Another one, scanner reporting 3.5 degrees knock at 7.6 (11.1) degrees timing interesting to know the load.Originally Posted by Molch
the below example shows its between 3500-4000rpm then flattens back out!
Another one, scanner reporting 3.5 degrees knock at 7.6 (11.1) degrees timing interesting to know the load.the below example shows its between 3500-4000rpm then flattens back out!
Tune with actual data not simulated data!.....Applied Road/track Tuning
Using the Speed Density guide written by DarrylC Load Load can be worked out at approx: 2.39
Load = AirMass / Engine Displacement.
Workings:
MAP in PSI = 36.7 convert to HG = 73.2967 - I know this from other logs at the moment running higher boost then intended. going to focus on this again shortly.
Revs are 3912
Cam Angle is 0.6
?MAP at Zero Airmass IMRC Closed? = 1.33 ~Approx (Units inHg)
?MAP per Airmass IMRC Closed? = 17500 ~approx (Units = inHg/lb)
Therefore Airmass = (73.2967-1.33)/17500 = 0.00411238286 (lb)
Engine Displacement is from the tune file under Engine -> General -> Engine Displacement and is 0.00172 lb/cylinder.
0.00411238286/0.00172
Load = 2.39092027
Last edited by Molch; 12-18-2016 at 03:09 PM. Reason: Wrong psi
I am thinking that the afr that looks lean is possibly spark blow out (the sensor reads the unburnt oxygen and shows a lean reading).
Yesterday I changed the injector data over to the standard ID1000 data, tuned the low slope and then the high slope and got everything running pretty well. I then did one run yesterday where the same lean sympton was shown except I also noticed there was a slight miss around 3500rpm. The logs show revs dropping slightly. This also points to spark blow out to me.
Its hard to narrow because my tuning effort is so slow and im learning at the same time. I think i'll focus on getting my boost levels where i want them as the fueling looks ok everywhere except for this symptom under wot.
19lb boost, potentially actual knock, difficult to tell given the knock trace is not on the graph. The injector pulse width also appears relatively constant through your your lean spot, if a miss or knock is causing your lean spot it is possible it would show up with additional activity by your knock sensor in that area, again no trace to see.
Last edited by Yortt; 12-17-2016 at 08:54 PM.
Tune with actual data not simulated data!.....Applied Road/track Tuning
If you think spark is blowing out, would either
- lowering boost a little OR
- closing the spark plug gap a little
Help to diagnose?
Exactly. Going to start with the cheapest option first which is lowering boost.
I'll pull up the trace and add knock line properly and report later tonight also.
Oh crap just realised I had the boost wrong it' that calculation, actually more like 22psi.. I'll edit the post and sort it.
ok so dropped the boost and the lean spot disappears so it looks like it was probably a spark issue. I've made some other change enabling closed loop boost control again.
Next problem im looking into is the lean spike when accellerate. Ive read posts and reduced my ol delay/blend without little change and have also dropped the OL TPS trigger really low so it just is almost always in open loop and the problem still happens even if OL is enabled way before I mash the loud pedal.
I am guessing one of two things, I either need to determine where/how tip-in fuelling is controlled or figure out if the lean spike is due to some other changes in the engine making the AFR read funky. (CAM angle changes for example??)
The below shows two graphs, graph 1 is the problem and graph 2 is without the problem with the TPS ramping up much slower.
Chart 1: Lean spike and knock on tip in:
Chart 2: No lean spike when TPS ramped on slower:
Also still noticing many knock events pulling spark. its a bit hard to determine what is real and what is not without a dyno and some headphones. If anyone has advice let me know.
The load,RPM and timing is changing while the knock remains relatively constant, very unusual.
I have removed 6 degrees of timing to get rid of a 3 degree knock with very little change in the knock reported while the wideband is reporting AFR as expected I am beginning to suspect an issue with the logging software.
Tune with actual data not simulated data!.....Applied Road/track Tuning
I thought that what its showing is just how the knock sensors are meant to work. So it see's a knock event and then pulls timing until the correct parameters are met which allow timing to be increased again. All of the parameters are customisable.
So the first chart shows a knock event (As detected by the knock sensors with factory settings) just before where i have marked the line. What we are seeing could be:
1.) Real knock caused by any number of factors - AFR to low or high / timing to low or high / heat in the cylinders etc
2.) Something else that the knock sensors picks up as knock. i.e knock sensors are too sensitive
I have lowered the knock sensor sensitivity in the tune effectively turning them off, made no difference, however my knock problem is at low to medium load lower rpm. When I am under WOT initially a small level of knock at 1500 rpm then nothing through out the WOT run.
Assuming your inlet temperatures and coolant temperatures are ok and running on 98 with an AFR of 12.1 and a timing delivery of 7 degrees @ 18.2 psi would indicate to me normally it should not be knocking particularly when logging on the road. In the absence of any other indicators this the leaves in my mind a question mark over how the logging software is processing the the data stream coming from the car.
Maybe those with a better understanding than me may be able to enlighten me as to what effect the logging software calibration has if any on the scanner reported knock.
Tune with actual data not simulated data!.....Applied Road/track Tuning
Reply With Quote
