FI Issues (Code 25)

tsperez

Joined
May 6, 2008
Messages
17
Location
TX
I have a problem that I need some help in addressing. A few months ago I had a weak battery and at about the same time I got a blinking FI light. The code is 25. I've had the bike at the dealer twice. First time they did nothing. Second time they replaced the left knock sensor. The bike ran for about 200 miles before getting the blinking code again. I did replace the battery even before I took the bike to the dealer.

I can take it back to the dealer but they seem very weak in addressing this problem. The first time they even told me that it was not a hard error and to just ride it the way it is! I refused to do that. Btw, I have the extended warranty on the bike.

I am posting here to see if someone has any ideas for a fix. Some details on my problem:
The blinking always stops if I shut off the bike.
Light comes on in 3 to 80 miles after I start riding.
I must go faster than about 75 mph for light to come on.
I rode about 100 miles twice at speeds below 70 mph and light never came on.
I *think* there is a hiccup when the light comes on.
My mileage seems to drop when the light is on.
I only use 92 or 93 octane fuel.

Any ideas will be greatly appreciated. I just bought about $1000 worth of stuff for the bike so I was planning on keeping her for a while (2003 ABS with 36,000 miles). It has been my bike since new.

Thanks in advance,
tsp
 
Sounds like a "soft code". When the bike gets restarted the system checks out ok but something goes funny while riding. That's why there's no codes stored in the system. Have you made any other mods to the bike, had something serviced recently, or were "in" the bike for any reason (even if you were just looking)? I have the manual at work and can look up the code tomorrow.
 
I know what the code is. Deals with the left knock sensor and I forgot to mention that the dealer replaced the sensor when I took it in for this problem about two weeks ago. Also the dealer said it was a soft code like you say.

I've got a lot of mods on the bike but all have been on the bike for 2 to 3 years. A partial list... Corbin electric seat, parking lights under the wings, map lights on bars, loud dual horns, XM radio, GPS, accessory harness for some of the above, BMW outlets, regular cigarette lighter outlet, headlight modulator, and a power distribution block. But all these items have been on the bike for years. I was hoping that the dealer wouldn't tell me to remove them all!

tsp
 
So, what to do when you get a code 25 left knock sensor failure and changing the knock sensor doesn't fix the problem? The obvious answer is to go after the wiring and connections. Even the manual says to check for loose contacts at the knock sensor, and check for grounds and continuity from the sensor connector to the ECM.

I would run the engine at 3900 rpm (per the Manual) and after 10 seconds (still at 3900rpm) I would wiggle the connector at the sensor and see if you can induce the problem. Also try wiggling the connector at the ECM and shaking the wire bundle. Unless wiring has been chaffing it's usually not the problem. I would be more suspect of a bad connection. If you don't have a manual I believe you can go to the Honda Owners website and download it. Not sure though as I always purchase the Manuals for ever vehicle I own.
 
There is apparently a problem in the circuit for the "Left Knock Sensor", as you're already well aware. Either the sensor itself, or the circuit/wire that connects it to the ECM.

Reading the service manual, leads me to believe that the ECM will only throw a code 25 if the engine RPM is held above 3900 RPM for more than 10 consecutive seconds, and it detects a problem (either an open, or short) in the circuit connecting the knock sensor to the ECM.... which might explain why you only see the code after riding above a certain speed sometimes.

Do you have a service manual? Do you feel comfortable doing a few continuity tests? Have a look at the diagnostic procedure on P. 5-45, continuing onto P. 5-46.

Note: The service manual (mine at least) is WRONG. The wire for the left knock sensor is red/blue, not blue. And the wire for the right knock sensor is blue, not red/blue. Either that, or my bike is wired backwards :confused:

Also Note: The test #3, "Left Knock Sensor Open Circuit Inspection", on P. 5-46, is also WRONG. :( There should be continuity. If there is no continuity, then there is an open circuit somewhere on the red/blue wire, between the sensor and the ECM.

Hint: The way that the wire is routed may have caused a rub/chafe. Follow the wire from where it attaches to the knock sensor, all the way up into the wire harness. There are places where it may have developed a rub, thus causing a short to ground. At least it looks like a good possibility on my bike.

If the diagnostic test procedure (noted above) doesn't uncover the problem, then I'd lean towards a bad knock sensor (not probable, since you said the dealer already replaced it)... Or worse, a problem in the ECM (dooopppp, did I say that... knaaaawww). But I'd be seriously suspicious of a problem in the wire and/or connecting points at each end of the wire.


Edit: What Phil said, too :D
 
Last edited:
Just having a closer look at the wiring diagram... it appears the wires for the knock sensors are routed thru a large, gray, 10-conductor, plug, on their way from the sensor to the ECM. There's a possibility that the connector pins in this plug could be contaminated too, thus causing an intermittent connection.

The location for this plug is nicely depicted on P. 1-35 of the service manual. It's referred to as the "ENGINE SUB-HARNESS 10P (GRAY) CONNECTOR" on P. 1-35. Oriented near the upper-left section of that drawing.

You should pull this plug apart, and have a look at the connector pins inside. They may be corroded/goobered up. If so, clean em up, and put it back together after injecting some dielectric grease.

Given the plug's physical location, I can see how it could quite easily get corroded/goobered up. I've attached a couple of pics of the gray plug's location (circled in red). Although the pics don't show the actual gray plug itself, it is just inside this opening between the tank, body-work, and frame-rail. In the second picture you should be able to see the wire harness side of actual bundle of colored wires, just above the gray plug itself (which you can't see in the pic).
If you've done any riding in the rain, or sprayed water directly at this spot (whilst giving her a bath), it could certainly have contaminated the connector pins with hard water.

Edit: You should be able to get at the gray plug without having to take the entire right-side mid-fairing off. If you remove the allen head bolt that holds the tip-over wing cover on, the two 8mm bolts under the tip-over wing cover, then remove the allen head bolt just above the engine's valve cover (the one holding the mid-fairing to the side of the upper frame rail)... you should be able to pull the mid-fairing away far enough to pull the gray plug out into day-light (it worked for me). I just added a third pic, clearly showing the gray plug.
 

Attachments

  • Engine-Sub-Harness-Plug-Loc1.jpg
    Engine-Sub-Harness-Plug-Loc1.jpg
    46.2 KB · Views: 152
  • Engine-Sub-Harness-Plug-Loc2.jpg
    Engine-Sub-Harness-Plug-Loc2.jpg
    54.3 KB · Views: 149
  • Engine-Sub-Harness-Plug-Loc3.jpg
    Engine-Sub-Harness-Plug-Loc3.jpg
    55.9 KB · Views: 182
Last edited:
Pat brings up an interesting point. What if Honda has the wiring crossed. Then the right one could be the problem and logging a left failure. The way to determine this for sure is to remove the connector from the left sensor while running the engine (3900 rpm) and the code 25 fault should appear. As Pat and I mentioned look at connections first. Giving the wiring and connectors a good shaking will often allow a bad connection or wire to show up.
 
curmudgeon said:
Pat brings up an interesting point. What if Honda has the wiring crossed. Then the right one could be the problem and logging a left failure.
Click to expand...

Excellent point! I should have thought to stress this. Talk about not seeing the trees thru the forest :bow1: duuuhhhh.

I'm gonna go try it right now... will post my results momentarily.

Edit: Ok, I disconnected my left-side knock sensor. Upon running at 4000 RPM for ten seconds, the FI light comes on, as expected. Upon reading the stored code, the ECM did indeed throw a code 25. So, it looks like Honda got the wiring right, but got the documentation wrong :(
Note: Contrary to what the service manual says, the left side knock sensor (on my bike) has a red/blue wire, not a blue wire.... (as described above). Boy, do the revs take a major dive, when the FI light comes on.
Also of interest - the right side wire is black (not blue).
 
Last edited:
First of all thank you to all that provided help and/or info related to this problem.

I really went for the 10P Gray Connector in the last hour. I was wondering the same thing - is there a connector between the knock sensor and the ECM? The 10P was fairly clean inside and all the contact pins looked very clean. I put dielectric grease on all the pins and plugged and unplugged it about 4 or 5 times. Did the run at >4,000 rpm and the light came back on. I did this about 4 times and every time I got the FI to come on after about 10 seconds. (at least I can duplicate the problem now!)

I have not removed the left panel yet so I cannot check the sensor connector nor the wire but since it was replaced by the dealer I figure it is good.

I guess my next step (later in the week) will be to check the sensor/wire and the ECM connections.

Once again, I appreciate all the help.
Tomas
 
pdfruth said:
Excellent point! I should have
<snip>

Edit: <snip> Boy, do the revs take a major dive, when the FI light comes on.
Also of interest - the right side wire is black (not blue).
Click to expand...


You are right on revs taking a major dive. I locked my throttle between 4,200 and 4,400 rpm and just before the FI light came on the revs dropped to about 2,400 rpm!
I figure what it is doing is retarding the timing but I would never have figured that it would have made that big of a difference!

tsp
 
tsperez said:
....I guess my next step (later in the week) will be to check the sensor/wire and the ECM connections.
Click to expand...

If I were a betting man, my :money1: would be on the connector (at the knock sensor), and/or the wire between the knock sensor and the 10P connector. Since it seems to be reproducible, it should be fairly easy to find, by doing a few simple continuity tests ;)

Can't wait to find out root cause on this one. Haven't seen/heard of this issue on our ST13's yet.

Keep us posted.
 
Ok folks, quick update.
I just took off the left side panel so I could get to the knock sensor. I unplugged it and checked the wire going to the connector on the right side and I have continuity. I then plug the connector and sensor back on and did the >4,000 rpm test. No FI light! I did it three more times for a total of 4 tests and still no light!
Both side panels are off. I am going to move the wires back into place and try it again. And then finally put the panels back on.

tsp
 
You may have had a dirty connector. Pulling it off and reinstalling it could have cleaned it up enough to make a good circuit again. We used to see this happen fairly often on aircraft connectors.
 
About a year later and I finally got this problem resolved! The only thing that was left untouched was the computer (DMU?). Got that replaced and all is well again.

tsp
 
You must log in or register to reply here.
Back
Top Bottom