blue86iroc

iTrader: (2)

I just can't seem to get this right. Last year I had my idle tuned "well enough" -- I eliminated any richness-induced stumbling due to the oxygen sensor picking up fresh air from my (small) 208/221 cam, but I still thought the idle was a bit too rich overall. This year, I'm trying to get it more in line... at least as much as possible with my MAF.

I lowered the oxygen sensor thresholds down to the 320 mV range to try to pull out some fuel during closed loop idle. BLMs went down as suspected, but the injector PW remained relatively constant between 1.85-2.10 ms. It's the same when the BLM was at 128 and when it dipped down to 108. LV8 seemed to climb as the BLMs decreased, so I'm not sure if that had an effect on the PW (it shouldn't, though, since LV8 isn't part of the calculation). MAF airflow was reading in the mid-upper 9's at 116-110 BLM (high, IMO) but went as far as 11.0 at 108 BLM. Shouldn't this be decreasing if the ECM is trying to remove fuel?

And yet it still smells like gas. Not enough to make your eyes water, but the odor is noticeable. I'm running somewhat higher timing at idle (26°) because the engine seems a bit more stable. My Hooker headers are around 550-600° F at idle, which seems hot. Any ideas on what I can do?

Grumpy

How is your O2 sensor seeing fresh air, from the cam??????????

Lowering the (lower) threshold is going to force it richer.

My EGTs at idle are 800-900dF.

Might try pulling 5d of timing out at idle and see what happens.

Adding timing, will generate a slightly lean error, so then you wind up adding what turns out to be excess fuel (in some appls).

The MAF PW cal is about how much air is being used by the engine. So if the engine *thinks* it's using 11 gm/sec instead of 7, then the idle will be richer.

The LV8 cal is done all the time, it's used in the timing routines.

blue86iroc

iTrader: (2)

Some of the intake charge appears to pass through the exhaust, unburned, at low (idle) speeds. The oxygen sensor picks up the oxygen, thinks the engine is running lean, and adds fuel to compensate. Many of us have adjusted the oxygen sensor constants (lower, upper, and threshold) to prevent the extra air from interfering with the sensor at low engine speeds (low airflow). It really does seem to work. I've lowered all three oxygen sensor thresholds at the 0 and 8 gm/sec airflow in equal amounts, which effectively lowers the BLM that I see (went down from 136 with last year's calibration to 128, then to 108, etc.). It's just that injector pulse width doesn't change for some reason.

I'll try lowering the timing like you suggested.

blue86iroc

iTrader: (2)

Well, I might be on the right track... I fussed around with the fuel settings some more and I think I know what the motor wants. Maybe.

I disabled closed loop, locked the BLMs at 128, and reset the ECM. Initially I lowered the timing (first to 20°, and then I used the stock calibration of 16°). Injector pulse width rose as timing advance decreased, as did MAF airflow. The smell of fuel was still there and the idle was quite shaky. The whole car was vibrating, at this is at 750 RPM.

Next I raised timing to 24° total advance and decreased one of the entries of MAF Table #1. Injector PW ranged between 1.90 and 2.05 ms with header temps around 450-500 °F. This seems to be the sweet spot... the fuel smell isn't obvious and comes only in sporadic whiffs (probably due to no catalytic converter), the idle is a lot smoother, and the engine can return to idle RPM from a rev and not stall out or stumble. I pulled plugs 1 and 4 and both look satisfactory, considering the extended idle testing.

But I'm not surprised. I knew all of that; I had it written down in last year's notes. What I can't achieve is this level of quality in closed loop without the BLMs and pulse width going all over the place. I've had the O2 constants down to 300 mV with no change in fueling. Lowering the stock values from ~660 to 450 mV does help matters, but it's not ideal (ideal being what I can attain in open loop). The easy solution is to disable closed loop totally but I don't think I'm ready for that yet. There are guys out there with much more radical cams than I have with nary a problem with closed loop idle. I'm at a loss.

blue86iroc

iTrader: (2)

I did some more experimenting and I can get a better idle at 650 RPM with a slightly tweaked MAF Table #1 and 32° spark advance. That might sound like a lot of timing, but I ran through the gamut and the idle is so much smoother... plus I can get by with lower pulse widths. Header temps were between 350 and 400 °F (not EGT, but the temperature of the metal).

That was still open loop, and it worked great. I invoked closed loop and drove the car around several times, adjusting oxygen sensor constants as before. The BLMs are now all out of whack in cells 0 and 1. Cell 0 is reading rich and is pulling fuel; cell 1 thinks it's lean and is adding fuel. I have been changing the last two entries (0 and 8 gm/sec) for the three O2 threshold tables. Lower is 100 mV, middle is 200 mV, and upper is 250 mV. I wish the values could be a little closer to what GM had there (~600 mV range) but I'm having no luck with anything that high. Should I move the lower and upper threshold values closer to the median, making it more precise? During OL testing the oxygen sensor was picking up 320-350 mV with a stable idle (but shooting for that value with the O2 thresholds made the engine run rich ). Now I see stuff like 60 mV idling in cell 0 and ~300 mV in the same cell when I'm going 20 mph. The sensor has less than 500 miles on it. What the heck is going on?