Ok, what is this?

[k0ua]

Remember that first test I told you to make? The start at the battery post one. If that fails to read 12 volts or so with the headlights turned on (remember under load) it could be that the battery has a high internal resistance. A battery tested with a voltmeter without a load placed on it could read just fine and still be totally worthless. A circuit must not only have the correct voltage, but must be able to deliver current at that voltage. This all is related to Ohm's law. If you don't understand Ohm's law, it is difficult to troubleshoot. Another "trick" is to measure with your black and red leads on your voltmeter on each end of a suspect wire. If you measure all but a small fractional voltage at all between the wire terminals under load, something is wrong with that wire. Yes it seems incredible to be measuring voltages on each end of wire by placing the black on one end and the red on the other, but it works. And it works because the "wire" isn't a wire any more. The "wire" is now a resistor. This can be done on the "ground" wire too. Or any wire for that matter.

Things are not always as they seem. A quick and dirty method that may or may not find the problem is to yank on the big fat cables going to the battery. If they fall apart in your hand, that could be a clue. It isn't as good as testing them under load, but it can be a 5 second diagnostic tool.

 

[CobyRupert]

We took them all out, put the meter on them and they show up good. I am about to start a triple check so see if we missed something. My friend brought his multi meter and we made sure both were working correctly.

Like others say: A cable (let’s say a #2 AWG) that carries a lot of amps could be 95% gone (now the equivalent of a #16 AWG) , but its ohms (resistance) would still measure very low, and you’d see continuity with the meter.
That is, it would be hard for most people to realize that the ohms measured are really 3-10 times higher than they should be because cable is compromised.
Connection points at the cable might even measure 12.6V volts because there is still a connection (and no voltage drop because no current is flowing when not trying to start).
Everything would meter fine until you try to start. Only then does the large current cause the voltage to drop IN THE CABLE (or at a bad connection), so that there’s not enough voltage left for the starter. (Or one could say there’s now too much resistance that restricts amps to the starter).
That is, test when starting/cranking. The voltage across the battery should be close to the voltage across the starter. Any difference is being lost in the cables and connections.

 

[DieselBound]

We took them all out, put the meter on them and they show up good. I am about to start a triple check so see if we missed something. My friend brought his multi meter and we made sure both were working correctly. After he checked things and I had already checked them, then I still have the feeling there is just some, little something we missed. Just finding it. He took the relays off and handed them to me. I cleaned the connectors good with a wire brush. I am about to start the process again. Never give up. Thank you!

Don't need to pull them. Further, it's probably best to not pull them: ONE: doing so runs the risk of breaking* (yeah, if they're risky then they're not going to get any less risky- always heat the engine up a bit before pulling); TWO: your circuit is now different- you're testing the GPs but not the circuit. Let me expound on the second part:

* I just got a VW TDI running which wouldn't start. The woman's son replaced the GPs (code popped up for one being bad) but one broke off on him. I found the real reason the car wasn't starting- unless it's insanely cold out these TDIs WILL start (hard to kill). GPs were pulled COLD, which is a sub-optimal condition in which do do this. Not a guarantee a GP won't break if doing it the way I suggest, but given the gravity of things when you break one off it's best to do everything as correct as possible.

Continuity testing should be done with the plugs in. Ohm/resistance between GP top/tip and head/block (or another part of the engine that is known to be a ground point). The GP installed grounds to the head. Although the probability of a good GP showing up as bad when installed is very low it's not out of the realm of possibilities. So, just test INSTALLED: a LOT quicker. But...

Folks have said that the ONLY true way to verify GPs are working is to (well, remove them) and put a load on them. It's possible that the fail with a load but with no load they show OK. <<- This is along the lines of what others are saying with regards to the need to apply loads in order to really understand whether a circuit is functioning (holding up to the load it's supposed to). Another PLUS for doing this is to see if a GP's glow tip is swelling up or not: if they're swelling up then they're likely going to break- sometimes they'll break and drop into the cylinder; lesser bad luck situation is that they stay swelled and make it hard to remove (they break off). Quality GPs are worth it: good ones just don't seem to present as many problems.

Sometimes relays can be opened up. If that's possible open them up and see if the contacts might be pitting or showing signs of sticking. There are ways of testing but it depends on what type of relay (ones with diodes, which are pretty common, require testing the diode- plenty out there on how to do this). Simple test to tell if they're activating is to hold you hand on them and activate the circuit (of course, keep your fleshy bits away from the terminals!).

 

[Boss 302]

Yup, bad connections on a relay is what it looks like.