Printable View
Has anyone found the best way to cure false knock? I have a feeling its probably because it's in an older car with solid rubber mounts.
Car is 1993 Fairmont with a high comp FG Engine(10.3:1) GT3582.15psi BP98 No matter what timing I put in it its getting KR through a whole pull. AFR is good at around 11.5
If I kick down to 2nd from a roll I get commanded. A pull from a stop it has KR.
I'm thinking of Halving the sensor gain is this the best way to sort it?
I have found when I was running E85 and experimenting with timing that the only way to get rid of the false knock was to disable the knock sensor.
I wouldn't try it on 98. 15psi is a lot of boost for 98. I don't know how people get away with big boost on 98.
I got HPTuners to add the TipIn retard table to my strategy and I set all cells to 0. Although I wouldn't recommend that unless using high octane fuel.
If you set the IAT and ECT multiplier tables to 0 you will notice a reduction in knock degrees. Its the same if you zero out the VCT borderline and lambda correction tables. I have noticed when the ECU logs knock retard through VCM Scanner, some of that value is the corrections from the other tables.
If you log spark source it may help you pinpoint the problem.
When I run united e85 and 19-20psi boost I usually run 17-18 deg total timing. If you look at the MBT timing table that should give you a rough idea of how much timing you can run without being octane limited. Now compare those values to your actual ignition timing and see what you are missing out on.
You must be only running 4-6 deg total spark with 15psi or so?
Just my observation!
FG Engine 10.3:1.15psi BP98 AFR 11.5..4-6 deg total applied spark. on road loading @ maximum torque area........sounds like it is in the ballpark!
FG Engine 10.3:1.15psi united E85 AFR 11.5 (petrol scale)..17-18 deg total applied spark. on road loading @ maximum torque area...... way too much!
That brick wall you hit with high static compression and boost.
Quote:
Originally Posted by Yortt
Interesting...
I am using a BA NA motor 9.3:1. E85 AFR 11.8:1 18-20psi 15 total and ran 123mph at willowbank
Got greedy and tried 18 deg timing and gained 2 mph (11.1@125mph), that was with 70 deg inlet air temps over the line 5950rpm in 3rd.
You gotta be careful cut/pasting timing maps when you aren't using the same engine as the ecu.
The MBT ignition timing for a xr6 turbo is for 8.7:1 (I think) so when using those maps with a higher comp motor you still need to take more timing out to suit. I compared the timing on a NA motor to XR6 turbo below 100kpa and the low comp turbo motor runs a bit more advance.
Sometimes too, I've found not to overthink it too much. I've got away with murder on e85 haha
Obviously to optimise the ignition timing under boost we must know exactly where the knock threshold is particularly on 98, however this still applies on E85 under moderate boost it will knock before optimum timing is achieved, with the GM knock detection system I am satisfied the system calibration is accurate this appears not to be the case with Ford.
As with GM my original approach to Ford was do not touch the factory knock sensor settings, however with the falcon knock detection (FG XR6T) it was over sensitive. I had a lot of problem with low level knock at low load 0-3000rpm on 98. After driving fuel and timing up to improve low end response and after confirming it was false knock I desensitised the knock sensor settings which eliminated the knock this made a noticeable difference in drivability and smoothness in this area, the top end knock sensitivity appears to be ok.
I think "overthinking" is good, provided it is tested with extensive logging particularly at the track where it is under actual operating conditions.
Hey thanks for the replies fellas.
I have made adjustments to get commanded spark, lambda table , iat correction , ect correction etc. at normal temps IAT 25-65 it will run commanded and normal coolant temp its running 0 correction etc, high load i've copied MBT and BLDN. (probably pointless but old habbits die hard)
I'm at the point where I might just ignore it. lol it still goes really well. I did a test where I adjusted all timing on boost to 3 degrees. still had knock retard as soon as I went WOT.
Ill go to willowbank too see where its at soon anyway.
heres the timing table.
Attachment 84767
Car has usual mods. 60lb injectors. 4" dump 3" exhaust custom cat ;) pac springs, pump gears, head studs. 3582. FG inlet manifold. 4 Speed auto 3.45 gears.
Also since changing to a BA ecu strategy HAANFHZ. VCM has next to no scalers to log.....
I ended up logging with Forscan. Can see more :/
Only way I calculated knock retard was by working out load point vs set timing at the point and max knock retard allowed = the timing it was logging.
Wish I could log KR better to see the exact point it does it. Used to be able to on the BF ecu. Can't see jack now :(
Why don't you just switch to E85/E-flex?
I remembered when I was going through all of this tracing false knock earlier this year, I did an experiment on petrol on std boost (about 5.5psi), logged ignition timing on a short drive around town then out to the back road for a 200m sprint.
Then I ran the tank out to almost empty then filled with e-flex, changed stoich value to suit then did the test again on e-flex with 5.5psi and the knock retard and timing I logged was the same! No difference. I thought theres no way this thing is pinging/knocking with e-flex on stock boost!
I found some other strategies on the repository and compared the knock sensitivity tables. I cut/pasted those from a F6 typhoon map. Didn't really notice much difference to be honest.
I tried them at 100, I tried them all at 0. Tried setting knock retard limit to 1 which helps. Only way I could find to truly get rid of the knock is to disable the knock sensor.
I think if you did this on 98 at 15psi you could say goodbye to your engine though.
If you haven't already heard it, have a listen to the power and speed podcasts when they talk to Greg Bannish (the master ecu calibrator).
http://readingtondigital.libsyn.com/...brated-success
http://readingtondigital.libsyn.com/...brated-success
All of this shit ford puts in the ecu's are their for a reason. We are fools if we think we know more than Ford. I am especially guilty of this myself. I admit though that by experimenting and learning by my mistakes is a good way to learn.
I've only recently switched from E-flex 12psi street tune back to 98 and stock boost 5psi. I'm tired of high fuel prices and 19-20L/100 city driving LOL.
I have left the ignition side of things all totally stock settings. I've only done the basic dialling in of the injector slopes.
I'd go e85 but its a daily. So not an option. Thanks for your input man. I'm going to just leave it. Goes great so I'll accept it as it is lol
PCMtec has a custom operating system that will allow you to run an E85 sensor this will adjust the custom tune automatically for any level of ethanol content in the fuel. This solves the everyday driver problem.
Yeah I've looked into that Yort. It's more that the car does so many k's if I run e85 ill be doing monthly services and filling it up twice a week.
I'm very content with the power its making on 98 anyway I'd just never had false knock issues before which got up my goat lolAttachment 84787
This is her. Understated :)
When I chased my knock issue down checked every bolt on the engine checked all the intercooler pipe work not torque toughing anything, changed the exhaust system, engine mounts and knock sensors made no difference. Reduced ignition timing by the amount of knock plus, still no difference, concluded false knock, decensitised the knock sensors in small increments... knock gone.
Generally speaking you can sometimes tell false knock by the knock trace on the logger if it knocks and decays as it recovers chances are that is real knock. If the knock trace bounces around erratically it more than likely it is false knock, not conclusive but an indicator.
Yep i agree if it is a daily driver 98 is the best way to go. .........By the way a nice little sleeper there!
Hey thanks mate :)
I did actually halve the knock sensors gain and it did dissappear. But being on 98 at that boost I didn't want to risk it not picking up a real knock event from bad fuel etc. So I decided to put it back to stock :(
Sadly I can't actually log KR with this BA ecu. VCM Scanner has very little PIDs available for scanning. I can only log actual spark and do calculations to determine it's from knock retard. Really sucks. BF ecu I could log knock fine and also didn't get knock not even at 8 degrees at 4500! But Ill play this safe. Goes plenty well enough :)
Thats a great tip! Thanks for sharing that one, handy to know.Quote:
Originally Posted by Yortt
Thats a good idea. Sometimes I think its smarter to leave a little power on the table for the sake of reliability.Quote:
Originally Posted by edfairmont4.0
I originally had a BA NA ecu in my ute for years before I decided to upgrade to BF. (mostly because I was sick of the long wait reflashing)Quote:
Originally Posted by edfairmont4.0
I couldn't log knock with it either. (HAANGF5)
I've always had dramas with both ecu's with hardly any PIDS to log.
I think the BA only had 19 basic pids on the left column, the BF ecu I'm using now has 23 (I think)
Theres a ton of pids to log under Graphs layout and parameter selection.
I've emailed support about that to see if they can get more PIDS for me but I've had no luck. Tried repolling several times with no improvement.
well to revive a semi old thread.
ended up getting the shits with being down on power lol
moved ba F6 knock settings into my tune and reduced the knock sensor sensitivity about 40%
false knock is now gone. car feels FAR stronger and smoother through the rev range. for added safety ill just throw octane booster in it for the drags JUST in case I get average fuel.
Timing table I finished on. Running 15psi max. Stock FG gate with max duty cycle above 4k lol Attachment 86011