johnf95111

I have a quick question: How are the knock sensors tested by the John S. Spanner or the Hammer?

Bill Ball hooked my GT up to his Spanner and it first reported a bad Hall Sensor. I found an open in one of the wires leading to the sensor. I replaced it and now we see a Knock Sensor fail.

I visually checked the sensors when I R&R'd the intake manifold recently. The valley was shiny and clean, and the sensors looked physically ok.

Is there a test I can do without using the Spanner or Hammer?

Thanks,

John Fagerlund
'89GT

John Speake

Hello John,
The EZK ECU detects whether there is residual engine noise as the correct level and frequency and stores a fault code if all is not as it should be.

It would be quite difficult to test them fully on the car. Do you have an oscilloscope ?

Sometimes spurious fault codes are reported/stored. Try doing a battery reset, and then a test drive. Then run a Spanner test again.

You should have noticed a significant improvment in top end performance if al Hall and knock faults are cleared. Otherwise you will get 6 degrees ignition retard at high loads.

Maybach_Man

can you test them off the car???

geoff

John Speake

Hi Geoff
You can test them with a 'scope, and a hammer...

johnf95111

Hey John,

I do have access to an Oscope. I'll check for shorts in the wiring then test them on the car at the connectors with the scope. I assume I just connect the scope and rap on the block.

Cheers,
-j

John Speake

That wouold work as a crude check.

One problem that happens is that the insulation crack with the heat, and then the braid on the lead scrapes aroubd on the head or block. This electrical noise is detected by the EZK as a spurious signal and so it defaults to the 6 deg retard mode.

I am not convinced that oil on the sensor inhibits their operation. However, they are very high impedance devices. But oil is a good insulator....

Jim bailey - 928 International

Knock sensors are maybe $50-60 each it takes two and we KNOW they degrade overtime. How about do a timed 0-100 run then unhook one knock sensor and do another 0-100 timed run. If the time is the same then you have a bad knock sensor because unplugged the ignition timing will be retarded and power down.

Bill Ball

John:

In this case, the Spanner reported the front sensor was faulty. I wiggled the connector a bit. It seemed OK. I cleared the fault, but we did not retest to see if the Spanner still reported it as bad. We should have. John, I can drop by after work some day and we can run the Spanner again.

More puzzling is that every time we run the Spanner on John's car, the Spanner kicks out to the welcome screen as it begins to conduct the ISV actuator test. The ISV seems to be working, although the idle is marginally high. We can't get beyond this kick out.

If the front knock sensor continues to test bad, then we need to raise the intake maifold and replace it. If so, we can check the ground continuity for the ISV, and all the other grounds under the intake.

Tom. M

We tested 3 cars the last time we were at Louie Ott's place and it seems they all came up with front knock sensor failures (even though some were actually pretty new or had been replaced recently)....None of the cars were down on power..

Later,
Tom
89GT

Bill Ball

I have tested a bunch of cars and not had front knock sensor failure messages until John's car. The one time I got a rear knocks sensor "missing" (note the wording) error message, I found the connector on the passenger fuel rail had come apart. I guess "missing" and "faulty" could mean two different things. Missing might mean no ground, no signal. Faulty might mean out of spec signal.

John Speake

Sometimes spurious fault codes are reported/stored. Try doing a battery reset, and then a test drive. Then run a Spanner test again. (repeated from above.....)

Many Porsche specialists will do a battery disconnect, then test drive the car, then read out the stored codes.

I think the possibility of spurious codes is the reason that Porsche arranged for the codes to be automatilcally deleted over 50 or so ignition switch off cycles (not sure of the exact number, and it was changed at some point to 100, again IIRC)

Bill Ball

Also, as I recall, the codes don't get stored unless you do a ride of some 6 minutes and various engine speeds. So, pulling the battery, then reading codes may give you a false sense of security.

John Speake

Agreed - I said " do a test drive" .......which would check the car under a variety of driving conditions.

A knock sensor fault will be detected in less than 4 seconds at anything above about 3000rpm, 40% load.

johnf95111

John S,

I'll clear the codes and take it for a spin then have Bill run the test again. I'll also to what Jim suggests with the 0-100mph runs connected and disconnected.

While digging into this, one quick question: I have disconnected the battery during maintenance. When connected back up, the car idles normally, then after a few minutes of driving, the idle moves up to just under 1k rpm. It ramains high until I disconnect/connect the battery again. I seem to be resetting something doing this and then the engine re-learns and makes adjustments (?).

Thanks,
John Fagerlund
'89GT

Bill Ball

John:

Did we test whether the idle was MAF related? Don't think so. We could swap in my MAF and see if it drops. Such things have happened due to MAF.....I do have a can of the so-called MAF cleaner.

Yeah, do a couple of 0-100s