understand it, the 'goo' is there to insulate the microphone so it doesn't pick up too much sound from the engine bay. As the goo melts out the knock sensor will read higher and higher until it puts your engine into limp mode (rotary gods, correct me if I am wrong ). I don't see any harm in leaving it in there until it actually goes bad.

Call Ray at Malloy Mazda and see what price he can get you, although I'm sure it is still not going to be cheap.

 

afterburn27,

TNX! (I wonder why they didn't use mat insulation (fiberglas, rock wool, etc)?)

 

Can anyone advise me on whether to replace the knock sensor by working down from the top of the engine or up from the bottom? Looks like I would have to remove the L & T plugs on #1 rotor if I tried it from the bottom. Any easier from the top?

 

Found the procedure for testing the knock sensor on p. F-171 of the '94 factory manual. Jumper "TEN" to "GND" in the Data Link connector, then put a voltmeter on the "MEN" terminal (chassis gnd on other lead). Turn on ignition, and tap the engine lift bracket with a hammer and observe a voltage variation.

Result: Nothing. The voltage on my digital voltmeter sits at +12.64 volts and doesn't budge. I tapped a fast series on the bracket in hopes of seeing something, but apparently this leaky knock sensor just doesn't work.

It would seem from the wiring diagram symbol in the manual that this knock sensor contains a piezoelectric crystal. That is a high-impedance microphone-type device with electrodes on opposite surfaces. When physically stressed (shock, noise, etc) it would produce a voltage transient. The "normal" engine operation (no knock) would then be simulated by this non-working sensor, even if there is knock.

I am wondering if I can just leave this sensor alone, or if there is a real danger to the engine (stock), considering it always gets 91 octane CA gasoline. Is 10 psi boost likely to cause knock in a stock engine? I've never heard any, but then I have no idea whether knock or ping in a rotary is as obvious as it is in a piston engine.

Advice on this please? The sensor is all but impossible to reach. Looks like no wrench clearance from the bottom through the suspension gap (like for replacing plugs), and either the power steering pump and reservoir, or both that and the A/C compressor, would have to be removed to get at the sensor from above (does anyone know?). The manual shows the sensor being replaced only with the engine out of the car!!

 

Back from the dead...

I noticed the same sticky goo on my front trailing plug today. wstrohm, what did you end up doing about your gooey knock sensor?

-s-

 

There's not much to do. When the knock sensor starts loosing its goo, it needs to be replaced. Its not a huge deal if using a PFC since the PFC only displays the knock reading. However with the stock ECU; as that goo falls out, the sensor will become more sensitive and can throw the car into limp mode when it shouldn't be.

 

I think I'll replace the sensor; my AEM EMS can add fuel or retard timing based on knock sensor voltage. I wonder if all of my datalogs for the past month have been bogus...



One more question: can the knock sensor goo get inside the combustion chamber? My front leading spark plug is covered with some sort of gunk as well. When I first pulled it, I was worried because I've heard detonation can cause a glazed appearance. But now I'm not sure if I should blame it on detonation or on this knock sensor...


Photos:
Knock sensor, and goo on insulation of trailing plug




Trailing plug; no goo visible on plug face (unused plug shown as reference for electrode wear)




Leading plug: is this knock sensor goo, or glazing from detonation? Notice the rear leading plug looks OK.



-s-

 

The hole for the knock sensor isn't open to the combustion chamber, so there would be no way for it to leak into the combustion chamber.

 

Those plugs look horrible. How was the car running with them in? How many miles were on them?

 

Those plugs have about 3-4k miles on them, some autocrossing, some commuting, and the car sat for about a month while I was out of town. Some of the recent mileage was pig rich, as I was (and still am) learning how to tune my EMS. I've also been suspecting a coolant failure for the past 10k miles, but the car was running well enough so I didn't worry about it too much.

I showed the plugs to a couple of knowledgeable co-workers, and they said it looks like rust on the front leading plug.


The car idles and runs well right now, with a bit of hesitation during deceleration or very slight throttle, but I suspect that's more of a tuning issue than a spark plug issue.

What is the likelihood that the rest of my engine will be damaged by driving like this? I was hoping to wait until early 2008 to rebuild or swap in a remanufactured motor.

-s-

 

Be sure you correctly torque the new one you install. It won't work correctly unless it is.

 

I recently tore down my dad's motor, which had coolant seal failure due to his poor maintenance. Not good news for you, his plugs looked similar to your leading plug