IcantDo55

iTrader: (11)

I'm building a 04 and sensor looked like it overheated and the epoxy type substance had driped out of it. Wanting to keep stock Mazda sensors I ordered a used one and it came in looking worse than the one it had. This happen to them? Are they still good??

18psi

iTrader: (76)

Most. Yep still good

IcantDo55

iTrader: (11)

Sending it.....thx

2slow

Yes, same here. I though I had an oil leak in 2002 NB from somewhere until i found out it was resin leaking form the sensor.

stevos555

Have been searching used sensors and keep asking about resin and they are all as you said already melted.

ridethecliche

iTrader: (2)

Is there a better alternative that's still compatible?

18psi

iTrader: (76)

no

Joe Perez

iTrader: (8)

So long as neither lead is shorted to ground (via the housing), and the sensor element is still rigidly connected to it, I see no reason why a partially-melted sensor wouldn't continue to function.

In theory, you should be able to test it by whacking the intake manifold with a hammer while watching the pulley with a timing light. Never tried this myself.

stevos555

Is there any other way to test with a meter ?

Joe Perez

iTrader: (8)

Not really. It's a piezoelectric sensor, so it's not like sticking an ohmmeter across it is going to do any good.

If you have access to an oscilloscope, you could measure the voltage across it (maybe 5-10mv/div) while you bang on it with a hammer.

stevos555

would this procedure from FSM help to test ?

Joe Perez

iTrader: (8)

Yes, that's exactly what I said a few posts ago, re: "So long as neither lead is shorted to ground (via the housing)..."

(OK, to be fair, they specified a resistance range and I didn't.)

codrus

AIUI, a knock sensor is basically a specialized microphone. I *think* this means you could do a basic function test by hooking it to the audio in on a PC, tapping it with a hammer while recording, and then play back the sound file to see if you can hear the "knocks".

(disclaimer: I'm a software guy, not a hardware one, so this might blow up your computer, set your house on fire, and trigger the collapse of the false vacuum thus ending the universe as we know it.)

--Ian

Joe Perez

iTrader: (8)

This is correct.

And I kinda feel dumb for not thinking of the "plug it into a computer" test.

stevos555

So if resistance checks out and only small amount of resin is gone, buying a used one may be ok it seems. The OEM sensors are ridiculously priced - $140 on Mazdacomp.

shuiend

iTrader: (24)

I pulled the water lines to the stock oil cooler/heater to stop from melting my knock sensors.

stevos555

That got me thinking that a hose heat shield maybe a good idea to wrap around the oil cooler hose and maybe a spark plug boot over the knock sensor body to keep some heat off.

18psi

iTrader: (76)

What’s the point of all that? Why waste time and effort fixing a nonexistent problem?

stevos555

So just let it melt?

18psi

iTrader: (76)

the sensor will not melt. the dollop of resin that does melt is the stuff that simply holds the wire in place. Have you ever even looked at the sensor in person?

I know you're enthusiastic about wrenching on your car, and that's good, but it seems like you're just jumping into topics completely clueless and starting conversations about fixing problems that are just not there instead of focusing on issues that actually need attention.

So to re-cap: oem nb knock sensor is a perfectly good solution for knock detection on our cars. Nothing "better" needed. The resin melting doesn't affect anything or you'd see hundreds of people replacing their knock sensors every 5 years and/or complaining about it.

So basically, this is not a problem