A 2005 GMC Envoy rolled into the shop on a tow truck. The customer complaint? Classic: “Cranks but won’t start.” But there was a twist. Sometimes it would fire right up like nothing was wrong. Other times, it would stall out without warning. No check engine light. No stored codes. No consistent failure pattern.
To make things worse, this SUV had already bounced out of another shop that couldn’t figure it out. They swapped a fuel pump and a battery, but the problem came right back. Now it was our problem—and I was already smelling electrical gremlins.
The Symptoms: Intermittent No-Start with Zero Codes
The first clue came straight from the customer. They mentioned the radio had randomly reset, power windows were flaky, and interior lights flickered every now and then. That's not the kind of thing you want to hear, because it screams intermittent electrical issue—a category of problems mechanics love to hate.
When I hopped in and gave it a try:
- The engine cranked strong, but didn’t fire.
- No misfire codes.
- No stored DTCs.
- No warning lights beyond the standard startup sweep.
That alone was suspicious. A healthy system should at least throw a soft code if something crucial is failing during crank.
First Round of Diagnostics
I pulled out the basics:
✅ Battery voltage was solid, no drop-out during cranking.
✅ Fuel pressure tested at 55 PSI—right on spec.
✅ Crankshaft position sensor output looked normal on the scope.
✅ Ignition switch passed testing and voltage checks.
✅ Fuses and relays were all good.
The fuel pump primed. Injectors clicked. Spark seemed present… when the vehicle felt like cooperating.
But the problem was intermittent—sometimes it would crank five times with no start, and on the sixth try, it would fire right up and idle like nothing ever happened.
That’s when my brain started whispering: control module problem.
Scanner Confirms a Bigger Problem
I hooked up the scan tool and noticed something odd: occasional communication dropout between modules. The PCM would briefly vanish from the network—just long enough to log a U0100 – Lost Communication with ECM/PCM code, then come right back like nothing happened.
That code disappeared almost as quickly as it appeared. If I hadn’t been watching live data, I might’ve missed it entirely.
And when the PCM cuts out, everything else follows:
- No spark
- No injector pulse
- No start
It mimics all the usual no-start suspects—crank sensors, bad coils, dead pumps—but the issue’s upstream. This was the digital equivalent of a brain misfire.
Digging Deeper: Ruling Out Grounds and Wiring
Before condemning the PCM, I had to be sure it wasn’t getting sabotaged by bad wiring.