Only injecting on one bank

[itiejim]

I originally posted about this issue here:

http://www.pistonheads.com/gassing/topic.asp?h=0&a...

Anyway, I thought a new thread might be more suitable for a full update.

I have the following issue with my 4.5 V8:

When I turn the ignition on and try to start the engine all 8 cylinders inject, however, as soon as it tries to start and I open the throttle the nearside bank of cylinders stops injecting fuel.

I have swapped the lambda sensors over - no change same bank fails to inject.

I have also swapped the feeds to the lambdas over in order to check the wiring - again no change.

I have swapped over the plugs between the throttle pots.

To my mind this should eliminate the lambdas and the wiring going to them and also the throttle pots and their wiring.

I am awaiting the arrival of a laptop so that I can actually get some readings off it, but can't logically think of what else can be causing the problem - any ideas?

Why would I get injection from a cold start but then failure to inject as soon as it tries to start?

[ridds]

You really need to wait for the laptop to get some idea of what the values from sensors are.

It sounds like it's just performing it's initial crank fuelling and then stopping.

You need to see the values from Lambdas, T-Pots etc to see if it is reading them ok and if the ECU is then giving out a signal to the Injector driver. If it is then it would be replacement ECU time.

[itiejim]

Cheers Ridds, hopefully the postman will do the business tomorrow.

I see where you're going with the ECU - I guess that if it only reads lambda and throttle position - and I've excluded those, then it can only be ECU.

I'm amazed though that it should be fine when I turn it off and no good 2 days later. Perhaps a dry joint?

(Or perhaps wishful thinking)...

[headcase]

Just had a quick look at one of the wiring diagram i have, im sure this has already been mentioned, the injectors fire 4 at a time, there are 2 ouputs from the ecu,
one output is labelled 1,7,3,9
and the other 4,6,2,8

so if your lucky it could be a poor connection on one output or the other, it is possible it could work for a few seconds before the connection broke down under load.

I also understand where ridds is coming from, if the ecu is getting all of the right signals and not outputting what it should be then it could be a duff ecu, even if it is, it could just be a dry joint within the ecu, your gunna have to take the lid off it to hook up the laptop anyways so have a good look at the connectors and the soldering inside the ecu when you have it in pieces.

[itiejim]

Cheers headcase. I'm about to go outside and check out all the connections again.

Hopefully my laptop will arrive tomorrow - then I'll just have to work out how to use it

I invisage another post on here tomorrow evening along the lines of "How do I get my laptop to talk to the ECU"

[HarryW]

headcase said:

Just had a quick look at one of the wiring diagram i have, im sure this has already been mentioned, the injectors fire 4 at a time, there are 2 ouputs from the ecu,
one output is labelled 1,7,3,9
and the other 4,6,2,8

so if your lucky it could be a poor connection on one output or the other, it is possible it could work for a few seconds before the connection broke down under load.

I also understand where ridds is coming from, if the ecu is getting all of the right signals and not outputting what it should be then it could be a duff ecu, even if it is, it could just be a dry joint within the ecu, your gunna have to take the lid off it to hook up the laptop anyways so have a good look at the connectors and the soldering inside the ecu when you have it in pieces.

Besides the V9 Cerb being an issue .
It is a valid point about the injectors, I think they are 14.5ohm types, it might be worth swinging through each of them in case one is shorting and dragging the bank down with it. Or more easily disconnect one injector on the missing bank at a time to see if that cures it. Getting the engine to run as a V7 is better than a V4 shirley.

[headcase]

LOL i love typos :P (and dont call me shirley)

Edited by headcase on Wednesday 13th May 08:02

[itiejim]

OK I have managed to get my car hooked up to the diagnostics and got the following results, can anyone help me to interpret them in the context of the problem I have?

Speed 686
Throttle 1 18
Injection 1357 1.2
Lambda 1 0
Throttle 2 17.3
Injection 2468 2.3
Lambda 2 1.0
Ignition 20.2
Water temp 48.4
Air temp 15.2
Adaptive 1 -93.7
Adaptive 2 -22.4

These readings were taken at virtually idle after a period of slow tickover. I noted that the lamba 1 reading remained at 0 which would indicate a problem I suppose, also the adaptives seem wildly out - given the experiments I have already explained in my opening post does anyone have a clue

[alinton]

I'm no expert, and I'm sure one will pipe up shortly, but if I were you...

I'd conclude that Lambda 1 is faulty, and is causing that wild -93 adaptive reading, which I think means very little fuel!

You could swap over the lambda sensor plugs, see what happens. If the far-out adaptive moves to the other bank, Bob's your mother's brother.

..just to give you something to think about until one more knowledgable comes along!

Andy.

[ridds]

I'd reset the adaptives first with a value like that. In all honesty I'd say it's a knackered lambda sensor but that is a massively out number.

How did you check the Lambda's before, just swap them over.

If you can try and take a screen dump (prtscn button) and post it up.

Does the lambda value change at all?

[itiejim]

I originally thought that it was a duff lambda so I tried swapping them over with no change (ie the same bank keeps failing to inject).

I will reset the adaptives and try again and post my results here.

I have also taken a screen print, pasted it onto a MS Word doc, but can't work out how to post it on here - any ideas?

[ridds]

Got a facebook or photobucket account?

If not, sign up for one and stick the images there. MS Word isn't the best format just use paint.

[itiejim]

Thanks Ridds - Word was my problem, paint has worked fine:

Also, to confirm Lambda 1 always reads 0.

[spitfire4v8]

something very screwy going on there. the adaptive trim is limited to plus/minus 37.5% in the set up of the ecu .. it should never get to 90%. I suspect a problem with your nvram where the adaptive values are stored. What happens if you clear the adaptive values and all fault codes and let the car relearn? does it do the same thing again? but before that bring up the two spreadsheets of adaptive trim values and let us see those. cheers

Edited by spitfire4v8 on Saturday 16th May 11:24

[ridds]

spitfire4v8 said:

something very screwy going on there. the adaptive trim is limited to plus/minus 37.5% in the set up of the ecu .. it should never get to 90%. I suspect a problem with your nvram where the adaptive values are stored. What happens if you clear the adaptive values and all fault codes and let the car relearn? does it do the same thing again? but before that bring up the two spreadsheets of adaptive trim values and let us see those. cheers

Edited by spitfire4v8 on Saturday 16th May 11:24

I thought that was the case when I first saw that number but he managed to have the value so I could hardly call him a liar!

Does sound like something seriously a-miss with that ECU.

[itiejim]

It works!



Clearing the adaptive seems to have sorted it completely. The log below shows that adaptive 1 is still a little different from adaptive 2, but looking at the throttle pot settings I think that's why.

Any reason why the car should suddenly do this, also, is it normal for the lambda traces to jump around like that?

Thanks to all for the input, help, and ideas.

I'm off to set my throttle pots now

ETA - adaptives now all stay within the predefined limits.

Edited by itiejim on Saturday 16th May 12:06

[jammy_basturd]

Yes, the lambda sensor should switch between 0 and 1 (possibly because of the exhaust pulses flashing over it?). So if you have one that is staying at one number, or at 0 then it isn't working.

By the way, your adaptives should be around +15 - +19 I believe.

Edited by jammy_basturd on Saturday 16th May 13:23

[spitfire4v8]

yes the lambda output will jump from 0-1.2 volts or thereabouts and back again as the feedback control makes changes to the mixture.