Another step forwards in solving the problem...now, Lambda sensors info please?
- Thread starter Fanum
- Start date
Fanum
Toubab
wot a nightmare..............
buy a 1200..................................
I'm neither desperate or stupid
Oh, guess what?
I found the missing two valves from the original engine nuking today
I had Charwee around to act as my spanner monkey.....the new lambda arrived so we took the cat and stubby off to get the dead one out, and the cat was rattling a bit......so we shook it about a bit, and here they are!!
Those have been kicking about in the cat after blowing through the piston crown....we thought they'd been vaporised or the valve pixies had stolen them, as there was no sign of them when Neil stuck in the new half-engine, but no, the sneaky little buggers had been hiding in the guts of the cat since then
Nicely rounded off and smooth as Pikeydave on a night out in Tel-Aviv, I'm going to dremel a hole in one and wear it
New lambda wired up, heatshrunk and sealed properly, made an extra 3 inches longer as well to avoid problems re-fitting.
Old y-piece and Remus can polished up....I'm going to see if I can find a responsible adult to supervise me tomorrow 'cos I'm going to have to make up some mounting brackets to bolt the 'zorst back (My old Wudo lowering kit that used to hand the remus below the pannier seems to be lacking a major bracket
) so that means getting the angle grinder out (hey, it's all precision engineering here!!) to cut a bit off next door's zimmer frame.My neighbours love me as well....I very nicely allowed them to hear the glory that is the sound of an 1140GSA with no cat or box on it for a few minutes.
The big ginger tom from down the road shat itself and then fell over dead with fright, so all in all, it's been a positive day
Just because the lambda voltage seen on a serial output is dead, it does not mean that the sensor is not working. Have you checked with a meter or scope between the black wire and earth?
If the ecu is looking for a co pot instead of a lambda sensor, the displayed output will stay constant.
Sensor outputs and serial data are not the same.
Lastly, if you must fit a universal sensor then at least fit a Bosch one. Your bike has a dreadful reputation for surging and you compromise on the most important part in the mixture control system under cruise conditions
If the ecu is looking for a co pot instead of a lambda sensor, the displayed output will stay constant.
Sensor outputs and serial data are not the same.
Lastly, if you must fit a universal sensor then at least fit a Bosch one. Your bike has a dreadful reputation for surging and you compromise on the most important part in the mixture control system under cruise conditions
Fanum
Toubab
Too late.....got a generic 'universal' one (mobletron branded but could be manufactured by someone else)For the price of a Bosch one, I could get 3 of these......and frankly, that's important right now.
EDIT.....Disconnecting the plug with the original sensor made a positive difference, but not sure what that indicates in the context of the above.
Your bike has a dreadful reputation for surging and you compromise on the most important part in the mixture control system under cruise conditions
My bike is a twin spark.......no reputation for surging at all.
Thanks for the input though....I should have it all back together again tomorrow and we'll see if it's made any difference to the running problem...if it hasn't, or you're right about the serial output part of it, worse case scenario is it's another thing ticked off the list and maybe I've got a spare OEM lambda sensor that we thought was dead
Will you please elaborate on this for educational purposes.
As far a I know it's a lamda sensor which measures the oxygen within the exhaust gasses on their way to the atmosphere. I presume a CO pot' has a dcV output which is proportional to the quantity of CO it detects.
True on this bike. I don't think there are any sensors clever enough to provide serial data to this bike and be read by the ECU.
As far a I know it's a lamda sensor which measures the oxygen within the exhaust gasses on their way to the atmosphere. I presume a CO pot' has a dcV output which is proportional to the quantity of CO it detects.
True on this bike. I don't think there are any sensors clever enough to provide serial data to this bike and be read by the ECU.
Don't allow any distractions about serial data transfer within your bike. It's not a Canbus in any way whatsoever. Your bike reads the voltages from sensors within the Motronic and then decides what to do thereafter.
AFAIK all the sensors are resistive or magnetic (Hall sensor) and provide direct inputs to analogue to digital convertors (ADCs) within the Motronic.
It is only "digital" within the Motronic to allow its "computer" to work out the correct amount of fuel to send to the injectors and maintain a satisfactory engine output.
There's more to that but don't be thinking about data, just the function of a component will do.
I'll be keeping an eye on this this because I have been here and have learnt a few things about what make it tick and which Motronic inputs have relevance.
PS. I'm going 30/70 on your lambda sensor sorting it. Sorry.
Last edited:
Fanum
Toubab
I'm not pretending to understand the difference between serial data and what the 911 was interpreting into its graph, I'm going to look it up tomorrow and get my head around it, but according to the 911 people:
I do know that the 911 software was showing a flatline, and should have been showing something more like this as the ECU received signals telling it to ferk around with the mixture though (you can tell I'm right up on this level of tech babble, can't you
)
EDIT...I guess that's pretty much what you just said while I was typing
And yes, I agree, it's unlikely to SOLVE the overall problem, but even if it doesn't, it's another box ticked off
I'm not pretending to understand the difference between serial data and what the 911 was interpreting into its graph, I'm going to look it up tomorrow and get my head around it, but according to the 911 people:
I do know that the 911 software was showing a flatline, and should have been showing something more like this as the ECU received signals telling it to ferk around with the mixture though (you can tell I'm right up on this level of tech babble, can't you
EDIT...I guess that's pretty much what you just said while I was typing
That lamda sensor graph was better than the last one I saw on my bike over 12 months ago. A geezer noticed mine was flat-lining and changed the sensor but it was the same afterwards; I had to telll him that.
Anyway, try it tomorrow and we''ll see. You might want to try some resistance measurements across your air intake temperature sensor to see what that's doing.
Neglecting your compression problem, the swapping or borrowing of some throttle body assemblies would really help.
I'm moving tomorrow but should be back in the evening if I can help.
ahutcheon
Registered user
I'm no expert, but I'm curious (and probably know just enough to be dangerous...)
Ditchwater, are you suggesting (as alternatives to a failed lambda sensor giving a constant output):
a) The Motronic is not reading the lambda sensor value, and is instead reporting a substituted constant (e.g. lookup table value it is defaulting to) to the 911 over serial diagnostics?
b) The Motronic is not running closed-loop, so isn't adjusting the mixture via the lambda sensor (which will probably produce a near-constant output depending on the open-loop mixture rather than the normal oscillation of the closed-loop feedback).
It should be possible to get the raw lambda output using a multimeter, but that's only interesting (if at all) if very different to what the 911 reads from the ECU. Do the diagnostics on the 1150 Motronic allow a 911 to tell whether the mixture setting is indeed closed-loop via the lambda sensor? Is there a fault code for lambda failure?
Ditchwater, are you suggesting (as alternatives to a failed lambda sensor giving a constant output):
a) The Motronic is not reading the lambda sensor value, and is instead reporting a substituted constant (e.g. lookup table value it is defaulting to) to the 911 over serial diagnostics?
b) The Motronic is not running closed-loop, so isn't adjusting the mixture via the lambda sensor (which will probably produce a near-constant output depending on the open-loop mixture rather than the normal oscillation of the closed-loop feedback).
It should be possible to get the raw lambda output using a multimeter, but that's only interesting (if at all) if very different to what the 911 reads from the ECU. Do the diagnostics on the 1150 Motronic allow a 911 to tell whether the mixture setting is indeed closed-loop via the lambda sensor? Is there a fault code for lambda failure?
Pretty much what I was thinking myself from Ditch's post.
Bill, if you've got an analogue multimeter (or much better, an oscilliscope), it'd be well worth a check of the lambda output to rule out a failed Motronic
My earlier post was not a discussion about digital sensors or, indeed, CAN bus systems. All modern engine management systems - and Motronic is one of them - have a variety of strategies to cope with failed ANALOGUE sensors.
For instance, in the case of an engine temperature sensor which fails, the ecu will sometimes calculate the temperature based on the amount of time the engine has been running.
As I said earlier, if one of these Motronic units is looking for a co pot - like the earlier UK spec bikes - the lambda reading on serial data will be constant.
Just to remind you, the serial data observed by GS911 or any other diagnostic platform is the ecu's INTERPRETATION based on its inputs. If it is having to substitute data then an error code will be flagged for the sensor concerned.
If you want to test a sensor or actuator, you test it directly using a meter or scope as I said previously. That is why the dealers will have a break-out box which plugs into the ecu.
BTW, I don't just make up this stuff as some seem to think, I studied electronics for 5 years at college and have been trained on engine diagnostics by Bosch, Snap-On, Actia, DEC and Pico amongst others. I do this stuff every day
For instance, in the case of an engine temperature sensor which fails, the ecu will sometimes calculate the temperature based on the amount of time the engine has been running.
As I said earlier, if one of these Motronic units is looking for a co pot - like the earlier UK spec bikes - the lambda reading on serial data will be constant.
Just to remind you, the serial data observed by GS911 or any other diagnostic platform is the ecu's INTERPRETATION based on its inputs. If it is having to substitute data then an error code will be flagged for the sensor concerned.
If you want to test a sensor or actuator, you test it directly using a meter or scope as I said previously. That is why the dealers will have a break-out box which plugs into the ecu.
BTW, I don't just make up this stuff as some seem to think, I studied electronics for 5 years at college and have been trained on engine diagnostics by Bosch, Snap-On, Actia, DEC and Pico amongst others. I do this stuff every day
Error codes have been flagged up for the lambda- 84 times for one particular fault, and 60 odd for the other.
Fanum
Toubab
No, because at that stage, the cat was away from the bike and the lambda sensor wiring cut to start making the new loom for it.
When a catalytic converter is 'blocked', it's rarely because it's physically blocked because of objects....it's normally chemical fouling.....these two parts of valve heads were in the bottom of the thing, almost certainly not causing much of a physical blockage (the cat has been slightly....er....modded for freer flow in the past
)The Y piece, remus and new lambda are going in today, so that's effectively doing the same anyway
I've just had a scan through this thread and the previous one to try and get my head around it.
Did you ever find out why the valves went on the previous head? To lose two probably means that they were both hitting the piston at times - this could only be due to a mechanical fault, probably in the valve train. Even if you had a serious management fault causing overheating and burning of the valves, they don't normally fall off.
The sudden change of tappet clearance is a worry, that's not bedding in, it's something moving or wearing out.
When you did the tappets recently, did you check the valve timing - and I mean by looking at the valves, not just the marks?
Apart from taking the heads off my 1100, I don't know the boxer engine well enough to think through the next bit. Is there scope for the valve timing to shift around on one pot - bearing in mind some bits have already been replaced? Another issue which is coming up on modern cars (probably not a GS though) is cam lobes which rotate on the shaft
causing no end of problems.
Despite my earlier comments, a dead lambda sensor shouldn't affect high speed operation. When you're 'making progress' the thing will be running open loop with additional fuel being demanded.
If the mechanical bits are all ok, you're left with the usual culprits. If the TB's are opening and closing together and the TPS is working you can rule them out. They're not complex enough to warrant further investigation.
Is the problem always on one cylinder? If yes, then it might be worth scoping the injector pulses - they should be the same at all times. If not, have you checked fuel pressure? Should be 3 bar IIRC. The short term fix by swapping an injector is a bit baffling.
It's probably quite easy to swap the ecu on an 1150, I don't think they're coded in. Someone on here must have one you could borrow. Failing that, someone like AutoMatec will test yours for 25 quid.
Out of interest, what were the actual faults recorded on the lambda sensor? Were they electrical / circuit errors or mixture control faults. The latter will be inevitable if you have any mechanical issues.
Good Luck,
Dick
Did you ever find out why the valves went on the previous head? To lose two probably means that they were both hitting the piston at times - this could only be due to a mechanical fault, probably in the valve train. Even if you had a serious management fault causing overheating and burning of the valves, they don't normally fall off.
The sudden change of tappet clearance is a worry, that's not bedding in, it's something moving or wearing out.
When you did the tappets recently, did you check the valve timing - and I mean by looking at the valves, not just the marks?
Apart from taking the heads off my 1100, I don't know the boxer engine well enough to think through the next bit. Is there scope for the valve timing to shift around on one pot - bearing in mind some bits have already been replaced? Another issue which is coming up on modern cars (probably not a GS though) is cam lobes which rotate on the shaft
causing no end of problems.Despite my earlier comments, a dead lambda sensor shouldn't affect high speed operation. When you're 'making progress' the thing will be running open loop with additional fuel being demanded.
If the mechanical bits are all ok, you're left with the usual culprits. If the TB's are opening and closing together and the TPS is working you can rule them out. They're not complex enough to warrant further investigation.
Is the problem always on one cylinder? If yes, then it might be worth scoping the injector pulses - they should be the same at all times. If not, have you checked fuel pressure? Should be 3 bar IIRC. The short term fix by swapping an injector is a bit baffling.
It's probably quite easy to swap the ecu on an 1150, I don't think they're coded in. Someone on here must have one you could borrow. Failing that, someone like AutoMatec will test yours for 25 quid.
Out of interest, what were the actual faults recorded on the lambda sensor? Were they electrical / circuit errors or mixture control faults. The latter will be inevitable if you have any mechanical issues.
Good Luck,
Dick
Fanum
Toubab
Ok, A question.....
IF the cat had been 'poisoned' by a long period of bad running, including several hundred miles of running on one cylinder only with the other half of the fuel going straight through the cat, would it be better to put the Y piece I have on with the new Lambda, assuming that there may be enough residue in the Cat to kill the new lambda?
It's a fair bit of mucking about to get the remus back on with the Y piece rather than using OEM cat and stubby, but I've been told by Rosie that she would prefer the remus on, and the possibility that the cat might damage the new Lambda sensor on top of that makes it worth messing about to do.
IF the cat had been 'poisoned' by a long period of bad running, including several hundred miles of running on one cylinder only with the other half of the fuel going straight through the cat, would it be better to put the Y piece I have on with the new Lambda, assuming that there may be enough residue in the Cat to kill the new lambda?
It's a fair bit of mucking about to get the remus back on with the Y piece rather than using OEM cat and stubby, but I've been told by Rosie that she would prefer the remus on, and the possibility that the cat might damage the new Lambda sensor on top of that makes it worth messing about to do.
I don,t think the cat is responsible for running on one cylinder, the loose tappets would not do this either. If you replace the Lambda and it still runs on one cylinder it will poison the sensor and back to square one. I would be tempted to find your other faults first.
Shouldn't, but it does.
And they've been cured by replacing the lambda, by me and BMW main dealers on a customers bike who were stranded in leeds with bad high speed running and another customer in switzerland who went to a main dealer. Both cured by new lambdas.
