Missing Knock Sensor

kenewagner

Registered User
I posted a tread earlier on my Idle problems and in the course or searching for the problem found a plug dangling next to the oil pan which in the course of time I have found to be the knock sensor plug. I pulled the car up on my ramps and found the sensor is not there. (see picture) The threads in the center go up into the plug about 3/4 of an inch and are rusty. It doesnt look like something was broke off inside. Does the sensor just screw in and have the correct connection for the plug? I have read that it goes into the water jacket. I have no idea how long it has been gone. I have had the car a little over a year and have run it hard and never had any detonation that I could hear.
 

Attachments

  • Knock3.jpg
    Knock3.jpg
    196.9 KB · Views: 244
kenewagner said:
Does the sensor just screw in and have the correct connection for the plug? I have read that it goes into the water jacket. I have no idea how long it has been gone.
Yes, it just screws in.. but you'll need to drain the coolant before you go to replace it. Most likely, whoever worked on it broke it off. I think you should get them to replace it. :cool:

Looks like someone has done a bit of customizing with your Oxygen sensor also; no auto manufacturer I know of uses unsealed molex nylon connectors next to a hot engine. Go look at the setup on the driver's side O2 sensor to make sure all of the wires are accounted for and hooked up.
 
Last edited:
That threaded part is just an adapter that goes into the water jacket. If the threads aren't too rusty, you should just be able to chase the threads with a tap or just screw in the new sensor. You would not have to drain the coolant in that case.

knocksen.jpg
 
TbirdSCFan said:
Yes, it just screws in.. but you'll need to drain the coolant before you go to replace it. Most likely, whoever worked on it broke it off. I think you should get them to replace it. :cool:

It looks like it was removed rather than broke off because the threads are good and there is no sign of material left in the hole. When buying a Knock Sensor do you just get the part that screws in to that hole or do you get the plug with the hole in the center as well. Is it the plug that seals the water into the water jacket or is the small center hole open to the water jacket as well?? :confused:
 
kenewagner said:
It looks like it was removed rather than broke off because the threads are good and there is no sign of material left in the hole. When buying a Knock Sensor do you just get the part that screws in to that hole or do you get the plug with the hole in the center as well. Is it the plug that seals the water into the water jacket or is the small center hole open to the water jacket as well?? :confused:

Thanks for the quick response back and the pictures. One last question if the Knock Sensor is not in place and hooked up Is the timing set back all the time or is the timing correct and the sensor has to be there and sensing dentonation for it to set the timing back. Hope that makes sense
 
When I rebuilt my 94 I forgot to plug in the knock sensor and I got a check engine light. Make sure yours is lit before you start up the car. Im sure this would have some how effected your performance as the computer dose not see the sensor and tries to adjust for it.
 
kenewagner said:
Thanks for the info: Now if I could only solve the rough idle problem I would be a happy camper
You need to pull codes first.. I really don't trust that funky O2 sensor hookup you showed us. It looks rigged for trouble.
 
scbird1 said:
When I rebuilt my 94 I forgot to plug in the knock sensor and I got a check engine light. Make sure yours is lit before you start up the car. Im sure this would have some how effected your performance as the computer dose not see the sensor and tries to adjust for it.

That is strange, I have not had check engine light come on even though I am having Idle problems. That is something nobody has been able to tell me, would my performace suffer. It doesnt seem like it should. The cars performace has been dissapointing. I have gone to a 10% pully a FMIC, raised top. The car seems to be running very rich, almost no idle. Everyone seems to think vaccuum leak but I have used propane to check all connections. I plan to replace coolant sensor as well. How about it george, will replacing the knock sensor make any diffrence in cars performace??? :confused:
 
Dose the light come on with key on engine off? If not some one has removed the bulb. This is common when some one dummies up a car to sell it. My 94 ran like crap untill the sensor was pluged in. Im not sure about your year.
 
scbird1 said:
Dose the light come on with key on engine off? If not some one has removed the bulb. This is common when some one dummies up a car to sell it. My 94 ran like crap untill the sensor was pluged in. Im not sure about your year.

Check engine light works. The exhaust tips are black inside and the smell of gas is very strong. It could be a faulty gas pressure regulator or an injector sticking open. I am going to replace Knock and coolant sensors and if that doesnt do it take it to the ford dealer and let them track it down for me. I can fix it if I know whats broke
 
scbird1 said:
When I rebuilt my 94 I forgot to plug in the knock sensor and I got a check engine light. Make sure yours is lit before you start up the car. Im sure this would have some how effected your performance as the computer dose not see the sensor and tries to adjust for it.


I am pretty sure that the early cars cannot tell if the knock sensor is connected or not. It's basically what you find in an "electronic" cig lighter. A Knock sensor is a piezoelectric device and does not have a resistance unless it is flexed (knock) at which point it sends a signal (a voltage spike) which the EEC see's as knock. Basically it is a Quartz crystal that is sized to react to a certain frequency, based on piston size weight and other factors.

I guess the later cars pulse the knock sensor to determine if it is there or not.

Piezoelectric devices work both ways. If you flex it it will produce a voltage and if you put voltage on the device it will flex. (ever hear of a Piezo tweeter?)

Jeff
 
guys, i checked this thread and the linked thread above and still havent found quite what i need....

first of all at the begining of this rebuild i ordered a new knock sensor from ford. they gave me the F3LY-12A699-A part number. this sensor has the same connector as stock and the threads match the threads on the block adapter.

tryin to find out exactly how the 94 style sensor reads and what the sensor output should look like when datalogged during normal operation and when sensing knock.


trying to get this thing tuned and i really think the knock sensor is causing the PCM to pull timing out under load. im not getting any detonation so im not sure why its doing it?

Thanks
Chris
 
I think what Jeff was saying is that all the SC knock sensors are the same, just that the 94's are pulsed to check if they are there or not. In either case if the sensor is not there it should operate just as if there were no knock since the sensor only sends a signal if there is knock.

The way to verify if the sensor is working is to run KOER test without attempting to engage the knock sensor. You should get a knock sensor code. Then run the test and tap on the exhaust manifold when directed. Code should do away.

I have nothing to show exactly what should be read on a DMM in the case of knock. Can't you just run a an LED in parallel with the circuit to monitor knock sensor action?
 
XR7 Dave said:
Can't you just run a an LED in parallel with the circuit to monitor knock sensor action?

yes but thats not my intent.. i believe that the computer is getting signal back from the knock sensor showing knock. in turn the computer is pulling timing out. i have a wideband with datalogging capabilities and i have an auxilary input setup as the knock sensor. im trying to interperet the readings im getting

Chris
 
XR7 Dave said:
I think what Jeff was saying is that all the SC knock sensors are the same, just that the 94's are pulsed to check if they are there or not. In either case if the sensor is not there it should operate just as if there were no knock since the sensor only sends a signal if there is knock.

The way to verify if the sensor is working is to run KOER test without attempting to engage the knock sensor. You should get a knock sensor code. Then run the test and tap on the exhaust manifold when directed. Code should do away.

I have nothing to show exactly what should be read on a DMM in the case of knock. Can't you just run a an LED in parallel with the circuit to monitor knock sensor action?

I'm not sure if the knock sensors are the same or not, just saying that early cars don't "look" for the knock sensor. In order to read the signal from the knock sensor you will need an input of about 400V, though I am not sure exactly high it is, it may be on the order of 1kv. In any case it's a high voltage. An LED (LED's can only handle about 2-3V and they must be current limited) could be used, but first you need a peak and hold circuit and a buffer of some sort (for the input). The length of knock is very fast and would be easily missed by the eye (anything over 20hz looks as if it is on all the time, our eyes respond very slow), so you need to streach this time out so you can see it. I am not sure if your datalogging equipment has a port for knock sensor or not, but if it's 0-5V or a digital input it isn't going to work.

Jeff
 
J57ltr said:
I am not sure if your datalogging equipment has a port for knock sensor or not, but if it's 0-5V or a digital input it isn't going to work.

Jeff

this is what i needed to know.. i am using a 0-5v auxilary input
on the graph it shows as an up and down pulse all the time and during a WOT blast the pulse will spread out get and get jagged.

i was interperetting that WOT blast as the computer sensing knock
 
I assume that is a analog input and not a digital input. Are they very short peaks (duration), but they ramp all the way to 5V then down to 0V. The slew rate will be very high and the pulse should last only a few ms.

When you say that it shows an up and down pulse all the time, you mean that when idling or driving normally it pulses like this? Are the pulses the same frequency (distance between peaks) when this occurs, or do they raise or lower (frequency) with RPM? Does it read these pulses with just the key on and engine off?

When you said that the pulses spread out and get jagged, Do you mean that the up ramp and down ramp get farther apart? Or do you mean that each pulse gets farther from the next pulse? Does the up and down ramp look like it is a series of progressivly larger spikes starting at 0V and moving right until it hits 5V?

Jeff
 
Back
Top