Can I get a link or explainaton please.

Simple they are piezoelectric microphones, which are gated on [alternate banks] for approximately 20 degrees of rotation every 90 degrees.........to listen for abnormal sounds which occur with spark knock [spark fires at 20-40 degrees before TDC and max pressure occurs at 17 degrees AFTER TDC.

A sound of ~~ 2-5KHz that occurs in the -15 to _+5 degree ATC position of EACH cylinder] is a knock. The frequency spectrum [from the microphones] is shaped and filtered by software to ignore other frequencies.

The sounds are accumulated as a voltage rise on 2 separate capacitors which continuously discharge [at a rate based on rpm].........knock counts per second.

This voltage is used by ecu to subtract ignation advance from a maximum number.......no knocks no subtraction and ignition advance is maximum preprogrammed value.

The are 2 x 2 accumulators [long term ones for those using regular fuel to sense all those extra knocks] and reset maximum to 2,3,4 less degrees vs short term to sense extra that average knocks.

Tricky to decide average vs peak vs rpm vs load and the knock sensors/ecu run out of calculation time as rpm increases with old style non obd2 ecu [90-95 models] so it gives up and just retards -3.........but this is set into software for WOT with RPM above 5,000.

So the knock sensors don't offer any protection above 4,000 rpm.

As engine ages noises increase from wear and knock sensors may get fooled.

Why you study WOT advance numbers to look for anomolies. Does the 6,000 rpm advance get to -28 degrees is the clue.

As KS age the sensitvity vs frequency curve changes from heat cracks etc.

As new knock control requires as new KS.......why they get changed every 100k.

A better more expensive way is to measure each cylinders peak pressure during combustion......what's used on race cars to set individual cylinder timing [ignition advance for each individual spark plug firing].

Excellent post dennis!!!!!!

What currious is that many engines use KS but not many of them relay on KS feedback to the extent that of Q45 sytem. I was surprized to see significant difference in performance change when KSs were replcaed. I can only think that on high compression engines such as 4.5 or 4.1 it is important to have KS noise feedback for prober ing timing!Cheers!

Thanks Dennis!!!

I love your explainations (I guess its the math)

My 1994 GreenQ45a is about to cross the 120000 mark.

Is it time? I was gonna do a complete under Plenum hose job this September.

While I'm there should I do the Knock's?

Is there a way they be done while the Plenum is on?

I just did the Alternator W\Belt, New Battery, New Active System CPU.

Last spring I did the Valve Covers, New Plugs, All Belts Throttle Body cleaning, and fuel filter and Breaks.

Tomorrow or Tuesday I do the water pump.

The reason I ask is my Fuel pump is getting noisy and my old 1993 Base Q45 the pump died @ 126,000. I don't want to buy the Sending Unit, FPCU Relays and Fuel Pump this time... Only the Pump... Damn the Dealer

Early Q [300zx/J30] KS were experiments to understand the future requirements of OBD2 [1996]. ad OBD3 coming soon.Just like Nissan uses wider more sensitive MAF control to set the fuel flow.

Many engineers were/are still stuck in old MAP ways.........actually you need both sensors [MAF and MAP] to do the job right.

Study the Karmen Vortex type MAF to see Toyota faults.

Also study why returnless fuel systems need to monitor fuel temperature [density]...........but that only works IF Fuel is standardized.........not so in US with 36 different blends each with a regular and premium version.

Why we will eventually get ethanol everywhere and every PARK will be turned into a CornField.

GreenQ45a wrote:
The reason I ask is my Fuel pump is getting noisy and my old 1993 Base Q45 the pump died @ 126,000. I don't want to buy the Sending Unit, FPCU Relays and Fuel Pump this time... Only the Pump... Damn the Dealer

Green- the 'green' pump relay is only $13 or so, so I'd say do it. as for the FPCU, I'll offer you a 'rebuilt' FPCU [ground path upgrades, new power transistors, and a few other fixes] for say $25+ shipping and your old core [so i still have a backup].

GQ Jay wrote:
Green- the 'green' pump relay is only $13 or so, so I'd say do it. as for the FPCU, I'll offer you a 'rebuilt' FPCU [ground path upgrades, new power transistors, and a few other fixes] for say $25+ shipping and your old core [so i still have a backup].

Relay I agree..Was the rebuild subject to pin 16 burnout?(little brown mark)For 25 damn man your more than fair

FarFetched wrote:What currious is that many engines use KS but not many of them relay on KS feedback to the extent that of Q45 sytem.

Actually old Ford EEC-IV systems use the same strategy - Back in 1985! Of course they have the same limitations on RPM and retard, but it's old and proven technology. In fact, Ford had a patent on the system for a while. Nissan may have licensed it.

Approximately 45 percent of new cars sold in the United States since 1990 have engines that are equipped with a device called a knock sensor.

http://www.freepatentsonline.com/4337641.html

Think you will find this 1980 Patent to Nissan superceeds Fords work however the first patent to GM was the grandfather in 1975-1977

http://www.freepatentsonline.com/400215 ... 53020.html

Anyway there was independendent work going back to 1967-1970 when tetraethyl lead was removed from gasoline and serious knock started.

Measuring knock was one thing but electrically varying the timing had to wait for ecu which need fuel injection to pay for it.

"Hot Wire Air Mass Meter - A New Air Flow Meter for Gasoline

Fuel Injection Systems_,Sauer, Rudolf, 1980"-------BOSCH

Every company rushed on the bandwagon to out tweak each others designs.

Good info, Dennis. I was going on SAE paper 790312, which I can't seem to find right now.

Back in the old distributor days with centrifigal advance [rotating weights and springs to move the point plate position] they had a vacuum advance for cruise which was cut off at idle.

Early experiments involved using this same vacuum cut off when knock was detected. didn't work well but it worked somewhat.

As we see today knock often occurs at mild acceleration at low rpms [2000] where modern system delay tip in throttle advance [sudden mild acceleration]...........drive by wire [ecu controls throttle opening] solves much of this as it can fix the advance before the throttle plate moves, so the knock cannot start.............vs trying to stop it after it starts.