Steve Saunders Goldwing Forums banner

1 - 14 of 14 Posts

·
Administrator
Joined
·
57,342 Posts
ran my 2002 with a "bad" left knock sensor for a long long time, with no ill effects.


other than, poor mpg.


the ECM goes into Fault Mode so the fuel mileage suffers.


BTW, the knock sensor itself is not bad.
The ECM is the culprit, and it is epoxy coated and non-repairable.


I was lucky, got on www.SearchTempest.com and found a 2004 ECM 30 miles from home. could not believe it. it works like a charm.


2001-2005 are all the same.


https://www.ebay.com/itm/283419216473?ul_noapp=true


this is the only one available as of today
 

·
Premium Member
Joined
·
12,683 Posts
Ron, I just had this problem with my 03. I was getting a code 25- right knock sensor. I pulled the lower pans and the front lower cowl so I could access the sensors. I unplugged the right sensor and tested it with a digital volt meter. By tapping on the block near the sensor I would get a volt reading . Sheesh, I dont recall if I was on AC or DCscale:(. But I did get a reading and moved to the left one just to compare and similar readings. I put it all back together to ride till I could source an ECM. Its been fine the couple hundred miles since. I may have been the lucky one in a hundred but it seems as though I only had a bad connection.
You might at least try to disconnect and reconnect to see if that helps.
 

·
Administrator
Joined
·
57,342 Posts
the knock sensor outputs an A/C waveform that is Very Fast.


an Oscilloscope works best, but a fast digital DVM might catch the readings.


DaveO430 used the Hertz ( Frequency Measurement ) setting on his DVM to test my bike for a bad sensor.


at first, we switched them from side to side, the code stayed the same, indicating it is the ECM that is at fault.


then, Dave tested both sensors for the A/C output waveform, and yes, both of them put out the same signal.


it is very small, and very fast. look close.
 

·
Administrator
Joined
·
24,334 Posts
the knock sensor outputs an A/C waveform that is Very Fast.


an Oscilloscope works best, but a fast digital DVM might catch the readings.


DaveO430 used the Hertz ( Frequency Measurement ) setting on his DVM to test my bike for a bad sensor.


at first, we switched them from side to side, the code stayed the same, indicating it is the ECM that is at fault.


then, Dave tested both sensors for the A/C output waveform, and yes, both of them put out the same signal.


it is very small, and very fast. look close.
As I recall I could never get anything other than a Hz signal from them.
 

·
Junior Member
Joined
·
103 Posts
As I recall I could never get anything other than a Hz signal from them.

Actually everyone is correct in their own way.



One test is done to test the Bias Line Voltage which is in DC volts measurement.


because the bike is in fact a DC electrical system.
Bias line voltage is a circuit integrity check for computer.




and then the signal produced when activated is an AC voltage, measured in HZ.


The Ac voltage runs along the Bias line voltage.



By tapping on engine block or near sensor it should send a signal and should be able to read on Ac volts scale as well, as HZ scale.

And Kudos to John, John you would be surprised how many times for example
Mechanics I have worked with will condemn an Ignition coil on a car with multiple coils.
Without fully testing it. They could have moved the suspected coil to another cylinder.
And see if problem follows the swapped Cyl. You John did this on your bike by swapping around Knock Senor to prove it was good or bad.



For that I give you a gold star....bravo :smile2:


:smile2:
 

·
Premium Member
Joined
·
154 Posts
Discussion Starter #12
knock sensor, update

Threw a #26 code so Lf Knock Sensor is replaced today.
Hope FI light stays off.
Replaced ECM 3 years ago so I hope I don't need another.
About 50 miles so far. Feeling lucky! :praying::praying:
 
1 - 14 of 14 Posts
Top