|
julex
|
 |
« on: December 01, 2011, 12:37:11 PM »
|
|
|
This is driving me crazy. I modified all the maps I could think of and RLMAX is still offset about 5% lower until the car hits 4k RPMs where is proceeds to properly fill RLMX limit.
From about 3.25k to 4k the engine is missing out on some power due to some adjustment in the system.
I am not overboosting
IATs sub 20C
No significant knock
Anybody encountered this is the past and know what am I obviously missing here?
|
|
|
|
|
Logged
|
|
|
|
|
julex
|
|
« Reply #1 on: December 01, 2011, 01:55:56 PM »
|
|
|
Would a work around strategy be to raise limiter higher and use KFLDHBN to limit boost to certain ratio like 2.50 to not cross the MAP sensor limit?
|
|
|
|
|
Logged
|
|
|
|
thom337
Full Member
Karma: +15/-1
 Offline
Posts: 81
|
|
« Reply #2 on: December 01, 2011, 03:59:52 PM »
|
|
|
What rl values in your KFMIRL correspond to 100% torque at those speeds?
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #3 on: December 01, 2011, 04:04:30 PM »
|
|
|
|
|
|
|
|
Logged
|
|
|
|
|
julex
|
|
« Reply #4 on: December 01, 2011, 07:50:02 PM »
|
|
|
it is related. After moving the cam shift to earlier RPMs, the notch moved accordingly to earlier RPMs.... Now the question is how to make the load even?
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #5 on: December 01, 2011, 09:19:43 PM »
|
|
|
smooth all three out  |
|
|
|
|
Logged
|
|
|
|
|
julex
|
|
« Reply #6 on: December 02, 2011, 06:34:06 AM »
|
|
|
ok, sanity check:
KFPRG - translating description means that this is a map that defines how much charge remains in the cylinder after completing the cycle, before new charge is pushed in? I am probably sadly mistaken what it is though.
KFPBRK - correction map for boost charge (multiplicative?)... when NWS not active so above 4k notch or any cell KFNW has "0" for?
KFPBRKNW - same as above but for region that has "1"?
KFURL - this one I can't figure out...
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #7 on: December 02, 2011, 10:29:06 AM »
|
|
|
KFPRG - translating description means that this is a map that defines how much charge remains in the cylinder after completing the cycle, before new charge is pushed in? I am probably sadly mistaken what it is though.
Sounds right KFPBRK - correction map for boost charge (multiplicative?)... when NWS not active so above 4k notch or any cell KFNW has "0" for?
KFPBRKNW - same as above but for region that has "1"?
Yea, both are req boost corrections dependent on NWS position. KFURL is part of the conversion from req load into req boost. |
|
|
|
|
Logged
|
|
|
|
|
julex
|
|
« Reply #8 on: December 02, 2011, 04:11:50 PM »
|
|
|
I think I found a good candidate to blame for load having a notch up @ 3.8k and peak at 5k...
So it happens KFURL, which is %/hpa, has this exact characteristic if you follow overlap angle of cams.
Slow sloping down towards 3.8. bump up at 4k where it switches to different angle when NWS disactivates reducing advance, slope up until 5k where the value maxes out and then down again.
Picture of my load shows this exactly.
Off to flash and test it.
|
|
|
|
|
Logged
|
|
|
|
|
julex
|
|
« Reply #9 on: December 02, 2011, 06:35:57 PM »
|
|
|
KFURL is the map. Once smoothed out, the load is constant and doesn't notch. I will post smoothed out map content later on for wiki purposes.
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #10 on: December 02, 2011, 07:27:54 PM »
|
|
|
HMM what I dont understand is why your req boost doesn't show the notch?
|
|
|
|
|
Logged
|
|
|
|
|
julex
|
|
« Reply #11 on: December 03, 2011, 06:23:54 AM »
|
|
|
Confusion of terms perhaps.
RLMAX is the requested load after corrections to maximum load RLMX which is nothing else than theoretical maximum load limited by either LDORXN or KFMIRL, right?
So the actual load RL should never exceed RLMAX which is what my RL perfectly shadows in this pull.
I am not sure why people don't pay more attention to this part of curve (3.8k cam switch over notch) is general, maybe because anything below 4.2k where FATs start they consider irrelevant... a mistake I think since earlier KFURL values still limit boost slightly up until 5k when it reaches max and then beyond it falls slightly again.
|
|
|
|
|
Logged
|
|
|
|
|
nyet
|
|
« Reply #12 on: December 03, 2011, 10:31:58 AM »
|
|
|
I always look at req boost below 3.8k.
In my logs, boost is roughly KFURL * req load + (other corrections), which generally means the notch was not in req load, but in req boost.
|
|
|
|
|
Logged
|
|
|
|
|
julex
|
|
« Reply #13 on: December 03, 2011, 12:36:38 PM »
|
|
|
I request quite a lot of load on my tune which would normally result in requested boost over MAP limit so I decided to control boost level via KFLDHBN map. So far it works, I have it filled with 2.52 for most of map and slightly less above 6k rpms....
it is either ingenious or foolish of me... but my tune allows me to actually control boost level via that map.
So far the tactic works very well. I will keep posted.
|
|
|
|
|
Logged
|
|
|
|
|
julex
|
|
« Reply #14 on: December 03, 2011, 08:30:03 PM »
|
|
|
Well, I think I know the reason why my tune was so backwards. Sometime ago I modified experimentally KFLDHBN (max pressure ratio) and had the same value in all fields not knowing how it exactly works. That's naturally before corrections which are very large negative in low rpms but switch over to positive in high rpms so I never really noticed problems in FATs RPM range since then my boost was limited by load in LDRXN and I was running fine tuned KFLDRL table to limit post KFMIOP LDR.
Lesson learned.
|
|
|
|
|
Logged
|
|
|
|
|
