Lean @ idle

I ran my Daytona today for the first time since installing a wideband. Car runs stoich for 10-15 minutes at idle then it goes lean, reading 16:1. I tried unplugging the CTS and it drops the afr to ~13:1, so I am confident the gauge & wideband o2 are working. Computer stored code 51. Fuel pressure is set @ 35 psi, with +20 injectors, stock 2.5 MT SBEC. FP did not drop when the wideband went lean. Idle gets lumpy and erratic when it goes lean. Tach not working, so I'm not sure of the change in idle rpm. Narrowband o2 is new. Anyone have any thoughts?

Narrowband O2 bad? EDIT: I see its new, but maybe something went bad with the install or something killed it

Do you have any way of confirming O2 sensor is swinging back and forth as it should? Worst case you could probe its output with a voltmeter. Best case hook up a OTC/DRB/MPSCAN

Any codes being set?

Code 51 is being set. I have a narrowband that can throw in to see what the computer is seeing. What's odd, at least to me, is that it happened 3 times. Start, stoich, then lean. After I pulled then CTS, and it richened things up, I plugged it back in and it went back to stoich.

your fuel pressure is set too low plus you're running a set of +20's on a stock computer. put the stock injectors back in and reset the fuel pressure to 55 psi or get a cal to work with the +20's and reset for fuel pressure to 55 psi.
dialing back fuel pressure to try and compensate for bigger injectors will just cause lots of problems- especially at startup and idle. the computer will never give correct fueling. the fact that you're seeing stoich at idle on startup , then after a while it goes lean makes it obvious what the problem is. on startup the car should run rich- warmup enrichment- and after the car has warmed up it should run around stoich as the computer bounces back and forth between rich and lean while in closed loop.

any time you make a mechanical change to an engine you MUST reprogram the computer to work with those changes or the car will never ever run right.

stock ecu should be able to handle +20's at the stock pressure since it can correct +/- 25% from it's preprogrammed curves at idle and cruise. So do an engine mod that increases fueling need 10% then the +20's will be right in their sweet spot.

Thanks for the feedback. I ran this setup, ported head w/Super 60, in the past - albeit without the wideband - before tuning was commonplace with good results. I think the computer can compensate - it starts off rich while in closed loop, bounces around around 14.7:1 while in open loop for 10-15 minutes, and then goes full lean with no bounce.I will drop in the stock TII injectors and raise the fp to stock & see what happens, and then troubleshoot the sensors.

Keep the +20's in it, and turn the FP up. What's happening is that the adaptives don't take over right away, so the car seems to run ok, Then the adaptives start trimming or trying to add fuel, when they get to 25% it freaks out and goes full lean.

I had this issue when I went to a ported head, big plenum intake, tube header. Fired it up, ran great, 15 minutes or so of idling or driving and it would lean way past 18:1 and die. I had to increase the idle fuel by 50% over where I had it. Once that was accomplished it ran perfect.

Great. Thx for the info. That's exactally what happens to me, including the car eventually shutting off. Much appreciated.

if you have a scanner you can watch the adaptives kick in and slowly climb to 25% before it freaks out. It's how I found my issue.

Brief update. After adjusting the fuel pressure to 55 psi the car would run better, then would go full rich. After chasing the a/f for while with fp adjustments I got it close and started driving it. The car would run good for a few minutes, then go lean, vacuum would drop a little, and it would not want to go into boost and then miraculously get better. The condition got progressively worse until the car undrivable. As it felt like a MAP issue I bypassed the barometric solenoid and it seems to have corrected the problem. I was able set the +20s at 40 psi and the computer is able to achieve ~ 14.7. It's a little rich - ~10:1 @ WOT - but I'm still running stock boost. Car ran clean enough to pass the state emissions test. I think the barometric solenoid was sporadically sticking when it was cycled every few minutes creating a bleed on the MAP. I tested it and the solenoid sounded as though it was working, but the effect by bypassing the solenoid is undeniable.