Recently, I repaired a 2015 Base Silverado with the 4.3 V6 engine. The complaint was SES on all the time. The truck started and drove into the shop normally with the SES on. Scanning the control modules for codes set, I saw this...

[2015 Chevrolet Silverado 1500 LS, ECM/Inputs/Outputs Scan Data]

The P2635 is for low pressure fuel pump performance. The U0074 and U18A2 show a failure with HSLAN communication with the FPCM. The FPCM is on the Powertrain Expansion Bus which is a stand alone high speed network between the ECM and the FPCM only. Since the Powertrain Expansion bus does not connect to the DLC, no scanner communication is possible with the FPCM. The DBDT also cannot be used for the same reason. This complicates data diagnosis.

Why is the P2635 set? If the ECM loses communication with the FPCM, it's commands to the FPCM will not be received and the ECM cannot maintain fuel pressure control. The FPCM will default to a safe higher pressure. Let's look at ECM fuel data.

[2015 Chevrolet Silverado 1500 LS, ECM/Inputs/Outputs Scan Data]

Look at the desired and actual fuel pressure. Then look at short term fuel pump trim. The ECM sees high fuel pressure and is commanding the FPCM to reduce pressure. Since this is not working, the ECM is commanding reduced fuel pump trim to try to reduce pressure. Since pressure is not being controlled, P2635 is set.

How do we test for a communication problem? Usually, we test power, ground, and data lines at the non-communicating module first. Since the module is located above the spare tire and the connector is not very test friendly, and I have only .3 to diagnose with, I took a different approach. I tested the battery and ignition power fuses in the UBEC to the FPCM. They both were good and powered up. The two LAN circuits go from the ECM straight to the FPCM with one connector, x125, in between. This connector is very available under the hood so I diagnosed from this point.

Unplugging the connector, I tested resistance between the two wires, first toward the FPCM and then toward the ECM. I saw 123 ohms toward the FPCM and 121 ohms toward the ECM. Then I tested resistance to ground on both sets. There was K ohm resistance toward the FPCM and Megohm resistance toward the ECM. This shows that my LAN circuits to both components are most likely OK.

With the key turned on and the connector still unplugged, I checked voltage between the two LAN circuits from the ECM. I saw zero volts. With a voltage differential of zero volts, no communication can take place. Either the circuits are dead or equal to obtain zero volts. Measuring voltage to ground, both LAN circuits read 2.457 volts steadily. Unplugging the X1 connector at the ECM caused both LAN circuits to drop to zero. I checked resistance between the two circuits unplugged to make sure that they were not shorted together and saw OL.

What does this show? My circuits are good, but I have no LAN+ and LAN-, only rock steady equal voltage. Could I test further? Sure, but it's time to pull the trigger and replace the defective ECM. With the new ECM installed and programmed, what does fuel control and pump trim look like now?

[2015 Chevrolet Silverado 1500 LS, ECM/Inputs/Outputs Scan Data]

The vehicle is in control and the three codes did not reset. But what about the history U0073 and U0077 set in the EBCM? It's very common to see these types of communication codes set as history with no evident problems. They can be caused by voltage surges or several types of glitches. At this point, I'm just clearing and ignoring them unless they reset on a road test.

James from Florida

Files Referenced: