knock sensors ? aren't supposed to be for exceptions?

[forecast]

I think this would be better as new thread, however it addresses Q45 drivability issues seen in both:

http://www.nissaninfiniticlub....21840

I was under the impression that the knock sensors are really on the car to protect the engine from detonation only under extrodinary circumstances OR if a another system was malfunctioning in a way the ECU could not detect.

But I read all the time here how so and so replaced the sensor(s) or harness and it took care of this or that problem.

I really need to hear from an expert like Q45tech on this, but I've been lead to believe that if everything (ignition system, fuel delivery system, EGR and MAF) is all within spec and 93 gas is used then cylinders should ALMOST never mis-fire - and the knock sensor shouldn't come into play.

So when my car mis-fires at low RPM under load (40 in 4th) it's an indication that one of the previous systems is at fault - bad timing, or a situation which creates as overly lean cylinder (EGR stuck open, clogged injector, low fuel pressure, MAF problem etc)

My car has one O2 sensor for four injectors, so if one is mal-functioning the ECU enriches all four until the o2 sensor is satisfied. I suppose if one is producing 25% less output than the others, the compensation factor is all are increased by 7% - so the normal injectors are running 107% and the malfunctioning one at 80% - still lean enough to cause misfiring at low RPMs.

More likely, I think is that all injectors are all more or less dirty with one more so than others.

Am I way off here - are the knock sensors supposed to be a daily part of the cars function?

dan

[forecast]

I suppose I should follow that with:

If I'm looking to solve reccuring drivablilty issues that happen almost everyday, then my thinking has been that replacing the knock sensors is just going to mask a still exisiting underlying problem. Me (or the customer) is happy that the car doesn't exhibit x behavior anymore, but is the problem really fixed? Won't gas mileage, emissions, and power all be taking a hit?

[Q451990]

The ECU always uses the knock sensors to "listen" to the engine. If the ECU doesn't get the specified resistance from the sensors at start-up, or on a regular basis, it sends the car into a light "fail-safe" mode and retards the timing advance. The theory is that if it can't monitor for knock, then it's going to protect the engine until it can.

On the early Q's (Federal ECU) the detonation sensors won't set off the "Check Engine" light. I think the assumption was that the owner would take the car to the dealership every 3 months or 3K for an oil change and a Consult check-up would be part of the package. About as pie-in-the sky for American consumers as brake pad, rotor, and tire replacement in 15K intervals! It amazes me how poorly Nissan understood American consumers even in 1990, after exporting cars here for decades! I guess they assumed that they'd get a better consumer on higher price point cars.

Heath

[forecast]

Heath, if I get what you're saying it's that - the knock senors are the microphones to hear the engine, if the ecu determines the mics are unplugged or shorted, it degrades the overall performance of the car, without notifying the driver.

An interestesting angle I hadn't considered.

This doesn't really answer my question - if the sensors are the mics - should the ecu be hearing anything if everything else is working right?

dan

[maxnix]

Dan,

In no way an expert, but as I understand it the ECU checks the KS for integrity, and if not present, deep sixes the timing advance. If integrity is present, and no adverse information is present from KS or other sensors, then full advance of timing is permitted.

[Q451990]

This doesn't really answer my question - if the sensors are the mics - should the ecu be hearing anything if everything else is working right?

I think there's always going to be a tendancy for spark knock, even if everything is clean and working correctly. It's just a function of how hot the cylinders get, air density, etc. I think what Nissan was trying to do is allow as much timing advance as possible, but use the sensors as a last line of defense.

Perhaps the ECU isn't quite powerful enough to calculate everything on the front end (timing advance, injector pulse width, etc) so as a last resort, it's looking for detonation as a final check. Sometimes it's easy to forget that the Q's computer technology that was developed 15 years ago - I think we had an IBM AT by then... with a 30 megabyte hard drive!

Heath

[Q45tech]

Cruise and very minor acceleration [small throlle opening] requires a lot of timing advance because the engine is operating at a very low volumeric efficiency. Something like 25-30% [compared to 100% at 4,000 rpm WOT].

Say the advance is 35-40 degrees at cruise [for fuel economy], when you depress the throttle - the ecu knows this by TPS voltage change [x amount of additional opening in y milliseconds]. The ecu tries to guess what you want to do so instead of dropping the 35 down to 20, it over compensates down to 10 momentarily then ramps up to 20 fast [this is called tip in acceleration retard].

The air flow into the cylinders at cruise is very slow [high vacuum] plus the engine is running at 14.7 AF vs a 11-12AF need for acceleration..........the transition period between 1500-2000 rpm cruise and mild acceleration is where most subaudible knocking occurs.

The knock sensors hear some knock but [kind of like a bucket with a big hole] the total knocks pour in but the system allows many to flow out the bottom. When the bucket overfills [the accumulator rising voltage] signals the ecu to cut timing..........the rate of knocks [you can only have 4 per rpm] at most and the amplitude of eack can vary......The is always one worst cyklinder that knocks FIRST and some that may never get a chance because the ecu is fast enough to start retarding before 2 rpm have past........relly bad knock from regular gasoline in warm conditions [or just a bad batch of gasoline ] can create conditions where the knock sensors and ecu cannot correct the problem .....thats when yo hear the knock [rattle on a more or less continuous basis on medium acceleration.

[Q45tech]

http://www.fs.isy.liu.se/Publi...J.pdfh ... AQ...lay=2

"Under warm cruise (closed loop) strategy, EEC will push the spark advance out to the point where trace knock is detected, and then back the timing off slightly, to keep away from pre-ignition. This strategy makes sense to get the most efficient burn and the most power. Under these conditions, you may hear some knock if you mash the gas, because it takes EEC a few CPU machine cycles and crankshaft rotations to drop out of closed-loop strategy and retard the timing.

Under wide-open-throttle (WOT) conditions, EEC is not looking at the knock sensor because the engine is simply making too much noise for the knock detection strategy to filter out the knock signal from the noise. Timing and fuel for WOT is derived from lookup tables that are vectored primarily by crankshaft RPM and engine coolant temperature, and to a much lessor degree by the mass air signal.

Under most conditions, EEC should be able to get the most out of any reasonable fuel that you pump into the tank, however, an especially bad load might be beyond the range that the spark lookup tables can compensate for, so pulling the octane shorting bar invokes a new set of tables, with less aggressive spark advance curves for WOT operation."

http://www.intersil.com/data/an/AN9770.pdf