I'm not sure what you're testing, but how do you get any voltage anywhere with the battery cable disconnected?
Wait a minute, I see the problem. Your positive probe is taking a reading related to a reference point, which commonly is ground, which is the negative battery post, but you removed that. This is like trying to measure barometric pressure with the gauge sitting in a jar pumped into a vacuum. There is some circuitry that is allowing the tiniest trickle of current through your meter, and that is what is allowing it to pick something up. You can prove this by measuring any of those voltages, then place a test light across the same two points. You'll see the test light doesn't light up and your voltmeter will go to 0.0 volts.
The only valid voltage readings are those taken with the battery connected. The battery does need to be disconnected to take some resistance readings unless the circuit can be made dead some other way, as in turning off a switch or pulling a fuse. Voltage in a circuit will cause totally incorrect resistance readings.
Terminals 30 and 87 are the contacts that are switched on and off. 12 volts on terminal 87 all the time is correct. That 12 volts will show up on terminal 30 when the relay is energized. Terminal 85 or 86 will have 12 volts when the ignition switch is turned on. I can never remember which one. If you have the relay plugged in and are able to reach the terminals with the meter probe, you'll find 12 volts on both 85 and 86. 12 volts is getting to the second terminal through the coil of wire in the relay. One of them gets grounded by the Engine Computer for one second after you turn on the ignition switch. It gets grounded again anytime the computer sees engine rotation, meaning cranking or running.
All of the testing you've done so far can be summed up with a single voltage reading. The relay has to be plugged in so you might have to fashion a thin piece of wire if necessary to reach the terminal. Measure the voltage on terminal 30. A test light works much better for this test for two reasons. First, most digital meters don't respond fast enough to catch quick pulses. Second, think of a compressed air line. All you need is a tiny pinhole through what is otherwise total blockage of the pipe, and you'll have full pressure on a gauge at the end of the line, but you'll never be able to get enough volume through there to run a tool. Similarly, all you need is any circuitry with really high resistance to let a teeny dribble of current through and the meter will pick that up as a voltage, but you'll never be able to get enough current through that resistance to do any useful work. A test light has very low resistance so it's going to draw relatively high current through the circuit. It can only do that and light up if the circuit is working properly. Also, we don't need to know the exact voltage at terminal 30. The test light is going to be off, (0.0 volts), dim, (insufficient voltage, but something), or bright, (enough voltage to prove the circuit is working).
There's two parts to this test. With the test light connected to terminal 30, it should be off. Now turn the ignition switch on. You should see a bright test light for one second, then it will go back off. You might hear the fuel pump's hum at the same time. That has just proven the circuit is okay and the Engine Computer has control over the relay. Next, crank the engine. The test light should turn on anytime the engine is rotating. If it does, move on to something else.
The only time a mechanic would do this test is if there was a fault code related to the "switched 12 volts is missing" and he wanted to verify that or see if the 12 volts was making it as far as the relay. That would instantly tell him if there could be a break in the circuit before terminal 87 or after terminal 30.
Tuesday, September 29th, 2015 AT 9:22 PM