Ludikraut

iTrader: (17)

The scenario:
my car is on a dyno running a conservative tune (10.5-10.8 AFR) and ... 3 out of 5 dyno runs the ECU pulls timing. Sometimes a lot, sometimes less, and 2 our of 5 times no pulled timing, runs great.

What has been tried:
- Ran the same map with racegas in the tank ... still pulled timing
- inspected plugs ... they are clean. So clean they look new.
- Ran even more conservative map ... still pulled timing
- set (gasp) the low octane map = to the high octane map ... still pulled timing
- disconnected coolant temp sensor ... major pulled timing (limp mode?)
- disconnected knock sensor and hooked up an external knock senser (read: wrapped in a towel off to the side) ... lots of pulled timing (limp mode again?)

So my questions are:

1.) What other variables besides knock sensor activity can cause the ECU to pull timing?
2.) How is the ECU determining to pull timing, and how much timing to pull, when the high and low octane maps are equal?
3.) Does anyone have info on what the knock filter maps do? (can PM if you don't want to make it public)

This is on a heavily modified Evo, although the car has been inconsistent over the past year or so (e.g. not quite as heavily modified). I'm thinking the ECU gets information from several sensors: MAF, barometric pressure, intake temp, coolant temp, knock mic ... am I missing any?

Needless to say, it's been a frustrating afternoon....

l8r)

djmikeymike

iTrader: (1)

are your low and high timing maps the same?

Ludikraut

iTrader: (17)

Yes, we tried setting them to be the same (both for fuel and timing maps), since we were sure that there was no knock. The ECU still managed to pull timing.

l8r)

djmikeymike

iTrader: (1)

i was under the impression that when the ECU heard knock it switched to the low octane map........... do you have a log when this is happening

MalibuJack

iTrader: (5)

There are several Knock filters which determine a noise and frequency threshold for it to determine if it thinks there's knock..

I don't see if you specified what RPM this was happening, but are you still running the XEDE or any other piggyback?

Anyway, I found if you scale your MAF reading with a piggyback (or in my case a blowthrough meter adjustment) you can get into regions of the stock-ish maps that have very low timing or negative timing..

I can consistently reproduce this on my car..

I set the car's idle MAF frequency to 30hz, everything works fine, midrange timing is around 5-6 degrees, redline is about 19 degrees, a nice mild tune.. (I ran out of injector so I am waiting on installing some parts) If I were to raise my idle maf frequency (its the same as scaling for injectors on the ECU+ or XEDE) It would globally make the car think there's MORE airflow all over the place.. In my case I'd go from 800hz at 4000rpm to 1300hz at 4000rpm, and would put me squarely in load cells of anywhere from -4 to +4 degrees at any point and the car would feel TERRIBLE..

If you have cams, and other mods, you end up having to raise the MAF value globally to get the car to maintain happy fuel trims (if you didn't reflash the ECU for one reason or another) and in return you would essentially be lowering your stock fuel cut values, place all the calculated load up a few columns in the Maps..

What does this all mean? Well if you look at the noise filters, the higher the load, the more sensitive it becomes and will start pulling timing at lower thresholds..

Although I'm not 100% sure how it handles timing, it does go between maps (high octane and low octane) using a calculated long term trim similar to fuel trims, but I do think it can and will instantaneously remove more timing under the right conditions then the maps are programmed for, only it doesnt remove it perminently, only for the incident..

That could explain why you are getting timing pulled when there are no instances of real knock..

RazorLab

iTrader: (8)

How much timing is being pulled per instance and at what RPM?

Ludikraut

iTrader: (17)

Thanks Malibu. I do not run any piggyback at the moment (wanna buy my XEDE? ). The tune is 100% ECU based. As a matter of fact the car behaves the exact same way with 2 different ECUs (both with the same tune). The MAF settings have not been adjusted. The only things that have been adjusted are:
- injector scaling
- idle target values
- fuel and boost cut values
- fuel and timing maps (both high and low octane)

Pulled timing ranges anywhere from -1 deg to -4 deg, IIRC. Timing is pulled at times across the entire RPM range, sometimes only across a portion of the RPM range - i.e. there were runs where it would pull timing up to ~6000 RPM and then suddenly pull no timing all the way to the end of the run.

Again the real mystery is why and how the ECU can pull timing above and beyond what is specified in the high/low octane maps...

l8r)

AutoXer

iTrader: (1)

How are you logging timing? Are you using OBDII? My OBDII timing logs have always been inconsistent.

MalibuJack

iTrader: (5)

Look at your calculated engine load values at those points.. I'm thinking with your mods that the MAF reading itself may be on the very high sides anyway.. Have you looked at the timing maps in the RPM range to see where a -1 or so would occur?

But if the ECU hits criteria for what it sees as knock or some sort of need to cut power, you would see the timing momentarily altered outside of the ECU's timing in its maps..

But I am thinking your calculated load column is outside of where your looking..

I can relate my experience simply becuse I have the level of control that lets me force the MAF readng into areas not normally used, and thats exactly what I would see, and it would manifest itself as what felt like stumbling or misfire.. When I looked at my data logs (in ECU+) the knock voltage was well below 2v, injector duty was 60% or less, AFR's were rich at 11.2 or fatter, the only thing was the timing and it was exactly as you described.. When I matched the MAF reading to a load cell, I noticed it was in a load column most would never hit (1300-1400hz at 4000-5000 rpm)

I hope maybe this can give you a little insight..

burgers22

Hi Malibujack

I to have ECU+ and Ecuflash. I am wondering how you translate MAS Hz to load cells. The how to, for this would be of great help.

MB

MalibuJack

iTrader: (5)

Honestly, I have just gotten to the point that I know roughly what the MAF value represents at a particular RPM..

Unfortunately those % values don't directly correlate to any particular Boost level, but a calculated load for each RPM step.. Since I have worked pretty deeply with the MAF sensor in the past, I am familiar with what the value should be roughly with relation to the stock cells tuned, and what bumps it one way or the other..

HOWEVER, most of you guys can simply look at the path in the MAP, its pretty easy to see where the tune travels through in the RPM and Load cells, If you make changes, the values would fall somewhere outside it, at that point knowing the MAF frequency and if its higher than stock, or lower than stock, will give you a hint to where to look.

But the short story is, I'm just used to looking at it...

taenaive

iTrader: (20)

the same thing happnes to my car too. It pulls the timing like you do. I think it might has to do with the inconsistent MAF reading + intake temperature and other environmental conditions and sometimes knock! . But Also, I am wondering it has to do with open loop load table. It has 2 tables, and If you look at it, one is set for high load at midrange and the rest of loads are the same on 05 ECU(03 is kinda same way but more different values accross the range). ECU may be switching those Open loop load maps depends on the moood.
I will try to experiment with this. unless this is just fixed value for high and low load, I might see some changes...

DynoFlash

iTrader: (91)

This is a fascinating aspect of the stock ecu flashing and one which had me dead in my tracks about two years ago. It took a lot of work and effort to internalize how the ecu works and what it is doing to then find a way to stop the ecu from pulling the timing that way.

At one point I felt that this "problem" could not be fixed and we started putting SAFC's in cars which had this condition as a band aid fix.

I guess I am just writing here to tell you that you are not lossing your mind, the ecu is reacting to something you are doing and you just need to take a great deal of time to work through the problem to find a solution.

Best of luck.

AlwaysinBoost

iTrader: (17)

wow that was helpful...

fsugatorbait

iTrader: (16)

lol...seriously.