ECM Hope?

Mellow

Mellow

Joe
Admin-dude
Joined
Dec 1, 2004
Messages
13,301
Age
61
Bike
'21 RT / '24 GSA
Sounds like a possibility on our bikes also. I hope this fix works.

Aniwack did you figure out which wire it is on our bikes already?
 
Wow!!! I just watched the whole 45 minute video!!! Promising!!

I hope someone here has knowledge like him to determine if there is a magic wire on ours as well!!
 
WØQNX said:
Sounds like a possibility on our bikes also. I hope this fix works.
Click to expand...


I may be misunderstanding what Dark Sun has done, but I am not sure it is a fix for a ST which would already have the code on.

He calls it "Zombie" fix because it could save an ECU about to die of 25/26 pandemic. And could be good prevention for an ECU that isn't
fried yet.

But main thing he does if the ECU is permanently damaged is disconnect the sensor wires from the ECU so it is not throwing the codes anymore (after reset), just like a poster once reported on ST-Owners somebody was doing for the Aqua Traxes with the same issue.

He is perfectly fine with running his Wing without knock detection, he says. Many here are not. But some are....

Looks like the WIngers have it worse than us, because their limp mode is not drivable, so they have to do the Dark Sun dance if they want to keep going (without knock detection) with a "bad" ECU. The ST only takes a much more manageable hit with the codes on and some are fine with keeping going like that, without doing anything.

If we can implement something similar for the ST, beside preventing the MIL to go off, it may also prevent to go into the mpg robbing pseudo limp mode 25/26 are throwing the bike in (but still w/o knock detection).
 
Well, Honda doesn't try to reinvent the wheel on every product... so, similar tech and thinking was probably used during the 2001 to ???? time frame when it came to putting ECMs together...

So hopefully, we can take advantage of this reusable thinking.....
 
Hi Midlife…not trying to counter you at all, but did you watch the whole video….on the Wing anyway, he shows you how to continue to have knock sensor coding using an isolated ground soldered into I think it was pin 9 attached to a lug on the engine in addition to cutting the logic wire. The EASY fix was cutting the logic wire which would simply bypass the knock sensing circuit…..at least that is how I understood it. I fall in the camp of not being super concerned with electronic knock sensing. I run 93 octane and I don’t ask a lot from the engine in my everyday riding. None of my other motorcycles have had it. But, if the ST ECU/M issue is a similar issue to the same year Wings, based on his video, you can still have knock sensing; it’s just a harder fix.

Again, not arguing, just having a conversation based on my understanding of his 45+ minute video.

Ray
 
Cutting the logic wire to bypass the knock sensing circuit, would be no different than putting a piece of electrical tape over the light on the dash.
The bike's ECU would still be confused, and not running correctly, and could roam between being overly rich, to dangerously lean etc.
:WCP1:
 
rdmnt2002 said:
Hi Midlife…not trying to counter you at all, but did you watch the whole video….on the Wing anyway, he shows you how to continue to have knock sensor coding using an isolated ground soldered into I think it was pin 9 attached to a lug on the engine in addition to cutting the logic wire. The EASY fix was cutting the logic wire which would simply bypass the knock sensing circuit…..at least that is how I understood it. I fall in the camp of not being super concerned with electronic knock sensing. I run 93 octane and I don’t ask a lot from the engine in my everyday riding. None of my other motorcycles have had it. But, if the ST ECU/M issue is a similar issue to the same year Wings, based on his video, you can still have knock sensing; it’s just a harder fix.

Again, not arguing, just having a conversation based on my understanding of his 45+ minute video.

Ray
Click to expand...


Absolutely fair. I did preface with "I might be misunderstanding", I am not an electron guy at all and Dark Sun is not exactly easy to follow as he seems to discover new things as he goes.

But he also spends time emphasizing that in his view, running without knock sensing is not a sudden death event. Hence my thinking that the rewiring on a ECU which 25/26 circuitry is already completely fried will mainly disable the MIL (but in the case of the Wing, apparently also keep it out of its very limp limping mode).

He makes comments like:

- Many bikes (even some Honda) do no knock sensing at all (Meaning, don't worry too much about it).
- He doesn't really think it is essential for himself, he is in the camp of "You run them til the pistons come out of the exhaust" or something like that.
- And he talks about his 5 mph uphill roll on test in fifth gear more than once, which is best for inducing pinging, and I take it as his attempt to show that the bike can survive some pinging.

But what I am mainly getting out of his thorough run through, is that he seems to think that the cause of the decay of the 5V Knock Sensing Circuitry is that it is sharing the same ground as some of the 12V stuff and inducing "noise" damaging to the Knock Sensing stuff over time. And he proceeds to show a ground location that would be more appropriate for the survival of the Knock Sensing.

I'd be interested to see how this translates to the 1300. Find a new grounding location for our 1300 equivalent of the Wing #9 wire. That might help extend the life of the early 1300 ECUs (with the "design flaw") which still have Knock Sensing in working condition (no MIL). And hopefullyhelp mine last a little longer.

But it may require Dark Sun taking a look at a 1300, as it seems there is a lot of thought going into what makes an appropriate grounding location or not.

Of course I'd love to think Dark Sun is resuscitating dead Knock Sensing circuitry.

But he managed to pinpoint which component on the circuit board does the talking to the Sensor and interfaces with the CPU (the little sausage looking one alongside the CPU) and gets bad over time (due to the "noise"). And did say that it is no longer available and could trace no manufacturer of documentation on it. Which explains why none of the most renowned ECU repair shops have ever been able to fix it: no replacement for this non serviceable dead component.

And I have a difficult time imagining how solely rerouting wires could replace the logic and function of this failed component.
 
Igofar said:
Cutting the logic wire to bypass the knock sensing circuit, would be no different than putting a piece of electrical tape over the light on the dash.
The bike's ECU would still be confused, and not running correctly, and could roam between being overly rich, to dangerously lean etc.
Click to expand...
While I don't disagree with your first statement at all, I have to wonder if the second needs to be the result.
I suspect that loosing knock sensor capability for any reason, intentional or not, causes the ECM to abandon its dynamic ignition timing advance due to a missing piece of information that it needs to calculate the timing advance dynamically. This means that the timing advance is now fixed based on RPM and throttle position just as it was before the days of knock sensors. This is most likely the same timing advance that is used as the baseline and is the starting point for the dynamic timing advance anyway. This will cause an increase in fuel consumption because less timing advance results in less power being produced from the same quantity of fuel consumed, but as long as the O2 sensor system remains functional they should keep the air/fuel ratio stable where it was programmed to be.

To bad that we can't connect a scanner to these ECM's, would be able to see the answers to these questions.
 
Igofar said:
The bike's ECU would still be confused, and not running correctly, and could roam between being overly rich, to dangerously lean etc.

1703962572838.png
Click to expand...


For the strangest reason, that seems to have happened only to bikes coming your way. Never happened to the few 25/26 STs my old timer mechanic is tending to and have been running like that for quite a while. He's been wrenching on Hondas for over 50 years, has many friends at Honda and when he inquired with them back then when 25/26 started raising his sinister head, they flat told him "Keep on riding, we ain't doing nothing about it". No mention of the above doomsday scenario!

I know very little about electrons, even less than you if you think it is even possible. I think current going one way while electrons are definitely going the other way is akeen to voodoo and too scary for me.

But still, what Mr. Shadow is most diplomatically offering in a prior post makes a lot of sense, even to me. I encourage your trying to digest it.

Mr. Shadow said:
While I don't disagree with your first statement at all, I have to wonder if the second needs to be the result.
Click to expand...
 
Andrew Shadow said:
While I don't disagree with your first statement at all, I have to wonder if the second needs to be the result.
I suspect that loosing knock sensor capability for any reason, intentional or not, causes the ECM to abandon its dynamic ignition timing advance due to a missing piece of information that it needs to calculate the timing advance dynamically. This means that the timing advance is now fixed based on RPM and throttle position just as it was before the days of knock sensors. This is most likely the same timing advance that is used as the baseline and is the starting point for the dynamic timing advance anyway. This will cause an increase in fuel consumption because less timing advance results in less power being produced from the same quantity of fuel consumed, but as long as the O2 sensor system remains functional they should keep the air/fuel ratio stable where it was programmed to be.

To bad that we can't connect a scanner to these ECM's, would be able to see the answers to these questions.
Click to expand...
Yeah, I would think that algorithm would be to pull timing and not make any adjustments to fuel mixtures at all. Unfortunately, ECU doesn’t speak OBD-2, it uses older analogue K-line serial protocol that pre-dates even OBD-1. Honda has dedicated pad/laptop diag device to scan these ECUs.

I did figure out Bosch Motronic 3.1’s process from stepping through the assembly code. Knock-sensing was only used on Porsche’s Turbo models as the ignition-retard module was part of add-on boost-controller (non-turbo models didn’t use nock-sensor):

1. IF knock detected on this crank-revolution
2. THEN retard ignition 3 degrees
3. go back to #1 for next revolution, up to maximum 38 degrees retard

No fuel-adjustments are done when knock was detected. Although if you sniff exhaust, AFR would appear richer when ignition was retarded. That’s not because ECU increased fuel (which it didn’t), but due to incomplete combustion from retarded ignition timing. Mixture was ignited later and hadn’t completed burning by time exhaust valve opened. This allowed more unburned hydrocarbons into exhaust and would be picked up by wideband as richer mixture even though ECU injected same amount of petrol as before. Different amounts of ignition retard would change amount of unburnt HC going out exhaust, giving impression of varying AFR. Although O2-sensor feedback would pull these errant mixtures into line soon enough.

This incomplete combustion also has effect of lowering BMEP and moving engine away from detonation & knocking.

I use Woolich software to re-program my race CBR600RR. Using 87-oct pump petrol, I can add 6-8 degrees additional ignition advance before getting hint of detonation under WOT (partial-throttle always Ok). With 110-oct E85, I can add 10-12 degrees more advance and get to diminishing returns with no detonation. More advance can’t make it knock, just no further improvements. So I would surmise it’s safe to run without active knock-sensors on ST1300 since they’re lower compression with lower redlines and aren’t pushing limits as much.
 
Last edited:
If anyone has pre-‘08 ECU with 25/26 code, please send to me. I’ll sniff exhaust with wideband when it goes into limp-home mode above 4000rpms to see how much AFR varies. Will also test the GL1800 solution as well as some ideas of my own.

For the effort involved in reviving these ECUs, I may just gut case and install Microsquirt motherboard and end up with programmable EFI configuration. That’s what I did when racing NASA 944-Spec decades ago. Rules only stipulated “original DME box must be used”, but didn‘t limit what may or may not be inside that box… heh, heh…Yeah, it’s a Porsche MoTeC ECU, yeah, yeah, that’s it! ;)
 
Last edited:
Igofar said:
Cutting the logic wire to bypass the knock sensing circuit, would be no different than putting a piece of electrical tape over the light on the dash.
The bike's ECU would still be confused, and not running correctly, and could roam between being overly rich, to dangerously lean etc.
:WCP1:
Click to expand...
I'm confused and my mouth is covered with duct tape.
Basically yes.
 
MidLife said:
For the strangest reason, that seems to have happened only to bikes coming your way. Never happened to the few 25/26 STs my old timer mechanic is tending to and have been running like that for quite a while. He's been wrenching on Hondas for over 50 years, has many friends at Honda and when he inquired with them back then when 25/26 started raising his sinister head, they flat told him "Keep on riding, we ain't doing nothing about it". No mention of the above doomsday scenario!

I know very little about electrons, even less than you if you think it is even possible. I think current going one way while electrons are definitely going the other way is akeen to voodoo and too scary for me.

But still, what Mr. Shadow is most diplomatically offering in a prior post makes a lot of sense, even to me. I encourage your trying to digest it.
Click to expand...
Not sure why your comment "that seems to have only happened to bikes coming your way" is directed towards me.
I'm also an old timer mechanic who's been doing it about as long as your friend, and also has friends in the industry.
Your comment about his People telling him to Keep on Riding it, we ain't doing nothing about it, made me laugh :rofl1:
I've heard so many master mechanics telling folks stupid stuff over the years...."that grinding noise will go away when it beds in"...."the rear brakes always smoke when new"....the brake system never needs to be flushed or bled unless the system has been opened.....just put a piece of tape over the warning light, etc.
I know even LESS than you about electrical stuff, which is no secret to anyone on this forum.
If I see a wire harness chewed by a rat, I go running down the drive way screaming like a little girl :run1:
However, many folks have tried to fix these PGM-FI units to no avail.
And most of the claimed fixed involved simply disabling the idiot light.
Which sounds to me like what cutting or disabling this wire would do.
I often wondered why they called them idiot lights...was it because the owners were so clueless that they couldn't tell the bikes/cars were running poorly for some reason, and needed an idiot light to bring it to their attention? or was it because when a warning light came on, and folks ignored it :rofl1:
I've been seeing as many as two to three ST's come through the garage every month, so maybe I just see more than your mechanic does.
Maybe I'm not of the camp to "rub some dirt on it, it'll be fine" keep using it group.
I don't ignore warning lights, and sure don't tell folks to ignore them and keep operating their bikes until they sort it out.
Maybe I'm in the camp of In a world of compromise some don't.
While I think its Great that folks who are VERY knowledgeable about electrical stuff are trying to help us, I just don't think bypassing a warning light is a fix, and more of a band-aid than anything else.
This is just MY opinion, and nothing more. Take it for what its worth.
Sure, the bikes can be ridden with the light on, or even bypassed, but you MAY/COULD be risking further or worse damage if you continue using it.
I've seen first hand, many melted knock sensor switches, wire harnesses, and melted body panels from excessive heat from riding with the light on and ignoring it.
Its up to the owner how he uses his idiot light.
Once a light bulb flickers and goes out, I don't have much hope or faith that someone will be able to fix it again.
I may be wrong, as I often am.
 
Honda’s PFM-FI PCM that monitors and adjusts fuel air ratio does not include data from the knock sensors. Those seem solely connected to ignition timing. This is from reading Hondainfocenter.com on the fuel delivery system in PGM-FI.

This supports engineering a “limp mode” fix that defaults the timing to a standard setting. Horsepower would be reduced but the engine would run perfectly fine for the design life of the engine. The mid 2000 Honda 3.2 liter V6 was allowed by manual recommendation to run on both premium and regular gas producing 270 and 256 hp respectively. The owners running regular gas in their bikes empirically prove the limp mode theory of shutting down the knock sensors in the system.
 
You must log in or register to reply here.
Back
Top Bottom