Ground 101

[James Scott]

Does anyone know if the ground in the engine compartment (battery, engine, transmission and body) should be different from the ground points (brackets, bolts, steering column) under the dash?

When connected (engine compartment to dash) by an ohmmeter - IT READS 15 to infinity ohms depending whether I have the IGNITION switch OFF, ACC or ON! [12 to 25, 35 to 60, 150 to infinity respectively] IS THAT RIGHT?

Please someone tell me what this means???

HEEEEEEEEEELP! (Or tell me what Beav means under "Bad Injector" thread?)

THANKS!

[lee]

I'll take a stab at this, but stand by for corrections as an electrical tech I'm not...

First, ground should be ground with little or no resistance from one place to another. That said, are the places you are checking really ground points or just some metal under the dash? I think it was mentioned in another post to check the battery to body ground strap to make sure that's OK. There's also some kind of ground connection (I think) at the back of the engine - tranny to body.

Second, I believe what Beav is trying to say is that measuring ground resistance from point to point isn't as good as measuring voltage drop across the system. Voltage drop is a natural occurrence of less than perfect conductors, complicated by induced resistance at connectors, because of corrosion, etc. A simple way to see it is: stetch out a couple hundred feet of wire from both posts of a battery. Test the voltage at the battery and test it at the end of the wires. There will be less voltage read at the wire end just as a natural resistance effect - less voltage drop for bigger wires, etc. Now in any real system, the voltage drops increase as various connectors, wires, switches, etc all add resistance to a circuit. Checking the voltage drop between battery and the end item puts all those resistance causing items in the path for measurement. At some intermediate level, voltage drops can cause erratic system behavior (when it's on the borderline between OK and dead enough to not work at all). Since I believe the ECU is operating in a switch mode (on/off, on/off, etc) the resistance in the total circuit may be just enough to be at that borderline erratic operation point.

This voltage drop thing is very real for me as an owner of an older British motorcycle. More than once I've had to take switch gear apart and clean contacts, etc.

[lee]

The addition of 1 or 2 ohms of resistance in the circuit caused by an oxidized connector is significant. For example, in a 12 volt starter circuit drawing 40 amperes of current, the addition of 0.2 ohms of resistance drops the current to the starter to 24 amperes. This drops the operating power from 480 to 288 watts. This is a 40% loss in operating power caused by a 0.2 ohm resistance. High current, low voltage circuits are resistance sensitive.

The best method of checking for resistance is by measuring the voltage drop on each side of the circuit while current is flowing. What you actually measure is the resistance present in the circuit but displayed in terms of voltage.

Total voltage drop is measured with the positive lead of the voltmeter prod toward the batteries positive post and touching the device's output terminal and the negative lead connected to ground. Current must be flowing in the circuit tested. If the meter reading is high (over .5 volt typically) then take separate measurements across each segment of the hot circuit; always with the positive prod toward the batteries positive post. You can check across solenoids, switches, lengths of cable, etc. Thus each segment of the circuit can be checked to find the fault.

A similar check can be done on the ground side. Touch one prod to the grounded battery post and the other to the components ground connection. For example, you can check the return path from the engine mount back to the battery by touching one prod to the grounded battery post and the other to the engine mount. With the starter cranking the voltage reading is the amount of voltage drop in the return circuit; typically less than .5 volt.

Similar check can be made on the alternator by turning on lights so that the alternator is producing appreciable current. Check total drop from alternator to battery. This could validate or refute those with claims on the alternator wiring upgrade mod seen elsewhere in this forum.

edit: I stole most of these words on this second post from somewhere back when I started doing my own work on my Norton, sorry I can't give credit.

[James Scott]

AHA! Now I understand! THANKS! However, two questions:

1. How would I practically do this in following the shop manual troubleshooting guide procedures? Are you saying that in the case of checking the ECU grounds I mentioned earlier, instead of removing and probing the terminal (14, 15, 24, 25 of connector B62 for example) in the harness connector while the other ohmmeter probe is at ground (should be 0 ohms), I should leave the (B62) connector plugged in (but tap the appropriate terminal wire now from behind with one voltmeter probe attached [14, 15, 24, or 25 for example]) and the other probe should be grounded [with the ignition ON or the engine at IDLE] >> Less than 0.5 volts= GOOD >>More than 0.5 volts is BAD. Is this what y'all are saying?

2. With all you and Beav have said, isn't it true that the underdash ground and the engine compartment ground SHOULD NOT BE DIFFERENT! What does that mean? ESPECIALLY, WHAT DOES IT MEAN THAT THIS DIFFERENCE IN RESISTANCE VARIES FROM 15 TO 300 ohms, DEPENDING ON WHETHER THE IGNITION IS OFF [15 to 25 ohms], ACC [40 to 70 ohms], or ON [150 to 300+ ohms] ???

THANKS for your thoughtful and time-consuming help!

[lee]

wow. lot's of questions. let's see if this helps...keep in mind I am not qualified to diagnose the ECU, just trying to give you some general answers.

1. I think the answer is yes. What I would suggest is what Trevor (I think) recently posted on the TPS...put a pin in the wire and connect one probe lead to the pin you just stuck in there to get a reading "in place". That way you get connector resistance included. edit: how the voltage drop thing works is to measure along the path of current flow from beginning to end. If there is no resistance, then the voltage will show 0.0V. If there's say 0.5V or more, then start checking segments of the wiring to see where the drop is occuring - voila, the resistance point is identified - and if it's in a place where resistance is not supposed to be - then trouble found and a fix is in order.

2. Ground should be ground everywhere. In any practical system there are always stray resistances that prevent this from being perfect, but 300 ohms seems too much difference to me. I have not a clue as to why you would see differences in resistance between Off/ACC/On positions unless it has to do with a floating ground in the ECU itself. What resistance measurement is changing - the difference between the battery ground and the ground you are using inside the car, or is this between the ECU and some interior ground? The former is a problem, the latter - I don't know the answer to.

Pin 14, as the "body ground", should be the true ground for the ECU. Tests between this pin and the negative post on the battery should show zero or VERY small resistance regardless of key position. The other grounds may well be floating and relative to the sensor/ECU/system voltage - just don't know enough about the ECU to even guess.

If this is still all centered around the injectors, you may well have to beg/borrow/steal an Oscope and watch the signals.

[James Scott]

THANKS AGAIN for your thoughtful detailed answer!

To answer your question, the ground difference I'm talking about is the engine compartment (engine, trans, and body) vs. the underdash ground points (large brackets, bolts, steering column, etc.) > > > > > > >

When I check the voltage drop at IGNITION ON - it is 0.6 volts!

The car has not idled rough or shuddered for 3 days > > It has fooled me before, so I'm not counting chickens yet ? ? ? ? ?

Please see my new thread "Manual Mania" for my latest Subaru concern.

Thanks,

[James Scott]

It just ran rough (not too much tho) for about 2 to 3 minutes this AM - "Never Give UP, Never Surrender" (Galaxy Quest)