Oxygen Sensor Replacement

[LifterEyes]

I've been reading up on F.R. Wilk's website about using a generic Ford oxygen sensor as a replacement for the 944 by splicing it into the old wire harness.

I'm assuming that all Lambda sensors will output the same range of voltages for any given air/fuel ratio but will using a generic sensor such as this one operate in the exact same way as the stock replacement?

Basically I am just wondering if there is any compromise for using a cheaper alternative other then having to splice the cables.

[John_AZ]

I have the same feeling about the cross reference. I have not tried the Ford or others but those that have say they work. I will stay with the universal Bosch.

Here are a couple of links to further explain the sensor and installation:
http://www.babcox.com/editorial/ic/ic10044.htm
http://reviews.ebay.com/UNIVERSAL-OXYGEN-SENSOR-INSTALLATION-INSTRUCTIONS_W0QQugidZ10000000000991989
and if you have emmissions issues:
http://forums.pelicanparts.com/porsche-924-944-968-technical-forum/424848-early-944-smogging-up-world.html#post4116429

GL
John_AZ
1988 924S + 1987 924S

[SuperXRAY]

http://www.asashop.org/autoinc/may98/techtips.htm

O2 sensors are not completely universal. Using the wrong O2 sensor will, most likely, NOT make the car not run, but it will not run properly. If you used a different type of O2 sensor, you would probably need the equipment to measure (the lambda sensor charts) and the know-how to program a new DME chip to compensate for the changes. Even though they all use Zirconia, if I were going to pick an aftermarket O2 sensor...I would sort through many of them to find one with the same spinell layer. Hope that helps.

[John_AZ]

Good read^,

Follow up:
In addition to the above mentioned, the costly part of the O2 sensor is the platinum applied to the tip as well. Platinum is still over $1000 an ounce and platinum is used on better sensors because of wear and conductivity. The cheap sensors on EBay may not have any platinum. Sensors are still expensive because of last years runup in metals prices. I want the DME computer to get the best signal possible.

I replaced the 2 sensors on my 1998 Cherokee for the first time at about 90K. Cost $140 +-
I also know the rear sensor on the Jeep is to monitor the cat and may not have needed to be replaced but I did both. Porsche and Bosch suggest to replace every 60K or so.

Gl
John_AZ

[SuperXRAY]

Sensors past 1999 should be rated for 100k miles. Those from 95/2 (in most cases) to 99 are usually rated for 60k. That isn't a blanket statement, but it is generality. Most replacement ones for pre-95 years are updated to 60k.

[nynor]

i used a ford sensor in my 951. it is a bosch. ~$40.

[HondaDustR]

See if you can find a Bosch 15725 sensor. I've seen it available at autozone for like $50. Read this thread on RL. http://forums.rennlist.com/rennforums/924-931-944-951-968-forum/42959-o2-sensor-cost-and-replacement.html

There's a Bosch guy that knows his stuff and figured out the correct universal sensor.

[nynor]

i think i bought part number 13913. but this guy is saying bosch part number 15725 is the way to go.

[LifterEyes]

Thanks for the link HondaDustR,

It seems like the Bosch 15725 is the 'ideal' direct replacement Sensor for the 85.5+ 944's. It almost sounds like it's even better made then the stock one according to that forum, haha!

I just want to make sure that, if I do choose to use the 15725 replacement, it will function exactly how the original replacement would. So it sounds, according to that forum, that this sensor has the same ceramic, heater, and protection tube as the original and it should work exactly the same if I were to guess.

Also, SuperXRAY posted a link to an article that mentioned that the number and the design of the slits are also a big factor as far as compatibility goes. Does anyone know anything about any differences between the slits on the original and the 15725?

[Dave L]

I just ordered the 13913 today and it should be in this afternoon. I decided to go with the FRwilk opinion that the 13913 is cheaper. I am perfectly able to join 3 wires together so the "universal" kit (which includes connectors) really isnt worth it to me. My car is still in the barn so install will be next month. For now it will be added to the small pile of parts to be installed.

With the money I "saved" I think I might buy some plug wires

[HondaDustR]

The advantage to the universal kit is the wire splicers are supposed to be waterproof. And I'm sure you probably already know not to solder the wires.

[Dave L]

I will use heatshrink tube and yes I know not to solder.

[LifterEyes]

How much will a bad O2 sensor effect Idle quality? I'm trying to diagnose the source of my rough idle issue and I am pretty sure that the O2 sensor needs to be replaced anyway, judging from the smell of my exhaust. I Just wanted to get some opinions before I start spending money.

[John_AZ]

If the oxygen sensor is bad and does not send a signal to the DME computer, the car will run "rich", idle poorly and fail emissions.

Another part to consider if you have idle issues is the engine DME temp sensor II (top of block behind/near the rubber intake boot. You can test resistance to check. It sends a signal to the DME computer to adjust fuel as the engine warms up.
http://www.clarks-garage.com/pdf-manual/elect-19.pdf

GL
John_AZ

[wild man]

Originally Posted by nynor

i used a ford sensor in my 951. it is a bosch. ~$40.

I thought ford used bosch. I remember that from when I had the 88 t-bird turbocoupe. So are the fuel injectors (but with ford #'s on them) And the AFM (ford called it a VAM, or vane airflow meter) is the same unit (slight # difference, probably because it had a round 3" inlet adapter attached) as on the 951. In summary, it used a bosch fuel injection system.

[LifterEyes]

I believe I've checked both temperature sensors in the past and, from what I remember, they checked out OK. I had an emissions test done not too long ago and it seemed to pass with no problems. I did change the sparkplugs recently and just checked them now after they have been in the car for a few weeks. The plugs had a pretty bad carbon buildup all over them already. I'm going to check the coil soon the make sure there is a strong enough spark. If there is a good spark i'll probably go ahead and replace the O2 sensor and see if it begins to run better.

Here are the results to my emissions test as well. Can this be used as any indication of the problem?

[HondaDustR]

I actually did a bunch of testing with a voltmeter hooked to the O2 sensor and driving both connected to the DME and not connected. The tests with it not connected were by far the most interesting, since the sensor gives a very stable reading when the DME isn't constantly compensating for what it senses. This test originated from the fact that I was driving with the FQS set to -3% fuel for awhile and just got curious. I tested with the FQS at normal, +3%, and -3%. What was pretty much universal was that a lean mixture gave a crappy idle compared to a richer mixture. -3% was registering only about 0.15-0.2v from the sensor (all tests were done with the heater energized). I got brave enough to switch the FQS on the fly, and the idle instantly improved when set richer. I saw about 0.25v on normal with a decent idle and ~0.8v on +3%. The 0.8v idled the best. This is all with the DME in open loop mode. When the DME got a signal from the O2 sensor, the idle wasn't quite as smooth, and proportionally less smooth compared to the degree of leanness the base maps were set to, i.e. with the more compensating the DME needed to do. Having the fuel mixture bounce around, which it does with the O2 sensor plugged in, will affect the idle somewhat (the O2 sensor returned both a very consistent, stable voltage and responded very quickly to changes when there was no DME intervention). It's just an inherent weakness of the system, especially since the updating rate of the DME is much slower at idle, since it just doesn't get as many crank signals from the speed/reference sensors as it does at speed.

What that boiled down to as I think about it, vacuum leaks can really screw with the idle, which is good, because they have to get pretty bad before you'll notice a problem at higher revs, meaning they're easier to diagnose by the idle. If the O2 sensor is old or incorrect, the response will be slower, which will make the DME take much longer to cycle rich/lean, meaning an even more erratic mixture. Keep in mind, when running in closed loop, the Motronic will swing lean and then rich and then back to lean, changing from one side to the other only when the O2 sends a rich, then lean, then rich signal. It is a slave to the O2 output voltage as long as it gets one. Without the O2 voltage signal, it just does what it was told to do by the fuel maps. The O2 system was tacked on for exported cars and not really well designed. It does not run excessively rich without the O2 sensor, but the O2 sensor can compensate fairly well for minor mixture problems, like vacuum leaks, clogged injectors, a dirty AFM, bad plugs or distributor points, etc. Exhaust leaks between the engine and the sensor can also affect it. I always like to run without it if I suspect anything's wrong. Another good indication is if you feel any discrepancies between high throttle and WOT, since the DME goes into open loop mode past ~2000 rpms once the WOT switch is activated, and will run either better or worse depending on the condition of the sensor and what problems it might be hiding. Check everything else to ensure it's working right, then suspect the O2 sensor.

[John_AZ]

Very thorough information on the O2 sensor and the signals sent to the DME.
I am still trying to picture you changing the FQS while driving.

Nice writeup.
John_AZ

[HondaDustR]

I didn't actually change it while driving, although I thought about it. Just while sitting and idling.

Another interesting observation is that it seemed that the DME also defaulted to open loop mode past about 5500 rpms even at part throttle with the O2 sensor plugged in.

[LifterEyes]

Yeah great writeup. I really appreciate the help. I just tested all the lines for vacuum leaks using the FR Wilk article as a guide and found no leaks. I also tested the spark and found it to be a little weak. I'm going to try to fix that first and see if that helps.

Thanks again!