moro

Hi!
I'm having a very strange issue I'm fighting with for quite a time, and seems like can't do without collective mind, so asking for help.

A IX I'm trying to map experience some weird ignition timing pull between 2800-3800 rpm at WOT. I've made dozens of different runs in the last month, and the manner is the same: absolutely every run some timing is pulled somewhere between 2800-3800 rpm. Timing before and after these rpm ranges is ok, i.e. what it is in the map. Knocksum is zero before the timing is pulled. In the logs you may see that between 3k and 4k rpm timing suddenly falls to -2 to -5 degrees. Additional information:
- I'm running Tephramod v7 88580714 on Euro Evo9
- both hi & lo timing maps are the same and don't contain negative numbers
- alternate maps are disabled
- EGR timing advance tables are zeroed out
- this happens when octane rating is 100% or lower, no matter
- I've mapped few tens of Evos using open source software, so I'm not a beginner, but this time I'm totally lost, so feel free to ask anything, I'm not trusting myself on this car anymore.

Car mods:
HKS free flow exhaust with O2 housing
BBK full turbo
HKS 272 exhaust cam
3-port solenoid
1000cc injectors
drop-in Cosworth pistons and rods
ACL main/rod bearings
ARP head studs
new Denso VKH22 plugs

Here and here are two logs - one with zero knock sum (usual situation) and one with some knock _after_ timing pull (rarer).
Timing map

I can post csv files if needed and I can PM you a link to the ROM.

Thank you for looking!

EVO IXMR

iTrader: (6)

Weird!The logs look clean!

CGMSTI

iTrader: (6)

im having this EXACT same problem.. i have no idea why either.

hediki12

whats ur launch map set to in that rpm range? (assuming thats not apart of ur alt maps) log clutch switch if it is and if it changes then thats ur problem.

merlin.oz

try adding some more spark maps / trim maps to your def:

address=59c1

address= 6765

address=65b5

moro

Thanks guys!

hediki12: a good idea actually. But in this case launch ign map is the same as main one in the relevant load/rpm area.

merlin.oz: hmm, these maps look interesting:
1
2
3
Especially trim #1. I will zero out all these, see what happens and post here as soon as possible. Too snowy right now unfortunately.

wht95scort

iTrader: (25)

One issue I see in your timing map and its around the RPM you listed yet at the start of seeing boost. Your scalling increments are a differance of 5. This is most likely shocking the ecu. You typically dont want more than a differance of 3 between neihboring cells. I'm not 100% sure this is your cause but might want to try and get the numbers closer to make a smoother transition.

moro

Finally had an opportunity to make a test drive.
I've set the #1 & #3 to main ignition map values and zeroed table #2. No change - timing still pulls to negative. I agree that map isn't smooth - but I still can't start tuning it. Of course I've tried to smooth it out a bit - but with no effect on the weird pull either.

I've lowered down values in Base WGDC to lower boost and then made a pair of logs: one with usual parameters and one including only RPM, knocksum, knock_base and knock_adc. Here they are:


This time ECU decided to pull timing even at 2500 rpm. I'm frustrated and have no ideas.



Looking at the knock log I'm a bit puzzled about knocksum going 2 at 3500 rpm - I don't see any reasons for that in knock_adc signal.