There is a number of things we have to consider before starting the diagnosis. First, if you were measuring that 6 volts at the fuel pump's connector with it unplugged, that may be confusing the real cause of the problem. This would be similar to standing on a garden hose and blocking ninety nine percent of it. You would still see full pressure at the nozzle, as long as it was closed. No water is flowing, so no pressure is dropped across your foot. Voltage is electrical pressure. With the connector unplugged, no current is flowing, so no electrical pressure, (voltage), is dropped across some undesirable resistance in the circuit. This would suggest everything in the circuit is turning on properly, but there is a high-resistance connection or break in that circuit. To avoid this confusion, voltage readings should be taken with everything plugged in. Some connectors make that very difficult, but most have rubber seals around the individual wires where you can slide the voltmeter's or test light's probe in. Test lights are more accurate too for this type of problem because they draw current from the circuit. A voltmeter will give a false 12 volt reading when there is a high-resistance break that is preventing the needed current flow. Voltmeters confuse the issue because you have to interpret the reading. We are looking for 0.0 volts or 12.0 volts. Which do you have? 6.89 volts is neither. With the test light, you are much more likely to find something or nothing. The only way to get a bright light is for the circuit to have no breaks or high-resistance connections.
The second potential clinker is that feed wire to the pump is tied to the Engine Computer. The computer turns the fuel pump relay on, then the 12 volts that gets switched on goes to the pump, and it goes to the computer to verify it got turned on. Often the computer will send out a small voltage, then watch the resulting current flow, to verify the integrity of the circuit. Once the fuel pump relay gets turned on, the 12 volts that shows up makes that little signal voltage from the computer irrelevant. You could have a relay that is not turning on, meaning you would have 0.0 volts at the pump, but the 6 volts you are seeing is coming from the computer.
The biggest issue is you must find 0 volts at the fuel pump when the engine is not rotating, (cranking or running). That is the safety feature on all cars with electric fuel pumps. If a fuel line got ruptured in a crash, the pump would dump raw fuel onto the ground, creating a serious fire hazard. Instead, the engine cannot run with no fuel pressure, so it stalls. Once stalled, no signal pulses show up from the crankshaft position sensor or the camshaft position sensor. It is the lack of those pulses that tell the computer to turn the fuel pump relay off. What this means for you is other than for one second when you turn on the ignition switch, the fuel pump will not get 12 volts until you are cranking the engine. In fact, that one second can provide a clue. If you can hear the hum of the pump for one second when you turn on the ignition switch, that entire circuit is working, and there is no point in diagnosing anything in it. Often you cannot hear the pump over the noisy chime.
One more important point is did you check for spark? Too many people get hung up on the first thing they find missing, and forget to check for additional symptoms. Check for that first. If you do have spark, we will need to find where the voltage is being lost in the fuel pump circuit. It is not practical to have a helper crank the engine while you poke around in the circuit. Instead, remove the fuel pump relay, then gently insert a jumper wire between terminals 30 and 87 in the socket. A small stretched-out paper clip works well. Be careful to not use anything fat enough to stretch those terminals. That will cause intermittent problems in the future.
The ignition switch does not need to be on for this. The jumper will energize the circuit, and the fuel pump should be running. If it is, this entire part of the circuit is okay. If it is not running, now is the time to work your way through the circuit with the test light.
Images (Click to enlarge)
Wednesday, October 10th, 2018 AT 5:58 PM