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

[garz]

I am using a Kpro on my K20A2 and the Kal has been tuned and uploaded to the ECU in Kmanager version 3.4.2.0.
engine is supercharged and runs JDM spec cams - but otherwise pretty standard and has been running fine.

But while investigating a separate issue that is now solved i noticed that something is causing the ECU to pull ignition timing for the first second or so after i open the throttle - i have seen a difference of up to 10 degrees in ignition advance between what the tables call for - and what the ECU is delivering according to the data log. This seems worst when i have coasted down for a while - then reapply the throttle slightly.

i have looked at
1, tip in appears to be set to off - can someone check is this is the case if i post a snapshot
2, IAT ignition compensation - but the IAT sensor is a fast type - but even so the reading does not change by anything like enough to to cause that according to the table
3, wondered if it was due to a gear compensation - but does it in any gear including 6th

tip in 2.JPG

it has been suggested that this is actually the stock knock control still operating even though it has been set to off in the Kal - and that this is a bug in the software in version 3.4.2

has anyone seen the same thing?

i am wary of updating the Kal to a later version of KManager and uploading that to the ECU - as i have heard this can cause issues and need to be checked carefully - but i must admidt i would not know what exactly to look for to be sure its 100% OK - and with these SC engines running close to the edge i am not keen to risk it

any suggestions ?

will post logs and kal if that would help

[Hondata]

There never was a KManager V3.4.2.0 - do you mean 3.2.4? In that case it is two years old.

i am wary of updating the Kal to a later version of KManager and uploading that to the ECU - as i have heard this can cause issues

There are no issues that we know of updating software versions. If you have a specific source, we'll investigate, otherwise I'd write that off as hearsay.

For any further analysis a calibration and datalog is needed.

[garz]

thanks for the reply

yes 3.2.4 - sorry was late when i was writing that

will upload a Kal file and a data log - this one i have to hand and was a gentle run out - there is a difference between table and log is around 4 degrees in many places - eg 2m59 and 9m31 - other logs have shown a bit more than that

note: i have tried altogether different kals - loaded to the ECU with the same version of Kmanager - also with stock knock turned off - and these do the same thing.

a friend loaded one of these Kals with an older version of Kmanager - and it did not retard on acceleration - hence the belief that its Kmanager

K20A SC orig knock off.kal

run out 10mind knock off.kdl

just looking for the explanation

ref using the latest version - numerous people have said there can be trouble when using a later version of Kmanager to load the ECU than the version that was used to originally map it - i think their meaning was that there may, quite naturally over time be new options or features added in Kmanager that may need to be set or adjusted to suit the vehicle before uploading a Kal written with an older version to the vehicle ecu. my concern is only that without some kind of checklist of what to look for i could be unknowingly getting a feature i didnt want or cause some kind of option or new feature or behavior to be active i don't need.

thanks very much for the support

ps - is the tip in retard function fully off ( inactivated ) by those settings - below

[Hondata]

You should set the acceleration smoothing gear retard to -100% for all gears. There's no downside to doing so.

It is hard to test under an old version of software, but I would say that it is applying the knock ignition limit tables which accounts for the retard. In this case, you options are limited to upgrading software versions.

[garz]

thanks for the reply

i will try setting the tip in retard to -100% first and report back when i get a chance to try that out - thanks for the clarification

ref the upgrade to later version - will see how the above goes but if its still doing it i will try upgrading - but seriously is there anything i should check after doing so?

i very much appreciate the help

[garz]

OK, so i have had a chance to adjust all tip in "Retard %" values to -100%

I have then done a similar datalog with these settings - and the result is the same

at 4m 44s, as i open the throttle, tables call for 23 degrees of ignition advance - actual advance in the log is 6 degrees momentarily - then around 10 degrees - and even 1s later there is still a gap between tables and log - so something is still pulling timing
knock retard in the log reads 0 for the entire log

malms with tip in off.kdl

I'm prepared to update the kal to the latest kmanager version - any tips on what to check to see that all went well would be appreciated

[garz]

OK

have now updated the laptop to Kmanager 4.0.8

I went through the old version verses the new version screen by screen to understand the differences
mainly there were new options in some screens - but most of these were not enabled - so that's fine as far as i can see
there are a few new options that are populated - eg overrun fuel settings, idle control settings - but these are not critical in the area I'm concerned with - and i can only assume if v4.0.8 shows them populated - they now only show the defaults that were already present in 3.2.4 just not visible/

One new option was however enabled - on the Fuel trim Tab - new tick box available in 4.08 - and seems to be defaulted to ENABLED/ticked
my kal was already set up to have custom values in the WOT table - some of them richer than 12.5:1 - and on the basis that my tuner seems to know what he is doing - i have unchecked this tick box - which i assume will mean that the ECU will follow the table values( richer/ safer) - not the default 12.5:1 across the board
be good it you could double confirm - as this is the kind of new feature that if enabled will cause the ECU to behave differently if Kal is uploaded with a later kmanager version - the kind of thing i was referring to earlier as a possible "issue" when updating Kals with latest Kmanager

anyway - with that done - i uploaded the updated map to the ECU ( after clearing security etc - so a full clean install of the map) and did a short datalog to check if the ignition timing is still retarded vs the tables

as above
tip in is all now zeroed ( set to -100%)
stock knock control is disabled
inlet air temp does not change enough( or quickly enough) to cause it

car drives OK - but still getting dramatic retardation of ignition timing vs the tables on applying even small amounts of throttle - can anyone explain it ?look at around 17seconds and 21seconds - the logs and tables seem to differ by 15 degrees

will attach log

short blast from mals 150415.kdl

and Kal files

hondata tps adjust, tip, knock 15042015.kal

[Hondata]

The difference appears to be from the 175 degree air intake temperatures.

17 seconds - table is 29 degrees, actual is 21 degrees, air temp compensation is -8 degrees, so within 1 degree.

21 seconds - table 23, actual 9, comp -15, so again within 1 degree

[garz]

thanks very much for the reply - much appreciated -

but somehow i'm not seeing the same data as you

if we take 21 seconds as the example
from the tables - i have 23 degrees ignition advance = agreed

ign.JPG

from the log i have an actual ignition advance of 8 degrees

log.JPG

23 - 8 = 15 degrees of retardation from somewhere
but the ignition compensation table shows 8 degrees of retard at 175 degrees F - under low load

comp.JPG

so there are still 7 degrees of retardation unaccounted for, unless i am reading it wrong .....

[Hondata]

Sorry, I used 15 degrees rather than the 10 degrees from the table. The difference is around 23-10-8=5 degrees. Knock ignition limit at that point is -6 degrees, so I'd say that is it.

[garz]

OK - thanks for that - i am now following

sorry - i thought i had turned off the stock knock for that run - looks like i need to rerun a log with that box unticked - will report back on that

question
is the ECU following the low load ignition temperature correction tables - or the medium to high load - where is the switch over point

i very much appreciate the replies

[Hondata]

Around 40-50 kPa

[garz]

OK got it - thanks

So i have now double checked that the "stock knock control" tick box under Kmanager "Knock/TPS" tab is unchecked - see below

knock off.JPG

and ran another datalog - but still getting this additional ignition retard

as an example see 2m12s in the log attached -

ign2.JPG

table says around 24 degrees of advance
log says actual is 7 degrees
difference is 17 degrees
IAT is 80 degrees - so IGN compensation table says an extra 10 degrees of retard - so still 7 degrees of retard unaccounted for .....

to double-double check that the knock control tick box is unchecked (to be absolutely sure stock knock control is set to off in the kal the ECU is running) i have downloaded the actual kal from the ECU before posting - i can confirm the tick box is unchecked.


will attach kal and log files below
kal

JRSC tps adjust, tip, knock 15042015 chk.kal

log

run to crick with knock off 180415.kdl

am i missing something - or is stock knock control still operating - even though its turned off in the kal?
if not that, can you help me understand what is causing the additional ignition retard?

once again - i do appreciate the support.

[phoenix5]

What type of TPS are you using? Just throwing an idea out there, the TPS may have a worn section in the resistor pad that could cause it to read wrong, which could cause this.

[garz]

thanks for the contribution - and i can see how that would cause erratic retard - but the TPS plots are smooth - so i don't think its this in my case - see below the TPS for the point in example

tps2.JPG

also the amount often seems to be 7 degrees rather than a random amount - so it is as if its looking up that value from somewhere ......
its a standard honda TPS sensor

i am however grateful for you input and willing to check out other possible causes as i do need to get to the bottom of this