2005 Chevy Colorado 3.5L no start after rain; ignition/fuel control timing prob

3.5L I-5 engine (similar to 2.5L I-4 and 4.2L Trailblazer/Envoy I-6 engine)
DTCs: P0300 (after it dries out enough to start), no other codes besides P0449 EVAP vent code that is unrelated.

This one is a head-scratcher! I have been working on it for weeks. After a hard rain, it will be a no-start for 1-3 days with barely any combustion happening, not enough to start the engine. Fuel pressure is fine. Fuel pump operation normal. Scoping various PCM inputs and outputs shows that spark and injection control is erratic, sometimes having no control and usually occurring at the wrong times, which is why there isn’t much combustion happening. Powers and grounds to coils and injectors and cam sensors are all good.

After 1-3 days, it turns into a "starts, then stalls after 5-10 seconds" condition. It will run fine for a few seconds upon start-up, then as the RPMs come down, it will start to run rough, then chug, then stall. If I touch the throttle at any point, it will stall immediately, because the throttle opens, but the PCM is not adding fuel to match the increased air flow, so it dies immediately. However, some of the time for as much as 1.5 weeks after the hard rain, it does not start -- it tries to start but won't quite start. It's erratic. On some crank cycles, there’s no combustion whatsoever. Spark and fuel injector control are erratic. Eventually, after a lot of dry weather, it will turn into a pretty good running truck until the next hard rain (or car-wash, a "no-no").

From the details you’ll see below, it seems like there’s an “outside influence” messing up the PCM’s ability to be aware of crank and/or cam timing. I’ve investigated radio frequency interference but think there is none. I’ve tried shielding the PCM with aluminum foil, disabled coil #1 that sits by the wiring harness that carries the CKP signal, disabled the alternator. None of this changes symptoms. RPM PID does not jump around while cranking, which according to the manual is the telltale sign of radio interference.

I saw on a forum someone had a similar moisture-related start/stall problem, and they said that the problem never re-occurred after they had the PCM tuned which included disabling anti-theft, as well as installing a cold air intake (probably unrelated). Hmmm. Could the problem be VATS? On this system, the BCM has to send a fuel-enable code via comm lines to the PCM. This system is known to be very problematic on these vehicles, and it’s common for people to disable them. The “description & operation” section of the Chilton manual for VATS seems to hint that the engine might start but then stall later due to VATS. Perhaps there is more complexity to this system than the manual indicates. I did spray mist the BCM, but that did not make the engine quit or misfire.

Some facts:
-This truck sat for 3 years until recently. After getting some repairs (unrelated to engine performance) to get it back on the road, it performed well until it went through the car-wash.
-PCM was replaced with a used one that was programmed to this VIN - behaving the same as the original PCM. Works fine when dry, no-start when it rains. I caulked the PCM so there’s no way any moisture is getting into it.
-CKP sensor replaced with an OEM one.
-I am confident the wiring and connections are all good. Power and ground feeds to the PCM were all load-tested at the PCM while the vehicle was a no-start. All good. CKP & 2 CMP sensors all have good signals, and I know they're making it to the PCM via various tests I’ve done. There is a 2-volt bias on both CKP wires making it to the sensor.
-The base timing is good. Crank and 2 cams match the known good waveform. The problem is definitely electrical.
-It’s not the exhaust camshaft actuator. Unplugging that does not help. There is no P0017 exhaust camshaft correlation code.
-It’s been 1 week since raining, and it runs well most ignition cycles, but some ignition cycles, it will be missing spark or injector control the entire cycle for 1-3 cylinders. One time, it was missing injector control on 3 cylinders, but when revving above 1,800 RPM, all of a sudden injector control came alive. I put in the other PCM and the same thing happened. Very strange.
-One ignition cycle, cylinder #1 had ignition control EVERY OTHER CAMSHAFT REVOLUTION, as in NO CONTROL every other spark event, consistently until shutting it off.
-The engine is able to start and/or run using only 1 timing sensor (CKP or exhaust CMP) when the system is performing properly. But I’ve noticed that when it’s struggling, it needs all 3 sensors, or sometimes it needs 2 instead of just one. Then, the next ignition cycle, it will change. Very erratic, but holds consistent within an ignition cycle.
-Sometimes it’ll be a 1-cylinder misfire, sometimes a 2-cylinder misfire. Sometimes 3 or 3.5 (especially early on).
-I cannot duplicate the moisture intrusion. I’ve sprayed down the entire engine compartment including the PCM. The wiring connectors are all clean – no moisture getting into them. I can wash the truck myself and it’s OK. It’s only a hard rain or the car-wash that makes it act up. I think it’s high humidity / mist that’s causing the problem.
-So the PCM is probably good, the crank/cam sensors are good, other engine control sensors are good, the powers and grounds are good, wiring and connections I think are all good. What else is there to controlling spark and fuel injection? Knock sensor retard PID is zero. Temp sensors, MAF, MAP, etc. all good. Unplugging MAF doesn’t fix it. Unplugging B1S1 O2 sensor doesn’t fix it.

Any suggestions? What about disabling the VATS?

Thank you!

What are you fuel trims doing when it's running normal?

Fuel trims:
I just tested it. At first, at idle, combined F.T. was +0-5%, at 2,000 RPM it was +5-8%. Under load (brake torque) it was +13%. Then after a minute they came lower. At idle and 2,000 RPM they were slightly negative. Then under load around 0%.

Looks like normal numbers to me. The evap lines on top of the gas tank on those trucks tend to get leaky. Which push lean when the evap vent solenoid opens up. I was just curious if it might be sucking water from the top of the gas tank, since the evap lines sit up in the low spot on top of the tank. Definitely a long shot. I just repaired a Colorado with evap lines leaking on top of the tank, and the vent solenoid stuck open. So that likely gave me bias.

"The evap lines on top of the gas tank on those trucks tend to get leaky. Which push lean when the evap vent solenoid opens up. "

Are you referring to the purge solenoid? As in, the water gets sucked into the intake manifold? The vent solenoid in the back is normally open and is only closed during leak testing.

How was that truck running after a rain?

I do smell some gasoline so what you described is probably the case here, but I know that's not THE problem because I have ignition/injector timing issues.

I will check to see if the purge solenoid is stuck open.

Sorry, yes. Purge solenoid, not vent solenoid.

You can always disconnect it from the intake, and cap both sides to test how it runs.

The purge solenoid is not stuck open, so water would only come in when the PCM commands it open, which should be at higher throttle.

Yesterday it had a dead misfire on cylinder #5 for several ignition cycles. I swapped in the old PCM and had the same dead misfire on cyl #5. Past testing has shown this to be a loss of injector or spark control, not powers/grounds. So consistently, both PCMs behave the same no matter what the issue is.

Any rain lately? Or unrealated to water this time?

We got rain last night but not a really hard rain. This morning it acted up for the first 2 minutes, and now has smoothed out. It's definitely moisture-related.

I'm sending the PCM out today to a tuner to disable the VATS (anti-theft) feature, as I just don't know what else to check. I saw someone post on a forum in 2020 that in 2015 after it would rain, their Colorado would idle, but then die when they tried to drive it, and disabling VATS seemed to be the fix.

That is so strange. And doesn't make sense to me as to why that would cause it. But definitely worth a try. Are you sending it off, or taking it somewhere local?

I'm sending it off. For everyone's benefit, here are 3 PCM tuners who are said on forums to be reputable:

-Lime-swap.com, Jeremy, in Georgia, 313-378-1547
-Supermodulation, James, in New Jersey, 732-690-6443
-PCMforLess.com, Keith, in Ohio, 330-329-5167

Recap: tuning the PCM to disable the VATS (security system) WAS INDEED THE FIX!

This is very strange how one of the components in the VATS system (probably the BCM) was sending a "flawed" signal to the PCM that caused the PCM to behave this way. The problem was not the PCM itself as both PCMs would behave precisely the same way at a given time. It was how the PCM was reacting to the VATS signal it was receiving from the BCM when there was moisture present after a hard rain / carwash or even many days afterward.

While it was acting up, I checked the data PID for VATS pass/fail, and it would say PASS (or whatever the parameter is) even while acting up. But the problem was definitely that signal to the PCM. The PCM was behaving in very strange ways because of this, and the trigger was moisture/humidity. I am confident the component that was acting up was in the cabin, not the engine compartment.

Usually when VATS fails, the vehicle won't start or it will start and immediately die. But in this case, the PCM's control of spark and fuel became very erratic and would change from one key cycle to the next. For example, on one occasion it was firing a certain coil only every other spark event, i.e. that cylinder was misfiring every other event. Typically, it would be a no-start for a couple days, then transition to a start and run for 5-10 seconds for a couple more days, then transition to being able to stay running but with various misfires where spark or fuel cut out on various cylinders or spark / injector pulse occurring at the wrong time.

I am so thankful this is solved! And I'm thankful for that one forum post where someone mentioned that their problems (not as severe as mine) went away after a PCM tune where the VATS was disabled.

Thanks to all who helped.