Hello all, i have a 90 model dx crx with a B16A1 swap, stock ecu to the best of my knowledge. and it is throwing a code 23.

I thought the simple fix was to replace the knock sensor. However i have tried 3 different knock sensors now and the code doesnt go away. I have pulled fuses after each sensor replacement in an attempt to reset the code. I have also ohmed out the wire coming from the sensor to the computer to check if the wire was bad but it does conduct current.

any suggestions?


FWIW the CEL light only comes on when i get up to temp. but the computer will flash 23 before that.

 

bump

 

From what you said it seems like you have the right idea. Be patient and thorough and test one thing at a time to determine the problem.

1. To reset ecu you can remove the 10amp "hazard" fuse from under hood fuse box for 10 seconds. You may also want to disconnect Neg terminal on battery to be absolutely sure.

2. Try to replace knock sensor with a known good knock sensor.

3. You tested for continuity on the wire, but what if it has an intermittent problem? I would just run a new wire and test it out. (don't bother routing this wire too carefully, it's just a test. drive the car and see what happens)

4. If after all that and you still have the code, try another ECU. If you still have the code after the new ECU, then it's either the sensor or the wire.

Hope this helped.

 

i appreciate the response and i will be running a test wire. i have forgotten to mention that i tried another ecu that my friend had. he doesnt know much about it but it was from some modded b16. i tripped the same cel code there as well.

as for using a known good knock sensor i dont know anyway to test it short of swapping it out of a car with someone else and i havent found any willing victims yet.

 

also make sure that the sensor is grounding to the block good because it gets its ground from the block and make sure all the grounds on the engine are good and clean and tight. If the plug isn't making good contact with the sensor or has a lot of oil inside it that could cause problems.

 

Alright putz, take yer knock sensor off and stick it up your *** for about an hour (just to warm it up) then make a ground wire and attach it securely to the body of the knock sensor attach the knock sensor wire and wrap the whole thing in foam or other such insulating material and secure it to the chassis in a way as to isolate it from any vibration. Clear code and drive normally (the way you were before when the code came on). If no code returns then you have an issue with the sensor either picking up noise from something loose and banging around in that heap, the sensor itself or your distributor might have an issue being retarded like its owner (I love abusing you!) Also a knock sensor is not that much IIRC? I had one at one point, but I can't find it in the mess I call a garage. You mentioned that you replaced the knock sensor, was that with a new one? If so ignore what I said about the sensor being bad

 

knock sensors are $108. i have tried 3 different in the car with no change in result, one new one used and this one.

ill wrap that sucker up in one of your adult diapers and see if it is still picking up noise. ill report back here.


is that a glock in your sig? fine weapon isnt it.

 

That's an HS-2000 The finest combat handgun ever produced!!

Look around for loose **** in your engine compartment, broken brackets and other such "rattled loose" items as that engine is a vibration machine with those mounts

 

i have an hs200 and i must say ever since i got it i havent had a problem with loose papers flying around my desk should there be a light breeze. it does function well in that regard.


i unhooked the knock sensor, grounded it out and left it plugged into the computer. then i wrapped it in some old towles. then drove gently down the road for almost 2 blocks before the cel came on again just like always. yes i did i clear the codes before checking.....


so what now babycakes

 

am i looking at having to chip this?

 

Have you tried a different ECU????

 

yes but it threw the same code

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by davidu &raquo;</TD></TR><TR><TD CLASS="quote">am i looking at having to chip this? </TD></TR></TABLE>

that would solve it, but there is some other problem that is at work here that should be solved without writing a program around it. Is the car OBD1 or 0 (I don;t remember) If it's 1 then just throw a P28 in and see if that fixes it (OBD1 P28=no knock)

Has this been doing it all the time or is this something new? In other words did this do it when you drove it home from here?

Also can you find a distributor to swap and try that first?

 

Alright I reviewed the manual in regards to the knock sensor and according to that there are only 3 things you have to worry about 1st the knock sensor 2nd the ECU and 3rd the wiring between. If 1&2 have been replaced then the only thing left is the wire between. Disconnect the wire from the knock sensor and the plug from the ECU check continuity between if thats good check to see if there is a connection to ground (if there is fix it)

Let me know what you find

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by davidu &raquo;</TD></TR><TR><TD CLASS="quote">i have forgotten to mention that i tried another ecu that my friend had. he doesnt know much about it but it was from some modded b16.</TD></TR></TABLE>

That hardly qualifies as a known working ECU.

Anyways, as I pointed out in my first post, its either the sensor the wire or the ecu. best of luck

 

" cool it runs, whats the check engine light?"

"thats because your to cheap to buy an o2 sensor"

"ok thanks for all the work you did on my car im leaving town now bye"


fast forward 3 years, i buy an o2 sensor the light doesnt go away, i pull the cover off the ecu and check the code. its the knock sensor.

so maybe it has always been this way.

el gringo es correcto the ecu i "tested" hardly qualifies. it seems odd that someone chipped an ecu and did not delete the knock. especially since there was a marked improvement in the cars performance when i used that ecu.

so to recap.

there are 4 things that could cause this.

1. back knock sensor
The only test for this seems to be finding a good car with a stock ecu to test it in

2. back wire
i ohmed this wire out and it tested clean

3. bad ecu
need a stock ecu to test it and unless its from a car i see running how do i even know its good?

4. my car is making a knock that is legitimately throwing the code.



so the way i read this i need to find some poor bastard with a obdo b16 and buy him enough beer so i can futz with his car.

or pony up $65 to get my ecu chipped

 

Do you have a PW0 or PR3?

 

pwo

also while reading up i see im supposed to have 2 oxygen sensors i only have the one. do you remember how it is wired up?

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by davidu &raquo;</TD></TR><TR><TD CLASS="quote">pwo

also while reading up i see im supposed to have 2 oxygen sensors i only have the one. do you remember how it is wired up?</TD></TR></TABLE>

I'm lucky if I remember to wipe my *** after shitting

Do you have a code 1 too?

 

cel code 1... no

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by davidu &raquo;</TD></TR><TR><TD CLASS="quote">...

1. back knock sensor
The only test for this seems to be finding a good car with a stock ecu to test it in

2. back wire
i ohmed this wire out and it tested clean

3. bad ecu
need a stock ecu to test it and unless its from a car i see running how do i even know its good?

4. my car is making a knock that is legitimately throwing the code.

...</TD></TR></TABLE>

1. To test the knock sensor put the + of a voltmeter on the part where you plug the knock sensor into the loom. Put the - of the voltmeter on the body. Tap the sensor gently with a hammer or something and the voltage should fluctuate. If you do this correctly and the voltage does not fluctuate then it is not working correctly.

2. The wire could be shorting out. To see if it is shorting out to earth disconnect the wire at both ends and probe it and earth, if there is continuity then it is shorting out to earth.

3. ECU is a possibility, I suppose.

4. If your car does have a knock you would probably know, and if the sensor was working and detecting this then you would NOT have a check engine light. AFAIK Code 23 only occurs when the ECU is not receiving the expected voltage from the knock sensor when the car is warmed up. This is why it comes on when the car is warmed up and the car is at an RPM where a voltage should be being created by the normal vibrations of the engine. If voltage is too high then the ECU assumes knock and that is when timing retard occurs, AFAIK this does NOT trigger a check engine light.

Hope this helps.

 

1. To test the knock sensor put the + of a voltmeter on the part where you plug the knock sensor into the loom. Put the - of the voltmeter on the body. Tap the sensor gently with a hammer or something and the voltage should fluctuate. If you do this correctly and the voltage does not fluctuate then it is not working correctly.

2. The wire could be shorting out. To see if it is shorting out to earth disconnect the wire at both ends and probe it and earth, if there is continuity then it is shorting out to earth.

3. ECU is a possibility, I suppose.

4. If your car does have a knock you would probably know, and if the sensor was working and detecting this then you would NOT have a check engine light. AFAIK Code 23 only occurs when the ECU is not receiving the expected voltage from the knock sensor when the car is warmed up. This is why it comes on when the car is warmed up and the car is at an RPM where a voltage should be being created by the normal vibrations of the engine. If voltage is too high then the ECU assumes knock and that is when timing retard occurs, AFAIK this does NOT trigger a check engine light.

Hope this helps. <TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by DeltaSlaya &raquo;</TD></TR><TR><TD CLASS="quote">

1. To test the knock sensor put the + of a voltmeter on the part where you plug the knock sensor into the loom. Put the - of the voltmeter on the body. Tap the sensor gently with a hammer or something and the voltage should fluctuate. If you do this correctly and the voltage does not fluctuate then it is not working correctly.

2. The wire could be shorting out. To see if it is shorting out to earth disconnect the wire at both ends and probe it and earth, if there is continuity then it is shorting out to earth.

3. ECU is a possibility, I suppose.

4. If your car does have a knock you would probably know, and if the sensor was working and detecting this then you would NOT have a check engine light. AFAIK Code 23 only occurs when the ECU is not receiving the expected voltage from the knock sensor when the car is warmed up. This is why it comes on when the car is warmed up and the car is at an RPM where a voltage should be being created by the normal vibrations of the engine. If voltage is too high then the ECU assumes knock and that is when timing retard occurs, AFAIK this does NOT trigger a check engine light.

Hope this helps.</TD></TR></TABLE>

thank you for your help, here is what i found

1. i tried testing both my knock sensor and 2 different used knock sensors of unknown origin. on none of them could i get a reading. i tried continuity (the little beep) and i tried to "ohm them out" i could not get current thru them regardless of whether i was knocking on them or not.

2. i rechecked the wire for continuity, it checked well. i disconected the computer and the sensor and i check for continuity to ground from either end. no current at all going to ground and the wire tests fine.

3. if my ecu is bad, would chipping it fix the "bad" or would i need to replace it with a new ecu regardless?

4. i hope you are correct it does make sense.


once again thanks to everyone for all your help. does anyone have a pr3 or pw0 ecu they want to donate to my charity of choice?

 

<TABLE WIDTH="90%" CELLSPACING=0 CELLPADDING=0 ALIGN=CENTER><TR><TD>Quote, originally posted by davidu &raquo;</TD></TR><TR><TD CLASS="quote">thank you for your help, here is what i found

1. i tried testing both my knock sensor and 2 different used knock sensors of unknown origin. on none of them could i get a reading. i tried continuity (the little beep) and i tried to "ohm them out" i could not get current thru them regardless of whether i was knocking on them or not.

2. i rechecked the wire for continuity, it checked well. i disconected the computer and the sensor and i check for continuity to ground from either end. no current at all going to ground and the wire tests fine.

3. if my ecu is bad, would chipping it fix the "bad" or would i need to replace it with a new ecu regardless?

4. i hope you are correct it does make sense.


once again thanks to everyone for all your help. does anyone have a pr3 or pw0 ecu they want to donate to my charity of choice?</TD></TR></TABLE>

1. Ok you put the - probe on the sensor body, and the + probe on the terminal. You are not looking for continuity or amperage, you are looking for VOLTAGE. So the multimeter needs to be on voltage, on the lowest scale if possible, as the voltage created is small. Tap on the bottom of the sensor.

2. Sounds like the wire is fine, unless its shorting out to another wire etc...

3. Chipping the ECU, if it is at fault, would not fix the bad, it would just mean that the ECU ignored all knock sensor input. If you do this, just make sure your engine is not actually knocking, because the timing will not be retarded.

4. Yip, I'm pretty sure that's how it works.

 

whats up DavidU? I met you at Advance Auto today. I was the guy with the black
90 CRX Si. Since I'm still a "NOOB" I wasnt able to send you an IM so I figured I would just find one of your posts. Hope to see that crx of yours kickin up some dust sometime.