Knock sensor issue

[alaw]

Gday all
Where having issues with my son's 1993 FZJ80R .
It's throwing a check engine light , diagnostic is showing code 55 which is knock sensor .
This is driving us mental.
We have tried new knock sensors , different ECM and re wiring
We've had no luck , please help

 

[Bambusiero]

Suggest pulling the spark plugs and reading the insulator color for signs of running lean.
Does your error code indicate bank1 or bank2 knock sensor? That will help your diagnosis.
Bank1 is the front sensor, cylinders 1,2,3.
Bank2 is the rear sensor, cylinders 4,5,6.

I had an issue for quite a while with one fuel injector partially clogged, so that one cylinder was running a bit lean at low altitudes.
As if the maximum flow rate was limited, maybe turn on was slow? Don't know.
But - knock sensor error code kept showing up. Sort of borderline, but it kept coming back.
Going up in altitude (somewhere above 1000 m or so), the knock sensor error code would go away, because the control system was leaning all the cylinders to match the lean one. Then could drive around in the highlands all weekend, no error code.
Back downhill, error code returned.
It took several trips to see the pattern.

It correlated with a lean looking plug.
Sending the injectors out for cleaning fixed. it.

 

[alaw]

Suggest pulling the spark plugs and reading the insulator color for signs of running lean.
Does your error code indicate bank1 or bank2 knock sensor? That will help your diagnosis.
Bank1 is the front sensor, cylinders 1,2,3.
Bank2 is the rear sensor, cylinders 4,5,6.

I had an issue for quite a while with one fuel injector partially clogged, so that one cylinder was running a bit lean at low altitudes.
As if the maximum flow rate was limited, maybe turn on was slow? Don't know.
But - knock sensor error code kept showing up. Sort of borderline, but it kept coming back.
Going up in altitude (somewhere above 1000 m or so), the knock sensor error code would go away, because the control system was leaning all the cylinders to match the lean one. Then could drive around in the highlands all weekend, no error code.
Back downhill, error code returned.
It took several trips to see the pattern.

It correlated with a lean looking plug.
Sending the injectors out for cleaning fixed. it.

Cheers for your reply mate
Code 55 is rear Knock sensor
Code 52 is front Knock sensor
Those codes 52 and 55 I believe are shown when there is a fault in the system
If there is a knock it throws code 53
I could be wrong, I'm hoping someone tells me otherwise.

I will check the plugs , cheers

 

[Bambusiero]

If all the cylinders are knocking, then the ECM has means of adjusting - overall fuel flow, spark timing, etc.
It doesn't think a knock detected anything is wrong, just an adjustment needed.
But, if only one cylinder knocks?
How can it distinguish between a sensor problem and many other possible causes?
It cannot, so it blames the sensor.
But, you've already eliminated that the sensor, or wiring, or ECM is faulty, right?
Sooo, it really is something wrong...

 

[alaw]

If all the cylinders are knocking, then the ECM has means of adjusting - overall fuel flow, spark timing, etc.
It doesn't think a knock detected anything is wrong, just an adjustment needed.
But, if only one cylinder knocks?
How can it distinguish between a sensor problem and many other possible causes?
It cannot, so it blames the sensor.
But, you've already eliminated that the sensor, or wiring, or ECM is faulty, right?
Sooo, it really is something wrong...

Makes sense cheers

 

[Bambusiero]

...and now it's up to you to call out that lyin' SOB of an ECM - blaming poor innocent sensors for all its troubles, just trying to do their job.

 

[alaw]

I pulled out the plugs
Cylinder 5 had a different brand plug to all the other cylinders.
Could this cause an issue? Maybe a knock?

 

[alaw]

Also , can't be injector's, as it does it on gas too

 

[Bambusiero]

I pulled out the plugs
Cylinder 5 had a different brand plug to all the other cylinders.
Could this cause an issue? Maybe a knock?

Could be, could be, but purely guesswork.
I would install a set of new proper plugs. Do a plug read after a few hundred km. Maybe will tell you something.
And, if luck is with you, problem fixed?

Also , can't be injector's, as it does it on gas too

I don't understand, but that's OK, I don't need to. Just keep an open skeptical mind.
You have everything I can contribute. Good luck.

Well OK. 2 more ideas.

1) Swap knock sensors from bank1 to bank 2.
If the problems says in place, then you have eliminated the sensors themselves.
But maybe you already covered this.

2) Put an oscilloscope on the knock sensor ECM connector pin, while running, just to verify signal integrity.
I have done this. It's a chore. Tracking down the oscilloscope, probing the connector pin from the wire side, getting a ground, etc.
Triggering the scope properly to get a stable waveform to look at is very difficult, because it's asynchronous.
The signal is a constantly varying amplitude stream of pulses from the 3 cylinders firing.
But, any signally looking signal tells you the whole sensor pathway is working.
But...desperate times...

 

[BILT4ME]

I pulled out the plugs
Cylinder 5 had a different brand plug to all the other cylinders.
Could this cause an issue? Maybe a knock?

That may point to a cracked spark plug, but that would make a rich condition unless the O2 sensors are leaning it out due to unburned fuel offsetting the readings.

 

[BILT4ME]

I would replace all plugs with new ones from Toyota. May as well do wires and cap and rotor as well while you're in there.

 

[alaw]

I would replace all plugs with new ones from Toyota. May as well do wires and cap and rotor as well while you're in there.

Cheers, we plan to put in new plugs

 

[alaw]

Could be, could be, but purely guesswork.
I would install a set of new proper plugs. Do a plug read after a few hundred km. Maybe will tell you something.
And, if luck is with you, problem fixed?


I don't understand, but that's OK, I don't need to. Just keep an open skeptical mind.
You have everything I can contribute. Good luck.

Well OK. 2 more ideas.

1) Swap knock sensors from bank1 to bank 2.
If the problems says in place, then you have eliminated the sensors themselves.
But maybe you already covered this.

2) Put an oscilloscope on the knock sensor ECM connector pin, while running, just to verify signal integrity.
I have done this. It's a chore. Tracking down the oscilloscope, probing the connector pin from the wire side, getting a ground, etc.
Triggering the scope properly to get a stable waveform to look at is very difficult, because it's asynchronous.
The signal is a constantly varying amplitude stream of pulses from the 3 cylinders firing.
But, any signally looking signal tells you the whole sensor pathway is working.
But...desperate times...

Cheers, the sensors are new .
We will put in new plugs and see what happens from there
Cheers for your advice

 

[alaw]

Put new plugs in , still getting CEL code 55
I'm going to clean the injectors,
Any other suggestions
Cheers

 

[WCBlueSky]

You said you tried re-wiring, but have you checked for continuity in the wire all the way from the knock sensor plug to the ECM? In case there is a short in the wire...

 

[alaw]

You said you tried re-wiring, but have you checked for continuity in the wire all the way from the knock sensor plug to the ECM? In case there is a short in the wire...

Yes we have, have opened up the ECM and checked from the knock sensor plugs back to the terminals inside the ECM.
Have tried
New knock sensors
Have tried mounting Knock sensors to inlet manifold via bracket
Replace spark plugs
Checked valve clearence
Tried another ECM

I'm ready to set fire to this thing

 

[WCBlueSky]

You only get code 55, which is rear knock sensor, so the front knock sensor should be all good right? I read a post here of someone who broke their rear knock sensor and in order to pass emissions they spliced the front knock sensor signal into both front and rear wires. While I'm not advocating you do that long term, it could be a good check to see if the code goes away by tricking the ECM temporarily.

 

[WCBlueSky]

Here is the thread I referenced: Another Knock Sensor Thread

Last year I had one bad sensor, throwing up a CEL light. I needed to get it "legal" in a hurry so I spliced the good one into both of the wires that feed into the ECU, and this solved the CEL light. This reminds me that I need to revisit the issue again to actually fix it. It may be worth it if all you need to do is pass inspection or something as I did.

 

[WCBlueSky]

One other thought: since the knock sensor plug only has one wire, it must make a good ground connection when threading into the engine block. If your threads for the knock sensor are rusty or oily or dirty it is possible there is a higher electrical resistance than what the ECM is looking for, leading to the CEL.

 

[bigredmachine]

when you say that you rewired di you run a new wire from the sensor to the ecu or just do a continuity check on the harness? if you did run a new wire the wire has to be shielded as it can pick noise and cause a bad reading. this is from a us spec ewd but I would expect that the aus spec would be similar