Honda CBR XX Forum banner
1 - 20 of 32 Posts

·
Registered
2000 titanium
Joined
·
20 Posts
Discussion Starter · #1 ·
I recently fell victim to the F1 light code 25 ECU fail noted in many threads on this site. I have done the loom and checked all the sensors hoping in vain that I might be the lucky one. Nope. So, I did a whole bunch of research and hunting salvage yards as the ECU is no longer made by Honda (38770-MAT-A02) us 49 states on a 2000 titanium bird. 32k on the odometer.

After trying CARMO (said it could not be fixed) and the other ECU doctor options with no success and hoping for a miracle before spending a bunch on an aftermarket stand along unit. I took a chance on a small company in New York state. G7 computers (on google and other search platforms) I spoke with the Tech directly and he seemed to think it was not a big deal and the ECU did not need replaced or rebuilt. He used the term tweaked.

So, I sent it off to them and paid the fee (225 dollars us). Turn around at their shop once received was 24 hours as advertised and I got the ECU back in 8 days (ground shipping).

Plugged it back in...no codes, no flashing...Fired right up... no codes ... Test ride .... No codes - running great.

I will see if it holds as I have about a 300-mile trip planned this week. I'll post the report after the long ride.

Let me be clear.

I am not advertising for them, and I am not affiliated with them in any way.
 

·
Registered
2000 titanium
Joined
·
20 Posts
Discussion Starter · #3 ·
Well, I can confirm that pipe dreams still happen. Failure after 50 miles. Same code same issue. So, no new fix here. (Most of you already suspected that.) The tech has agreed to dive deeper and take another look. In the meantime, I'm off to search the internet for a miracle.

Sorry if anyone got their hopes up.
 

·
Registered
Joined
·
1,266 Posts
Well, I can confirm that pipe dreams still happen. Failure after 50 miles. Same code same issue. So, no new fix here. (Most of you already suspected that.) The tech has agreed to dive deeper and take another look. In the meantime, I'm off to search the internet for a miracle.

Sorry if anyone got their hopes up.
Does the bike run ok, full power etc, with the code 25?
 

·
Administrator
Joined
·
9,634 Posts
Well, I can confirm that pipe dreams still happen. Failure after 50 miles. Same code same issue. So, no new fix here. (Most of you already suspected that.) The tech has agreed to dive deeper and take another look. In the meantime, I'm off to search the internet for a miracle.

Sorry if anyone got their hopes up.
Sorry to hear that outcome.

You can run an L02 as well as an A21. With the A21 you'll have to change out the FI light to an LED. It's what I had to do to run a Dynojet PCV in my 2K (That's a whole nother story)
 

·
Registered
2000 titanium
Joined
·
20 Posts
Discussion Starter · #6 ·
Does the bike run ok, full power etc, with the code 25?
Fizzy, in answer to your question no. Sluggish above 3000 and a notable rich condition. If i flash the codes, clear and run bike it runs great for about a mile then Light and a noticeable reduction in power. (Always at about 3000 RPM)

The tech at G7 is going to dig deeper // thinks it's a latent issue in the code under certain load then does not appear under normal testing // said he had some ides so I'm sending back to him for another go at it.
 

·
Registered
Joined
·
1,266 Posts
I went through this code 25 nonsense. My ultimate solution was another ECU. In my case, my bike retained full power, just had the FI light on.

My understanding of how it works from memory: Upon startup the ECU will look for issues with the knock sensor between 3000-5000rpm (or thereabouts). Engine needs to go up and down thru this range at least 1x before code 25 can be triggered. Upon triggering, ECU sends engine into get home mode, which is what you experience. In my case, she never went into get home mode.

Recommend you revisit the loom fix, or if not done then do it. If any wires have corrosion then strip back until 100% clean and extend with new. Also go over all connectors and clean + inspect all ECU wires at the 2 big multi pins for hidden damage. Check the knock sensor wire sheath for ground.

Test knock sensor. I performed a rudimentary test with a hammer and drift and a multimeter.
 

·
Registered
2000 titanium
Joined
·
20 Posts
Discussion Starter · #8 ·
Fizzy, I have already done the loom revisit- solid. Did checks on wires and continuity - replaced knock sensor just because i enjoy spending money I don't need to. (Wishful thinking) Cleaned everything. It's the ECU and that's that. Can't just go replace it (discontinued by Honda) so I have been calling salvage yards and picking the brains of Circuit board repair geeks in hopes of a better solution. Wife told me to just buy another bike. (she's the best). Still have some options which may prove fruitful so I'm still at trying to find someone to repair it. Time is something I have right now. I'll keep up the thread as new attempts are made. Eventually all of us are going to need a solid repair option. Might as well find it now.

Wish me luck.

Oh, wife did tell me, after explaining the situation in detail, "you have a better chance finding a unicorn taking a shower in the bathroom." At least she understands.
 

·
Registered
2000 titanium
Joined
·
20 Posts
Discussion Starter · #11 ·
so far it's still at G7 he is working into the circuit board and trying to isolate the issue. His latest thoughts are alone the line of too much noise from EMI or RF getting into the circuit. Going to be doing a Feritte core choke test this weekend on the knock sensor shield wire. Will keep posting as new developments occur.
 

·
Registered
2000 titanium
Joined
·
20 Posts
Discussion Starter · #12 ·
Ok... after doing research and spending hours going over the problem and reading other riders descriptions of when the failure occurs and why I have come to some new conclusions that maybe all of us should consider. So, I'm putting this out there for comment and retrospect.

Failures the first time are sudden with common points being // Hot temperatures // in traffic // fan coming on and loading down system - and FI light comes on . these seem consistent in most accounts of the ECU failures

ECU failure throwing knock sensor code is due to a moffett failure from excessive RFI interference transmitted thru the knock sensor wire. Source of the RFI is ... The aging fan motor brushes. They can transmit enough RFI interference to overload the circuit and cause an overload explosion see link


"Noise suppression for an electric motor"

this link also shows how to make a simple noise reduction filter on the fan motor wires to eliminate the problem before the ECU goes belly up.

Lets hear what ya think.
 

·
Administrator
Joined
·
9,634 Posts
What if you put and EMI shield over the wire and grounded one end to the case similar to a coax cable?
 

·
Registered
2000 titanium
Joined
·
20 Posts
Discussion Starter · #14 ·
Interesting possibility. That should make a difference. Anything to reduce the RFI from the electric motor from being transmitted to the knock sensor wire. I am also looking at a ferrite choke bead at the ECU end of the knock sensor wire to reduce the signal just before it enters the ECU. I have the luxury of an oscilloscope and looked at the noise on the shield and WOW. Bounces all over the place- saw over 8v spikes. Kinda like blowing out an ear drum that's designed to listen for noise. No wonder the circuit blows up. This is all speculative, but the more testing i do the more sound the evidence is. Poor grounds are a big factor which is way cleaning and running new grounds works in the short term. trying to find a more bullet proof answer to protect the ECU for longevity. Any other electricians want to weigh in?
 

·
Registered
Joined
·
1,408 Posts
From the schematic in the manual the knock sensor wire is shielded already. Maybe the issue is that the shield ground is corroding over time and becoming ineffective?
 

·
Registered
Joined
·
1,266 Posts
From the schematic in the manual the knock sensor wire is shielded already. Maybe the issue is that the shield ground is corroding over time and becoming ineffective?
Easy to check.. The knock sensor wire shield is visible at the ECU end. You may have to remove a little wire wrap. It is grounded at a connector under the tank.
 

·
Registered
Joined
·
8 Posts
So... read this over my cup of coffee and thought about it all day....

For history... Bought my bird in July of last year... it sat for a month, rode it a month or two fine, then code 25 popped up... more or less ignored it for a month or so and winter hit, light goes out... rode all winter without a peep out of it. Spring/summer started rolling in and a month ago I took a 2 hour each way out of town trip and on the return the light kicked on half way back. Since then its been... relatively intermittent but always comes on when the bike gets hot.

After reading your idea, on my commute this morning I intentionally kept the bird as cool as possible, moving quick and changing my route to miss lights... Temps never really crossed 195-200F and no light for my half hour commute. On the ride home it was significantly hotter outside but temps stayed 185-195 through the whole commute without a light. At the last stretch to my driveway temps creeped up and the second it hit 210F (roughly where my fan comes on) the light decides to come on. So I think you're on to something about the fan noise.

Since I was bored and had a brand new box of them laying around, I installed a pair of ferrite chokes on the wires right at the fan, and a single one at the ECM for the knock sensor wire. I also verified with my meter that the shielding on the knock sensor is grounded. Hopefully my ECM isn't too far gone, but I'm curious to see if it will make a difference.
 

·
Registered
2000 titanium
Joined
·
20 Posts
Discussion Starter · #19 ·
SGT - please post on your findings on a hot ride - that is the same idea I am trying this weekend on mine. I am very interested since you have the intermittent light to see if it has a positive effect. I expect it will eliminate the light at high temp all together.

As for the knock wire being shielded, This is true for all harnesses. The ground can be weak and faulty which would add to the noise getting in but the excess noise source is outside of the knock circuit. The cooling fan is a basic electrical brush motor that is exposed to weather, rode grime, and time. Shielding the source as SGT described is best option to eliminate the noise from getting out to the knock circuit.

Fizzy- good thought on the fuel pump and I see what your saying- However the fuel pump is shielded somewhat in its housing and constant on which would lead to an F1 light that would be on all the time with the engine running. I'll check the leads for noise. The common thread I noticed was the light comes on when the bike gets hot so I started with the fan first to find the source. Keep the ideas coming- I'll test out the fuel pump and relay for noise tomorrow.

I would like to hear from anyone who added a fan on / off switch to manually engage the cooling fan. Any ECU issues associated with switching the fan on and off or in constant run?
 

·
Registered
2000 titanium
Joined
·
20 Posts
Discussion Starter · #20 ·
Leighb - In answer to your question you are absolutely correct. a full ground check of all the wires check for proper grounds is step one. Once your sure they are solid. then look to the fan motor. (if this tests true) The noise (RFI Radio frequency interference) was getting through the shielding even with solid grounds - RFI increases dramatically (on the oscilloscope) when the cooling fan circuit is energized which lead me to where we are now.- Good thought and I fully agree the knock sensor shield ground is a vital part of the whole issue. If there is no ground to the shield at the engine block the system is wide open to all RFI sources.
 
1 - 20 of 32 Posts
Top