Strange O2 sensor graphs Jeep 4.0l, ideas on diagnosis?

[Duncan]

Hello, I am working on a 1996 Jeep Grand Cherokee with a 4.0l in it, and I don't understand what I'm seeing on the )2 sensor graphs that the MODIS is producing. Here are the details of the vehicle:

This Jeep has a "lower mile" 1992 4.0l that was recently swapped in, and ever since the Jeep has had poor mileage and less power. The replacement engine was pretty bare, so the Jeep still has all the wiring, sensors, manifolds, injectors, fuel pump, exhaust, and accessories that it had before the engine swap. It does not pull any codes now, nor has it ever pulled any codes.

Both engines were/are stock and unmodified, except for having a throttle body with the restriction bored out, and 4-hole injectors to replace the OEM 1-hole injectors. The throttle body and injectors were changed years ago and they did not affect mileage or power noticeably at the time.

A look at the O2 sensor output graph showed a strange pattern with a lot of latency and irregular crosscounts, so I installed a brand new O2 sensor of the same brand (NTK) and part number. Surprisingly, the O2 sensor output graph with the new O2 sensor installed is only marginally better, if at all! I have uploaded videos of the OLD sensor graphs and the NEW sensor graphs here: https://photos.app.goo.gl/AkQnQegdCZft1KvF7

I should note that the STFT does hover around 0 (whether at idle or higher RPMs), and the LTFT ends up around 8%, which I don't think is too strange given that it doesn't have OEM injectors and that's less than 10% off. Perhaps I am wrong to dismiss this info though??

What do these O2 graphs mean, or, what else should I look at to diagnose the problems with this engine? Thanks!

[greasybob]

It's kind of hard to see the 02 pattern on the scanner because of the scale. It starts at 5 volts which makes it hard to see things switching at 250 to 750 mV. I noticed in shop key that a 92 4.0 is listed at 8.8:1 compression and a 96 is listed 8.7:1 compression. Maybe that's nothing but are the cylinder heads and the cam profiles the same ? It would be better to scope the 02 for a more accurate view of things. To test an 02 sensor you could try to force the engine rich (with propane or similar) or lean ( forced vacuum leak) to see if the 02 responds.

[greasybob]

If low power or fuel mileage are concerns I would be more interested in seeing the 02 and fuel trims from a test drive when the engine is under load. If That scanner is capable of recording that, the data could be uploaded here as a picture or a file using shopstream.

[Duncan]

Hello Greasy Bob, and thanks for your reply. I have forced the mixture rich with propane and it worked great, the STFT responded instantly and of course the LTFT slowly responded too. And then I took the propane away, and I could easily see everything settle down, back to normal.

Yes, I wish I'd reset the MODIS at some point to imrpove the resolution of all graphs!! A mistake for sure, but I was trying to capture the complete cycle from warm-up to warmed up, revving and at idle. But yes, it makes it hard to see anything when the scale is so large.

The cylinder head profiles are pretty similar between the 92 and the 96 from what I understand, but the cams are quite different. The 96 is a dual-pattern cam that makes max torque at 200RPM lower. Either way, I figured that the O2 crosscounts should be a lot faster, and the PCM should be commanding "LEAN" and "RICH" states equally, rather than the "lopsided" crosscount distribution seen in the videos.

That said, the fuel trims while driving are essentially the same as I have posted in those two videos. I can save files on the MODIS but I don't know how to transfer a file from the MODIS to anywhere else. I've never used shopstream before, so I'll have to install it on a laptop and see if I can figure out how it works.