I guess this is more a question for LD9 pcm swapped Eco guys. We are working on Evilution's M62 LK9 swapped cavalier running a 2000 LD9 pcm with 2-bar reflash. The car has a 3.1" pulley, ZZP S3 heat exchanger, dual pass, LS1 TB, full exhaust, 60# injectors, and racetronix pump.
We have the fueling for low load/low rpm pretty much dialed in and the car is very "drivable", but as soon as we get into any kind of positive pressure, the knock sensor goes crazy. We are running 11.8:1 afr to 3200 rpm then dropping to 11.5:1 above that. At as little as 2-3 psi, we are getting 10* KR. Even after globally pulling 13* advance we still get the same KR. At this point the total advance was -10* in some cells. Now, the trigger wheel likely is not dead on perfect, but there should be no reason it is off enough to warrant the need for pulling that much timing and not improve things. If it was my car, I'd plow through the gears at WOT just to see whats going on, but it's better for now to ask questions lol. I've read that LD9s are known for false/phantom KR, and that people have desensitized the base knock retard to overcome this. I always assumed that this was a property of the motor, not the PCM. However from this experience, the PCM/OS seems to be guilty of the false knock. FWIW, there is no audible pinging/knocking even when the KR is 10*.
The LK9, essentially being an LSJ, should take boost and timing like a champ; LSJ M62 guys typically run 20+ degrees advance on 93 pump and a reasonable pulley/cooling situation. Any input from LD9 PCM swapped guys or even LD9 guys is much appreciated. If it would help to see a tune and/or log, I can make a thread on HPT forums too. Thanks guys!
A few questions.
what year LD9 harness?
Have you done any changes to the knock sensor settings?
We did not get any or much knock on Roberts cars, but I'm going to say his is not the best to compare to.
I have seen 6 degree's of knock retard on stock LD9's, and taking timing across the board not make any difference. This includes my car. After making some changes to the knock sensor settings it really helped.
FU Tuning
I believe it's a 2000 harness and it's definitely a 2000 pcm. I have not changed the knock sensor settings from the stock reflash settings (I assume you mean the base retard and baro/ECT modifiers). I bumped the maximum PE retard like I usually do on boosted cars to 10*.
So people usually desensitize the sensor via the base retard table?
I used the knock fast attack rate vs. RPM. This was with Ryan's help.
FU Tuning
wahts an LK9?
I choose the technology built for land speed records... not the technology built to save on emissions and gas mileage.
Z Yaaaaa, it's a saab 2.0 ecotec.
Addicted to meth wrote:I used the knock fast attack rate vs. RPM. This was with Ryan's help.
How much did you reduce it? I see the LD9 reflash goes from 0.375 at 800 rpm up to 0.625 at 3600 rpm to redline.
It was a percentage that Ryan told me to do (I do not remember what
, but it is .174 all the way across.
FU Tuning
Addicted to meth wrote:I used the knock fast attack rate vs. RPM. This was with Ryan's help.
Mine and Brads are also "turned" down using this. Brads and mine are I think 25% less sensitive, and I think the file John sent me was 35-40% less sensitive.
PRND321 Till I DIE
Old Motor: 160whp & 152ft/lbs, 1/4 Mile 15.4 @88.2
M45 + LD9 + 4T40-E, GO GO GO
I can double check if you like sometime.
PRND321 Till I DIE
Old Motor: 160whp & 152ft/lbs, 1/4 Mile 15.4 @88.2
M45 + LD9 + 4T40-E, GO GO GO
That would be great! I've logged ONE stock LD9 and it showed KR all over the place, so the more data and resolutions, the better. The fact that pulling that obscene amount of timing didn't help (and made the car slow and sluggish almost to the point of misfiring) kinda confirms that it's false.
My biggest concern with the car now, being boosted, is to dumb down the sensor enough to eliminate the false knock, but still be able to pick up real events.
I redid to show what I did.
Brad I do not remember how much we removed on yours....
Mine was pulling the most timing from 4000-to about 4500 RPM. It was knocking about 5-10 no mater what I did there. Even with 110 unleaded race gas. P.S. I love $7 plus a gallon for that...
Changing mine to 25% less killed 95% of it. Every auto LD9 I messed with knocked somewhere between 3500-4500rpm in 2nd, and a little less in 3rd.
John's varied from about 35% less on the 800-1600 RPM to 41% on the 3600-6400 RPM. The center was like 38% or something like that. I'm sure Ryan did more research and testing than I did.
Edited 5 time(s). Last edited Monday, October 29, 2012 8:33 AM
- Attachments
- stock99zknockattack.jpg (94k)
stock99zknockattack-25percent.jpg (87k)
stock99zknockattack-40percent.jpg (92k)
PRND321 Till I DIE
Old Motor: 160whp & 152ft/lbs, 1/4 Mile 15.4 @88.2
M45 + LD9 + 4T40-E, GO GO GO
That is a big help - we'll start with minimum reductions on Matt's. It's amazing - i think that table is listed as a min/max of 0-32* yet such a small change can potentially correct the problem.
And actually, Matt's file has values from 0.375-0.625 stock (as opposed to a max of 0.750 in your first file). Again, it is the 2000 LD9 reflash from the tune repository.
Is it the one I uploaded?
PRND321 Till I DIE
Old Motor: 160whp & 152ft/lbs, 1/4 Mile 15.4 @88.2
M45 + LD9 + 4T40-E, GO GO GO
Snowman wrote:Any updates?
Not yet, we haven't had a chance to get out and tune this week so far. Damn hurricanes.
oops lol it hadnt crossed my mind, till you said that.
Snowman wrote:any updates?
We'll be tuning tomorrow morning
We did some tuning yesterday morning. The first flash I dropped the table by 5% and there was zero knock retard, although timing was still very very low (single digits!). I think it may have been a fluke. BTW Matt, I'll text you later, but I think we should go back to the first flash we did yesterday and test it out more. If it was a fluke, it'll knock, if it's not, I need to figure out why that worked and dropping it back didn't.
After trying to ease the table back (4% drop, 2.5% drop), we get nothing but knock city no matter what. I dropped the table down to almost nothing (~80% drop), and it still behaved the same way. Once it crosses into PE and about 110-115kpa, it lights up the KR to 10*. At this point i think matt is going to swap the knock sensor out and we need to recheck the timing of the trigger wheel.
Sure your not getting any cross talk of anything between sensors?
PRND321 Till I DIE
Old Motor: 160whp & 152ft/lbs, 1/4 Mile 15.4 @88.2
M45 + LD9 + 4T40-E, GO GO GO
Not much to report since last time...i swapped in a new knock sensor just in case mine was old age affected, and we discussed that the trigger wheel may not be dead on...i never took into consideration the difference in stroke with the 2.0 and that the pistons stay at tdc a bit longer...i had the wheel set at 0* right as piston 1 hit tdc...not realizing that the duration of tdc equaled about 1.25" on the trigger wheel which looks to be just shy of 20*.
I haven't had a chance to log anything with my Torque app til Ryan and I can get together with hpt, but i did take a quick trip around the block and punched it to about 10-11 psi...felt better and didnt hesitate so my butt dyno says retarding the wheel a bit seems to have made a difference. Im gonna take a short trip tomorrow with Torque and see if i can confirm any of that.
Of course the car also decided to start munching on a wheel bearing last weekend, so i want to change that before we do any more tuning.
Yea, kind of a wild card in this equation - the trigger wheel and timing an LK9/LSJ to TDC. With respect to spark advance, is "0" when the piston first hits the highest point in the cylinder, halfway through the rotation at the highest point, or at the end of the highest point. It sounds like there is a lot of room for variation, although in theory a shorter stroke should mean less dwell at TDC.
Clearly though, it is a little less forgiving than an LE61 hybrid lol.
Update for those interested...
I started to think maybe the trigger wheel's adjustment was off a little bit...or figured it was at least worth a shot, so I clocked it back to retard it just a little bit. Logged it with Torque right after and saw maybe 5-6* KR max while boosting to 10psi (previously saw an immediate 10* KR at 1-2 psi). So seeing a step in the right direction, Ryan and I headed out to log and tune last night, and after he changed the knock sensor's desensitization more, I'm now seeing 1* KR or less at full boost.
So I think it was a combination of the wheel slightly off (definitely seemed more forgiving on Ryan's Le5 bottoms) and just finding an acceptable adjustment for the knock sensor. We still have a ways to go, but it was definitely a huge step forward. We still have some playing around to do with the knock sensor, it's severely desensitized as of now, so we want to see if we can pull it back little by little and still not get such high false results.
Timing is also greatly reduced right now while were experimenting, so I guess next time we go out, we're going to start adding some in and see if we can do so without reflecting more KR.
It's a shame the LD9 PCMs are so touchy with the knock sensor communication. It's great having a speed density OS available, just sucks having the trade off being rolling the dice on how accurate you leave the knock sensor.
John, Robert used a 2200 pcm on his build, correct? Did you have to scale back his sensor at all or did that PCM seem to play more nicely? I'm curious if the false knock is limited to the LD9 pcm or if also the 2200...I'm assuming they are the same PCM/OS from factory, just calibrated differently to suit their respective motors. Makes me wonder if the problem is hardware or software related. We did notice that the sc reflash already has the knock sensor desensitized in comparison to the stock tune, which makes me think it's hardware related.
Glad to hear you got it working a little better. It sure seems like a hardware issue if GM decided to desensitize it for the sc flash anticipating problems. Really wish I was down there more often to hang out when you guys do this stuff. Anyway, keep us posted on progress. I'm definitely interested in a swap like this. Plus, if you keep doing all the leg work and pioneering for me I can just bring my car over and have Ryan square it all away in a few hours
"In Oldskool we trust"
