What kind of scan tool are you using? Can it read "history" or "pending" codes?

What condition is the battery and grounds in? It is 10 years old and corrosion sets in.

 

Cel

If your check engine light isn't coming on until you rev the engine, first thing I would do is check the diodes in the alternator. It's an easy test to do. Take your DVOM, or digital multimeter, and set the scale for Volts AC. yup, you read that correctly, alternating current. Then, with the engine running, your air conditioning turned on, blower motor on high, your high beams on, and your rear defrost on, heated seats, radio, interior lights.. all turned on, put your test leads on the negative and positive post. A reading of zero is perfect, remember, you're measuring alternating current, not direct current, so 0 would be perfect... now.. you're not likely to get zero, I expect somewhere between .2 and .6 or .7 at most. have someone rev the engine for you and watch your meter. if it goes above 1 or 1.2 volts, you should start shopping for an alternator. AC ripple will confuse the heck out of the modules in your car. btw, does your scanner have the capability of reading vehicle specific codes? or just obdII generic? Check back and let us know what you find and we might be able to go from there.
-matt

 

Apparently the lamp went out on it's own during a drive today, but then lit again during the same drive. If it was a misfire, would not a code be set and stored?

Anyway, readings at the alternator post:
475.9 mvAC peak
66.4 mvAC idle
(average was 313.4 mvAC)

Just for kicks, I pulled out the silly scope to have a look...

My OBD2 reader is an Equus Innova 3100. I've had it a number of years, and yes it does view stored, current and pending codes.

So after checking the engine to firewall wire which appeared to be in good shape and only dropped 3Ω across it, I replaced it with a 6AWG welding cable and ran it to the same chassis point that the battery jumper connects to. No change in the diagnosis...

 

So after checking the engine to firewall wire which appeared to be in good shape and only dropped 3Ω across it, I replaced it with a 6AWG welding cable and ran it to the same chassis point that the battery jumper connects to. No change in the diagnosis...

Ok, "main" in my world had always been the "starter" ground, which is the largest of the conductors. Battery is brand new as of this last spring, terminals are tight. How many other satellite grounds are there so I can check them individually to begin with? I would think my 300$ Fluke RMS DMM would be good enough to check ground straps for the moment. Weird thing is, is that I've seen other vehicles with several ground points rotted right through and never had a problem. Computers suck...!

Which is it? You changed a satellite cable or the battery ground?

And even if you replaced the battery ground, 6 is too small.

You haven't posted which engine you have, but the H4 and H6 are the same when it comes to ground lugs.

Batt to Block - MAIN
Batt to Fender - next to battery
intake to firewall
intake harness ground is on the intake right side of the engine on a runner (this set is the main ground for everything that is in the engine harness, including the AF sensor)
injector covers to fender
body harness grounds on the firewall at 3 or 4 points, depending on year, around pitch stop bracket
ABS to engine/or fender

You should also check the ECM harness connector for corrosion. If any moisture got onto the plugs due to HVAC box holding water, windshield leak or spillage, then corrosion is a possibility.

There are also ground lugs inside the car at each kick panel. They are the green bolts.

My battery puts out 800 amps at the post. I get 800 amps at the block.

 

. . appeared to be in good shape and only dropped 3Ω across it . .

For the ground link between the engine and the firewall, 3 Ohms is like an open circuit.

If it was that bad, there could well be other grounds that are equally bad, or worse.

Check the ground losses:

Engine running above idle (1k+), some electrical loads turned on such as lights, heater fan. Use the DVOM to measure the D.C. Voltage between the battery negative post (not the cable clamp) and 1) the engine block, 2) alternator frame, 3) car body. All should be low; below 0.2 Volt is okay, but less than 0.1 Volt is better. High Voltage indicates bad ground connections or wiring.

Then measure between the battery positive post and the alternator output post. Again, Voltage drop should be very low.

And yes, if the CEL was turned on by the ECM, there should be a code stored. (It's my understanding that the CEL is turned on only when a "hard" code is stored. A pending code, however, might not turn on the CEL. But the CEL coming on without any apparent memorized code is strange.

 

And yes, if the CEL was turned on by the ECM, there should be a code stored. (It's my understanding that the CEL is turned on only when a "hard" code is stored. A pending code, however, might not turn on the CEL. But the CEL coming on without any apparent memorized code is strange.

Not if a bad ground is effecting the ECM memory. Which in this case it seems the ECM is resetting every time he shuts it off.

Which brings up monitor status. Can your scan tool see monitor status? Are they resetting with the MIL?

EDIT: Nevermind about monitor readiness status info, I looked at their website.

 

Try adding a 4 gauge cable from the neg batt to the AC bracket. The AC comp bolt is the perfect size to hold a lugged cable. You may have a bad main ground cable. They tend to corrode within the insulation towards the block end by the starter.

 

Is the main ground not the one I replaced?


I'm sorry, I missed a decimal point. The OE cable only dropped .3Ω across it, same as my new 6AWG cable, so I didn't think anything about it.

Even when I do a continuous scan WHILE driving, I never see any stored, current or pending codes. If I go under the hood and unplug a random sensor, I can see the code that is set almost immediately, and it stays there after a restart. I have readiness icons on the meter, yes. Do they change after I clear a scan? Not sure, they're small and hard to read in the dark while driving... I'll have another look tonight...

 

The OE cable only dropped .3Ω across it, same as my new 6AWG cable, so I didn't think anything about it.

Measuring low resistance can be problematic, even with digital multimeters. The test leads themselves introduce some resistance, and the point of contact between the test lead tip and the part being measured is uncertain.

Even so, a 0.3 Ohm resistance in a cable carrying 10 Amp will drop 3 Volts there alone. So with a battery post Voltage of 12.7 Volts, only 9.7 Volts will be available at the other end of that one cable.

That's why I suggested the Voltage measurements between the critical points -- it's looking at the actual losses in real-time, and a digital multimeter is more likely to provide a reliable Voltage measurement. cardoc uses a Midtronics conductivity tester that uses a different technique, but also identifies current carrying capacity losses as one moves away from the battery or alternator, losses that a simple resistance measurement might not reveal.

 

Even so, a 0.3 Ohm resistance in a cable carrying 10 Amp will drop 3 Volts there alone. So with a battery post Voltage of 12.7 Volts, only 9.7 Volts will be available at the other end of that one cable.

I disagree that my 6AWG cable, 16 inches long, will drop a whole 3 volts at 10A. Voltage drop should be less than 0.01 I believe...

 

Ok, "main" in my world had always been the "starter" ground, which is the largest of the conductors. Battery is brand new as of this last spring, terminals are tight. How many other satellite grounds are there so I can check them individually to begin with? I would think my 300$ Fluke RMS DMM would be good enough to check ground straps for the moment. Weird thing is, is that I've seen other vehicles with several ground points rotted right through and never had a problem. Computers suck...!

 

My meter leads alone are 0.2 Ohm. The wire with the 2 crimped terminals metered out at 0.5 Ohm. The difference being 0.3 Ohm at the what, 1.5v that the meter checks with? I haven't had this meter calibrated in a great number of years. Could it be off? Sure. Do we all agree that the loss should be almost insignificant for this particular piece of wire? Yep. I'm back under the hood of this thing tonight to see what else I can('t) find...

 

Resistance from alternator case to negative battery terminal: 0.2 Ohm.

Resistance from negative battery fender stud to negative battery terminal: 0.3 Ohm.

Voltage drop from alternator case to negative battery terminal: 0.025 ~ 0.050 VDC varies with RPM.

Voltage drop from alternator case to chassis: 0.052 VDC varies as well with RPM.

Car starts and runs as fine as it ever has, but will no longer pass VT inspection due to the CEL lamp. Getting a second opinion with a different OBD2 reader tomorrow hopefully...

 

The resistance readings are misleading. Probably not getting good contact between the test lead probes and the surface being measured. And typically the test probe tip is a tiny point which can have significant resistance when measuring at these low levels.


But the Voltage reading between the battery neg post to alternator case isn't bad at all. Need Voltage measurements from battery negative post (not the cable clamp) to all the points that cardoc identified as grounds.

If all are good and low, I'd say there's a good chance the grounds are not the problem, but let's see . . .

(Need to find a lower cost conductivity tester like the Midtronics that can be used to do these checks as simply as cardoc can.)

 

Well well well... I borrowed a friend's OBD2 reader. Next model up from mine: Innova 3110. It fed me the P0032, upstream o2 (a/f) sensor. Looks like my reader is having issues. Picked one up on the way home tonight. Expensive little buggar...! Apparently all I need is a 7/8" open wrench and a lot of offensive 4-letter words strung together in a nasty run-on sentence to replace it. Now I have no CEL! Yay! Thanks to all those who helped!

 

Great news.

We do tend to put a lot of trust in the test equipment, but who thinks of testing the test equipment?

 

The weird part is that if I unplug any random sensor under the hood, the reader picks up on it immediately. And it works just fine in my 2003 Tacoma as well. Who'd of thunk it?

 

p0032

well.. looks like the heating element in your sensor went bad. not sure why the scanner isn't picking that up, but I wouldn't condemn the scanner just yet, it's possible they use different j protocols to connect? or maybe the computer just set the code when your friend's was connected? I don't know, but o2 heaters go bad a lot, hope you've got this one taken care of :roadtrip: