O2 sensor testing with multimeter

[JHZR2]

Hi,

In trying to get to the root of my NOx issue, Im trying to get a grasp of every issue possible. My BMW has a 4-wire O2 sensor. I can get a reading, but it never changes at any frequency.

At full hot, It will read about 0.7V. If we fast idle it at 2000RPM, it leans out, as one would expect with a no-load situation, and reads down to about 0.1V. Let off, and it comes back.

I know that you really need a scope to see cross counts, but from my reading, it seems that you should be able to get some reading that fluctuates notmally on a multimeter as well - not the best, but at least something that can be used for analysis, and supposedly see the reading move across 0.45 or 0.6V. I can't get any fluctuation, with a cheap analog, a cheap digital, and my fluke 179 true RMS meter...

Any thoughts why? the voltage cvhanges, but I think Im measuring it wrong somehow...

Ive tested it disconnected and jumpered to the ECU, no difference.

Thanks!

JMH

 

[shoz123]

I measured mine by back probing the PCM connector and ground. With the meter on millivolts it recorded quite well a fluctuating reading. Also con firmed a max rich setting.

 

[alanu]

Propigation delay in a digital multimeter is way too slow.

Those tests are meaningless. Scope it and that'll tell you whats really going on.

 

[brianl703]

Use a multimeter with a bargraph.

 

[kc8adu]

sounds like its going full rich at idle.
post a scan of the test report.
if it is going full rich at idle it may have melted down the cat.

 

[andyd]

The original 4wire O2 sensor in my 528e pretty much hovered around .45-.55 volts measured by my Fluke 77 DVOM. The new O2, speced for a 5l Ford 3 wire bounced between .1-.9 volts. This is how the Bosch O2 sensor site said to test it. The 4th wire is just a drain wire. To adapt it, I simply hose clamped the original drain wire to the new base. I saved about 50$ with this adaptation. 2yrs and 50k miles later the car passed MA emissions with the original cat at 330k miles.

 

[JHZR2]

Quote:

---

sounds like its going full rich at idle.
post a scan of the test report.
if it is going full rich at idle it may have melted down the cat.

---

The O2 sensor that is in there is VERY new - 20-25k at most, and is OE. I doubt that it is causing some sort of issues - as NOx was an issue before putting this one in, and when I had the cat cut out and replaced, the OE one looked to be in perfect shape.

Unfortunately I dont have a scanner at home. The reports are as follows:
For reference, in 2003, my emissions (dyno test) were as follows:
NOx 1114 (standard is 1134)
HC 97 (standard is 146)
CO% 0.33 (standard is 0.82)
CO2% 14.1
O2% 0.2

Failure in 2005:
NOx 1167
HC 96
CO% 0.35
CO2% 15.3
O2% 0.3

Replaced cat converter, reading that passed in 2005:
NOx 1074
HC 133
CO% 0.34
CO% 14.8
O% 0.2

My failure earlier this week:
NOx: 2126
HC:161
CO%:0.54
CO2%:14.1
O2%: 0.5

And then after swapping in the OE chip, I got:
NOx 2700
HC 206
CO% 0.63
CO2% 13.9
O% 0.7

Thanks!

JMH

 

[alanu]

Do you notice your CO is increasing.

Drill a hole before the cat. Get a 4 gas analyser probe in there and at 2500 rpms you should get aproximately .5 CO (before the cat). Perhaps your car has a crude method of a/f management.

Your running rich at higher rpms under load. Your cat sounds like its NO GOOD.

With that same probe the shop can do an Oxidation efficiency test. It still may be OK in oxidation. A graph from the smog test should show the NOX is clipping going over and not getting reduced.

Look at your 13.9% CO2 that is a crude way of indicating efficiency of the cat.

Whats your base timing?? does it advance to "spec"????

Notice your HC's are getting high too. I bet your cat is getting below 80% efficient.

 

[unDummy]

http://www.troublecodes.net/articles/smog.shtml

 

[tmorris1]

Quote:

---

Use a multimeter with a bargraph.

---

A digital multimeter is probably too slow to see this changing signal. Use a scope or an old analog meter and watch the needle.

 

[brianl703]

That's why I suggested a multimeter with a bargraph. That bargraph can respond to signals a lot quicker than the digits can. It's at least as good as an analog meter, if not better because there's no inertia of a needle to worry about.