1990 Cadillac Brougham with the 350 TBI Chevy. It has to be really warm to show the symptoms, like driving for a half an hour+, then it will surge under load and have low power.

I have replaced/done the following: rebuilt the TBI unit with mods, spark plugs, cap, rotor, new exhaust, O2 sensor, fuel pump, EGR valve... I replaced/did these things as regular maintenance and also in an attempt to fix this recurring issue.

The only code that comes up is: 44 lean O2 sensor. It runs GREAT all the time except when it is hot and under load, like climbing a hill or overtaking, those are the only conditions when there are any symptoms.

Any suggestions? I was thinking maybe the ignition module that drives the fuel injectors, or the injectors themselves? It is hard to diagnose, because everything looks/works well on the scan tool and inspecting under the hood, but then after driving a while the symptoms return...

Read more: TBI lean surging low power Page1 - Chevy High Performance Forums at Super Chevy Magazine

you check fuel pressure? inj spray pattern look like a nice solid cone? replace the fuel filter?

I didn't check fuel pressure before or after replacing the fuel pump, but more than likely it has adequate fuel pressure since it did this before and after replacing the pump, rebuilding the TBI unit. (including replacing the fuel press. reg.), and before and after replacing the fuel filter.

I have nothing to compare the injector spray pattern to, but to me it looks like a decent conical shape. I would guess that the spray pattern is good, at least while the car is idling in the driveway and I am able to look at them...

Is your 02 sensor leaning out when this happens? Whats your fuel trim counts at? Vacuum leaks?

check for vacuum leaks by using carb cleaner, run the engine and listen while spraying around the tb, manifold edges, the egr, all the different places it may leak vacuum

I checked for vacuum leaks high and low on a couple separate occasions and found nothing. I was pretty convinced myself that that had to be the culprit, but I never found any leaks. Plus, wouldn't it show lean symptoms all the time?

I am not sure about the fuel trim or the O2 activity when it happens, as I haven't had the opportunity to observe the scan tool while it happens since it doesn't happen all the time.

Today I felt a slight power surge while accelerating on the way to work. Suspecting my MAP sensor for giving false readings and leaning out the system intermittently, I unplugged it for the drive home. It ran great. I could smell some excess fuel but it ran well, except for when I was going 40 or so under no load with slight, slight throttle it bucked pretty violently. I am guess this is normal since I have a major fuel management sensor unplugged... I suspect the MAP sensor as the culprit, but I realize the possibility that having it unplugged enriches the system and may just be masking the real issue. It'* hard to say.

These systems can be real tricky, particularly with intermittent problems like this.

if you unplug the map sensor it runs in open loop and that doesnt really tell you anything

Yeah, you're right. I just came to learn that lol.

I am stumped! I don't know what else to do or check. I reset the ECM again to get rid of any codes from my experiments, such as running it with the MAP unplugged. So, I need to put some miles back on the car to see what the block learn is, and what codes return. I'm sure code 44 will turn up again.

The block learn is 128, but maybe that is because I just reset it and only put 20 miles on it (went to work and back). How long does it take to acquire accurate block learn data?

I have 21.5"Hg steady at idle. If I had a vac leak, wouldn't this number be lower?

There are a couple more symptoms that I didn't mention but could be relevant. It seems to have an extended cranking time, most of the time. Also, once it starts, most of the time there is a white cloud of smoke, varying in size, from the tailpipe.