PDA

View Full Version : Burst knock....B6212



hquick
January 18th, 2008, 07:13 PM
Well......
after alot of reading...again :Eyecrazy: I decided to try '0'ing all the values in table B6212 to see what effect it would have on the amount of KR I have been seeing.
Surprisingly..it made a huge difference.
Previously I have been getting KR of 8.0 or so for no apparent reason.
VE was good, out of boost, no serious changes in rpm.
Once B6212 was '0'd most of my KR disappeared and now I have a new feature....'KR on command' :) .
Basically...if I stab the throttle I can generate KR....so atleast I know where to start looking now.
On that note...would adding more fuel via B3647 be the way to go?

From this:

http://img.photobucket.com/albums/v35/hquick/B3647b.jpg

To this:

http://img.photobucket.com/albums/v35/hquick/B3647.jpg

cmitchell17
January 19th, 2008, 04:30 AM
I would also get knock with the stock tune with burst knock.

I zeroed it and now I can't even try to get knock. Maybe its becuase my truck never warms up to temperature (I think thermostat).

With burst knock on it would give me random 2-3degrees knock and then shifting into second it would pull like 4-7degrees for no reason.

And anytime I hear audible knock in the knock region(4000-5000rpm) it never shows up on the scanner?

hquick
January 19th, 2008, 10:37 AM
That's exactly what was happening with mine.
Here you can see I was getting KR with no apparent reason behind it.

SV8346
January 19th, 2008, 12:51 PM
i was getting a very similar knock with mine untill i started playing around with commanded fuel vs RPM. I found that getting the fuel mixtures correct for the given load was the trick to it. Also what might help you a little is to set your lean cruise all to 0 then create lean cruise within your commanded fuel vs RPM map.

What i have is from 20kpa-65kpa in the 1600rpm-2400rpm range I'm running 15.8AFR. That way when your just light throttle cruising it isn't dumping a bucket load of fuel into it. Excess fuel can create knock for no apparent reason.

Tordne
January 19th, 2008, 02:37 PM
In your last image you will see that the KR line comes in high, then decays, and then spikes up again. This generally indicates that the knock is real, as the timing is increased the knock sensors trigger again.

I used to get a lot of KR on throttle transition. Initially the few things I did were:
- 0 out the Burst Knock table(s).
- Installed a PCV catch can. This stops the oil from the crack case venting back into the throttle body and diluting the charge.
- Change to a spark plug one heat range colder.

As SV8346 says, I also believe that too much fuel can induce KR, as can too little fuel.

hquick
January 19th, 2008, 02:56 PM
Thanks for the suggestions guys.
I'm already running a temp colder on the plugs (as recommended by most running a supercharger...and new installed a few weeks ago).
I was unaware I was running lean cruise. As my truck uses the US PCM I didn't think it was possible?
Here's my current tune and the resultant log. If anyone can please have a look for me and tell me what they think..that'd be greatly appreciated.
So far I've run more than 100 logs in the past few months trying to sort this out.

Thanks

SV8346
January 19th, 2008, 03:25 PM
ok, no your not running lean cruise but you have got it set to closed loop for while your cruising. You also running very rich in the top end of the B3647 map, AFR's at some points are going to 11.1 on your plx while on small amounts of throttle and has a maximum of 10.0 AFR, that is way to rich. Also a few other things i have noticed, what injectors are you running and what fuel pressure reg are you running. I dont think your injector flow rate table should be flat like it is, it should be referenced to manifold vac/boost.

i have create a quick edit of you tune, give it a go and tell me what its like.

hquick
January 19th, 2008, 03:30 PM
Thanks for that....I'll try it.
I'm running Bosch 42lb green top injectors in the mercruiser marine manifold.
I'm running the stock reg...so..58lb fuel pressure.
I was of the understanding that as the reg is manifold ref'd it should be a straight line?

Tordne
January 19th, 2008, 04:03 PM
I was of the understanding that as the reg is manifold ref'd it should be a straight line?

Correct. That opened my eyes the first time I saw it :eek:

SV8346
January 19th, 2008, 04:12 PM
yeah as tordne said, that is correct. As the reg is manifold ref'd it should be stright lined.

hquick
January 19th, 2008, 04:18 PM
Thanks again guys,
I tend to get audible ping around 3000rpm so that's why so rich....from memory...trying to combat that. When audible it doesn't show as knock whilst scanning?

SV8346
January 19th, 2008, 04:22 PM
you could be getting audible knock from allready having to much fuel in there, so by adding more fuel will only make it worse.

hquick
January 19th, 2008, 10:44 PM
Thanks again,
alot of this is new to me so I'm learning and trying to understand the whole KR thing.
I have noticed in the past that my AFR's have hit 10's and less (showing RCH...rich on the PLX guage) whilst under accel, high boost and load.
I'll try this tune in the morning and report back with a log and details.
Thanks

hquick
January 20th, 2008, 12:10 AM
Just out of curiosity...should I be looking at B6245?

http://forum.efilive.com/showthread.php?t=3036&highlight=B6245

Remembering that my Truck WAS running a 98 PCM before the upgrade to the 0411 and COS3?

hquick
January 20th, 2008, 07:33 AM
Hi Guys,
SV8...I ran that tune this morning.
Attached is the resultant log.
You can see that when I ease off the pedal I get AFR values of as high as 19+.
I also heard a fair amount of pinging (light crackling). Also knock was still present as I stabbed the accelerator.

SV8346
January 20th, 2008, 09:27 AM
Yes on back of while still moving at speed your wideband will register very high AFR's, that is the DFCO working. I think your VE tables still need a bit of working to, By looking at your Main VE Table-PLX wideband BEN averages you still have a lot that are out.


After having another look at your log your DFCO's dont seem to be working, i'll have another look at the tune this arfternoon and see if i can find out what is wrong.

hquick
January 20th, 2008, 10:10 AM
Thanks again.
I am using a roadrunner and ran a couple of hour long logs to try and auto tune the VE tables.

Thanks

SV8346
January 20th, 2008, 10:51 PM
been having a look through the tune i sent you and your original tune and cant seem to work out why your getting the knock. Can i ask how your keeping your ECT down to low 70degree temps, personally i think thats to cold. Mine definitely seems to run smoother and creates better horsepower and torque at 89-91degrees. I hope that for now you have uploaded your original tune till i have more of a look at your tune and work out whats going on with it. But i still definitely think your original tune is way to rich.

hquick
January 20th, 2008, 11:07 PM
It was a cool evening when I logged that data and it was freeway driving.
I have a 160 degree T/stat installed.
On the hot days we had recently I was mostly around 90oC and in traffic got to 99 or so.
Yes....I took a step back and installed the previous tune.
Thanks for helping me out..this is driving me crazy.

SV8346
January 21st, 2008, 12:09 AM
ok here is a couple of questions that i'm not sure about(not to familiar with the holden suburban)

Q. Does your engine only run 1 Knock sensor? I was under the impression that all LS1's used 2 knock sensors.

Q. how much boost are your running.


try the following tune and tell me how it goes, i have altered allot of the knock sensor settings.

hquick
January 21st, 2008, 12:14 AM
Hi,
Yes I only run one knock sensor. Motor is a 5.7L Vortec L31.
I'm running at most about 7psi..I think. Most Kpa I've hit is 140kpa manifold pressure.
Motor is non-intercooled.
I'm running 98 octane petrol
I'm running NGK TR6 (standard copper - 1 step colder than stock) plugs as suggested by 'everyone' running the Whipples on the Chevy trucks.
Thanks

SV8346
January 21st, 2008, 12:18 AM
ok dont use that tune..........yet

the setting are wrong for single knock sensor

SV8346
January 21st, 2008, 12:20 AM
ok tune updated, if you have allready downloaded that tune then change B6201 to 1 not 2 sensors.

hquick
January 21st, 2008, 12:23 AM
OK...thanks

SV8346
January 21st, 2008, 12:25 AM
will you be able to test that tune tonight or would you rather wait till tomorrow

hquick
January 21st, 2008, 12:29 AM
Just going through that tune...wow! You've been busy.
So as to assist me in learning...what exactly have you done and why?
I mean...I can do the compare...which I just have...and I can see the desensitizing of the knock sensor but how do you know how far to go when adjusting these parrameters? If you don't mind explaining?
I'll run a log in the morning and report back.
Thanks again

SV8346
January 21st, 2008, 12:35 AM
ok alot of the settings i have used are from the 2002 camaro factory tune which is the same op system as what you are using, i noticed that in that tune they had allot of the knock sensors turned up really high. So it has made me think that if they can get away with running the sensors fairly high on them why cant you on your tune.


Its a bit of a long shot but its worth a try. Boost tuning can be VERY tricky sometimes.

hquick
January 21st, 2008, 12:41 AM
Thanks.
I'll run the tune in the morning. I've been up since before 4am and am likely to be doing the same tomorrow (on call). One good thing about being on call is the early morning logging sessions are free from other vehicles on the roads.

Thanks again

SV8346
January 21st, 2008, 12:45 AM
well i'll be back online from 6.30>7.30am, just go easy to start with and listen to the engine very carefully it can be risky turning knock sensors up high. If you can hear any audible knock at all then back off and go easy and post up the log in the morning when you get time.

hquick
January 21st, 2008, 08:31 AM
Ok...here's te log I just took.
I marked the area's I'm pretty sure I heard 'pinging' with a '1' so if you're in the dash B page you'll be able to see it as a dotted line by pushing the number 1. There's a few spots so if you keep pushing it, they'll all scroll accross.
Thanks

Oh....my WBO2 didn't want to work at first...that's why it's at 10 in the first part of the log.

SV8346
January 21st, 2008, 09:11 AM
ok, i relise your using roadrunner and your doing autoVE, but you VE table is still out by a fair bit, If you have a look in your scan tool, goto to your maps(F11) then open a new map(SHIFT+CTRL+M), then open the map "B0101 Main VE PLX wideband BEN" then also open "B0101 Main VE PLX Wideband 2Bar BEN". With both of those maps click on hide empty or low cell counts(CTRL+ALT+H). Those figures need to be all 1.00 the further they are away from 1.00 the further your VE table is out.

On the Knock side of things i have altered the prg, wht i have done is started to increase the sensitivity of the knock sensors. Let me know how this one goes.

hquick
January 21st, 2008, 09:20 AM
Thanks again.
One of the other members on here (AussieLS1evil) loaned me his RR (in the Burb at present) and I used it to do the AutoVE thing. Even with the RR I still seem to be having trouble getting the BEN's in line. I have that map opened up and I can see what you mean....I just can't seem to get it right.
Everytime I've had KR...I've never heard it (inaudible). This morning with the small amounts of 'pinging' (audible) it never registered. Actually....I've never seen 'audible' knock register in any of my logs.
Thanks again for all the help you're giving me.
Seems my Burb is somewhat of an anomaly for many people :Eyecrazy:

Howard

hquick
January 21st, 2008, 09:25 AM
Oh...I just changed the name of the file to 'temp2' so as I don't get confused.

Thanks

JezzaB
January 21st, 2008, 05:39 PM
Howard. The reason your Commanded and Actual at 14.63 is going up and down on the PLX is because it is now in Semi Open loop. Its the O2 sensors cycling. You VE was bang on so we put it into Semi Open ;) Thats why you can see the STFTs now.

You will have to make them 14.5 (instead of the 14.63) again if you feel like you need to dial in your MainVE table again but it was very accurate before it was put into Semi Open Loop.

Jez

hquick
January 21st, 2008, 05:47 PM
Thanks Jez,
I don't understand why I get such rich lean variations on large cell counts.
Is that also because of the 'semi open' setup?
SV8346: Here's todays log after running the above posted tune.
I heard small amounts of ping and once again marked them with the '1' key.
Unfortunately the freeway was jammed up with morons doing 90 in the right hand lane :bash: so I couldn't really kick it down to check thiongs. I did try a couple of short bursts to get it to knock.

Thanks

SV8346
January 21st, 2008, 09:01 PM
that is very weird that it is now detecting 2.2 and even less in KR but it wouldnt detect the audible knock that you have marked. It has me beat as to what is cause that knock but one thing i can see is that in most place where you have mark it, usually a couple of frames after you mark it the PLX AFR is allot richer than the commanded AFR. How far down the exhaust pipe is your PLX AFR. There might be a delay from the Commanded AFR to the time the PLX AFR reads it.

hquick
January 21st, 2008, 09:05 PM
The PLX O2sensor is right next to the standard O2 sensor.
Here's a pic of it's location.

http://img.photobucket.com/albums/v35/hquick/4L60e%20to%204L80e/P9290010.jpg

SV8346
January 21st, 2008, 09:10 PM
thats should be perfect. This has really got me buggered as to what it could be.

hquick
January 21st, 2008, 09:13 PM
Lol! Story of this truck and it's tune.
I knew being 'different', 'unique'...etc wasn't all it's cracked up to be.