FI Code 25 Knock Sensor Issue - HEALED!

Khav

Khav

Joined
Jun 21, 2013
Messages
19
Location
West Bridgewater, PA
Bike
2004 ST 1300
Rather than tack this on to any of the many threads on the issue, I thought I'd post a new one as this event seems worthy.

So...about a year ago I noticed that the FI light would come on anytime I ran my '04 above 4000rpm. In addition, the bike would lag.

Digging into the issue, I found that it was throwing a 25 code, so I scoured the forum for potential issues, fixes and advice. From ECU's to harnesses to sensors, a variety of potential causes presented in the forums...the most likely from my horribly optimistic point-of-view was the wiring. I had always noticed that the cowling seemed to hold massive amounts of heat, so it made sense. The fix was also well below the cost of a new ECU, so I decided to try that first. I ordered the sensor pigtails (the cheaper automotive part) and set them aside for a sunny day.

Fast forward to a couple of weeks ago when I pulled the Duchess from storage and went about firing her up. No pulse. The 6 year-old battery finally went to AGM Heaven, so I ordered another as well as a new solenoid relay...it had been stubborn at times the previous season, so I figured I'd switch it out.

I replaced both yesterday after work and took her for a spin. After a little warm-up, I let her breathe a bit and opened her up...and much to my surprise, I was turning 5000 rpm without a hint of the FI light coming on. I ran her up several times for extended periods and...nada. No FI light or hesitation. The engine seemed more responsive, too.

Aside from Divine Intervention, the only explanation I can come up with is voltage...or lack thereof. Either my old battery was not producing sufficient, consistent voltage or the relay was sapping it somehow and perhaps the ECU was becoming unhappy at the slight.

Or...maybe it's just an anomaly and the FI code will return tomorrow. Dunno.

So, as well as wanting to share my cautious elation, I also wondered if anyone had experienced a similar miraculous healing or had any insight.
 
did you ride for a year with this condition
caused no other problem?
I can go hundreds of miles with no light as long as I don't hit 4000 for more than a few seconds
very good speed limiter device
my license likes it ha ha
 
axel3570 said:
did you ride for a year with this condition
caused no other problem?
I can go hundreds of miles with no light as long as I don't hit 4000 for more than a few seconds
very good speed limiter device
my license likes it ha ha
Click to expand...

LOL...EXACTLY! I ran it that way all last season because there was too much going on elsewhere in life to address it. It made me pay close attention to RPM and, as mentioned, it ended up being a speed limiting device...and a buzz kill...and ultimately as much of a distraction as a 'safety' device. Top end was ~70mph which corresponds to max speed on the interstates here. On the odd occasion when I would cross the threshold and notice the cylinder drop, I would simply 'kill and bump' while rolling to reset. I have noticed no resulting issues.

I will not presume to know all there is about it, but as numerous threads here have discussed, it is a common knock sensor issue that has been attributed to the wiring harness and/or sensor connectors deteriorating as well as the ECU sometimes being suspected as the culprit. Perhaps we can now add voltage to the list of suspects. Let us know how it turns out...I will post if anything changes on my side.
 
Well, that's great news Khav - its always good when a fix comes through at a manageable cost without a great deal of angst.

In my experience, electronic devices seldom fail. The problem is nearly always either a bad/corroded/loose/dirty connection (most often a ground connection rather than the "live" side) or a low voltage issue as a result of some other easily fixable problem in the circuit. It appears that yours was the latter.

Pete
 
A low battery can make your FI/EI bike do weird things.
A bad ground, even weirder.
Always check them both before going further into electron land. One of the many things I have learned the hard way.
 
That's good to hear. I had code 26 for quite awhile and tried several fixes with no success. In the end mine ended up being the ECU but I at least know that several parts have been replaced and should be good for a long while.

Skip
 
*Update*

There was a break in the dismal Western PA weather, so I made a run to the grocery between torrential downpours yesterday. Unfortunately, I didn't fully manage the between, and I was met by rain and pellet-sized hail on my way home. I can deal with rain, but God bless anyone who has to maintain highway speed while getting pelted with ice.

Anyway, I took the Duchess well above 4000 rpm on the way out and there remained no trace of the dreaded FI fault. I would definitely add an under-performing battery to the list of suspects for such codes.
 
had knock sensor code for 3 months read all the posts finally got used ecu off ebay 200 bucks hooked up test ride no light did nothing else
happy now hope info will help someone else go figure
 
FWIW, on CBR1100xx forum, code 25 is 95% caused by faulty ECM. The other 5% is the ECM isn't strapped down. I would think someone must work on these ECM's by now with all the common failures.
 
jmcarruth said:
A low battery can make your FI/EI bike do weird things.
........
Click to expand...

I have an '06 Yamaha FZ1. Injected and very high performance. I installed 2 accessory outlets. One time on a trip I have my GPS plugged in, heated grips on, and decided to plug in my phone to charge. Bike started running like crap. Stumbling, missing, weak running. Stopped for lunch, pulled my phone off. After lunch bike ran great. Guessing it might be the phone charger I plugged it in while riding and it ran bad again. FI is very, very, sensitive to voltage draw.
 
I've looked over this sight to solve problems in the past, both with a st1100 and my current st1300 .I'm not really a joiner so here goes... I'm an electronic engineer with 35+ years in machine control and repair .I've followed the code 25/26 thread out of curiosity-and I may have noticed the common thread in the failure. Higher operating temps lead to greater fuel trim / a thinner mix at speed leads to higher snap KV requirements with less soak between pulses(4k rpm commonly noted). If the filtering on the supply to the ecm becomes degraded then noise from the ignition secondary could invade and partially swamp out the a/c signal coming on the knock sensor input / by piling the noise onto the ground reference.I would recommend a shunt on the ecm ground to observe outside interference(scope) and choking or filtering any circuits with similar frequencies to the noise .An additional filter or choke to recover faded noise rejection would be cheaper than a new ecm. High difficulty level with potentially high reward. I live in cleveland area and would be interested at seeing this problem 1st hand .Good luck/testing and filter design are electronics 101.
 
pearlsteinp said:
I've looked over this sight to solve problems in the past, both with a st1100 and my current st1300 .I'm not really a joiner so here goes... I'm an electronic engineer with 35+ years in machine control and repair .I've followed the code 25/26 thread out of curiosity-and I may have noticed the common thread in the failure. Higher operating temps lead to greater fuel trim / a thinner mix at speed leads to higher snap KV requirements with less soak between pulses(4k rpm commonly noted). If the filtering on the supply to the ecm becomes degraded then noise from the ignition secondary could invade and partially swamp out the a/c signal coming on the knock sensor input / by piling the noise onto the ground reference.I would recommend a shunt on the ecm ground to observe outside interference(scope) and choking or filtering any circuits with similar frequencies to the noise .An additional filter or choke to recover faded noise rejection would be cheaper than a new ecm. High difficulty level with potentially high reward. I live in cleveland area and would be interested at seeing this problem 1st hand .Good luck/testing and filter design are electronics 101.
Click to expand...
Welcome! The sharing of ST knowledge and general BS makes this a great site.
 
Can you still pull the code from the ECM memory or is it gone there too? Normally the ECM will store the codes for some period of time, even if the FI light goes off to help diagnose intermittent issues.
 
Or you can do a reset and clear the code. The process is in the manual and I've had to do it several times after thinking I've fixed my nuisance warning. Still get it but will look into actually swapping out the knock sensor(s) next time I'm in there. Until then, it's just a nuisance but since I ride highway to/from work (majority of miles on the bike) it is annoying for sure.
 
skipcurt said:
That's good to hear. I had code 26 for quite awhile and tried several fixes with no success. In the end mine ended up being the ECU but I at least know that several parts have been replaced and should be good for a long while.

Skip
Click to expand...
Yep, me too. Checked the wiring, all good (even though the RH knock sensor plug is ‘partially’ melted. Replaced the knock sensors with a couple removed from an ex-police scrapper. Finally buckled under and went for the ECU (tried to find a used one but nada). Paid ridiculous amounts of money for the ECU but it’s now fixed and the world is good .
 
I’m gonna say this with all the wistfulness and staring off into space that I can muster. But while reading about the FI light someone mentioned disabling code 25 and 26. Other than sticking duck tape over the FI light is it possible to disable these codes.? The codes come off of Maps that the microprocessor reads and when it gets either a lack of or too much of a 5 V signal it gives you the mil code in the form of two quick flashes and five or six flashes after that.
I understand that by disabling these codes it would put you in Dutch with the Honda motorcycle manufacturers. but we have all complained mightly about these error codes and the only point where they stepped up to the plate was in 2008 when they improved it. Since then no error codes. Why can’t they fix 2002 through 2007. Or why can’t we disable the error code. Has anyone ever had a knock sensor really go out?
 
Kurt0721 said:
I’m gonna say this with all the wistfulness and staring off into space that I can muster. But while reading about the FI light someone mentioned disabling code 25 and 26. Other than sticking duck tape over the FI light is it possible to disable these codes.? The codes come off of Maps that the microprocessor reads and when it gets either a lack of or too much of a 5 V signal it gives you the mil code in the form of two quick flashes and five or six flashes after that.
I understand that by disabling these codes it would put you in Dutch with the Honda motorcycle manufacturers. but we have all complained mightly about these error codes and the only point where they stepped up to the plate was in 2008 when they improved it. Since then no error codes. Why can’t they fix 2002 through 2007. Or why can’t we disable the error code. Has anyone ever had a knock sensor really go out?
Click to expand...
The ECM needs to see the knock sensor respond to the ECM over advancing the timing that way it can keep the spark timing at the most efficient setting. Lets say total advance is 34 deg btdc when the fault is present it goes into default of (just guessing) 25 degrees.
 
Al st1100 said:
The ECM needs to see the knock sensor respond to the ECM over advancing the timing that way it can keep the spark timing at the most efficient setting. Lets say total advance is 34 deg btdc when the fault is present it goes in
Click to expand...
Al st1100 said:
The ECM needs to see the knock sensor respond to the ECM over advancing the timing that way it can keep the spark timing at the most efficient setting. Lets say total advance is 34 deg btdc when the fault is present it goes into default of (just guessing) 25 degrees.
Click to expand...
so there is one sensor for 2 cylinders. How can it do such a critical function I figured it was an alert and less of a calibration probe.
 
You must log in or register to reply here.
Back
Top Bottom