ECU Repair?

Joined
Apr 2, 2017
Messages
55
Location
Newfield, NJ
Wondering if anyone has ever tried to open up their ECU and re-flow the solder joints and traces with an iron.

I have an intermittent 26 code, seems to come back after 300+ miles after a hard reset.

I have a post in the wanted section looking for a known good unit, but it looks like I will have no choice but to buy a new one. Considering the winter is approaching here in the northeast, I am considering pullng my ECU and seeing what is involved in opening it up and going over all of the soldered connections and traces. I know this has been done in the past over the years with other bikes, and I have a lot of experience with a soldering iron.

The problem is, it will take a while to know if I actually fixed anything, for I don't often ride my ST, partly because of life, partly because of my Electra Glide, and as stated above, winter is fast approaching here.

If anyone has attempted to crack one of these units open to repair one, please share your experienxce.
 
I'd wonder if its a broken pc conductor, and not a failed chip or other component. How often do these traces break that they can be fixed by simply flowing solder over them? Could it be a software issue?
 
Wondering if anyone has ever tried to open up their ECU and re-flow the solder joints and traces with an iron.

I have an intermittent 26 code, seems to come back after 300+ miles after a hard reset.

I have a post in the wanted section looking for a known good unit, but it looks like I will have no choice but to buy a new one. Considering the winter is approaching here in the northeast, I am considering pullng my ECU and seeing what is involved in opening it up and going over all of the soldered connections and traces. I know this has been done in the past over the years with other bikes, and I have a lot of experience with a soldering iron.

The problem is, it will take a while to know if I actually fixed anything, for I don't often ride my ST, partly because of life, partly because of my Electra Glide, and as stated above, winter is fast approaching here.

If anyone has attempted to crack one of these units open to repair one, please share your experienxce.

Older posts on this site and the European site and the Australian site all show major ECU repair shops declining to repair this one.

Some ST-Owners have also recently reached out to Gen-Tune about it but he seems to have given up as well.

 
I hope that one day the fair repair Act will make possible to get access and repair these units.
The lack of information about the software and hardware is somehow not acceptable from the manufacturer especially when you consider the number of cases involved.
What about repairing something existing instead of having to replace it just because its just voluntary locked by design.
Car makers are starting to build cars where the parts are locked together by an encrypted serial code on a chip, no interchange possible without the manufacturer software.
Apple, Tesla, John Deer and a lot of other manufacturers are going this way.
What a waste and so much frustration!
 
The main problem is the unit is completely sealed in the manufacturing process. The ECU was not intended to be a serviceable part.
 
Car makers are starting to build cars where the parts are locked together by an encrypted serial code on a chip, no interchange possible without the manufacturer software.
Apple, Tesla, John Deer and a lot of other manufacturers are going this way.

The writing is on the wall for independent auto repair businesses. Unless consumer protection shows it's head, expect to pay handsomely for repairs through the manufacturer. Guess who has more lobby influence. Because of this trend, my "pre-connected" autos are becoming more valuable to me.
 
I'd wonder if its a broken pc conductor, and not a failed chip or other component. How often do these traces break that they can be fixed by simply flowing solder over them? Could it be a software issue?

If you think about it, the conditions aboard a motorcycle are far from ideal for a pc board. Between the range of temperatures, vibration, and the jolts, poor or cold soldered connections could start to fail over time, and cause intermittent problems like I am experiencing with this 26 code. I worked for RCA/GE in the late 80s - early 90s manufacturing PC boards for the navy. I would love to see if re-flowing as many connections as possible would rectify this issue, but as aniwack mentioned above, getting the unit open without damaging it could present a challenge. Avoiding heat sensitive parts could also be a problem.

I understand why indies aren't interesting in fixing these things. Unless they can pinpoint the issue, it would be impossible to guarantee the repair.

Because a few hundred miles pass before my light comes back on, that makes it very unlikely the sensor has failed, or charred wires are causing my issue. Because I have already read many of the threads on this forum concerning the 26 code, and ECUs, I am going to replace my ECU. If I end up buying a new unit, I will try to open the one I am replacing, and I will go from there.

The one other thing that makes me think this could be the result of a failing solder connection or a brittle trace is because when my light and code returned last week ( after 343 mles) I had just gotten on her pretty hard and took her well over 100mph. Once I slowed back down to cruising speed I looked down and saw the light. That makes me believe vibration could be a factor.

Of course, I certainly didn't get to where I am in life by being right all the time.
 
The writing is on the wall for independent auto repair businesses. Unless consumer protection shows it's head, expect to pay handsomely for repairs through the manufacturer. Guess who has more lobby influence. Because of this trend, my "pre-connected" autos are becoming more valuable to me.
I purposely bought a 2005 Subaru Outback a few years ago. The reason at the time was that I knew it would sit in the driveway 90% of the time, and it has. But like you, I appreciate the minimalist amount of electronic sophistication it has.

Chris
 
I weighed in big on this bike, I paid a lot for a bike that is twelve years old and discontinued. Replacement parts are expensive and well, I dunno.

I've learned a lot from the contributors on this site about the bike, and learned that it's overall more durable than I had imagined, notwithstanding if I could get my money back, well, it's a bit like that scene from No Country for Old Men, "If you knew that everything you did brought you to this moment... would you still do it?"

It's a good question, and hard to answer; whether these microprocessors are set to eventually fail after so many cycles; maybe ultimately run some BIOS line where the machine cripples itself, not entirely, but just enough to produce the desired objective of planned obsolescence, get you back out there in the market, but not so pissed off that you don't maintain your brand loyalty, I don't know.

I kind of doubt it with this run [02 - 14] of this particular Honda product; these machines come out during what I'm going to call the waning years of the golden age of motorcycles; Honda really wants this to be the comparative reliable, durable standard of utility; they're going up against BMW, and after the unfortunate weave problems that occur they resolve to address those engineering shortcomings and substantially maintain the platform for what, another ten years?

Every Ignition Module that I know of eventually fails; that Honda didn't make it a discreet component separate from an ECU that looks after everything else is too bad, but at the same time, there must not have been enough in warranty failures, and that's when guys were winding the odometers.

If the last few years of production, Honda by then knows this thing is maybe too durable.
 
It seems to me that someone with the necessary expertise could develop a jig to test the ECU and perhaps locate or pinpoint the problem rather than trying to tear one apart to see if they can find a cold solder joint would be more appropriate. Honda still produces similar ecu's for other bikes, so it is understandable why they are tight-lipped about releasing info about them.
 
solder over a joint usually repairs a driver circuit that has a burnt foil. The ecm only reads the knock sensor and it's very low voltage so I don't think that the problem
 
It seems to me that someone with the necessary expertise could develop a jig to test the ECU and perhaps locate or pinpoint the problem rather than trying to tear one apart to see if they can find a cold solder joint would be more appropriate. Honda still produces similar ecu's for other bikes, so it is understandable why they are tight-lipped about releasing info about them.

I agree.......but

Unless this ECU is used on different models, nobody is going to make the up-front investment in R&D for a motorcycle that was discontinued several years ago.
 
solder over a joint usually repairs a driver circuit that has a burnt foil. The ecm only reads the knock sensor and it's very low voltage so I don't think that the problem

Think about the nature of intermittent issues similar to what I am experiencing. The level of voltage is irrelevant if it is interrupted.

As for the first sentence, never mind.
 
One more point folks.

I am surprised at the replies I have received in this thread. I have been hanging around motorcycle forums for about 25 years (BMWMOA, Concours.org, HDforums.com, KLR650...etc)

I paid $3500 for my bike in 4/2016 with 26K, and I still, as recently as 3 weeks ago at a memorial/charity ride have people ask me if my ST is new. The bike owes me nothing, and I am in now way complaining I have to replace a ECU in a 18+ year old bike. The only thing I will complain about is the $813 cost for a new ECM.

That's where these forums have served me very well over the years. I have saved a boat load of money, and time because of the folks on these forums.

I am surprised at the replies I have received telling me why this won't work. I'm not stupid, I already know this is a low percentage play. But, what f I am right? I happen to know of two motorcycles that were repaired this way (one by yours truly), and a buddy and I repaired the ECU in his Yamaha wave runner this way. Maybe I can save the next guy $800 who comes to this forum to look for a solution for his intermittent 26 code.

My only decision now is whether I am going to pull my ECU and crack it open, or spend $800 for a replacement, then crack mine open. It's already a paper weight; why not, right? I guess the third option would be to replace mine, throw the bad one in the trash bin and go riding.

The original purpose of this post was to hopefully find someone who has attempted to open one of these.

I would like to add one additional question. I know some folks have come across burned/charred knock sensor wires; but has anyone actually had a bad sensor?
 
If it were me, I'd de-pot it and take a look. Like you say, the only risk is wasting a few hours :) In fact, if it were me, I'd probably pull it apart before buying a new one haha (it'll still run the bike in an emergency without the potting, although it'd be fragile)

De-potting might be tricky (chemical or thermal options...) finding the cause of the 25/26 might be straight forward (burnt component/bad solder joint?) or rather more difficult (out of spec component, failing memory, cracked trace?) but would probably be easier to trace with a bare ECU, running bike and a 'scope (compare signals from left to right, tracing from input 'upwards').

If you do open yours up, hi-res photos would be really appreciated :)

re: knock sensors: They're just piezo modules. Incredibly robust, although you could probably break one if you tried hard enough. I can't imagine many possible failure modes while fitted to the bike. Heat won't hurt them, corrosion possibly (but they've usually got a brass backing), mechanical impact my do it, but would probably also result in damage to the block...
 
Earlier this summer, there was a thread regarding another failed ECU. At that time one of the posters mentioned that used police ECU's will work in civilian bikes (a direct plug and play) and they can be found on ebay for around $100. I'd guess word might have leaked out and the price might have climbed but you might check around for a used unit. You suggested your bike is an '03, any used pre '08 ECU, police or otherwise will work for you.

I just looked: Note, I did not even read the description, for all I know these too, are paper weights.

 
Last edited:
I am surprised at the replies I have received telling me why this won't work. I'm not stupid, I already know this is a low percentage play. But, what f I am right? I happen to know of two motorcycles that were repaired this way (one by yours truly), and a buddy and I repaired the ECU in his Yamaha wave runner this way. Maybe I can save the next guy $800 who comes to this forum to look for a solution for his intermittent 26 code.

My only decision now is whether I am going to pull my ECU and crack it open, or spend $800 for a replacement, then crack mine open. It's already a paper weight; why not, right? I guess the third option would be to replace mine, throw the bad one in the trash bin and go riding.
I have repaired all kinds of things by simply re-flowing solder joints. Sometimes I was not even able to find any solder joints that visibly looked bad, so I took a chance and just re-flowed every solder joint that I could find and guess what, the problem was fixed.

I haven't opened up any ECM's but, if I were you, I would certainly crack it open and have a look. If you are not going to send it somewhere to attempt a repair, and if you have already decided that you are going to replace it, then it isn't worth anything to you the way it is anyway, so I say go for it.

Make sure to let us know what happens, I'd like to see the insides of one.
 
I have repaired all kinds of things by simply re-flowing solder joints. Sometimes I was not even able to find any solder joints that visibly looked bad, so I took a chance and just re-flowed every solder joint that I could find and guess what, the problem was fixed.

I've never attempted that but I'd also be impressed if you were to discover the weak link associated with the black box, why you could be famous! Nothing ventured nothing gained.

I've condemned many boards in stationary equipment, just on the logic in my head based on Inputs and Outputs but never really being sure I did the right thing until whatever it is has worked afterward. I know it's apples and oranges

It seems that just from reading submissions that in some instances the faults occur at a specific load / rpm and in others they might run for some duration and reoccur. I wonder if it could be temperature related and since it makes such a poor door stop if one couldn't open or ventilate the enclosure somehow, get some air across it and see if that changes anything.

On another topic, as far as R & D investment feasibility, I'd try what Honda recommends for their 92 Accord when you get to the last page in their telephone - book sized manual: "Substitute known good one"

Swap it into your buddy's bike and see if you can export the fault, just tell his wife you missed his birthday and you thought you'd surprise him by doing something nice and taking his bike and giving it an oil change and a good polish; then when he sees it on Saturday, he'll be so happy he can take her out for lunch instead of working on his bike, it's a win win!

And if in the meanwhile, it turns out that it is the ECU? Well that could present you with a moral conundrum. I mean, you know how hard they are to clean.

I guess the only two reservations I'd have with that idea might be whether or not I'm running the risk of creating a greater risk of component failure which could result in a fault while you're riding, and if the repaired ECU might cause a problem with the test bike.
 
Back
Top Bottom