Simon I have monitored the lambda sensors in the past live with a code reader. It seems strange to me that starting the van from cold both bank 1 and bank2 sensors show 0v leaving. This is when the van was ticking over, even when the engine was revved up. It took about 3mins before bank 1 sensor showed about 0.7 down to 0.04. At this time the bank 2 sensor began to read about the same. As the engine heated up to normal working temp and switched to gas the bank 1 sensor dropped out and the bank 2 sensor was working only. The readings where about in the range as when it was running on petrol.
I had a fault code come in before stating a low variable on bank 2 sensor. So I have replaced that sensor and the state of play where the readings above I got after the replacement. I have seen the lambdas on other cars on petrol have a reading on the sensor instantly when starting from cold. So why does my first bank 1 sensor not read any voltage at all ? This does not seem right on my van !!!
Does bank 1 sensor going out on gas indicate that the rpm/ lambda filter is receiving this info and putting it into the LPG ECU ?
Do not know if this makes any sense to you. Would like to know your thoughts.
Cheers
Vince

The sensors need to be hot to work. They incorporate electrical heaters to speed the process, problems could occur with either the heaters or the wiring to them but that is rare. They tend to taker longer as they age. If you think one or both is slow to warm, renew it/them. These sensors are consumables just like spark plugs. That is probasbly unrelated to the more serious sysmptom though ...


Somethings wrong; perhaps clagged slots in your distributor causing lean running on one or more cylinders. Or possibly a clagged passive injector.

_________________
www.blazegas.com

You read the Lambdas with a code reader but that's not the same as reading them directly using a voltmeter. For example, for sure on certain Vauxhall factory fit systems (that use the Koltec system), the Lambda signals shown by a code reader will be emulated figures not real lambda figures! The LPG system feeds an emulated lambda voltage to the petrol ECU just to keep the petrol ECU happy and that voltage is the voltage that the code reader will show. The Transit factory system you have may or may not do the same, some Transit factory systems show actual lambda on a code reader and some don't. On the versions that show actual lambda the petrol ECU fuel trims are unchanging even when lambda is always high or low and of these versions the code reader may read the fuel system status as closed loop (as normal for a petrol vehicle), open loop or even open loop due to a fault (even though there is no fault the open loop fault reading can be normal when running on LPG). So, to be sure you're not reading an emulated voltage instead of actual lambda voltage you need to go directly for actual and connect a voltmeter directly to the lambda probe signal wire.

The cutting back to petrol under certain conditions symptoms I believe you described (without re-reading full thread to remind myself) are exactly what can be expected from a low for too long or high for too long lambda reading and the faith you're placing in the code reader for showing real lambda values is a common mistake when it comes to diagnosing that problem! Not saying this is your problem for sure, but it's likely!

Simon
Lpgc

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

A quick update. I was looking at some of the info that was posted in this thread that listed common problems with a guy who mended fleets of Transits. I ran down his list and eliminated things that I had done to my Transit. I looked once more with the code reader at the faults logged on my ECU.
Taking Simons suggestion with the Lambda readings, lead me to the think that the Lambda reading was not getting fed to the LPG ECU. Also with the revs running high on tick over, approximately 1400rpm. This put my thoughts to what was the common denominator ( The Lambda/ RPM filter ).
I ordered one up and fitted it. The van is now running sweet as a nut, and changes over to gas smoothly. Have ran it for a week and no problems so far. Read the petrol ECU yesterday and there is a fault code for limited flow in the egr valve. DPFE sensor might be the problem on that one. Have a few codes in the past with codes for the egr valve so DPFE might be playing up.
Will keep you posted on my findings.
Once more thank one and all for your help in this problem.
Hope 2014 is better on LPG.
I like to wish every one all the best for 2014.
Cheers
Veeman

Glad to be of help Veeman,

EGR type faults can also cause idle high rpm problems. I've since read the other guy's post you mentioned and he's written some good advice on EGR type problems on Duratecs and some good advice on Transit factory systems in general. Strange that he didn't mention the lambda mis-match points, though.

Simon
Lpgc

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

I know this is an old thread but,


upstream oxygen sensor is disabled on lpg


yes
yes, not so much when hot, more so when cold



yes, if its an outright failure, but whilst breaking down causes various intermittent symptoms


be careful testing these relays, there is a diode across supply connectors to protest from back EMF to ECU


this is best practice for dealing with Lambda faults when used in conjunction with diagnostics live data


I could go on quoting from through out the forum on this tiny little black box that causes so many problems, have suffered all of the usual symptoms and it was the intermittent problems that lead me down the repair path of this little piece of sh1t.
Testing posed a problem with no hard data to go on, if your confident enough to go down this route, be very careful with your probes.
I have re-manufactured 3 of these now, successfully, two by repair and one I am rebuilding with premium components, however, not all dead units are salvageable. I don't know who manufactured these units, but I have seen better quality boards come out of China.
Will post later, on a repair service for these units so keep hold of your dead ones, that is if is acceptable on this forum.

Depends on what you mean by 'disabled'. Upstream probe is the one in the manifold or between the manifold and cat, this probe is used by both the petrol and LPG system as primary means of detecting lean/correct/rich mixture. Petrol systems can run (improperly) without this input (cold start conditions, open loop conditions, limp home conditions) but factory LPG systems need this input to function, if the input doesn't respond properly for more than a few seconds the vehicle will switch back to petrol (except on some factory fitted Volvo's - which will continue to run on gas usually with a very lean mixture with the potential to cause engine damage).
But you might mean 'disabled' in the way I described above - The petrol system (on factory Necam LPG equipped) vehicles is either fed an emulated upstream lambda signal or is made to run in open loop mode while the vehicle is running on LPG.

The downstream probe isn't of much consequence on many vehicles, pretty much there to check if the cat is working properly, but on some vehicles errors regarding this probe can force the petrol system to run open loop / limp home mode - In the case of a factory LPG install, though, when running on LPG it will continue to run closed loop regardless of errors brought on by the post cat probe.

Autogas2000 sell (or at least did sell) replacement black boxes (rpm filter / lambda emulator / switching units) but where these are not available your services in repairing broken units might come in handy. On a fairly recent thread, your services may have been of interest to BrianH but I think he's got around the problem already or got rid of the car.

Simon

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

Still got the car, but its not got the koltec system on it anymore. I replaced it with a KME Nevo kit instead as it will be reused in due course when the rust finishes the rest of the car off (which won't be long). I had got the car upto 230k before having any major problems with it, only had a reducer upto that point (which was a used one anyway) But had just got completely sick and tired of messing with it and felt any more money spent on it was just wasted as it wouldn't be reusable kit.

I did get the last of the filter units that autogas2000 had - it didn't help, I think my main issue was down to the relay board - equally badly constructed (and doesn't plug in, so to remove it you have to cut wires to it first). The tracks on it look like the kind of thing you'd expect in a hobby pcb etching setup, not a part produced in any quantity.

Theres different versions of those boxes - Autogas2000 show one with 2 connectors on it that does rpm and lambda (or did when i was looking, thats still available) the one i had is just rpm - part number 238000-024 i think. On that particular setup the lambda sensor is fed through the gas ecu and cut off when on gas (open loop due to error shown on diags) It also intercepts other bits of wiring - injector positive and either the idle control valve or tps, can't remember which when I was removing it, suspect TPS now i think of it.

Ahh yes Bri, bit of a memory lapse on my part there

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

ohhh @!##
that should be: diverted to lpg ecu
sorry about that

when testing, it is essential to probe sensor directly in conjunction with running diagnostics live data,
after change over to lpg, diagnostics live data shows no oxygen sensor output,
for those unfamiliar with this system this can lead to a mis-diagnosis of a faulty sensor/ecu/....

Autogas 2000 supply the units at £63, Power Torque were £46 and can't remember what Nicholson-McLaren charged.
The later two no longer have any dealing or stock, our European friends charge about £45 excluding postage.

There are several versions of the Necam filter 238002, for lpg integration into various manufacturers(Vauxhall, Subaru..._)
I have no knowledge/experience of these what so ever, only the 238002-050 for the Transit.

If you have a dead or suspect one, don't bin it just yet.
I can do a passive test but it's not reliable, it will only show a component failure/anomaly, it does not indicate any problems where there is progressing failure, which can usually be attributed to the poor quality assembly, components and pcb.
Will be offering a repair service for these later in the year.

Well... LPG ECU may or may not receive actual lambda signal all the time (on LPG or petrol).. There is no problem with ECU's reading correct lambda voltage when multiple ECU's are connected to the same probe signal and many converted vehicles are wired like this.

From the point of correct running on petrol there can be problems with this sort of setup because petrol fuel trims can drift leading to fuel trim errors, limp home modes and engine warning lights or at least poor drive-ability when the vehicle first runs on petrol after running on LPG. Far less commonly, from the point of correct running on LPG, some petrol ECU's hold lambda voltage at a pre-set figure during certain conditions, so this could upset fuelling while running on LPG. In both cases some lambda signal diversion / emulation might be necessary to prevent the LPG ECU reading the petrol ECU's preset voltage instead of actual lambda.

On Koltec systems where actual lambda signal is disconnected from the petrol ECU while running on LPG, the petrol ECU is then sent an emulated lambda signal. The point of disconnecting petrol ECU from actual lambda is to allow the emulated signal to be forwarded to the petrol ECU while allowing the LPG ECU to continue to read actual lambda. The petrol ECU's emulated lambda is steered to keep petrol fuel trims in check and prevent the petrol ECU generating error codes etc.

On many Transits the petrol ECU continues to receive actual lambda (even while running on LPG) but switches to open loop mode. In open loop mode the petrol ECU fuel trims won't change anyway, so it's just another way of achieving the same as above (no fuel trim drift/errors while running on LPG using a none slave type LPG system).

Whether manufacturers choose the emulated lambda or open loop method will probably boil down to whether there's an easy way of forcing the petrol ECU to switch to open loop mode and whether the petrol ECU holds the lambda at a certain voltage under certain conditions... If, say, can just pull a petrol ECU pin to ground or +5v, then that will make for a much easier, neater and less expensive way of preventing the petrol ECU picking up mixture faults etc when running on LPG than the emulation route (which must divert lambda signal, connect to the petrol ECU via canbus, read petrol fuel trims and generate an appropriate emulated signal to keep petrol fuel trims in check). If the pet ECU holds lambda at a certain voltage under certain conditions then for correct LPG system function (while that is happening) it is necessary to disconnect lambda from the petrol ECU.

I found the thread I mentioned in a post above, you may be able to help identify the mystery electrical components... http://www.lpgforum.co.uk/viewtopic.php?f=28&t=14056

Simon

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

It was a case of necessity that drove me to investigate and resolve the numerous and varied problems attributed to this component, namely a random misfiring, unexplained switching to petrol and poor lpg fuel consumption(18mpg), I don't bother to monitor petrol consumption.the lpg is now returning a healthy 25+mpg, not too bad for a vehicle with the aerodynamics of a brick.


As far as I ascertain, the oxygen sensor signal is diverted to the lpg ecu by 238002-050 on fuel type switch over, there does not appear to be any emulated lambda signals outwith this or the ecu, there may additional connection(s) to the oxygen sensor within the loom but have not identified any. It would be interesting to check a wiring diagram, if one exists.


Had a quick look at this thread, and as usual, a number a people guessing what the components are or could be.
Best guess or consensus of opinion is definitely not the way to go in reverse engineering a specific component built for a specific purpose.

No wish to come across as confrontational, just that lambda signal is pretty fundamental to correct running on petrol or LPG so generally needs to be considered when diagnosing most problems. Where it earlier seemed we disagreed on Transit lambda handling it didn't inspire me with confidence in your services, though I suppose rpm module problems are a different issue that don't need to involve lambda and, anyway, maybe we were really always of a similar understanding regards lambda handling..

Yeh, as said I don't believe the Transits do emulate lambda when running on LPG, on most Transits I've repaired the petrol ECU continues to receive actual lambda voltage just the same as the LPG ECU (i.e. no diversion of lambda signal, both ECU's constantly get actual lambda) but the petrol ECU is forced into open loop mode when running on LPG. Both actual lambda and open loop mode can be confirmed via the usual method of live data via OBD2 (when OBD live data reading is the same as multimeter reading connected directly to lambda signal).. However, since the petrol ECU doesn't care about lambda signal during open loop mode it wouldn't matter if lambda was diverted to the LPG ECU only, just un-necessary. Yours seem the same as those I've described and fixed?

In fairness though, fairly informed people who're into electronics and ran a few tests in effort to determine electrical components inside the part in question.. A circuit diagram of the part listing components would be ideal but in absence of such is there another method for back engineering parts, such as the part you offer to repair (and just to confirm is that the rpm module)?

Simon

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

Identifying the components should be a lot easier if you can get the potting compound off of the board etc without destroying the components in the process. Never seen a schematic or parts listing. Suspect they were never widely available.

It's the general lack of info on the system I had that led me to to conclusion it wasn't worth spending money on. Bits of it work. But there seems to be a fault on the electronics side which in fairness is going to be down to age of the system. It would now be around 16 years old after all.

I figured at least spending money on new components that could be used elsewhere and would be known OK was a better idea. Just got the level sender to sort and it would seem the thermostat needs swapping. As it's only managing to get to around 65 degrees at the moment.

This unit consists of two completely separated circuits on the one board, fortunately each circuit terminates on an individual connector and slightly minimises the possibility of probing across the two. Each circuit is at two different base volts along with differing signal volts compound this with each of their 0v(lets not call these earths or neg's, which they are in essence, as there are +/- volts monitored between these and the vehicle earth/battery neg)
Hence the previous warning about probing and the possible dire consequences for the petrol and lpg ecu's.
On the small sample of only 3 units I've repaired, 3 had lambda cct faults and two of these also had rpm cct faults, so really difficult to speculate which is the most likely to fail. The poor quality pcb, assembly and components are the problem; as the only commonality between these cct's are both ecu's, I would go as far as to say that it is unlikely one faulty cct will not cause the other to fail.

The O2 sensor signal monitored with diagnostics live data on the petrol ecu and confirmed with multimeter at startup and/or running on petrol is present. When change over to lpg occurs, having been switched either automatically or manually, live data O2 volts drops to zero, probing for O2 signal with a 2nd multimeter, to find a matching signal at the petrol ecu side proved fruitless. I did not find any components that might emulate lambda under the dash and drew the line at slicing open any of the looms to follow wires to a possible alternate location.
At change over from petrol to lpg, the units lambda circuit diverts O2 signal to lpg ecu confirmed with multimeters.
A wiring diagram would have saved many hour's of probing.

The RPM circuit was less time consuming; signal sent from the petrol ecu, modified and onto lpg ecu.
The consequences/effects on running, when the rpm signal to the lpg ecu is lost or outside parameters, is up for debate.

Although I didn't spend too much time on the other thread of little interest to me, there were more guesses than tests.

The module referred to here is the, Necam Koltec Ford Transit "RPM Filter + Lambda" 238002-050 and no other.
I do not have a FINIS or Ford Engineering number.
Yes, I can test, but it has it limitations and only confirms a dead unit/component.
Yes, I can repair, but only if pcb is salvageable

I would suspect that the rpm side of that filter is the same as the one i was looking for - the one you mention is still available from Autogas2000 http://www.autogasshop.co.uk/necam-ford-transit-rpm-filter-lambda-442-p.asp , the one I had (238002-024) isn't any longer available, and wasn't the cause of the problem. The 58xx module would be more useful to have a repair service available for, though again the vehicles in need of it are reaching the end of their life as well.

238002-24 RPM filter is completely different, only the output signal from the filter would be the same in each case, this to match the signal required at the lpg ecu
238002-24 available at £28 + p&p

Bit of a difference regards what happens with O2 signal to the Transits I've fixed..

Bit of a summing up:

We both find Transit petrol ECUs run open loop while running on LPG.

We both find there is no emulated lambda for the petrol ECU when running on LPG (which we might expect anyway, as there should be no need for emulation while petrol ECU is running in open loop).

I find Transit petrol ECUs continue to receive actual lambda while running on LPG, you reckon actual lambda is disconnected from petrol ECU when running on LPG (you refer to the diversion). Does your diversion disconnect lambda from the LPG ECU while running on petrol? Due to the same point as above (i.e. no need for emulation while running open loop), if the petrol ECU is disconnected from actual lambda while running on LPG it begs the question why.. Can only think of a couple of reasons why ths should be so and I've never known them to be so on Transits. Reasons would be the petrol ECU holding lambda at a certain voltage, perhaps during engine warm-up or limp home type conditions (again not seen this on Transits), or the combined rpm module / lambda electronics on your Transit is more similar to the type of units fitted on Vauxhalls (which do need to disconnect actual lambda from the petrol ECU while running on LPG so emulated lambda can be fed to petrol ECU).

When repairing both Vauxhalls and Transits I used to connect a multimeter directly to actual lambda voltage, I still do that on Vauxhalls but for a long time now when repairing transits I haven't bothered, as I've found actual lambda can be read via live data method. Both Transits and Vauxhalls will switch back to petrol (due to LPG system having detected a fault) after several seconds of receiving none flicking lambda signal. On factory systems with distributor unit the usual cause of none flicking lambda is either a problem with the reducer (pressure, flow or leaking diaphragm issue) or a problem with the distributor unit (usually plunger or stepper issue)... So, when I see a system that will at least switch to LPG at all (perhaps necessitating multiple engine starts or battery disconnection for a while) and I read lambda voltage (via OBD and/or at the probe itself) at 0v while running on LPG I assume lack of gas flow to the engine causing actual lean mixture, then go on to check the reducer / distributor. Only if the system wouldn't switch to LPG for a period of less than say five seconds would I suspect the rpm module (or other electronics in it). I recognise that if your unit switches lambda between petrol and LPG ECUs there is the potential for shorting lambda to ground but I've sorted maybe 50+-transits and only ever replaced 2 rpm units, the others had problems with reducer, distributor, EGR etc. If you've only fixed 3 Transits and all 3 due to rpm module faults, don't know what's going on there! Problems running on LPG and lambda reading at 0v you say...?

Simon

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240

I didn't find signal at ecu, yes, yes, don't know
Have you got pin details for petrol ecu/O2 input at the ecu, I could not find a matching signal with one DMM on O2 sensor and one DMM for probing, if I get time to setup scope next weekend I should be able to compare signals and see if I've missed the plot. I did mention a possibility of an additional, possible parallel, connection in the loom, didn't find one by probing, might show an error in my methodology of recording results, or it maybe just an unsuitable comparison method altogether.


Never claimed to have only fixed 3 Transits, just the 238002-050 modules.
Numerous other members here and on the Transit forum have had problems attributed to these units. You've replaced 2 modules in 50 repairs, that equates to a 4% failure rate, it might not be high up on the check list for ongoing problems, but, for the intermittent problems and poor fuel consumption I was experiencing(see post) it was the solution.
I'm not so sure there is a ground(as in vehicle, battery, body) in the related wiring to short out to, there are obviously return path wires to the ecu's, referred to these previously as 0v from the external(to the ecu) components, probed potential differences between these and vehicle ground.
As you rightly point out, there could be X number of other reasons for problems with Transit lpg system, but here I'm only dealing with what I've found regarding the combined RPM and Lambda Transit module, and not the Vauxhall unit which is a different kettle of fish. Please read from multiple previous posts in case you missed it.

Anyway, for those who wish to "play" around for themselves, if you HAVE unexplained intermittent problems and poor fuel consumption it's worth swapping out for a known good one even just to test and confirm, I'd rather spend 15 minutes trying this than spend hours checking, testing and monitoring other parts of the system.

I never needed to know which petrol ECU pin is lambda because I always found lambda can be read via OBD2, but shouldn't be hard to find pinout info via Autodata or Google. If I were to look up that kind of info I'd also look up how the petrol ECU is switched to open loop while running on LPG, something I never had to dig into because I've never seen a Transit petrol ECU not switch to open loop mode when running on LPG.

Although I only found a 4% failure rate of RPM modules, many owners arrived here half expecting their fault to be the RPM module and no doubt some of them got the idea from other forums. I've known about the chats about RPM modules on other forums for some time (been a while since I've been on many other forums) but I never read much on other forums about the systems switching back to petrol due to lack of lambda flick etc. The lambda flick time-out is reset when the throttle position is moved, so it is possible to drive a factory converted Transit on LPG for prolonged periods all the while with incorrect mixture if the driver keeps moving the throttle, only with incorrect mixture the system will switch back to petrol at some point soon after constant throttle is maintained.. Unless the switching back to petrol due to no lambda flick characteristic is known about, the symptoms of poor mpg and switching back to petrol of it's own accord (especially when following an initial switch back it refuses to switch again to LPG) might seem due to an electrical problem, when actually the real problem might be incorrect reducer pressure or distributor problem. A 15 minute drive watching lambda voltage via OBD live data should be enough to rule out / confirm switching back to petrol due to mixture problems - this is, of course, only true if I am correct regards the petrol ECU continuing to receive actual lambda voltage while the vehicle is running on LPG and if there isn't some kind of problem where lambda voltage is pulled to ground..

It isn't unusual for a factory LPG system to stop working or to miraculously start to work again after having work done on the alternator, starter motor, a new battery fitted, etc... the common thing here is battery disconnection, which might also be done during removal/refitting of components such as the rpm module.
When the battery has been off for a while the LPG system's learned fuel trims are forgotten, which can lead to the time-out on mixture, especially where the learned fuelling is some margin from the default settings, such as when reducer pressure is lower than spec due to the spring in the reducer going soft over the years - which is usually fixed by adjusting reducer pressure a bit and then driving not long enough at constant throttle for the time-out to occur but long enough under various constant throttle positions for the various fuel trims to be re-learned.. This procedure alone fixes maybe half of factory fitted Necam systems but I haven't seen this mentioned on other forums. On the other hand the miraculous cures can come about because any trouble code that the LPG ECU had learned (which might have prevented the system switching to gas) is forgotten when the battery has been off for a while.

Of course there can come a point in diagnostics where an issue with the module needs to be ruled out or confirmed, and they do go wrong.. Next time I come across one that's broken I know where to send it for repair. Just in my experience factory converted Transit LPG problems are far more likely to be due to other than the module (20 odd times more likely in my experience and out of those 20 times, 10 times would be the simple pressure issue).

Simon

_________________
Full time LPG installer
Servicing / Diagnostics / Repairs to all systems / DIY conversion kits supplied with thorough tech support
Mid Yorkshire
2 miles A1, 8 miles M62,
http://www.Lpgc.co.uk
Twitter https://twitter.com/AutogasSimon
07816237240