Hi all,

After opening a thread on rangerovers.net where I got absolutely no replies whatsoever, I was advised to come here and talk to you about my P38's issues !

So let's hope I'm in the right place !

I noticed a very poor mileage, loss of performance and unburnt petrol (and its smell) spat out of the exhaust.

Hooked up the Torque android app, P0174 - Bank B too lean.

Yet my P38 is overfueling badly and therefore running extremely rich (trims at 25 on bank B, constantly enriching to the max).

I changed the 02's, nothing changed ...(they had about 70k miles on them so...couldn't hurt anyway)

I check the voltages, Bank A does vary between .1 and .9V, bank B stays fixed at 0.04-6V so very low...

I'm suspecting that it's a wiring issue.. shorted wire etc...

How can I test the wiring harness ?

Thanks a lot !

Vico

I forgot : rover is a 99 4.0L Thor equipped with a Prins LPG System.
Only 2 precat 02 sensors, nothing post cat.

And the LPG computer in the engine bay has 2 wires soldered directly on the 02's signal wires about an inch before the engine ECU.

I would have replied over there but RRTH banned me for life so bugger him. 0V or a very low reading means a lean mixture so you will get a lean mixture error code. The ECU will be enrichening the mixture to try to correct it but the lambda sensor still shows 0V so it still thinks it's running lean. What you need to do is find out why you are getting the low reading. First thing I'd do is rule out the lambda sensor by swapping them side to side. If the problem stays with bank B it isn't a dodgy sensor. Then try it with the sensor unplugged, does it stay at 0.04V or drop to zero? Also check for a short to ground on the signal wire while the sensor is unplugged to check for a wiring error. You can always try the same checks on the other bank to compare what you find. If you suspect a short to ground in the wiring, unplugging at both ends (sensor and ECU) will confirm that or not.

The lambda signal connections for the LPG system aren't needed for it to run, just so the lambda signal can be displayed on the LPG diagnostic screen, so disconnecting them won't have any effect on the LPG running but will also rule out a short on that side of the wiring.

How stupid of me not to have thought of swapping the sensors...

Will do tomorrow.

I will also disconnect the LPG wires. If for some reason indeed there's a problem with the wiring or even the LPG ECU, this could be the cause.

I'm a bit of a tu*d with everything electrical... So, when I unplug the sensor should I be looking for a drop in voltage to 0 or hope it stays at. 04?

Also are there any checks or tests that can be performed while the car is off? I wouldn't want to ruin my cats (3000 miles only!)

Thanks a lot for replying!!

Richard might correct me on this but would think a 99 Thor might have a 5-0v sensor? Though OBD sometimes recompiles this to reflect 0-1v, but then on 99 year would expect most OBD2 readers to show 99% slow trims anyway...

Can we confirm this is a Prins sequential system, not just a Prins single point closed loop system....? The latter wasn't a patch on R90 and AEB based single point closed loop systems...

As memory serves (and I've used this disclaimer too often recently!) the most OBD2 compatible Rover V8 equipped vehicles I've seen have been post 2000 Disco's with EGR, these do show trims correctly over OBD2 in the usual +-25% range.

Simon

Glad to see you made it here, Victor. Told you you would get some answers here. You'll find riddlemethis is based in Dijon, don't know if that's far from you.

Gordon, am I on commission for the number I bring over from the other side? I make that 4 now lol

Gems have Titania sensors which need a 5v supply and work on a reference basis to that supply and show (in perfect working order) a range of volts from lean~0.4v to rich ~4v. Heater resistance should be around 5.7 ohms.
Thor have Zirconia sensors which generate their own voltage from 0 to 1 volt. Some OBD readers have problems with the GEMS (Titania) sensors and do odd things like transposing voltage to the 0 to 1 volt range, just to confuse things! Torque via Bluetooth and ELM used to do that on my GEMS.

Gordon, am I on commission for the number I bring over from the other side? I make that 4 now lol

Oh man wouldn't that piss on RRTH's chips, an Achievement Badge system for poaching his users!

With the sensor unplugged, especially if you unplug it from the ECU too, you should see 0V as all you are looking at is a length of wire that is disconnected at both ends. However, some cheaper digital multimeters will always display a voltage with the probe connected to nothing, the only way you can make then read zero is to ground both probes.

Simon and OB both got it wrong. GEMS Titania sensors are 5-0V with high, around 4.8V, being lean and 0.2V being rich (that's the range mine switch over anyway) whereas Thor (the bunch of bananas engine) use 0-1V Zirconia sensors with 0V, or pretty close to 0, being lean and 1V being rich. A 99 model will be XA VIN so will be a Thor (but could well have been built any time after September 98).

OB was right on the problem with GEMS and generic OBD readers though. As the standard expects a signal from 0-1V, that is what they display. Some simply display a 5V signal as 1V which can make you think that a lean signal is actually rich while others drop the displayed voltage and transpose it too so it is right. Thor, being fully OBD compliant will always display correctly as RRHSG has confirmed, low volts saying it is lean and an ECU that is trying to correct the problem by throwing more and more fuel in.

gordonjcp wrote:

Oh man wouldn't that piss on RRTH's chips, an Achievement Badge system for poaching his users!

I've already got the Achievement Badge, it's a lifetime ban from RR.net.......

That's the story of my week so far. Why break the habit! :)

Gilbertd wrote:

... OB both got it wrong.

OB was right on the problem with GEMS and generic OBD readers though.

Thank you all !

So in order :

• Yes, it's a Thor (plenum looks like 8 bananas)
• Yes it's a XA VIN
• Yes LPG has sequential injection, Prins VSI
• Yes they are 0-1V sensors, 0.1V being lean, 0.9V being full rich. 4 wire bosch sensors to be exact.

It's freezing today so I'll wait for tomorrow though.

Dijon is a bit far, I'm in Versailles (10 miles from Paris)..about 220 miles from Dijon

This is such a frustrating fault...A stupid electric fault on a perfectly working car making it useless...... Damnit

Been to Versailles a good few times. Beautiful little town.

I've driven past it about 120 times but I think I did stop there once (intentionally, not due to a breakdown). When I was at my mates last week we worked out that I've driven to his place just outside of Nice around 60 times in the last 23 years. At just under 2,000 miles round trip, that means I've done 120,000 miles in France and paid out just over 11,000 Euros in motorway tolls......

OldShep56 wrote:

Been to Versailles a good few times. Beautiful little town.

It's about an hour from my parent's place, really beautiful but not popular with the French!

HI all !

Did a bit more testing today.

I disconnected the wires from the LPG ECU that were soldered onto the signal wires coming from the lambdas.

Voltage from sensor A does vary correctly. Sensor B however...did go as high as .6V (which is new!), but then it drops back down and stays down...as show on the graph below.

(all my files are stored on my bar's website. 100% safe I promise :-) )

GRAPH

And here is the values registered with TORQUE, let me know what you think !!

Values

Ah, and I forgot, when I disconnect the sensor, the voltage drops to 0.00V and I get a code (forgot which one) that says "02 sensor, no activity detected). So that's good.

The fact that the readings are permanently low and when disconnected you get an error saying it isn't detecting it would all point to a dead sensor even if it is new. Presumably you didn't get time to swap them? Either that, or knowing how cold it's been here today and that the Paris weather is usually much the same as ours, it was too bloody cold!

But, the values from TORQUE look odd. The Lambda sensor outputs for bank B look how would be expected with a dead lambda sensor, as do the long and short trims (both going very rich), but why is bank A going fully lean? Do you have any means of resetting the adaptive values?

One other thought, is it running on all 8? If one cylinder on bank B isn't firing, the air being drawn into that cylinder isn't going to be burnt so the O2 sensor will see far more Oxygen than it expects (as they only detect Oxygen and not fuel) and assume a lean mixture. That will make it throw more fuel into the cylinders on that bank. So as well as having one cylinder being fed with fuel but not burning it and spitting it out of the exhaust, the other 3 will be getting far more than they want so the excess will also end up in the exhaust. Result will be poor economy, rough running and a smell of fuel in the exhaust. With 3 cylinders running very rich it will be running so rough you may not even notice that one cylinder is doing nothing. Spark plugs and leads would be the likely cure here.

Just noticed the word bar. That gives me an excuse to take the westerly route on the Periphique when I'm next driving down south. Although looking at the map, on the one occasion I went to Versailles, I parked up just around the corner (while the girlfriend of the time went to look round the gardens).

Gilbertd wrote:

The fact that the readings are permanently low and when disconnected you get an error saying it isn't detecting it would all point to a dead sensor even if it is new. Presumably you didn't get time to swap them? Either that, or knowing how cold it's been here today and that the Paris weather is usually much the same as ours, it was too bloody cold!

But, the values from TORQUE look odd. The Lambda sensor outputs for bank B look how would be expected with a dead lambda sensor, as do the long and short trims (both going very rich), but why is bank A going fully lean? Do you have any means of resetting the adaptive values?

One other thought, is it running on all 8? If one cylinder on bank B isn't firing, the air being drawn into that cylinder isn't going to be burnt so the O2 sensor will see far more Oxygen than it expects (as they only detect Oxygen and not fuel) and assume a lean mixture. That will make it throw more fuel into the cylinders on that bank. So as well as having one cylinder being fed with fuel but not burning it and spitting it out of the exhaust, the other 3 will be getting far more than they want so the excess will also end up in the exhaust. Result will be poor economy, rough running and a smell of fuel in the exhaust. With 3 cylinders running very rich it will be running so rough you may not even notice that one cylinder is doing nothing. Spark plugs and leads would be the likely cure here.

Just noticed the word bar. That gives me an excuse to take the westerly route on the Periphique when I'm next driving down south. Although looking at the map, on the one occasion I went to Versailles, I parked up just around the corner (while the girlfriend of the time went to look round the gardens).

Yeah It was about - 2°C today...pretty cold to be fiddling with wires in the engine bay outside the house ! And definitely way to cold to get under the car and touche those freezing pipes and sensors... eheheh will do as soon as temps get above zero !

It does look like a dead O2...I'll switch em and see.

Obvisouly I checked the plugs. On bank A they're good. On bank B, they're black...like nasty dirty black...so I bought some replacement irridium NGK. I was just waiting to see of the problem was related to the wiring before putting them in as I didn't want to destroy them....

And I still don't know whether the foul plugs are the cause, or the consequence of something else... !

God this is annoying (although I do enjoy the diagnosis part of things)

Cheers

RRHSG wrote:

Gilbertd wrote:

The fact that the readings are permanently low and when disconnected you get an error saying it isn't detecting it would all point to a dead sensor even if it is new. Presumably you didn't get time to swap them? Either that, or knowing how cold it's been here today and that the Paris weather is usually much the same as ours, it was too bloody cold!

But, the values from TORQUE look odd. The Lambda sensor outputs for bank B look how would be expected with a dead lambda sensor, as do the long and short trims (both going very rich), but why is bank A going fully lean? Do you have any means of resetting the adaptive values?

One other thought, is it running on all 8? If one cylinder on bank B isn't firing, the air being drawn into that cylinder isn't going to be burnt so the O2 sensor will see far more Oxygen than it expects (as they only detect Oxygen and not fuel) and assume a lean mixture. That will make it throw more fuel into the cylinders on that bank. So as well as having one cylinder being fed with fuel but not burning it and spitting it out of the exhaust, the other 3 will be getting far more than they want so the excess will also end up in the exhaust. Result will be poor economy, rough running and a smell of fuel in the exhaust. With 3 cylinders running very rich it will be running so rough you may not even notice that one cylinder is doing nothing. Spark plugs and leads would be the likely cure here.

Just noticed the word bar. That gives me an excuse to take the westerly route on the Periphique when I'm next driving down south. Although looking at the map, on the one occasion I went to Versailles, I parked up just around the corner (while the girlfriend of the time went to look round the gardens).

Yeah It was about - 2°C today...pretty cold to be fiddling with wires in the engine bay outside the house ! And definitely way too cold to get under the car and touch those freezing pipes and sensors... eheheh will do as soon as temps get above zero !

It does look like a dead O2...I'll switch em and see.

Obvisouly I checked the plugs (yet not the leads...I did however test compressions on all 8, all good).

On bank A sparkplugs are good. On bank B, they're black...like nasty dirty black...so I bought some replacement irridium NGK. I was just waiting to see of the problem was related to the wiring before putting them in as I didn't want to destroy them....

And I still don't know whether the foul plugs are the cause, or the consequence of something else... !

God this is annoying (although I do enjoy the diagnosis part of things)

Cheers