ECM FAILURE POLL AND DISCUSSION
- Thread starter Mellow
- Start date
askadolph
MM2(SS)
has anybody used carmo electronics in uk to repair ecm
OP
OP
Not that I know of.
Knowing ECMs can be repaired would lessen my fear of pre-'08 1300s.
Yes.
But as it looks like the repair would require replacing a chip that is no longer available, they had no better luck than all the other shops that have been approached all over the world.
From the Carmo site:
Jambriwal
2016 ST1300PA
Lightning hit a tree in September on my property. Took out the circulation control on our boiler. Found out in December. Brrr! Fortunate to get a rebuilt aquastat board because they are obsolete.
Is there anyway I can get a failed ECM and wire harness so I can build a breakout plug. so it will be easier to troubleshoot the electronics. At this point in time the wiring is aging and intermittent problems can be hard to find with static testing.
I can tell you that the harness and ECM will not survive the build process, all I am after is the male and female ends with a chunk of wire on it.
My intention is to build a loaner tool and it can be sent anywhere...... unless there is one available out there now. Which means .......nevermind
I can tell you that the harness and ECM will not survive the build process, all I am after is the male and female ends with a chunk of wire on it.
My intention is to build a loaner tool and it can be sent anywhere...... unless there is one available out there now. Which means .......nevermind
Stu Padazole
Wayne
Is this what you want to make?
At one time I knew the Honda part # but now I've lost it
At one time I knew the Honda part # but now I've lost it
Andrew Shadow
Site Supporter
Honda P/N 070MZ-MCA0100, ECM Test Harness.
Two required, only $2,000.00 each.
Last edited:
askadolph
MM2(SS)
got it ....will try all options first before i set the bike on fire
I've put 13,000 mile on mine with the code 26 light on, seems fine(I just leave it in diagnostic mode). I do have another harness and a good ecm to try out when I get the time.
Now you need to identify all the circuits with specifications, 5v ref wires, sensor signal wires, KOEO, KOER . Along with resistance specs. Not a fun task If you can find them all. Then if the intermittant problem isn't there when your testing what then?
Your testing under operating conditions with the breakout so you will see what the ECM sees, the intermittent problems might be able to show up with wire wiggle and you can also look for low voltage readings during operating conditions. Frequently bad wire plug ends show up as ok with OHM testing but fail during operations. And yes I have run into this exact problem. The intention is to make the pigtail long enough to be able to ride and possibly duplicate the failure ( with the memory on the volt meter only, not watching of course) The expectations on the sensors are similar across almost all FI systems. The circuits are identified on the wiring diagrams for each year and can be separated from the harness and directly accessed for troubleshooting.
The nice thing about the ST is the back seat can be removed and ridden without having to throw a towel over the front to ride.
Intermittent problems crop up and be difficult to find, but the threshold for low voltage and ok voltage reading can be hard to read also. Everything at your disposal helps
The nice thing about the ST is the back seat can be removed and ridden without having to throw a towel over the front to ride.
Intermittent problems crop up and be difficult to find, but the threshold for low voltage and ok voltage reading can be hard to read also. Everything at your disposal helps
Last edited by a moderator:
....but not what the CPU sees.
As suggested by "Dark Sun" on the GL site (see post from @Mellow on that) and illustrated by some of the instructive schematics that @DannoXYZ had posted (but now gone), it appears that a discrete component on the ECM board is causing the malfunction: The chip that pings the Sensors and passes the Sensors info/status on to the ECM's CPU.
The CPU is doing its job: Triggering a "Knock Safe" default mode because it is being fed garbage info from the failed chip.
For troubleshooting, you'd have to tap into the traces at the board level, to check the signal flow between this component and the CPU.
Once the shape of this internal signal on a healthy ECM board is known, it can be compared to the one of a malfunctioning ECM.
And then what?
All we would have achieved by then is confirm we have a chip that needs to be replaced.....and there is no replacement available anymore (hence none of the most reputable ECM repair shops are able to fix the issue).
Even if trying to implement a "Ping Chip" emulator, it would still need to be connected at the board level, to talk to the CPU. Then we are better off with what @DannoXYZ once talked about (and suggested by others): Replacing the Honda ECM with a MicroSquirt type device.
@DannoXYZ indicated he once packed a MicroSquirt on one of his bikes into the bike's original ECM case, that would be a dream solution for the ST.
I haven't found yet a way to check your conclusion without disassembling the ECM and hoping that I covered all the inputs to the ECM. How many ECMs have been replaced 1st before any complete troubleshooting, this offers a way to test circuits without disturbing most of the wiring as well as checking the load drop. The sensors in the ST are all the same as in any Fuel Injection systems and don't offer anything unusual in their readings, even the outputs don't differ from other systems. being able to watch the knock sensor to see if it actually triggers a code.
I cannot see a downside to troubleshooting a system without back probing or piercing wires to get voltage sensor readings or drop readings. Simple OHM reading can be misleading
I have always found it easier to fix problems when they are identified completely' The knock sensor is just a small part of the fuel injection system and the last thing I want to do is concentrate on one circuit. I have run into too many parts changers that will go after a problem by buying parts. For instance a speed sensor can go bad if it gets enough build up on the end, an out of range coolant sensor will cause bad fuel mileage. and the list goes on.
I cannot see a downside to troubleshooting a system without back probing or piercing wires to get voltage sensor readings or drop readings. Simple OHM reading can be misleading
I have always found it easier to fix problems when they are identified completely' The knock sensor is just a small part of the fuel injection system and the last thing I want to do is concentrate on one circuit. I have run into too many parts changers that will go after a problem by buying parts. For instance a speed sensor can go bad if it gets enough build up on the end, an out of range coolant sensor will cause bad fuel mileage. and the list goes on.
YES THAT IS IT EXACTLY. I looked for one and could not find one for the ST, maybe I will have to drag my butt deeper into the web, can I get a Doctor to connect a probe to the back of my head to make it easier to search!s this what you want to make?At one time I knew the Honda part # but now I've lost it
Andrew Shadow
Site Supporter
I posted the Honda P/N for it in post # 131 above. It would probably be a good idea to check to see if there has been a revision to the P/N.
I came across a site a few years ago where someone has already done a lot of the research to make one of these test harnesses. They posted what connectors were needed to make one yourself. If you poke around long enough on the internet you will probably find it eventually.
Found a couple of breakout boxes on Ebay, they call them pin-out harness. One was 297 Terminology matters on search. Missed a few steps on my searches. Thank you for the bump in the another direction
Not my conclusion. But rather what "Dark Sun" was reporting on the Gl site, as posted by @Mellow, related to the similar Knock Sensor Code apparently commonly thrown by the 01-05 GLs.
I wouldn't know of a way either. Once you've determined that the ECM is not doing its job as a black box, you'd have to look into it at the board level.
There is no flash memory on this unit (or at least nobody to this day has been able to access it), so a fix would have to be a hardware fix.
Dark Sun thinks he has traced the failure to the chip that processes (filters) the noise recorded and passed on by the Knock Sensor. He has done a lot of work to get there and it seems it could make sense to build on that, checking if what was found to happen on the GLs could also be what is happening on our STs.
On the GL site, it is suggested that a "clean" ground for the Knock Sensing Circuit could help extend the life of their ECM. I was hoping something similar could be implemented on the ST, to help extend the life of our earlier ECMs........before they all crap out!
Interestingly, after Dark Sun published his findings, somebody else reported that, not knowing about the noisy Knock Sensing Circuit issue, he had cleaned the grounds and added another "cleaner" one to his bike and done the same on a few other 01-05 GLs, and none of these have thrown the Sensor Code yet.
Has anyone tried to use 2008 or later ECM> I think it should work fine but would require the connectors to be re-pinned and upgraded to 33 pin style.