AC Delco Original Equipment versus Professional

[AtlWrk]

I finally replaced my marginal thermostat (no codes, but big temperature dips on acceleration and during warm-up that take a long time to recover). However, I think I may have waited too long.

I scoped my upstream O2 sensor last night and I'm getting <1/2Hz (more than 2seconds/cycle).

I know O2 sensor replacement has been discussed ad nauseum but:
I'm shopping for a replacement (I know to stick to AC Delco) and I see to two part numbers:

2133539 - Original Equipment
2134348 - Professional

What's the difference? Catalog says both apply to 2006 MY. Does the "professional" sensor justify twice the cost? Is it a durability issue?

Also, would a clogged cat converter itself cause the sluggish readings (and the sensor might be fine)?

I plan on doing an exhaust back pressure test while I have the sensor out but I'd like to have a replacement sensor on hand in case my only option is cutting the wires (split socket slipped once already--gonna PB Blaster it for a few days).

 

[Mounce]

The upstream sensor is supposed to fluctuate. Not sure if you're worried that it's fluctuating too much or not enough, they are known to get lazy as they age though. Other than that, I don't think a clogged cat would make it read different unless it just didn't let enough air volume through to get a good reading.

Down stream sensor should be pretty steady though.

 

[AWD V8]

I've had very good success with NGK or NTK O2 sensors. Never a problem.

I'm having concerns over poor quality on some AC Delco replacement parts and I'm not likely Bosch at all anymore.

Your O2 should graph like this. Old O2's do get lazy and the response time slows down.

 

[Robbabob]

I wouldn't replace an O2 sensor unless you are getting a code that it is bad. I would check the flow through the cat, though, and be prepared to replace it.

 

[AtlWrk]

Sorry, I should've been clearer.

My O2 sensor is fluctuating at the proper levels (~.1-.8+V) it's just doing so too slowly. Everything I'm reading says the signal should fluctuate around 1Hz or faster. Mine is only running at 1/2Hz (a little more than 2 seconds/cycle) telling me that it's getting lazy.

I didn't really mean to turn this into another O2 diagnosis thread.

Really what I was getting hoping for was any insight on the "Professional" versus "OEM" AC Delco parts??

 

[bobdec]

Strangely. The 'OEM' are GM parts and sold by GM Service Centers/Dealerships. 'Professional' are aftermarket and are generally directed at parts stores. So you would think Professional would cost less, just looked at Amazon and the O2 Pro is double the price of the OEM. Both are warrantied for 12 Months.
Also noticed Denso who I have always used for my O2's (They actually manufacture the Delphi and Delco O2's) now has 'OE' and 'OE quality replacement' part numbers for the Trailblazer. So even they have gone to this confusing process.. Denso prices are only a few bucks difference between OE and OE replacement.

 

[AtlWrk]

Thanks Bobdec That's exactly the kind of info I was looking for.

 

[bobdec]

Forgive me it this is redundant...Not to many people actually scope their O2's. I have an old beater upper I keep in the garage..

The frequency you are looking at on the scope is a combination of the O2 switch speed and the PCM adjusting the mix to keep the voltage average at about .450 volts. Under normal operation the PCM tweaking the injector duration keeps the frequency at about 1Hz. However a very slow switching O2 (hanging in high or low voltage state) could slow the frequency by delaying reporting the fuel burn to the PCM soon enough.

Below is the specifications for Delphi O2's . Notice the O2 switching response times, they are less than 100ms/80ms so in essence they are a 10% contributor to the 1000 Ms (1 Hz) voltage frequency. The PCM adjusting the injector pulses controls the other 90%. If your scoping the actual O2 pay attention to the slope of the plus and minus wave forms, an extended slope between 200 and 800Mv is an indication of a lazy O2. If the slope has steps or hangs at a up or down level under 'stable RPM' conditions that also indicates an O2 going bad, it s/b acting like a slow analog binary switch, either on or off, with smooth transition between states as the PCM changes the mixture.

Note this a scope procedure, a scanner is not as reliable as it redraws the picture from a digital file dump (log data) from the PCM, collates and organizes the digital data and then sends it to a display to re-paint an analog picture of the O2 signal. Fast scanners are better, the torque app is slower than most. the conversion can distort the actual O2 frequency, signal and amplitude.


Rich voltage >720 mV
Lean voltage <120 mV
Rich to lean response time <100 ms
Lean to rich response time < 80 ms
Static lambda 1.005 (± 0.003)
Sensor light-off (450 ° C, as above) <12 seconds
Sensor time to activity (cross 300 mV) <8 seconds
Heater power at 13.5 V, 450°C exhaust 7 watts
Heater in-rush current at 21°C, 13.5 V <1.7 amps max

 

[AtlWrk]

Fantastic information!

That makes perfect sense about laziness showing up as a slower transition rather than overall frequency. I re-ran it again last night after a little more research and I'll probably do it again tonight with this info in hand. I'll figure out how to grab screen shots too.

I was originally scoping at idle but when I bump it up to 1500-2000 rpm the trace is just about perfect IIRC(amplitude, frequency, shape all spot on with no blips). When I drop it back down to idle the slow response returns (even though the transitions still seem as snappy) but I also get blips that look like missing pulses from the square wave (both high and low). Maybe the PCM manages idle differently?

 

[bobdec]

Love theoretical questions

Now that you mention it most O2 info I've seen/read usually shows engine at 1500-2000 RPM. Hardest part is trying to look at the results without having the software design documentation available. I seen some reverse engineering on timing control, and the flow chart some guy developed by picking through the code ended up with 20 pages of how a PCM adjusts timing. Things like it takes 256 knocks detected and pages of flow charts to bump the knock count 1 increment. PCM speed to engine fueling control is like us watching a snail cross the street.

Here is my WAS (wild ass scientific) guess... at 600 RPM that's (600) / (2) = (300 power cycles per cylinder per min ) / (60) = (5 power fires per cylinder per second) X (6 cylinders) = (30 total power strokes per second). Usually the PCM does things incrementally to avoid any sudden changes and to prevent oscillation back a forth caused by over corrections. At 600 RPM the PCM only has 30 shots (only 5 per cylinder) to tweak injectors to go from rich to lean and back to rich if it was to maintain the 1Hz frequency. IMO maybe the code sees the engine changing so slow it just slows down the injector pulse width increment/decrement process to get more samples and wait for the results. This would have to be verified on a few different 4.2's to be of any validation.