is this a bug in KManager? tip-in, stock knock. etc

[Hondata]

At the point 2:12 the ignition timing is fairly flat at 9 degrees and the table is also fairly 20 degrees. There is 10 degrees of retard from the IATs.
I would zero the knock ignition limit table.

[garz]

thanks again

perhaps i did not pick the best example - but i am getting sudden steps in ignition timing in that log in lots of places - even at pretty much steady load conditions
example 9m18s

knocke.JPG

can this really be the knock function kicking in (let us leave aside the issue that the feature is clearly turned off for the moment) as at that point the knock limit table says -0.5 both immediately before the big step in ignition timing - and immediately afterwards. not 9 degrees per the sudden step.
IAT is flat
ECT is flat
MAP is flat
CAM Advance is flatish
no gear changes

what else can it be doing ?

[Hondata]

I see the 10 degree step in places. I believe it is the knock ignition limit tables, as the step occurs at the half way point between cam angle tables.

Here is what I would do:
- use the latest software version (currently 4.0.8)
- reset all the stock ignition limit tables
- reset the knock sensitivity tables
- switch the ''use stock ignition logic' back on
- enable closed loop P0135
- the cam table are unusual, but should not cause a problem

The elephant in the room is the very high IAT. I can look at my non intercooled supercharged datalogs, but I don't think I saw 190F, even on the dyno.

[garz]

thanks again - i think we are making progress

i have just done a short run with all the knock ignition limit tables set to zero as you suggested earlier - will post that up in a minute

i can confirm - all the kals i am making/testing now are with the latest 4.0.8 as we discussed

ref IAT - yep, for some reason the IAT sensor does seem to be reading a bit high - its the delphi type - but i note that even when left for days and ambient is around 15 degrees and ECT reads around 15 it reads 38-39 ( Celsius ) so something not quite right there - but its more a case that the sensor is giving high readings than the intake air is actually super high - i do have a spare brand new one - but have to dig out as i think its in storage.
so well spotted - but its not erratic - just consistently high - which means the ECU will be consistently pulling timing for that according to the ignition compensation table - but the iAT readings are steady - so this will not cause big steps - unless there is some ECU behavior i don't know about.
The error will be on the safe side - which is fine till we figure out what the other reason for additional retard is.
The engine is intercooled by the way and the IAT is in runner number 1.

The logs now show the ignition values following the table very closely (within 1 degree ) when i open the throttle - and no big steps in the trace as before - so it looks like the stock knock control is active in version 4.0.8 unless the knock limit tables are all leveled.

knock lim.JPG

the good news is its very good to finally know what was going on - the bad news is it would seem this makes my current kal / tune obsolete as all the tuning was done with this bug present. Has no-one spotted this in the 2 years its been like that?

kal for reference

JRSC tps adjust, tip, knock 23042015 chk knk zrd.kal

log ( had to shorten it a little )

mals to pur knock lim zero short.kdl

ref the below could you perhaps elaborate a little
- "reset the knock sensitivity tables" do you mean adjust them all so they do not activate ? is there any need now that we can see stock knock is no longer acting ?
- "switch the ''use stock ignition logic' back on" Could you help me understand your reasoning
- "enable closed loop P0135" - again - can you explain why i would do that
- "the cam table are unusual, but should not cause a problem" - do you mean the Cam angle tables - if so yes, i know about that

thank you for the support

[Hondata]

I do not believe it is a bug when your problem stems from having ignition knock limits which are radically different between cam angles, so when the actual cam angle oscillates between the two tables the ignition goes all over the place. Also the sensitivity tables are all 100%.

> "reset the knock sensitivity tables" do you mean adjust them all so they do not activate ? is there any need now that we can see stock knock is no longer acting ?
Just right click and 'reset to default'. The reasoning is that they are currently almost random.

> "switch the ''use stock ignition logic' back on" Could you help me understand your reasoning
To see how it affects the ignition behavior.

> "enable closed loop P0135" - again - can you explain why i would do that
With a working o2 sensor it should not be switched off.

You need to fix the IAT sensor. Obviously it is not compatible with the stock IAT sensor. The OBDI bolt in sensors work the best with forced induction.

[garz]

ok - just to be clear - i am not here to knock the Hondata product - its a good product and works well - i am a supporter

i just need to understand what its doing when its behavior is less than obvious so i can decide what to do to fix my engine/tune/driving experience.

may be i am misunderstanding something still:
the issue i was referring to as a bug is when the "use stock knock control" feature is turned off ( i.e. tick box unchecked ) - i assumed this would deactivate any application of what was in the knock tables - regardless of their value - which, as we have established in the above tests - it does not do. Have i got it wrong? - does this tick box load a set of defaults over the top of whatever is in the tables, rather than disable the function altogether - or how does it behave ? clarification would be much appreciated

>> "reset the knock sensitivity tables" do you mean adjust them all so they do not activate ? is there any need now that we can see stock knock is no longer acting ?
Just right click and 'reset to default'. The reasoning is that they are currently almost random.
understand that - thanks - will do this as another trial

>> "switch the ''use stock ignition logic' back on" Could you help me understand your reasoning
To see how it affects the ignition behavior.
understand that - thanks - will do this as another trial

>> "enable closed loop P0135" - again - can you explain why i would do that
With a working o2 sensor it should not be switched off.
OK - i have noticed my narrowband o2 sensor seems to have v low voltages - but seems to work OK anyway - at least it seems to allow the ECU to make fuel trims - could the P0135 be causing the low voltages from the o2

>You need to fix the IAT sensor. Obviously it is not compatible with the stock IAT sensor. The OBDI bolt in sensors work the best with forced induction.
yep - i will get to that - but at present its just adding a bit more predictable retardation which i think is helping keep the engine on the safe side while i run trials. By the way - many people use the delphi fast sensor in SC and charge cooled K20A applications without issues - about a 100 people in the UK i think. will figure out whats going on with mine - others do not read high

thanks and once again - i am a fan

[garz]

hello Hondata

clarification on the above points would be highly appreciated

kind regards

[Hondata]

If you take the standard tables and radically modify then sometimes to ECU will detect things are way out of bounds and change behavior. In summary: fix the IAT sensor, put the knock ignition limit and sensitivity tables back to close to stock.

[garz]

OK - i have some mechanical issues to take care of first so it may take me a while - but i will come back

thank you very much for the support thus far - it has been very helpful