knock on AP v2 Stg1, 93 oct. logs included, help?

hey all. i didn't want to keep derailing the "knock, do you have any?" thread so i made this new thread:

 

i was on a 93 octane stage 1 cobb ap v2 map and i started logging with RR and here's what i got for my two pulls in 3rd at WOT a few days ago:

 

http://www.legacygt.com/forums/attachment.php?attachmentid=70617&d=1247675776

http://www.legacygt.com/forums/attachment.php?attachmentid=70650&d=1247723684

 

i know, the parameters i chose aren't all needed, but it was stuff i tried to remember off the top of my head.

 

fast forward to today, i had the opportunity to change the maps to try both a "stock" style map from cobb, and also a stage 1 91 octane map i've tried to label the files as descriptively as possible. i did two runs for each of the maps. also, i don't know if this is very important, but i did not do a full reflash. i just had the AP upload a realtime map over my base map (which was still 93 octane).

 

SO... moving on. strangely enough, today when i did the run for the 93 octane map, only one of the runs had knocking in the 6000 RPM range. also, in stock run 2, pay no attention to the knock at the 6500RPM range.... the knock sum jumps from 31 to 34, probably because i bounced off the rev limiter, oops!

 

now, if i am interpreting this right, the stock and 91 octane maps are not knocking whereas the 93 octane map was knocking. safe to say that i should switch to the 91 octane map?

also, am i right in understanding that the -1.41 i see consistently in the 4KRPM range is something that's been learned by the ECU and stored, as evidenced by the learning view screenshot i took? that should get wiped out as soon as i do an actual reflash of the base map with a 91 octane, right?

 

and honestly, should i bother with the 93 map? i have accesstuner street here, so i can modify the 93 octane map once i learn how to do it. but it seems like the 91 octane map is just fine, so maybe just use that? not really even worth it for the trouble since i'm getting a custom dyno tune (stage 2) in another month or two at most.

 

thanks. i know i'm asking a lot and sort of rambling, but i wanted to be clear.

day 2 - stage 1 93 oct 3rd wot (run 1).csv

day 2 - stage 1 93 oct 3rd wot (run 2).csv

day 2 - stage 1 91 oct 3rd wot (run 1).csv

day 2 - stage 1 91 oct 3rd wot (run 2).csv

day 2 - stock map 3rd wot (run 1).csv

day 2 - stock map 3rd wot (run 2).csv

i made another screenshot after having run the car for 2 days on the 91 oct map and a new air filter (old one was pretty dirty). this screenshot is BEFORE i reset the learning view...

 

after i took the screenshot, i reset the ECU using LV and drove around for a bit... only about 30 minutes. i'll see how LV looks again in another day or two after having driven more.

 

i'm also tempted to toss the 93 octane map back on and fill up using 94 octane gas from Sunoco and see if that prevents knocking.

Both maps (93 and 91) are knocking way too much...and high in the range which is not good.

 

I think you should fill up with 94, reset the ECU and reflash the 91 map. It could be that you just got a bad batch of fuel, but the knock you are getting is dangerous.

 

Your WOT logs should be entirely knock free, if there is any knock i would not use the tune.

where's the knock in the 91 run? i see the -1.41, but i thought that was just "left over" from the learning from before...? the knock sum doesn't increment during the 2 pulls for 91oct.

The knock on the 91 is reflected in the LV...althought he WOT logs are clean, the LV is showing knock in the high load 4000+ RPM range.

The fact that the pulled timing is being recorded on the LV tells us that there is some pretty consistant knock in that range..not just one isolated event.

hrm, ok. well like i said, i only just reset the ECU today. how long does it take for the ECU to really be 'fully' learned, whatever that means? my work commutes are pretty short, about 11 miles each way only.

 

i only drove a short distance today after resetting the ECU... maybe about, 18 miles total of mixed city/highway. and i didn't really get a chance to really get on the gas. when i got home and ran LV again, i didn't see anywhere near as much stuff in red. actually, i don't think i saw anything in the high engine load columns. only the first column (0-<0.85) had one entry in red. wish i had taken a screenshot of that, but i just assumed that the drive was too short to have had the ECU fully learned. i'll pull another LV tomorrow i guess after driving some more?

In flkc the ecu is reacting to knock, and trying to prevent it. So lets say one knock event caused timing being pulled in that region of the map, it needs time to make sure that it is okay to add timing back in. The time it takes is probably dependent on how much time you spend driving in the 4K plus range... I would guess a couple of days should be enough time for flkc to figure out where you consistantly knock and pull timing accordingly.

 

What is interesting is that both your LVs looks the same...meaning timing being pulled in the same regions. The first LV is with the 93 octane map?

The first LV is with 93 octane map only, yep. Immediately after I took that screenshot, I put on a 'realtime map' (didn't do a complete flash.... do I need to do that? if so, I will) of 91 octane, and drove the car about 2 days. Not a lot of driving.

 

I did NOT reset the ECU at any point between those two LV screens, so whatever was in the ECU remained, and I guess it made adjustments and started pulling less timing? It looks like the 2nd LV is 'better' than the 1st.

 

I only just reset the ECU today, so none of the images I posted are from a fresh 91 octane map with a reset ECU.

 

Do you know if I actually need to reflash a 91 base map, or is the 'realtime' map sufficient? My understanding was that the realtime map only updated some tables, but not all... wondering if the other tables are relevant.

 

BTW, really appreciate your help so far. I was going to ask on the RR forums too, but their new user reg system is borked at the moment.

Glad I could help...when it comes to tuning, we are all learning.

 

I think the RT is enough, but I went OS about 7 months ago and forgot how it worked with the AP (real time vs base map). It is probably overkill but I would flash the 91 base map just to be sure.

 

You next LV (after the reset) will give you a much clearer picture of what is going on.

 

If you haven't so already, you should look under the hood and make sure nothing is rattling around under there. Sometimes (below 3000 rpm) the knock sensor inerprets underhood noise as knock...but the knock sensor is pretty good at detecting true knock above 3000 RPM.

at the rattling under the hood. I'll definitely check for that. Will also do a full reflash to 91 to see what happens.

 

I was considering going OS but the RT map uploading feature of the AP is nice to have, just in case I get a bad tank of gas and need to change things on the fly, or I drive somewhere without 93 or only crappy 91.

 

I would absolutely love the AP if they added some of the knock parameters to the logging. It doesn't log any sort of knock at all!

 

The other reason I'm staying with the AP is that my friend's tuner (he's going to introduce me) only does Cobb Protuning.

Unless you changed the setting, which you haven't the ECU pulls 1.4 degrees of timing for every knock event that it has decided to use FLKC for.

 

So that -1.4 is from one knock event. The ECU will then use .35 degrees of timing to be added in at a time if it doesn't see any knock for a specific period of time (we can change that period of time, but AFAIK nobody knows what the ECU value equates to)

 

Also remember that the LV he has starts the last column at 1.6+ load, so that knock could have happened at a much much much slower pace on a non WOT run (where his load would be closer to 2.4 or so)

 

Besides, with his OTS map FLKC will not even be used above 2.3 load (gotta make sure your tuner changes that!!!).

Besides, with his OTS map FLKC will not even be used above 2.3 load (gotta make sure your tuner changes that!!!).

 

I never knew that..wow.

Here is what a version 1.16 map looks like, and everything I have seen points towards COBB not changing it on their later revisions.

 

http://i248.photobucket.com/albums/gg194/littlebluegt/COBBvs116flkcload.jpg

 

FLKC turns off at 2.2 load, and turns back on below 2.1 load.:eek:

 

Brutal!:eek: That means your only form of knock control above 2.2 load is FBKC, which is never learned, and will never affect IAM (or DAM).:eek::eek:

 

I should e-mail COBB, they have changed things in past on my recommendation (didn't think they would listen to me, as I was just repeating what I had learned from RR) maybe they will do it again?

Yep, you're right, they didn't change it in the 1.18map. Just pulled it up in AccessTuner Race and it says the same thing.

 

Also, here's an updated LV after having driven... about 25-30 miles total since resetting the ECU on a 91 octane map. Didn't do a datalog yet.

Much, much better looking LV. Looks like you had one knock even while under flkc..and it is already burning off as timing is being added.

 

One thing to remember about the LV is that it will only show pulled timing while in flkc mode. So you could have had 5 knock events that were dispatched by fbkc it will not show up in the LV. That is why when I see one knock event in the flkc table in the > 1 load/>3K RPM range I get a little concerned. It is possible that you are knocking more than 1 time being that if you knock in fbkc mode it will not show up in the LV.

 

Anyhoo...this LV looks good.

Awesome. Uhm, this might be a stupid question, but could a really dirty air filter be causing all the knocking I saw on the 93 map?

Were you running the stock drop in filter or an aftermarket? Either case i doubt it makes a difference but if you are running an oiled aftermarker filter, I suggest you get some MAF cleaner and clean the maf.

 

I cleaned mine out and I had to rescale my maf because it was so dirty..you won;t have to worry about rescaling your maf since you are running the stock airbox.

stock subie filter. stock everything on the intake end.

 

also, i'm really tempted to mess around with the throttle duty cycle table. the pedal as it is seems a bit too sensitive. makes moving around in the occasional bumper to bumper traffic a pain.

You could wait a few days and then try the 93 map again if you want.

 

Keep in mind that the difference between the 93 and 91 maps is very very small...especially at stage 1. I would just keep the 91 map on and buy 93 octane; it is better to lose the 5hp and have the extra safety margin.

yeah, true true. i definitely don't want to risk engine damage. and honestly, it's not like i'm speed racer here. also, my butt dyno is definitely not that sensitive.

did a datalog today. actually, i did a couple of runs, but i get this weird problem where sometimes under WOT, the data stops reading and the laptop loses connection with the ECU. it used to happen with the AP too and the only way was to unplug the USB connection and plug it back in. i thought i had fixed it with the AP and AccessTuner Race by using a different USB cable to connect the dongle, but now i'm getting the problem in RR using a vag-com cable.

 

maybe it's a EMI/noise issue. i notice that the new usb cable i have has a ferrite core. maybe i'll try to clip one on the vagcom cable. anyone else get this problem?

 

in any case, here's the log:

day 3 - stage 1 91 oct 3rd wot.csv

I occasionally have some freeze up issues at WOT with RR and tactrix..no clue what causes it.

 

Log looks clean..you have some learned knock at 5700 at very low load. I am guessing the learned pulled timing may be a result of shift knock. Nothing much you could do about that though...other than shifting slower.

hey, do you have an account at the RR forums? can you PM an admin there and let them know their new user registration system is broken?

 

also, does the tactrix cable have a ferrite core on the cable?

Sent a PM to a RR admin.

 

I no ferrite core on the tactrix...unless I am misremembering.