Is my ECU broken?
Is my ECU broken?
Author
Discussion

brogenville

Original Poster:

941 posts

231 months

Saturday 13th September 2008
quotequote all
It seems as if this is the only possibility left, but I still can understand what might have happened to it.

Basically I've had a problem with Lambda 1 which I noticed whilst setting up my new engine. The voltage just sits at either 0.00V, 0.01V, or very occasionally 0.02V. Lambda 2 was switching fine between 0-1.5V.

First check was to make sure all connectors were tight together- no change. Second was to clean the sensor by heating it with a torch for a minute or so. Again the same result. At this point I thought that the semsor must be fooked, so I ordered up two new lambdas, just for good measure. Well bugger me, its still doing the same thing. frown

So now Im sure it must be a bad connection somewhere, but having found the lambda 1 input in the ecu, I find that there's perfect continuity between that and the plug to the sensor!

Does this not just leave me with the ECU that could be at fault? I cant understand what could have happened to it though, as it seemed to be working fine before the engine swap.

I suppose what I need is to be able to try my ecu in another car (or my car with another ecu).

Anyone able to offer any help?

Nickccc

1,682 posts

278 months

Saturday 13th September 2008
quotequote all
reset the adaptives, what happens when you try to download the adaptive maps? Does one bank show all zeros? If so, then unplug the lambda connectors and switch them over as they are back to front.

trackcar

6,453 posts

256 months

Saturday 13th September 2008
quotequote all
getting the two lambdas plugged in the wrong way round is very common .. the plugs are makred: O means odd bank not offside as many people seem to think, e means even bank. the odd bank is passenger side, even bank is driver side.
next is to make sure teh throttle pot on the non-working side is steady as if it is not steady the ecu wont learn anything as th eu needs to see the mixture at each load site for a certain length of time to start trimming it.
the adaptives are held in the NVRAM and sometimes adaptive trim problems are solved by changing that.
It couldbe that you have such high airflow on the side not working that it is in fact working, but has already trimmed to plus 37.5% and there isnt enough headroom in the adaptive positive trimming to get a lambda=1 mixture.

check everything again before you condemn the ecu, theyre pretty hardy pieces of kit and you would be unlucky to have an actualy ecu problem.

< FORZA WEST

1,038 posts

239 months

Saturday 13th September 2008
quotequote all
It is very unusual to have an MBE fault but common to have a new lambda that doesnt work....as above,swop the plugs over and check to see if the fault swops sides.

If it is an MBE fault be careful where you get it from...someone had a snapped pin due to water ingress to the plug so they bought one from ebay...it had suffered severe water ingress to the lower part of the curcuit board...the Eprom etc were U/S....fortunately I was able to make a good one out of the 2

brogenville

Original Poster:

941 posts

231 months

Saturday 13th September 2008
quotequote all
Nickccc said:
reset the adaptives, what happens when you try to download the adaptive maps? Does one bank show all zeros? If so, then unplug the lambda connectors and switch them over as they are back to front.
The adaptives were very high on the bank with the non-functional lambda (about 30 I think), and quite low (about 4-5) on the other bank. I did try resetting the adaptives, but the engine really didnt like it, as was determined to stall straight after.

Unless the lambda plugs are marked up wrong on the loom, which I couldnt really believe either, then I am positive that they are plugged in correctly. I might just try them round the other way just for the hell of it.

brogenville

Original Poster:

941 posts

231 months

Saturday 13th September 2008
quotequote all
trackcar said:
It couldbe that you have such high airflow on the side not working that it is in fact working, but has already trimmed to plus 37.5% and there isnt enough headroom in the adaptive positive trimming to get a lambda=1 mixture.
Ah, so low lambda means that I have a lean mixture? With that being the case, I should be able to artificially richen it up by blocking some of the airflow into that bank?

brogenville

Original Poster:

941 posts

231 months

Sunday 14th September 2008
quotequote all
With help from the above I figured that I must have an air leak somewhere, and sure enough the inlet manifold wasnt sealing properly. With that all done up both lambdas are working away fine, and no doubt the old ones were working fine.

Even better, the strange pinking noise has gone too. So with new injectors wining their way to me from America, everything's looking good for my my track car dyno session.

Anyone want to buy a couple of as new lambda sensors?