NefMoto

Hi all,

Having this issue with tip-in/transient knock. I have coded a knocklight into ECU to blink MIL when average retard is above threshold and
I can make it blink when i touch accelerator pedal little quicker than very slowly. Doesn't happen all the time, but more evident when engine is up to temp.
It happens even during idling and also <2000rpm range. Above that it's clear all load/rpm ranges. WOT is clean as a whistle.
I have googled and found it to be a common issue and somewhat dark matter. Nobody knows for sure... Maybe you guys do.

What have I done so far? Retarded timing by 5-6 degrees around the problematic area. Increased acceleration enrichment. No help still.
I was thinking limit throttle opening at lower rpms to counter sudden airmass spike?

It's definitely worrying.
I found only one log quickly where it is recorded at idling.

Log AFRs, see how lambda behaves around the problematic area, play with %ESUK. Most likely too little "short-term" transient correction, so it doesn't enrich quick enough.

I have played with ESUK maps. According to WB gauge, AFR drops to about 13ish and then resumes at stoich. In logs there are a few mildly lean spots sometimes.

Is there a delay parameter in ESUK? Quick glance at FR didn't tell about it.

Depending on your lambda request, that might be too lean overall (not enough enrichment generally), but it definitely looks like there's actually not enough long-term compensation.


There are two components to the enrichment/enleanment basically: short-term and long-term.
Firstly the total amount is determined, then the distribution between long and short-term (KFABAK out of memory, could be wrong).
FR tells quite nicely about it, try a "long" glance at that section. ;D

The real reason for this, is that your timing map on middle loads is too advanced.
Knock during transients is usually almost purely due to incorrect ignition timing. The fuel mix does not affect it at all when the engine is not super hot.

I have reduced timing at 700-2500rpm and from low to mid loads. Currently ranging from 3-6degrees reduced. Just tested, during warmup its perfect, as soon as engine temp needle stops at middle, i get blinking MIL...
This is an 04 Volvo S60R if it matters btw.

Any more ideas guys? I rarely ask for help, I try to figure stuff out on my own. But I'm truly helpless here.
Found a value DUKK in %ESUK, defined it it my file, it's 1.5ms. not sure if it matters to be honest.
Reduced timing even more. Delta from stock values is up to 12degrees.
Added more enrichment for acceleration. I'm starting to get inefficient, lots of pops and bangs. Fuel mileage agrees...
Still problem, once coolant temp reaches 80ish C, shit hits the fan.

Get some detcans and listen to the engine.
Maybe something is banging against the block when the engine moves in the engine bay, and actually there is no knock... ecu just sees it because something is hitting the block.

I was thinking that (along with getting and setting up detcans), but would it not register retards at any engine temp?

Seems everyone is ignoring the obvious here.

Look at dwkrz, not all retard is the result of actual knock.

It should jump out just based on the behavior in your graph.

This?

I'm confused. Is it not for circuit error? B_krdws I mean.

Yes, I thought that too. The other path I posted is one I don't know anything about.

I'm open to the idea of it being some safety mechanism, but it puzzles me why is it active only at warm engine?

There are also:
1) Knock adaptation value
2) Pre-emptive ignition retard during tip-in.

I thought you made sure it is actual knock.
Look at knock voltages and individual cylinder trims, not just average.

Do the knock voltages spike? If not, this is not coming from actual knock.

I will search and log those variables.
Can you tell me please what should I search for in FR to find info regarding pre-emptive retard?

Hi all

Just to bump this thread and maybe help someone facing similar problem.
I have fixed the issue now. Volvo S60R has two CVVT camshafts. There is allowed adaptation range for mechanical and installation tolerances. This adaptation was -6.xx degrees for exhaust and -9.xx degrees for intake. Both should be zero ideally, but impossible to achieve. ECU should trigger a fault if deviation is larger than +-8 degrees. IDK why it didn't.

Anyway, turned my cams a little and reset the timing. Both cams are -1.xx degrees and a positive side effect is that engine has a lot more torque and feel much more responsive. I always thought it was warm weather why it felt so slow this year. Actually it was wrong camshaft timing.

Thanks to all for thinking along!