Just a couple of questions if anyone can say. Since posting this I've [re]read everything available here and through some of the other relevant threads, revisited it a few times to edit my questions.
Has anyone determined specifically how feedback from the knock sensor is utilized by either the first or second version ECM.
Is feedback processing delayed until a specific rpm threshold is passed, then used [along with other sensor inputs] to sequentially trim the advance, set the [25 or 26] fault code then revert to a secondary timing mode if a reduction in advance fails to reduce the knock signal.
Or is the signal processed after a specific rpm threshold and then used only to set the fault code, revert to secondary mode based on intensity and / or duration of the feedback signal.
From what I recall seeing with some of the charts and information I'm guessing that it's later ECU programs that incorporate Knock Sensor feedback dynamically; substantially manipulating advance. Do all ST ECUs work about the same this way.
I certainly wouldn't attempt or recommend this, but has anyone with the fault attempted to relocate a knock sensor or hang the sensor away from the engine. If you haven't, don't... but if you did and it didn't blow anything up, did anything change.
Could real detonation caused by carbon fouling be the cause. If you bought the bike new, along with the service history, you'll know what grade of fuel you've used and what rpm / load you've operated the bike in. But if you bought used, it might be anyone's guess.
Is there any consensus on assessing how much carbon buildup might exist in the combustion chamber; if reading the spark plugs, a compression test and leak down test indicate carbon, would show that. Then, could running a cleaner after the fact, over some duration clean out some carbon deposits. I guess the question is has anyone resolved a 25/26 fault and attributed it to a reduction in carbon build up.
Would it be possible to monitor advance and injector pulse width / start time under load with a rising rpm and through a fault event, and could a comparison be made between one that faults and a known good one. If tests like this could be done, then we could also probably test to see if post 2008 ECUs have any manufacturing dated variances. It might be found that while all post 2008 ECUs are physically the same there are differences in knock sensing and / or secondary mode after a specific date of manufacture; could it be possible that a new replacement ECU might have the same part number and appearance but be a revised version that relies less on the knock sensors. This would make sense from a manufacturer's position, even if the real problem is burnt sensor wiring or something else, a dampened detection circuit or elimination of dynamic advance but no more faults might keep the buyer [who just spent $1800] happy.
It's not part of normal engine-operation in any way, unless it's a diesel.
