poacherinthezoo

Recently I've been noticing intermittent, large (+10 to+15%) jumps in short term fuel trims on throttle tip-in which return to ~0% immediately after. Previous smoke tests have failed to find any vacuum leaks, but using isopropyl alcohol I was able to find, and fix, a small leak in the upper intake manifold.

While looking around under the hood while waiting for the oil to drain, I found a second vacuum leak, this time in in VFAD hose, which I subsequently fixed. Initially, there was a modest improvement in MAF readings (average of 4.7 grams/sec, compared to 4.4 grams/sec previously - fully warmed up engine at idle). After another 5-10 minutes of driving the short term fuel trims shot from ~0% to +25% and stayed there. When I got home, I let the car idle in the drive way and watched the MAF reading. The MAF was wildly inconsistent, starting out around 5.0 grams/sec, and then dropping to 4.7 g/sec for awhile before taking a dive down to 3.7 g/sec (engine RPM 820-850). After shutting the car down I checked for any pending codes and found a P2097 pending.

I'm starting to question if the MAF may need to be replaced, as cleaning it did not result in any change. [Side note - can I use a series 1 MAF to see if it rectified the issue in my series 2 car? I'm not aware of any local series 2 owners, only Series 1.]

Ideas, suggestions, and comments are greatly appreciated. I'm game to try and double check anything before taking the car to Mazda, since the last time all they did was give me was a big bill and couldn't do anything to solve the issue.

MikeTyson8MyKids

iTrader: (1)

P0172 present yet? May take another drive cycle or two. What does measured afr read on a highway cruise?

poacherinthezoo

Logged a few conditions (idle, steady state cruise) to get the lambda values for AFRs (my old generation of the ultra gauge can't pick up lambda).

The attached log (see image) of the commanded AFR and measured lambda values at a steady state cruise at 40 mph (my log at 55 mph didn't get recorded - accidentally hit the record button twice). The recorded lambda values are stable at 1.0 during both cruise and while sitting at idle.

Interestingly, the previously pending P2097 is now gone, and the short term fuel trims are no longed pegged at +25%. Now they are sitting around 0%.


Taken together, the elevated long term fuel trims (+10%) and MAF (4.7 grams/sec) suggest that I still have more vacuum leaks to find.


Thoughts and opinions?

MikeTyson8MyKids

iTrader: (1)

I had an intermitent front o2 sensor failure that caused similar problems, but it set the other code. This is just my guess at this point.

TheRedRotor

im pulling a p2097, pending p2097, p0031 and p0137 (not too sure about the last code i didnt write it down) i feel like changing my front 02 sensor should rectify the situation but not too sure. any suggestions?