killav

I recenlty switched to 30lb from 24lb injectors and have been slowly working on the tune. To date, I just have not been able to get the car to run as smoothly at all RPM's with the 30lb injectors. But I am getting really close.

So, last night, after going on a short 30 min. drive and tune, I pulled into the driveway to do my normal saving and closing everything down before I shut off the car. At this point, the car starts running pretty rough, and I see that the A/F ratio has gone lean, around 17.5 to one. It has never done this before. The weird thing is, the O2 sensor adjust is all the way at -20 and just staying there, causing the lean condition as shown by the wideband reading on the ST screen. The LH should have been adding fuel, but it was pulling out the maximum amount of fuel causing the car to run lean. When I cracked open the throttle, everything would return to normal, with the O2 adjust hunting right around stoich like its supposed to.

Some things to note:

1. I only have one O2 bung, so I must use the NB simulator from the LM-1/ST setup to feed the LH that data.
2. I shut the car off, re-started and went for a drive. I didn't change anything, and the car was idling normal after that.
3. I drove my car to work today-aprox. 23 miles in heavy traffic, and it ran fine, no lean condition at idle or anywhere else.

I have a spare O2 bung that came with my turbo kit that I never installed. I could drop the Y-pipe, drill a hole and weld in the second bung so that I can run my actual NB sensor to the LH instead of the simulated signal.

Thoughts?

One more thing, not sure if a intermittent failing idle switch would cause this funkyness, but when data logging on the way to work, I was seeing "1" at idle like I would expect to see when completly letting off the gas.

TIA

John Speake

Mike - did you optimise tune with using the SharkPlotter program, or the ST2 Autotune ?

After tuning, during which you should have inhibitted O2 adaptation in the Fuel Parameters screen, you then re-enable and drive the car for a few miles for the O2 and idle adaptation to fully complete.

blown 87

For what it is worth, I have had issues in the past trying to get the NB sim to work with the LM-1, but on a 928 it seems to work well.
But something does sound strange, a neg 20 at 17-1 it should be trying to add fuel, not take it away.

What kind of knocks are you seeing and what knock sensors are you using? the gold plated ones or the stock ones?

Good luck.

killav

John,

I am to the point where I am strictly using the sharkplotter and data logging. No autotune anymore as I am pretty close and just fine tuning things. And no, I did not re-enable the O2 adaptation last night after I shut the car down and re-started, as I continuted to log data last night and this morning. I will enable the adaptation here at work and drive the car home today. It only did it that short time and has not surfaced again.

Greg, knocks are fairly low. I do get some, but they are at heavy load, with lots of boost. For example, I had about six knocks on the way to work today in 23 miles with a few spots to really open her up. They were all centered at 3000 to 3500 rpms and only backing timing out on a few cylinders here and there ranging from 1.7 to 2.7 degrees of retard. Cylinder 7 seems to be more sensitive to knock than the others for some reason. I might get maybe three cylinders backing timing out just sligtly, but cylinder 7 always has to back more timing out than the other two. I never see knock in all 8 cylinders ever. Just two or three. And almost always between the torque peak area. I have the timing backed all they way down to 10.1 degrees at that point on the ignition map.

blown 87

I have some thing going on with mine, it now has the gold plated knock sensors and with as little as 3 degress of leed I am still getting knock counts, and with no other signs, no audible knock, plugs look good I have to be getting false counts.
I had so much timing taken out of it once that a POS 90's four door Honda just flat out ran me pulling from a traffic light, I went back home and put the stock tune in it as far as the ignition goes.
I am also using a EZK out of a 89, but that should have no impact on it I would think.

Sounds like you are getting close with yours, bet that thing will really haul now.

killav

It has been so friggin hot here that I have gone on strike and just refuse to go out in the garage. We are up to 22 or 23 days in a row over 100 degrees - basically all of July and no end in sight. Plus, I think almost half of June was over 100 degrees. I have seen triple digits here regulary in the middle of September.

I still have the methanol kit to install. That should really wake things up and allow me to be a more aggressive on the timing. Trying to get no knocks at all has only served to make less power than I should be making like you said.

Edit: My knock sensors are stock, but were replaced right before I got the car in 2006.

Double Edit: I know Andrew Olson was having crazy knock counts but I don't remember what the solution was.....

John Speake

I wouldn't think non matching knock sensor connector material should cuase a problem. Long term there may be some potential corrosion risk. Roger Tyson stocks the mating gold female connectors if you want to match your cableform to the later sensors.

It maybe we are being to careful with allowable knocks ? IIRC the old limit Porsche used with their diagnostic tester knock count routine was less than 50 knocks per 10,000 engine firings. i.e 2500 engine revolutions.