O2 Sensor Question
 

I just used an ohmmeter to check continuity between battery ground and the exhaust system. Nada, zip, no continuity. I'm pretty sure that on a three wire sensor, the O2 signal return path is grounded through the sensor threads much like a spark plug's return path. I'm also pretty sure the O2 sensor heater circuit is isolated from the sensor circuit (ie. heater has an isolated ground circuit).

I think I'm going to install a ground directly to the collector flange to see what happens. If the exhaust system is supposed to be grounded for the sake of the O2 sensor signal return path, then this is should be yet another thing to add to the poor gas mileage troubleshooting list. I have never seen this topic covered ANYWHERE.

Any thoughts?

hmm... can't hurt.

however, i am having a difficult time believing that the exhaust system is not grounded, via the engine and its ground(s).

I'm pretty sure you are correct on the things you were pretty sure about. I'm with Nynor, it's hard to believe the exhaust has no ground. Do you have continuity between the exhaust and the negative side of the battery? ...could be some corrosion on the threads of the O2 sensor.

The O2 sensor is a new Ford unit spliced into the factory O2 sensor harness. Bosch PN 13953 per Wilk's website (BEFORE THE INCIDENT).

I measured for continuity between the negative battery post and:
1) the collector pipe - no continuity
2) the exhaust manifold - no continuity

On a side note, but related: I added a separate ground wire (8 AWG) between the battery and the cylinder head when I installed a MSD-6A unit to insure a predictable ground path for the spark plugs. I never even gave grounding the exhaust system any thought.

on an ohmmeter doesn't "0" indicate no resistance, which means infinite continuity.

On an volt/ohmmeter set to the ohms scale, "0" ohms is a dead short between the two points you are measuring. In other words, perfect continuity. No ohms, or infinite ohms, means there is an open circuit beteen the two points you are measuring - no connection.

One would think that if the head were grounded, the exhaust would be as well. Seems like the exhaust manifold bolts would take care of that.

Something is wrong here. Check your meter.
Make sure the meter probes are scratching thru the rust and crud.

Jon

Yes, I did forget that iron oxide is not very conductive, HOWEVER, when I scratched through to bare metal on the exhaust manifold I read 2.1 OHMS which is still unacceptable. Should be VERY close to zero ohms. I didn't check the collector pipe where the O2 sensor is mounted, but I'm sure it was at least that. This is EXACTLY why I added a dedicated ground to the cylinder head for the spark plug return path. You can't trust 20 year old head bolts to maintain a perfectly clean ground path.

But I digress ... so I added a brown 10AWG ground wire with crimp-on connectors that were big enough for the bolts because that's what I had in stock. No solder here, it would definitely melt. Now I have zero ohms. Throttle response is definitely snappier now and the surging at WOT is gone! (Just like when I had the O2 sensor disconnected BTW)

The slight hunting at idle remains, however...

I will trace this systematically from scratch again tomorrow. You know, first, look for vacuum leaks, bla, bla, bla, etc, etc, yada, yada...

Hey Edik, did you use insulated wire, won't that melt as well? I've never thought about this kind of problem. I've powder coated my headers...I better do something similar before I put them on. One to definitely consider for a poor running engine? Did you check the reading from your O2 sensor before and after? It would be interesting to hear that.

Yes, the wire is plactic insulated. It is only a temporary sacrificial circuit for troubleshooting purposes. It will be replaced with a suitable high temperature wire or all metal ground strap with stainless hardware very soon. BTW, after only about ten miles the plastic yellow band on the crimp is toast as I suspected it would be, but the insulation on the wire is suprisingly intact. It's rated 105degC, (or 221degF).

The O2 sensor voltage reading was sluggish to respond before the temporary fix. I have not yet measured it since installing the fix, but judging by the test drive, I believe I've solved a problem that wasn't even on my to do list!!

The slight hunting at idle remains, however...