Code 26- New Sub Harness Install

Clay010

Joined
Jul 9, 2026
Messages
1
Age
39
Location
Florida
Bike
2005 St1300
Morning All,

I recently bought an ST from marketplace and by the time I rode it the next day the code jumped up on my dash. Since then I have ordered a new knock sensor sub harness and installed it. Taking off the old one, the ride connectoin to the knock sensor was so brittle that it basically disintegrated in my hand. Rewired everything and put all the plastics back on and yet..... The code remains.

I really do not want to start stripping down the bike again, especially since its my main mode of transportation. Was hoping someone could tell me a step I missed or point me in the right direction. Thank you!
 
Typically, but not always, it means the ECU is damaged and the only way to clear the code is to get a new ECU.

I personally believe, but there's no proof, that heat degrades the wires and shorts out the ECU. Some have replaced O2 sensors and fixed the issue but that is a rare successful solution.

Many just ride with the code and ignore it as it shouldn't hurt anything, or it has yet to be proven if it does.
 
If the problem is a defective ECM most commonly the code sets after maintaining engine speeds in the vicinity of 4,000 RPM or higher for around 30 seconds or more continuously. There are a couple of simple tests that can be done to gather some data.

Test # 1:
Go through the code clearing procedure regardless of whether the FI light is on or not.
With a fully warmed up engine ride for several minutes while continuously maintaining a sustained RPM close to but not exceeding 4,000 RPM. I would not exceed 3,800 RPM just to allow a little bit of a buffer.
Note whether any codes set and what they are.

Test # 2:
Go through the code clearing procedure again, even if the FI light did not illuminate, to ensure that there are no false test results.
With a fully warmed up engine ride for several minutes while this time maintaining a sustained engine speed of 4,000 RPM or higher for several minutes continuously. I suggest maintaining a minimum of 4,200 RPM to allow a bit of a buffer. The RPM must be maintained above this threshold continuously for a minimum amount of time before the ECM sets the code.
Note whether any codes set and what they are.

Test # 3:
This is not quite as easy to accomplish as the first two tests.
Swap the knock sensors from side-to-side.
Perform test # 1 and test # 2 again
Note whether any codes set and what they are.

If code 26 set in test # 1 there is a chance that the problem is somewhere other than the ECM. The diagnostic procedure will need to be followed to try to determine the cause.

If code 26 set in test # 2, and it did not set in test # 1, the likelihood is that the ECM is defective. There is no known ECM repair for this problem that I am aware of.

A couple of notes:
1- People have reported the code 25 and/or 26 setting at different engine RPM's but it is always around the 4,000 RPM mark. This is why I suggested a minimum of 200 RPM as a buffer.
2- The length of time that the engine speed must be sustained above 4,000 RPM before these codes are triggered can also vary. Maintaining the engine speed above 4,000 RPM continuously for several minutes ensures that sufficient time was allowed for the ECM to react.
3- If the code changes from 26 to 25 there is the possibility that there is a knock sensor problem, assuming that all wiring, connectors, etc., are good. If the code does not change, it remains code 26, then the ECM is again suspect.
 
Same problem here. I agree with the previous solution "Many just ride with the code and ignore it as it shouldn't hurt anything". I believe when the code appears, the ECU falls into "limp mode". Someone suggested "permanently" installing a jumper in the diagnostic plug. This seems to avoid the limp mode, but the FI light stays on. I have been riding my '06 like this for a couple months now and have not noticed any other effects.
 
eyrepb said:
Same problem here. I agree with the previous solution "Many just ride with the code and ignore it as it shouldn't hurt anything". I believe when the code appears, the ECU falls into "limp mode". Someone suggested "permanently" installing a jumper in the diagnostic plug. This seems to avoid the limp mode, but the FI light stays on. I have been riding my '06 like this for a couple months now and have not noticed any other effects
Click to expand...
If the FI light on the panel bothers you, that is fixable with black electrical tape.
 
There was a post and link not long ago. The knock sensor amplifier internal in the ecm fails. It listed a part # and video to change it, not for the faint of heart. I can't find it.
 
Al st1100 said:
There was a post and link not long ago. The knock sensor amplifier internal in the ecm fails. It listed a part # and video to change it, not for the faint of heart. I can't find it.
Click to expand...
For the GL1800 ECM but maybe the below is the video that you were thinking of.
If not it would be the same idea for the ST1300 ECM but I have no idea if the part being replaced is the same or not.

GL1800 ECM Failure
 
Many just ride with the code and ignore it as it shouldn't hurt anything.
Click to expand...
Many have, and will, disagree with the above statement. If all other possibilities have been eliminated and the problem has been diagnosed as definitely being a defective ECM, and a properly functioning replacement ECM can not be sourced, my question to them would be what alternative do they have to offer.

Obviously replacing the ECM with a new or a known good one is the best option. However, if ever I face this dilemma and I can't find a good replacement ECM, I figure that there would be six options available to me.
1- Sell the motorcycle as is to some unsuspecting person to whom the defect is not disclosed.
2- Sell the motorcycle disclosing that it has a defective ECM.
3- Sell it whole as a parts bike.
4- Dismantle it and sell the parts.
5- Stop riding it, park it, and let it rust away in to a heap of scrap.
6- Ride it with the defective ECM.

I would never do #1 so that is not an option and is off of the table. # 2 would be a hard sell and would likely not yield very much money. I know that I wouldn't pay a lot for a known defective motorcycle that I know can not be repaired. # 3 would probably be the easiest solution but would also yield very little money. I can't be bothered with # 4 for the little return that it yields vs the time and effort required. # 5 doesn't achieve much of anything and seems rather pointless to me. So, unless I was desparate for the paltry sum that selling it whole or in parts would yield, that leaves me with # 6 as the best option for my circumstances.

So the way that I look at is that even IF the defective ECM will eventually cause further problems as a result of riding it with this defect- so what. While it may be true that I may have to alter the way that I ride it, e.g. attempting to remain below 4,000 RPM more often, it still remains very rideable and I will still enjoy riding it. At least I will be able to use it and enjoy it for whatever amount of time and miles until something more serious develops. I see this as a far better option than not enjoying it at all and letting it sit to die a slow death from the ravages of rust and time.
 
I believe when the code appears, the ECU falls into "limp mode".
Click to expand...
I have often heard this mentioned and I am not convinced that this is the case when code 25/26 is triggered on an ST1300. I guess it could depend on what one considers limp mode to be.

In the automobile industry limp mode is not entered every time that the check engine light illuminates. It depends on what the failure is. Limp mode is only used when a problem that the ECM has detected can not be reliably overcome by it and that is serious enough to trigger such a severe response. Under those conditions it defaults to a pre-set suite of values that are programmed in to the ECM that it uses to manage the engine's functions. This is not only to allow it to continue to function but also to ensure that it continues to functions until it can be repaired. Many of the dynamic changes that the ECM normally continually makes in response to real-time conditions based on inputs from various sensors cease to a large degree depending on the fault that has been detected. If the problem is serious enough to warrant it they cease all together and the ECM runs the engine on pre-programmed defaults with no dynamic changes. In essence it reverts to operating the way that an older carbureted engine with no computer does. In the case with this code 25/26 issue I don't believe that this is what happens. Honda indicates that it does not go in to any kind of limp mode as they state that the engine continues to operate normally. Below is what I think is happening based only on my past training and experience with automobiles as I have no particular knowledge of what the ST1300 ECM does and Honda famously does not divulge this information.

The knock sensors are part of what allows the ECM to continually dynamically alter the ignition timing advance based on current conditions. The ECM wants to have as much timing advance as possible within its programming limits and without causing excessive detonation, commonly called pinging. Inputs from the knock sensors is a critical piece of information needed to accomplish this. If the ECM receives data from the knock sensors indicating that there is excessive detonation it will do things like retard the ignition timing advance and/or alter the air/fuel ratio in response. It will continue to do so until the detonation either comes back within allowable limits or it stops. When there is a code 25/26 the ECM can no longer monitor for detonation. It can no longer safely dynamically adjust the ignition timing because it can no longer monitor for the occurrence of detonation, so it no longer knows what a safe amount of timing advance is. As a result it defaults to a pre-programmed ignition timing advance curve that it knows is safe and is based on more basic parameters like engine RPM, throttle position, coolant temperature, etc.. This explains why many people report a loss in power and an increase in fuel consumption when the code 25/26 condition is present. Both of these are expected outcomes if the engine is operating with less than optimal ignition timing advance.

Will riding the motorcycle under these conditions cause it harm? Many people have been riding their ST1300's like this for years and for tens of thousands of miles because they can not find a replacement ECM. While it certainly will not do it any good, the engine is not going to suddenly explode because of it. If I was faced with this problem and could not find a good replacement ECM I would keep riding it while more closely monitoring for any possible developing issues until someone provides me with another better option.
 
Sounds accurate Andrew.

My bike definitely feels and sounds different when the fault is active; the throttle response is very 'sharp' compared to normal riding mode. I usually can tell when the fault triggers before looking down at the light to verify.

I do not just let the bike (motor) remain in this condition but instead reset the fault and the performance, sound and feel all return to normal until the next trigger point is reached.

Riding at less than 4k rpm OR monitored time momentarily above 4k rpm is not difficult and keeps the fault from reoccurring. Yes, it is an altered riding style and top (steady) speed limited to around 70mph, but for many, that's still sufficient to enjoy the bike for many miles without sourcing yet another ECM. My $0.02.

I still have the bike but upgraded to a 2010 after trying 3 ECMs (two of them did not exhibit the fault for many months of use before then doing so) and grew tired of not being able to maintain a higher steady speed with healthier bikes on group/long distance rides that I plan on doing more of now that I'm fully retired.

If anyone is looking for an 04 in good shape (and with a McCruise installed) that is fully rideable with this fault, hit me up!
 
You must log in or register to reply here.
Back
Top Bottom