Lambda unsuitable for LPG?

[Morat]

I need a new 02 sensor for the P38 Range Rover, I was looking for a Bosch but saw this on the page
"Unsuitable for vehicles running LPG" Is this just a way to wiggle out of any warranty claims or is there something in it?
http://www.island-4x4.co.uk/oxygen-lamb ... -7771.html

I don't see what difference LPG could make to an 02 sensor... but I've been wrong before!
Does anyone know the truth of this?
thanks!

[Gilbertd]

It's a get out. A single point system Tee's into the lambda sensor output and uses that to control the amount of gas. A multipoint like you have lets the petrol ECU control the fuelling and the LPG ECU just adds a fiddle factor to the injector pulses so isn't actually using the lambda sensors, the petrol system is.

[LPGC]

Agreed with Gilbert that it's a get out.

But it does beg the question why the company see a need for a get out... I wonder if they got fed up of customers who found fitting new lambdas didn't cure their LPG problem wanting to return them. In which case the customer may also wish they hadn't bought a new lambda unnecessarily but had instead had a diagnostic done by a decent LPG pro to identify and sort the real problem! Not aimed at you or anyone we know Morat, just a bit of general advice for any others reading this who might otherwise have taken their LPG issues to a none LPG specialist. Lambda probe issues are one of the most common to be misdiagnosed, the first part to be swapped willy nilly by those clutching at straws.

Simon

[Gilbertd]

Morat's P38 is the later Thor engined version that uses 0-1V sensors so wouldn't be a problem anyway, but the earlier (up to 1999) GEMS engined cars used 5-0V Titania sensors which don't play nicely with most lambda controlled single point systems. So anyone ordering new sensors for a GEMS thinking it will solve problems with a lambda controlled single point will only make them worse. It'll run fine on petrol as 5-0V is what the GEMS ECU is expecting to see, but like a dog on LPG as it isn't. Hence the logic behind the get out.

[LPGC]

I nearly mentioned the same, Gilbert, but didn't because (as read above), Morat's P38 runs a sequential system, it may well have 0-1volt lambda's anyway, most recent single point ECU's can use the signal from 0-1, 0-5 or 5-0 volt lambdas (some can use 0.8-1.6 lambda signals too). Relatively simple to make a little electric circuit to give 0-1 volt output based on 5-0v input for controllers without 5-0v capability (and easier for you or I to do that than weld and wire in a 3rd pre cat lambda). On that subject I also nearly said I wondered if Island 4x4 had taken advice from RPI, who at least at one time used to include a 0-1volt lambda probe and boss to weld into the exhaust with their single point P38 kits..

Good idea of Landrover's to Titania sensors on a 4x4 as they don't need at atmospheric oxygen reference (so no hole to let in water to damage the probe). Been a while since I changed a lambda on a P38, changed dozens on them in the past but didn't think it was as clear cut as all Gems using 0-1volt and all Thor using 5-0volt.. If you know different I'll take your say so over my partial memory on this one!

Simon

[Gilbertd]

but didn't think it was as clear cut as all Gems using 0-1volt and all Thor using 5-0volt..

Wrong way round, 5-0V on GEMS and 0-1V on Thor but there are two different ones on a GEMS, '95-'97 (ERR1834) and '98 (AMR6244) although both are 5-0V. No idea what the difference is though other than the colour of the plug which is the easy way to tell one from the other. You can't get them mixed up with the later 0-1V Thor ones though as the GEMS have a 12mm thread and the Thor an 18mm thread.

[LPGC]

[quote="Gilbertd"Wrong way round[/quote] Yeh I got them the wrong way around.. Maybe due to a bit of sheep syndrome, I think someone else recently got them the wrong way around. Checking lambda voltage at the ECU in it's box near the battery, orange and blue wires middle of the pinout 5v lean 0v rich, only relevant on Gems because on Thor the ECU is on the nearside, it's loom uses different colour sheathed lambda wires, more difficult to access. I have the info on which wire corresponds to which cylinder bank / side of engine stored on laptop and phone but got another phone just the other day and another laptop a few months back.. Mind you, time was I was changing a fair few lambdas on P38's so I scratched the info on the rear of the garage door, still have a couple of different types for P38's on the shelf. Owners are often unaware a lambda has gone as UK spec P38s without EGR have the MIL light disabled and they seem to run at least acceptably well (on petrol at least) with all sorts of problems.

Started conversion of yet another Rangerover L322 this morning, this one for a Londoner who's a mate of the Yorkshireman who's L322 I converted week before last.

Simon

[Morat]

Heh, seems my P38 is getting a bit well known!

Simon, if you remember you said that there were some mixture related codes pending when you converted it. Your best guess at the time was that the MAF might be on its way out but you said I should check it on dedicated diagnostic gear.

Since then I've had some fun and games with the immobiliser which prompted me to buy a Nanocom Evo. This is a great little tool which can read everything on the vehicle. The Nanocom came back with some "adaptive values out of range" codes and O2 sensor fault on bank 2.
This is nothing to do with the LPG conversion, these codes were generated while running on petrol after wiping all the existing codes.

Anyway I reckoned a new o2 sensor would be the correct course of action. Once I've got the Petrol side running properly again, I'll give Simon a bell to see if the LPG map would benefit from a little tweaking to take account of the different O2 values and petrol adaptation.

[LPGC]

Doh! Sorry Morat, it had slipped my mind I'd converted yours, not long ago too! Converted a couple more Rangerovers since then, L322's though (just doing the paperwork for the last one)...

As I remember on yours I found long term fuel trims maxed out positive but plenty to go at on short term trims.. I know P38's aren't all fully OBD2 compatible and to expect unusual readings but high positive trims seemed true on yours as the petrol injector duration was also higher than usual at idle (no engine probs it seems, as it idles at around .4 bar manifold pressure as they usually do).

First thoughts were air flow meter but I think I later told you it wouldn't be due to a problem with the air flow meter, it would be a problem with petrol pressure.... This is because if the problem were afm related, the engine would still need the same as usual petrol injector pulse time at idle but fuel trims would be incorrect. Your fuel trims are indeed incorrect but since so too are petrol injector durations, and all the way through the rpm/load range, on both banks of cylinders, this points more to low petrol pressure. Not a petrol injector problem, as all petrol injectors seemed to flow around the same (when switching individual cylinders between gas/petrol - a check I usually do to check LPG injectors but works equally well, in reverse, for petrol injectors).

To achieve smooth changeover and fuel trims that shift as little as possible when switching between fuels (for accurate fuelling immediately after switching fuels), I calibrated your LPG system just slightly on the rich side.

The 'adaptive values out of range' error is the same error I will have seen and would have picked up on in any case (even without an error code coming up) due to high positive fuel trims. I don't remember noting a problem with an O2 sensor (unless I said different at the time?).. The O2 sensor error could point to a problem with an O2 sensor or could be an error generated due to the fuel trim issue.

Despite the issue/errors, I think I said your engine was running closed loop as it should (so running correct mixture), and was going rich at high load as it should, on either fuel... Sorting the issues/errors probably won't be of any practical advantage except maybe for your peace of mind! It would be a bit of a pain to change the petrol pressure regulator, it is inaccessible without removing the top half of the inlet manifold.

Slightly off topic - MIL lights are disabled on all UK spec P38's, at least all but the very late models... and a fair percentage of none LPG converted P38s seem to have some sort of issue that would turn on the MIL light if it were not disabled! Fuel pressure problems are not common but air flow meter and lambda issues are.

If you do effect anything that sees fuel trims go back to standard (perhaps by changing the petrol pressure regulator) it would be a good idea to come back for a check/adjustment of calibration - I'd calibrate it so trims are exactly the same between fuels, rather than slightly on the rich side for LPG. Worth saying here that where I say 'on the rich side', your engine isn't actually running rich.. Due to the closed loop system it is running correct mixture on either petrol or LPG. But since long terms fuel trims are so high, it wouldn't take much of a change to any aspect (sensors etc) to see short term fuel trims also maxed out and if that happens the engine could run lean. By calibrating the LPG system slightly richer than petrol, the engine still sees correct (not rich) mixture on LPG but fuel trims don't need to be as high to achieve correct mixture - In effect, I just calibrated on the safe side, so running on LPG the engine is less likely to run lean in the event of any changes to sensors etc than when running on petrol.

Slow fuel trims allow plus or minus around 20% tolerance to mixture (for stuff like differences in fuel pressure and sensor etc), then on top of that fast fuel trims allow about the same.. The fuelling seen by the engine is the slow trim plus the fast trim, so there's about a 40% tolerance in total. Your slow trims are maxed out but fast trims are far from maxed out - not an ideal situation, and OBD may report errors as a result, but P38s are older vehicles and don't go into 'limp home mode' or 'open loop mode' (cease trying to correct mixture) when such errors occur... With the issues seen, your engine continues to run correct mixture.

Regards,

Simon