What makes an O2 sensor always read low voltage?

[Gary Allan]

I got the guinea pig back. The wrench did an emissions test/inspection on it since it was due. Needed an O2 sensor. It passed, as always. I do an on board diagnostic a few days later and I get a bad MAP code and a "O2 always reads low voltage" code. I read that the low voltage can be because of a whacky MAP sensor in the notes on allpar....so I swap out the MAP ...disconnect the battery ..run it a while ...no codes. Feel all warm and fuzzy. Go and get gas (6 miles = $0.20 cheaper

) ..have to wait in line (because it's $0.20 cheaper

) ..so I do another diagnostic (turning the key on-off-on-off-on and read the flashes on the CEL) ...O2 always reads LOW. So, just in case the wrench put in a defective O2 ..I put a new O2 in (I would get $$ back from him if it was indeed defective) ..same deal ..no code ...after a longer drive same code.


Now I'm stumped of where to look next. Some suggested that a vacuum leak will make the O2 to low voltage ..but I would assume that low is "lean" and one would think that the MAP would be effected and read it as a wider throttle and the PCM should act accordingly. I reason that a manifold leak won't be interpretted any differently than a throttle opening.

Any ideas?

 

[punisher]

Absolutely, a vac leak/unmeterted air is the most common cause of lean/lack of O2 switching and MAP codes. If your TP voltage still shows idle and your engine is lean, your PCM knows there is a problem.

If you have a scan tool with a data logger look at the fuel trims, pay attention at idle. Lean trims at idle progressively getting more even with throttle application/load= vac leak. Conversely, rich trims at idle moving toward normal under enginme load/throttle = faulty MAF or stuck fuel pressure regulator.

 

[unDummy]

http://www.asashop.org/autoinc/june/Tech2tech.htm

 

[mechtech]

The poor O2 is just telling you what it sees.
The problem is upstream.

 

[brianl703]

Does this vehicle have an air pump?

If it does, the control solenoids for the air pump could be sticking so it's always pumping air upstream..which will make the oxygen sensor read lean.

I had this problem on my 1988 Mustang GT and the code it was giving after running the self-test was "Bank 1/2 lean" (it has two oxygen sensors)

I did the diagnostics to verify the cause of the problem (I wasn't sure at first that it was a thermactor problem but I started to get suspicious when I pulled the vacuum line off the thermactor control valve and felt a bit of vacuum), and ordered new thermactor air solenoids, but in the meantime I had gotten 8 new fuel injectors (some of the old ones were leaking externally) so I installed them, along with a new fuel pressure regulator "while I was in there".

After I put it all back together, I reran the self-test and instead of getting a lean code, I got a code "thermactor air always upstream".

I suspect the old injectors were clogged in varying amounts, not enough to fail a cylinder balance test but enough that the computer couldn't detect a problem with the thermactor air system.

 

[Gary Allan]

Thank you undummy for that informative link!


This doesn't have an air injection system ..but there is the possibility that I'm maxed out on injector pulse width. I'll do another fuel system cleaning. I'd hate to pull the intake plenumn to send the injectors out to be serviced and flow matched on an exchange basis.

I did a shallow examination of the harness/pigtail to the 02 sensor. It appeared intact. I guess a fuel pressure check is in order ...

This is OBDI .. I wonder if my friends Snap-On® scanner can see if the adaptive cells are max'd out for pulse width.

I'd hate to go to the dealership to get the DRB scanner hooked up

 

[brianl703]

One thing you might want to check (if you can reach them) are the connectors to the injectors--perhaps one of them is making a bad connection. That can cause a lean condition.

 

[Gary Allan]

Well there are a couple of differences in my 3.0. It only has one O2 sensor ...no differentiation between left or right. I imagine that something like a Sun Interogator could disable one injector at a time and read the rpm drop and tell me which injecotor(s) is/are lame. I have no EGR valve.

The thing runs flawlessly ..so this isn't some major apparent problem. It's just that something is "unright" ..and it bugs the crap out of me.

 

[brianl703]

I think you could accomplish the same thing by unplugging each injector in turn and watching the RPM drop on a shop tach. Although you might need to disconnect the idle air control valve or equivalent to keep the computer from upping the idle as it drops.

 

[1sttruck]

Sears has a very good mechanic's stethoscope for about $15. It has a decent diaphragm and sonic isolation, and a nice long probe that enables you to touch each injector and listen to it working. With that, a fuel system test using a fuel pressure gage ($40 at Sears), and a 'dead cylinder test' you can verify most of the injector functionality. The dead cylinder test can be done with short lengths of 1/4in id fuel line placed between each plug wire and the distributor cap, and monitoring the rpm drop when each short length of fuel line is touched with a grounded probe.

Watching the engine run at night to look for plug wires arcing, and I haven't tried it but maybe complementing with a grounded probe run along the length of the plug wires, will check for insulation problems. Each wire is supposed to have something like less than 30k ohms or resistance.

 

[Gary Allan]

Remember ...knowledge is my tool :^) Labor is not ...

anyway...there's no pulling the injector leads unless you pull the upper intake plenumn.

I have no misfires ..at least no detectible misfires. The plugs are new. The wires are ...hmmm..not that old...but maybe old enough to change.


Like I said, it runs flawlessly. My only fear is running too lean and burning holes in the pistons or burning valves (that already have 170k on them). This won't be a "tearig into it until I find the problem" type thing. It will be a "well, I tried this ...now next Tuesday I'll try that" ..until I stumble across the issue.

 

[brianl703]

I think before it even comes close to running lean enough to cause damage, you'll have TERRIBLE drivability problems that you can't possibly ignore.

You also mentioned that it passed an emissions test..typically a lean mixture will make it fail due to NOx, but they only test for NOx if it's tested on a dyno.

 

[1sttruck]

A check engine light came on yesterday, after an initial attempt at unplugging a cat. I unhooked the EGR and EGR feedback sensors, poured some distilled water down an EGR fitting which feeds into the exhaust, and at least partially unplugged the cat. I was worried about trying anything stronger as the exhaust does feedback into the engine thru the EGR. It worked well enough to also 'unplug' an exhaust gasket at the back of the cat, which during a road test triggered a lean EGO code. From what I understand an exhaust leak can trigger the code as well as a vacuum leak.

Sources of possible vacuum leaks are numerous; the evaporative canister (the vacumm tubing was rotten on mine), brake booster lines (also had rotten lines), the EGR position solenoid, the PCV tubing, the air cleaner to throttle body tubing (a fitting where a line attached was cracked on mine), EGR stuck open or opening early, and hardened lines that don't fit well and leak. I have a few handfuls of hose clamps on my lines now, replaced a lot of line, and am still working on trying to minimize loss on manifold, especially during start up.

 

[1sttruck]

Oh, forgot to mention that some codes distinguish between banks of cylinders, so swap the sensors to see if it follows the sensor.

 

[Gary Allan]

This was a dyno test. This passed with typical good results. For NOX ..the timing had to be retarded a little. This has the orginal cat.

I'm not going to sweat it too much. As you pointed out, it would be probably showing symptoms if anything was too out of whack. It's just that nagging knowledge that something is (as I said) "unright".

 

[1sttruck]

As I recall O2 sensors will also read lean if they're cold, so if it runs 'lean' at higher rpm the EGR system needs a closer look.