FI Code 26

Joined
Jan 15, 2013
Messages
373
Location
Prairieville, Louisiana
Bike
'15 FJR ES
Code 26. Two long blinks followed by 6 short blinks. Right Knock Sensor.

Well, maybe. I originally piggybacked this onto another thread but now I think my story deserves its own thread. No, all stories don't have happy endings.

Running the engine above 3900 RPM for 10 seconds will generate this fault code. Instantly power is cut, and fuel mileage plummets. 3900 RPM looks like 72 mph on the speedometer. It is 68 mph on the GPS. This fault will give you 37 mpg at 73-74 GPS mph on the Interstate.

First, if the red FI light in the dash comes on you must read the code. Reading the code is very simple. If you know how. Engine running, FI light on, bike in neutral, lower the kickstand. Then count the blinks. A long blink (1.3 seconds) counts as 10. In my case two long followed by 6 short equals 26.

Your Genuine Honda Service Manual will have a chart telling you what the # of blinks translates into. My bike is an '04 so this is found on pages 5-14 thru 5-16.

A quick and desperate check of both ST forums told me that most often the problem was not the sensor. It is the plug and or wiring breaking down from the intense heat of its location. Cutting off the plug and replacing with a suitable spade connector solved almost everyone's issues.

I removed all the Tupperware from the right side. I unplugged the knock sensor. It is inconveniently located between the exhaust header pipes and the engine block. I checked the plug and it looked fine.

The troubleshooting section of the Honda Service Manual says to check for continuity to ground from the plugs on either end of the wiring harness. Then it says to check for continuity between the two plugs. Basically, you are checking to see if the wire has been damaged and is going to ground or has burned/rubbed/broken in two between the sensor and the ECU. If it is not going to ground and the wire is intact, the problem must be the sensor.

*Problem*
The Honda Manual is WRONG concerning the wire colors for my '04 ST. It clearly states on pages 5-49 and 5-50 that the wire colors are:
Left Sensor Wire-Blue
Right Sensor Wire-Red/Blue
This is not correct for my bike. The manual has two sections for two different year groups. The first is for '03-'07. The second is for '07 and later.
My ST is wired like the '07 and later version. My sensor wires were:
Left Sensor Wire-Red/Blue
Right Sensor Wire-Black

Using the wire colors the manual stated for my year I decided the wire between the sensor and the ECU was bad. I ran a temporary wire and cut/spliced into the wiring harness just ahead of the ECU. Tried running the bike, still got the code. Purely by accident I looked at the section for '07 and later and was able to re-diagnose the wiring.

I wound up removing the left side Tupperware during the confusion phase so I tried running separate wires under the bike to "switch sides" with the sensors. I stuck the stripped ends of #14 wire into the female sensor plugs and used spade connectors to attach the other end of the wires to the sensors. No code. Re-connected the plugs. Got Code. Stuck wires back in, Ran bike. No code. R-connected plugs. Got Code.

Sigh... Plugged left sensor back in. Jammed wire back into female plug on right side and plugged spade connector onto right knock sensor. No code.

Decided the connectors inside the female end were "sprung" out and not making good connection to sensor. I found that the plug (sensor end) was not making a good connection at the sensor. I attempted to use a tiny screwdriver to pry the metal connectors out and...

The plug shattered in my hands. The more I tried to repair, the more the brittle plastic crumbled in my hands. I replaced the plug with a tiny spade connector.


Ran the bike on the centerstand, no code. Fixed it!

Or not...

Over the next couple of weeks the 26 Code would re-appear. It was intermittent, it was unpredictable. No correlation to humidity, weather, temperature, or bike temperature could I find. I lost confidence in my ST and in my own troubleshooting skills. I began to doubt myself and my bike. I need to have confidence that if I have to I can get on my motorcycle Right This Minute and ride it Coast to Coast. If I can't trust the reliability of the bike, I don't need it.

I removed the Tupperware several times to check my connection. I removed and re-connected the ECU several times. I double checked everything.

I decided that the resistance of the wire between the new spade connector and the plug at the ECU was too high. It is at most 5-6 feet of wire but there was about 4 ohms of resistance in that short length of wire. I would have expected 0.4 ohms, not 4.0 ohms. I replaced the factory wire by running a separate length of wire with a high temperature insulation. I felt confident. Ran on centerstand, no code! Success!

Well, no. Continued to get code intermittently.





Now I am so depressed and disgusted that I am making everyone around me miserable. The ST is such a huge part of my identity and thinking that I am more affected than I ought to be. Yamaha announces the '14 FJR in Candy Red. Factory Cruise Control, electronic suspension, traction control, heated grips... Oh man, things are getting bad. Mrs. Redfish finally says, "I know you don't like giving up but...have you considered buying a new bike?"

I removed the Tupperware. Again. I poked around the sensor connection. Again.

Definition: Stupidity- Doing the same thing over and over expecting different results.

Then I remembered. I am an I&E Technician. I routinely troubleshoot and repair equipment that costs more than my (wife's) house and can affect millions and millions of dollars worth of production. I am supposed to know things.

What is a knock sensor anyway? It senses vibration. It would be called a vibration sensor in an industrial application. In the presence of vibration a millivoltage charge is generated by the sensor. The computer reads that millivoltage signal and interprets the amount of vibration present.

I had already installed two excellent test points at the ECU plug in the form of two butt splices. I connected my meter and started the bike.

I found that both the right and left knock sensors generated about -114.5 millivolts with the bike idling. Both generated about -122 millivolts at 4100 RPM. No appreciable difference between right and left sensor. No spiking, no dropping, no PROBLEMS!

Yet, I am still getting the code. Now, I get it all the time. I check the connectors in the ECU plug. I gently pry (with a fish hook) to get a bit more tension. I am sure there is no problem with the plug.

I cut the wires. Again. I use wire nuts to temporarily switch the left and right wires where they go into the ECU plug. I run the bike.

Still getting the 26 code. If the problem was in the wiring or the sensor, the code would have changed to the Left Knock Sensor, Code 25. But it didn't.

At this point I have determined the ECU is faulty. The PGM-FI is about $825 online.

I'm going to go drink a beer now.
 
I found that both the right and left knock sensors generated about -114.5 millivolts with the bike idling. Both generated about -122 millivolts at 4100 RPM. No appreciable difference between right and left sensor. No spiking, no dropping, no PROBLEMS!

Did the knock sensor light off during this test ?

If you switched the wiring side to side and it still says the same side is knocking, then I would think
it's the ecu..

Perhaps you can borrow and ecu to test out .

The ecu's are usually available on e-bay for around a c-note.

( More wrecked bikes, than bad ecu's )

Maybe short out the leads (maybe a small resistor to force a no-knock reading )
 
Yes, the FI light came on even though there was nothing wrong with the signal coming to the ECU from the knock sensor. Both knock sensors are generating nearly identical millivoltage signals, there is nothing coming to the ECU to indicate a problem. Yet, the ECU indicates a problem.

I am not understanding what you mean about shorting the leads or using a resistor.
 
Just a thought but could the sensor be detecting an actual knock or vibration? Engine knock (
ping) caused by bad (low octane) fuel or pre ignition is often imperceptible by ear. This however usually happens under moderate to heavy engine load. Consider another source since this seems to be happening at light to no load. This engine has numerous balances, timing chains and idlers that can induce a knock of sorts. I even saw this same issue on a Honda civic with a bad alternator bearing. Just food for thought. How many miles on the ST? My apology if I missed that earlier.

Chuck
 
Last edited:
That is an excellent question and certainly one I considered very strongly. However, if you look toward the end of the story (if there is one) you will see that I read the millivoltage signal from both knock sensors with my multimeter. If there had been anything out of the ordinary it would have showed up then. Since Honda did not include any info on what that millivoltage reading should be I had to check both sensors to get a baseline. They are reading identical. Thanks for the question though. I ordered a new ECU today. Let's hope I am half as smart as I think I am...
 
I am dealing with the same aggravating issue except mine is code 25. Which is the other knock sensor code. Like yours it is not constant but usually comes on after riding the bike a couple of hours. I took the fairings off and checked both knock sensor connectors. Both of them appeared to be undamaged and the connections look fine. I also did the first test suggested in the manual and it passed so I didn't go any further. I was hoping it might just be some bad gas causing an actual but not noticeable knock so I ran some Seafoam through the motor on a five hour ride this weekend but it came back on after about 3 hours. I can't really detect any difference in the way the bike runs when the light comes on. I suspect mine might be an ECU problem as well. I hate to start throwing parts at the problem like a new wiring subharness if its going to end up being an ECU. I guess I am going to have to spend more time dealing with this.
 
When I get my ECU in and verify SUCCESS! I will be happy to ship you my old ECU so you can see if your problem shifts to code 26. Mellow was kind enough to give me a link in my ECU Questions thread to another thread with the Code 25 problem. That was an ECU issue as well. If I can help you, let me know.
 
Thanks for the offer. I may take you up on it if I can't resolve the problem. There is a thread by stgolfer that is pretty lengthy. He ended up getting a new ECU as well. I also know a local guy who has the same problem. He replaced the subharness which didn't fix it and ended up buying a used ECU of eBay. I may be able to convince one of the local ST riders I know to let me swap there 2007 and earlier ECU for a test ride or put mine in their bike. I am going to do another inspection of the wiring and maybe do some more disassembly to try and get a better view of the engine sub-harness. I'm still hoping it may just be a bad contact with the plug connector to the knock sensor. I guess I am willing to try repeating what I did before at least one more time. It would almost be better if it was constant so I could swap wires between the left and right sensors in the garage and see if it has any effect on the code numbers.
 
I tried everything I could think of and spent some miserable hours thinking about it. I did not know and was unable to find anyone who had suffered an ECU failure until after I had already narrowed it down to that. It is not the end of the world if you have to replace the ECU. It is only money. You don't even have to disassemble anything. Just remove the seat and you can fish it out. I was to the point of shopping for a new bike. If this takes care of it, I am in good shape.
 
Pretty hard getting those sensors out, isn't it?

My right side wiring was crumbling a bit, but I wrapped it in tape, then aluminum foil. All wiring tested fine and it was the ECM - $105 on ebay from a wrecked bike. Plug and play. Just make sure you get one from a 2003-2006. Not sure about the 2007, but the newer ones won't fit.
 
I have not had to take the sensors out yet I guess I am assuming that is not the cause since they apparently have a low failure rate. I have a trip scheduled to the Barber Vintage Bike Festival in northern Alabama coming up next week. If it is still doing it I will fix it when I get back. I still think it is a false reading, the engine seems to run fine.
 
When the FI light illuminates the power loss is noticeable but at first I was unsure. The throttle hand compensates so quickly and instinctively that I was not sure at first. Could it have just been a wind gust from an 18-wheeler? I was easily convinced when I saw my fuel mileage plummet. I know for sure that both of my sensors are sending the same millivoltage to the ECU. I read them repeatedly. The Service Manual does not list that as a test but it is the most useful part of the whole troubleshooting process for me.
 
Redfish - Yours is almost certainly a bad ECU, if your wiring is all good. Check on ebay - you should be able to get one for $100-$200.

Ridingfl - I'm in Palm Beach Gardens and there are a couple other guys in Lantana. I have an '03, which I believe has the same ECU as yours, if not, Jamie in Lantana has a '10.

Taking out the sensors is a major pain. I took mine out and switched them to see if the code would change to 25 from 26. It didn't. First step is check the wiring. If not that, probably the ECU.
 
Thanks Stgolfer I probably won't have time to do much about this before I leave next week. But I will contact you when I get back if it is still a problem. I would like to rule out the ECU one way or another.
 
Thanks I have the manual. The fairings are off right now I decided to install some driving lights and rewire some of the farkles. I will check the knock sensor connections again. I am also going to pull the plugs and see if they look ok. It is interpreting that it seems a lot of us having this problem are in Florida makes you wonder if it is really is related to heat and humidity.
 
Well, PartZilla never delivered the new ECU. When I contacted them I was told it was on back order. Honda does not have one. I made a trip to my local Honda/Yamaha dealer today. The ECU is on back order, no one knows when or if it will be available. If they could get it, MSRP is $1007! After taxes I would be at $1100!!! It was fortunate for all concerned that there was not a 2014 FJR on the sales floor.
 
I thought I had fixed my code 25 problem by checking the knock sensor connectors and using some dielectric grease. Drove from West Palm to Tifton GA up the turnpike and I-75 light never came on even though I was well over 4,000 RPM most of the ride. The next day it came on again. Then today it stop coming on after a fast ride on some curvy roads with a lot of down shifting etc. Then after a gas stop came on again. Now I am starting to think it is related more to engine RPM and vibration and may be a damaged cable or connector that is effected by the movement of the engine when it is working hard. The ride on the interstate kept the engine at a relatively constant speed with not much down shifting Etc. or it could be the ECU. I guess I will check the sub harness next. Sorry to hear about the back order. May be worth trying to get one on EBay.
 
Mine is intermittent just like you describe. At times more frequent with no consistency that I could tie to any one factor related to temp, speed, humidity, or any other condition. I ran an entirely new wire from the sensor to the plug at the ECU. It changed NOTHING! I confess that I never considered the ECU to be the issue when all this began. I have a line on a used one right now. I am patiently waiting (not really!) on the owner to return from the Blue Ridge Gathering.
 
I just returned from my trip. FI light is coming on all the time now. What ever causes it has apparently gotten worse. I think this will make it easier to diagnose the problem.
 
Back
Top Bottom