Join the Car Talk Community!

Discussion Rules

Welcome to the Car Talk Community!

Want to ask a question or join the discussion? Great! Join now.

Sign In Register

To replace a knock sensor or not?

I have a 2004 Nissan Xterra that has been chugging for a few months now. The problem is intermittent and only happens about once or twice a week when driving, but when it starts to chug (feels like it is running out of gas, jerky, will not accelerate) it is unsafe to drive on the freeway. I have taken it to several places and no one can figure out the problem. It was pulling two codes the speed sensor on the transmission PO500 - and the knock sensor code was coming up along with that. I had the speed sensor replaced twice and checked the wiring. We replaced the fuel filter, checked the fuel pump with a fuel gage and fuel lines, replaced the distributor cap and rotor, put in a new MAF sensor and still nothing is working to remedy the chugging. My last stop is replacing the knock sensor for $600, the problem is that the Nissan Mechanics (I visited two dealerships in different cities) told me not to replace this as it would not cause chugging. Several other mechanics around town tell me that they are wrong and I do need to replace it. Does anyone have any input on this? Please help with any advice.


  • Lets see - "the" knock sensor code? No such thing.

    P0324 Knock Control System Error
    P0325 Knock Sensor 1 Circuit Malfunction (Bank I or Single Sensor)
    P0326 Knock Sensor 1 Circuit Range/Performance (Bank 1 or Single Sensor)
    P0327 Knock Sensor 1 Circuit Low Input (Bank I or Single Sensor)
    P0328 Knock Sensor 1 Circuit High Input (Bank I or Single Sensor)
    P0329 Knock Sensor 1 Circuit Intermittent (Bank 1 or Single Sensor)
    P0330 Knock Sensor 2 Circuit Malfunction (Bank 2)
    P0331 Knock Sensor 2 Circuit Range/Performance (Bank 2)
    P0332 Knock Sensor 2 Circuit Low Input (Bank 2)
    P0333 Knock Sensor 2 Circuit High Input (Bank 2)
    P0334 Knock Sensor 2 Circuit Intermittent (Bank 2)

    Perhaps you could report the specific code. Note that NONE of these say "bad knock sensor" - the computer is clueless about the actual condition of the sensor itself. Its just a box of wires hooked by wires to other wires. That's about it.

    I doubt your symptoms are from that though. You should troubleshoot the speed sensor code. Note that I said troubleshoot the code. There's also no code that says "bad speed sensor."

    But you also need to sort out the knock sensor issue - knocking is very bad for the engine.
  • The faulty knock sensor will not cause the Check Engine light to come on nor will it cause the driveability symptoms you describe. The cost for knock sensor replacement is that high because it involves removing the upper and lower intake manifolds for access. But replacing it will almost certainly not correct your problem. There are many things that may cause the symptoms you describe, but without being able to see the car/duplicate the complaint, any guess is just that--a guess.

    Your problem should be easily found if the mechanic is equipped with proper diagnostic equipment and the problem occurs when he is driving your truck with the test equipment hooked up. If the problem only rears its head once or twice a week you may need to leave your truck with the shop for a week or so.
  • Here's a little more information--- sorry I am trying to learn more about cars- PO328 code was coming up- the mechanic wrote on my estimate that "nox sensor shorted inside sending high voltage to computer 100% of driving time" he told me that he believes that this is causing the speed sensor code to come up from the vehicle jerking.
  • Any opinions or advice given here amount to "armchair quarterbacking" since we don't have access to the car. But I can say that the scenario for a shorted knock sensor (a common failure on Nissan V-6 engines) to cause a fault code P0500 for a Vehicle Speed Sensor would involve a rodent eating wire insulation off of your engine harness causing some crossed wires. Or some other strange thing. Not impossible, but certainly far from probable.

    But once again, a faulty knock sensor will not, can not, on your car cause the symptom you are describing.

    Here's another point. Since this problem is intermittent, it may take some time for road testing to properly diagnose. Remember, whenever the guy is working on your car, the clock is ticking at $90/hour (or whatever the going rate is where you are).
  • This mechanic I just took it to drove my car around for over 20 miles and a full tank of gas after a week and all he came up with is the knock sensor. If it were you...what would your next step be. I've taken it to over 5 different mechanics which includes 2 nissan dealerships only to have them throw up their hands and say they either couldn't figure out the problem, tell me it was the knock sensor, speed sensor (which was replaced twice and wires checked and ECU checked) or tell me to come back when the CEL was on again because each time the mechanics clear the codes out. Please any suggestions would be greatly appreciated. I've spent over $300 now on diagnostics alone and can't seem to get anywhere with any of the mechanics.
  • I'm in agreement that the knock sensor is not causing this problem. Maybe there's an intermittent fault with the fuel pump and that is certainly not uncommon.
    The perception is often that a fuel pump will quit and stay dead and that's not always the case. Sometimes they get balky and act up followed by straightening out for an undetermined amount of time and then going stupid again.

    Unless the pump acts up while the fuel pressure test gauge is in place and being monitored it can be pretty much a wild guess about this problem although replacing a pump on a wild guess is a bit distasteful.
    I would add that if the fuel filter has not been replaced regularly that is something that can contribute to pump failure because a partially clogged filter (which may not have any noticeable symptoms) can cause the pump to work harder than it normally would.

    I've diagnosed a few pumps that suffered this kind of problem by driving the car around and sorting it out on the side of the road. In the case of a Mitsubishi my daugher owned, I drove the car around for a week before it acted up and quit running on a deserted road with the ambient temperature about a 100. At that point I determined the pump was the problem and lucky for me about half an hour later the engine started and got me home with nothing more than a few minor hiccups on the way. I replaced the pump the following day and the car was never a problem after that.
  • It's very frustrating, isn't it?

    I tell customers that I'm as patient as they are. If I need to keep your car for a week, I will. We'll use it (with your ok, of course) as a shop vehicle, using it to drive to the bank, pick up lunch or parts, run shop errands, etc. Doing that, it should be possible to duplicate the concern. A dealership technician probably doesn't have that kind of freedom. And an experienced tech can probably tell by feeling the concern if the problem is related to fuel supply, ignition, fuel delivery, engine sensor, or even electrical.

    As far as testing, a fuel pressure gauge is no longer the only way to test a fuel pump. We regularly test the electrical condition of the pump using a "digital storage oscilloscope" and a low-amp probe. This gives lots of information to someone who knows how to interpret it. The fault could also be with the ignition system (plug wires, failing coil), a failing sensor (throttle position comes to mind), an electrical component (like a fuel pump relay), perhaps even a fuel injector failing only when hot. We also use a scan tool that can record all engine data like a movie that we then look at frame-by-frame back at the shop to help us find the fault.

    Needless to say, this takes time and money. But then so does working down the list of "likely causes" until you find the right one.

  • edited May 2012
    I disagree with the tech's thinking about the knock sensor high input code. A shorted sensor will not cause a high voltage reading to be seen by the ECU. If anything, the reading will be near zero volts due to the short on the sensor line. I also agree with the others that the knock sensor code won't cause the trouble you are having and what may really be going on is, the problem you are having is causing the knock sensor code to occur. The sensor is putting out a higher than normal signal due to the trouble with the engine performance. I would leave the knock sensor alone. I don't think it has a problem but other areas do.

    As far as the speed sensor goes it appears it has to be okay since it has been replaced. I assume the wiring to it has been properly tested for any faults like bad connections, shorted wires, and faulty grounding. If that has been done correctly then the trouble may be within the ECU for that circuit. Another possibility is if this vehicle has a second speed sensor that has been overlooked.

    My "gut feeling" about this problem is the real issue may be with the TPS or throttle position sensor. It may have some bad spots on it that is causing the intermittent trouble.

  • One possibility that I like except for the expense is to try changing the knock sensor. There was also a tapping noise from a loose lifter that triggered the knock detection (on a Toyota years ago) which then caused the computer to retard the ignition timing. That did affect performance in a bad way. Somebody should listen for tapping noises just for the sake of being thorough. It didn't make the engine chug though.
  • You should know the actual code it's throwing for the knock sensor. Also, a decent scan tool should be able to log whether any 'knock events' are recorded, to further verify whether the sensor is bad. But I agree with Cougar: a bad TPS could cause this. Unfortunately there's a lot of other things that could cause this too.
This discussion has been closed.