Has anyone actually tried to scope or use a voltmeter to read the upstream sensors in a 2000 XJ8? When scoping all I get is lots of noise around 50 mV peak to peak. When measuring from chassis ground to either of the sensor, not heater, pins I get 3.8 volts DC, regardless of engine RPM. IN fact I get the 3.8 with the engine off, and the ignition switch in position 2. Disconnecting harness plug EM-82 at the ECU removes the 3.8 volts, so it appears it is not a harness leakage problem.
I saw in another post that someone said the upstream sensors in this year car are a constant voltage type, but when I removed them and tested with a propane torch, they behaved just as any standard 02 sensor would, giving about .9 Volts when hot.
Any help is appreciated, as I'm trying to determine if my ECU is kaput.

 

I used my Tek 7704A to no avail. The upstream linear sensors are constant voltage in the Jaguar implementation, a scope will show you only the noise you have seen. You could try a current coupler but the best way would be to use a diagnostic tool which can show the upstream O2 sensor responses graphically. It might also give you some codes to give you clue (if you haven't got them already). I used an ELM-323 interface from Gendan and a laptop. Worked a treat (no affiliation).

Measure the black wire couple with an ohm-meter for about 6 ohms or so (you can do this at the plugs, don't need to remove anything - oh, you did already). Open circuit heaters is not uncommon. Don't measure the blue/white pair as you risk damaging a good sensor with the wrong kind of meter.

HTH

Jim

 

Jim, thanks for the reply. Well I took the upstream O2 sensors out and tested them as if they were the common, non-constant voltage types, and when heated with a propane torch, they put out about .9 Volts, and of course drop off to zero when cooled. I don't think the constant voltage (wideband) types would behave that way.

ideas anyone?

 

OK, can we get a bit more info? What makes you think your ECU might be kaput? Any diagnostic codes? CEL? What colour wires are on the sensors? I'm not sure linear sensors themselves are constant voltage - it's the implementation of the ECU interface which determines this. The fact you get volts when hot doesn't necessarily mean that they are faulty or the wrong type. Can you post a pic of the sensor and I'll compare it with my old ones?

Jim

 

Hi,
The main symptoms with my 2000 XJ8 are that one bank runs about 22% rich and the other runs about 21% (longterm fuel trims) lean. There are random misfires on all cylinders. Fuel pressure is good. Vacuum is about 17 inches at idle and steady. Compression varies from 165 to 185 on all cylinders. All 8 COP have been substituted with known good ones at once. No vacuum leaks, no injector problems. Both upstream O2 sensors have been replaced with known good ones from my 98 XJ8. I also tested the original 02 sensors out of the car with a torch and though one might be sluggish they still pass the test.

Trying to backprobe and measure the upstream O2 sensor voltages gets me no where, as they always read 3.8 VDC with respect to chassis ground, regardless of engine speed. This is on both sensors. I think these are not referenced to ground or even a signal ground.

Do you agree? Does this ECM rely on a change in 02 current instead of voltage to determine mixture condition?

Using the A bank as an example, and looking at the ECM pinout, at EM82-4 I see that it is an input and EM82-10 the other connection to the A bank upstream sensor is labeled as an output.

Trying to scope from ground to any of the sensors just reads noise.

Disconnecting either ,or both upstream sensors I still measure 3.8 volts at the sensor socket. Disconnecting plug EM82 at the ECM and remeasuring at the 02 socket, the voltage is now zero, indicating there is probably no leakage in the EM82 harness section.

Any advice is greatly appreciated. I bought this car 4 months ago and have been working on and solving problems since, while slowly going broke.

 

... but do you get any diagnostic codes? Have you measured the sensor heaters for continuity?

Just a few thoughts as I am just on my way out. Please sanity check first.

Don't think '98 and '2000 upstream sensors are interchangeable. If they look the same one set of them is possibly wrong. I know my '99 Sovereign sensors were different from '98 ones - Jaguar changed the sensor arrangements in the 99MY updates to constant-voltage linear response sensors. The old upstream (step response) sensors were moved downstream of the CAT. '98MY sensors WILL NOT DO. You should confirm you have correct sensors BEFORE you do ANYTHING else. Again - please provide a picture if in doubt for confirmation.

LTFT is as much the responsibility of the downstream sensors and their interaction with the upstream ones from the ECU's point of view (the delta is also used to indicate CAT efficiency). Might be worth factoring the downstream ones into your thoughts once you actually know what is fitted upstream. Previous owner cheap skated perhaps or just got it wrong? ECU does have a fail-safe for duff upstream sensors.

Changing over the sensors between banks (i.e. crossing the plugs over on top of the auto box) and seeing if the LTFT shifts bank will give you a clue (might try this on downstream pair as well but not at the same time).

Trying to diagnose with volts on the upstream sensors will get you nowhere. Occams razor - the simpler cause is the most likely one. Sensors degrade over time even if they don't fail and although ECU failure is not unknown sensor problems are more likely (provided of course that if you find you have incorrect upstream sensors they haven't toasted the ECU).

Jim

 

It would make sense, as these are constant voltage sensors, that the ECU applies a constant voltage to the sensor (3.9V, as you have described), and then measures the current flow thru the sensor...

The 0.9V shift you see when heating the sensor is into a very high impedence, and will likely just change the current flow a bit when driven by a low impedence voltage source (such as the ECU).

Try measuring current with a VOM or with a current probe on your scope.

 

The codes are P0300 thru 0308 random misfires, and 0171 bank 1 lean, and 0175 bank 2 rich. Sometimes 1314 cat damage bank 2 ( when I took the cat off, it rattled, so I'm sure it is shot), and once in awhile 1316 either misfire exceeds emissions or injector monitor. I think 1316 only occurs after I clear the other codes, before the injector monitor has a chance to run through its cycle.

Today I finally got smart and went to bosch.com and confirmed what you've said about the upstream O2 sensors in fact being wideband. No where else could I find this, as the on-line parts suppliers incorrectly listed the upstreams as being standard step type.

So I've ordered 2 from Amazon for $63 each (almost $600 each from my local Jaguar dealer, which by the way would not call me back and verify that they are indeed widebands).

I would post a pic of the originals but I don't see a way to do it here.

Thanks again to those of you who have responded with ideas, and I will let you know what happens when the new ones arrive later this week.

Bob

 

My 2 upstream wideband (constant voltage) O2 sensors arrived from Amazon, and they fixed the final problem! Since then I've put on a few hundred miles and the fuel trims move around slightly above and below zero as they should. NO DTC codes of any sort. So what's the lesson? Well one lesson is that the ECM in the 2000 XJ8 VDP does not always diagnose a defective O2 sensor, as it never gave me a DTC with two defective upstream O2 sensors installed. Another lesson is that one can't tell by outward physical appearance whether the sensors are the old step type, or the new wideband type. JTIS gave me no indication either. Yet another lesson is that parts jobbers list the wrong sensors for this car. Bosch.com correctly listed them. Finally, heating with propane and then cooling the old defective sensors showed them behaving as if they were the old step type, causing me no end of confusion!

 

Thanks alot for the info linuxenigma. It is good to know. Just to make sure of one point though - the upstream o2 sensors are constant voltage wideband and the downstream are the older style step type - is this correct ?

 

Glad you are sorted!

Just FYI, both of the O2 sensors were u/s on my Sovereign when I bought her and there was no ECM warning light displayed either. It greatly puzzled the local ******* as well. Never did find out the reason.

The Denso wideband sensors do have a different can with a different hole pattern in from the downstream 'step' sensors so you can tell if you know what to look for. Not sure about Bosch or others. They have grey plugs as well (I think the downstream ones are black but not certain). Again if you don't know it would be easy to get the wrong ones.

As you say they are often listed incorrectly except by parts shops or breaker who know what they are doing. Even then you can get caught - I bought a replacement pair of 'used but good' sensors which had been tested with a propane torch and voltmeter. The test was invalid although the sensor part was probably OK except they didn't think to test the heaters and they were both o/c so of course the ECM would just have disabled them, output a P1646 and P1647 again and fallen back to the default map.

Jim

 

To clarify, on my 2000 XJ8 VP with AJ27 engine, the upstream O2 sensors are wideband constant voltage types. As far as I know the downstream are step type, and looking at the ECU schematic pinout seems to confirm that.

 

Can anyone shed any light on this situation… I have a 2000 XJR & I keep getting a P1646.This is the only code. I have over the past year replaced all of the O2 sensors & knock sensors. The check engine light will not go off. We have tried swiping out computers & sensors from other like cars and it still will not go off. Once reset, it will stay off for a while then after 1-2 starts, it comes back on. We’ve also tried disconnecting the battery as well. I’m at a loss…..

Please share your thoughts.

 

Some thoughts in NO particular order:

You have an ECU problem

You have dodgy wiring for that sensor

If you have changed for an exact match (not a cheap eBay universal O2 sensor) you have a problem with one bank which is contaminating the sensor

Swap over the sensor plugs and see if the code changes to P1647 (the other bank) or not

Check the heater relays for that sensor (perhaps swap with the relay for the other bank and see if the fault code changes)

Jim

 

Thank you Jim,

We have preformed all of the following:

Replaced ECU, All O2 Sensor's 4 times, Replaced all of the relays...the only thing that i'm seeing that we have'nt tried is replacing the wiring.

 

Hi, I saw this and thought of you and your problem, might give you another approach.

Error code P1647 - Result!

I don't generally cross-post links but since you seem to have hit a dead end....

Jim

 

The MAF sensor and O2 sensors work together to control emissions. Even though you don't have a MAF code, buy a can of MAF cleaner and remove and douse your MAF. Use only MAF cleaner and be sure to clean to clean the two platinum wires that are in the MAF. The ECU in these models often don't show a defective or dirty MAF code.
Good luck!