Please help, Left bank cutting out

[Bensonv8]

I've been having a problem which has been slowly getting more regular.
and more frustrating.
Sometimes within a few hundred yards from home.
but also at irregular intervals.
My 95 Chim 4.0 stops running on the left bank.
Pull any plug lead off the right bank it slows down,
pull any injector plug off the right bank it slows down.
pull any plug lead off the left bank no change,
(holding the plug cap close to the plug you can hear the spark in the cap)
pull any injector plug off the left bank no change.
i checked the injector leads with a noid light and it was flashing,
so it is sending the pulse to fire the injector.
i removed the left lambda sensor to find the cap missing and whatever should be under it had melted away !!?? and it was very sooty.
having replaced the lambda sensor i thought my troubles where over.
But it is still happening.
As it sometimes happens soon after start up,
i thought it might be something to do with the temperature,
so i changed the engine ECU temperature transmiter....
NO change.
(when i tested it off the car as per the steve heath "Bible" there where some discrepancies from the quoted figures and the new one, but nothing major.)
When the left bank does stop running it will idle fairly happily,
but when i turn it off and try to re-start it won't start.
I have to blip the alarm on, then off and bingo, it will now start and run properly.
My suspicions are now turning to the one or more of the injectors on left bank.
Any on what to do next would be much appreciated.

[spend]

Repeat tests at the ECU connector, it is better to do all tests in situ as you will also be testing the wiring.

Dave

Its late... Do check the main relay - 2nd switched circuit(twin make/break) powers the injectors IIRC & ensure there is no great voltage drop across relay and fuse. Fuel relay is actually for fuel pump & purge IIRC ~ dont confuse the two of them.

Edited by spend on Sunday 27th May 01:15

[Bensonv8]

spend said:

Repeat tests at the ECU connector, it is better to do all tests in situ as you will also be testing the wiring.

Dave

Dave,
I assume you mean the temperature sensor tests ??.
I can't think of a way to pull the plug/injector leads off at the ecu ??
Can you test at the ECU with the engine running ?

As only the left bank is affected,
doesn't this indicate something on the left ??
unlike temperature and fuel pressure which are the same for both banks.

Steve

[spend]

Sorry, when you said you had tested I assumed you had done the std Efi tests (like google rrm_fuel_injection.pdf ((the estonian goodfellas seem to have been made on Google!)))

Of course it sounds like your injectors / that part of the ecu... Not even the spark making bits could only affect 1 bank ~ only the banked injection could produce this fault IMO. Why are you even thinking about the temp sensor?? Injectors are powered by main relay (along with ecu) and grounded by ecu IIRC - the only significant sensor relative to bank is the Lambda (you have replaced it ... is it good??) checking at the ecu may help you to narrow down the fault - it may even be that you have fried the ecu... Alt is get nearest Landy dealer / indie to read your fault codes.

Dave

[Bensonv8]

Thanks Dave, found the document.
It's going to take me a while to work out whats what.
I think i'll try and get the codes read first.
racing green are just down the road from me.

I tried the temp sensor because it seemed to cut out as it started to warm up, and i could get it from work for a couple of quid

thanks for you help
Steve

ps - i was hoping for a magic (simple) fix

[GreenV8S]

I suggest you do a resistance check of the two injector outputs at the ECU. The fact that you're getting a response on the NOID light is a good sign but you can't be sure they're getting a good enough signal to fire the injectors. You can also find that an injector fault can affect other cylinders at low revs / low throttle due to blow back sharing the fuel between cylinders.

Have you checked the plugs on the duff side? Are they wet or dry? Misfires caused by a weak spark plus dirty plugs may be mixture sensitive i.e. change with temperature due to cold start enrichment.

[spend]

The pre-cat is not blocking the manifold is it?

Dave

[E2DJH]

Hi Steve, sorry to read about your troubles.
I'm no good mechanically, but if there is anything i can do to help (bring mine along so you can compare anything?) just give me a shout.

[blitzracing]

Get your hands on a colour tune, and you can at least see a spark ( or not )to start with or any combustion at all. I had to use an osciloscope to trace a what I thought was an injector pulse fault, so I could see the injector timimgs. The pulse width controls the amount of fuel going in, and is fixed for the first couple of seconds of engine running. After that pulse width is reduced in relation to the air flow reading. This means an engine may start, but then die if theres an air leak in the plenum as the air flow meter reading is reduced. If you disconnect the air flow meter, the pulse width remains unchanged, and the engine runs very rich at idle, but it will run (on all 8 !), Proceed with caution as you may damage the cats with the extra fuel however if you let it run for any period, (ie when the cats are hot).

[Bensonv8]

Thanks for the suggestions guys,
I will work my way through them and i'll report back

All i need now is some dry weather and a breakdown somewhere convenient
so i can investigate

Spend,
Bit of dum question,
when you said "The pre-cat is not blocking the manifold is it?"
do you mean like a catalytic converter turning inside the exhaust box and blocking the exhaust and air flow ??? if so, where is the pre-cat ??
cheers

David,
Thanks for the offer, no visual checks needed at the moment.
Have fun with the at brands hatch tomorrow, i am coming down after work as long as the weather is ok'ish.
My rear tyre's are approaching the end of there life,....
so if it rains like it did for the bank holiday, i think i'll pass.

Steve

[spend]

Bensonv8 said:

when you said "The pre-cat is not blocking the manifold is it?"
do you mean like a catalytic converter turning inside the exhaust box and blocking the exhaust and air flow ??? if so, where is the pre-cat ??

Yes, the pre-cats are located just after the 4 pipes are blended into the large collector - immediately before the lambda sensor.

Dave

[Bensonv8]

I've just contacted the guy who hires the fault code reader and it's on it's way as i
He does advertise in the classified ads, under "parts/plates" - "TVR"
If anybody wants his email address PM me.

Steve

[Bensonv8]

OK, I've got hold of a code reader

I connected it up and went for a 40 mile drive on saturday.
After about 10 miles it cut out on the left bank, but no codes where generated
When i got home I tried pulling the fuel pump relay, thinking it would register a fuel pressure error code,
but all i got was an 03 error - Stored data corrupted since last trip.
I could not clear the code (following the instructions in the "Bible" )
The only way i could clear the code was to un-plug the ECU.
Which generated the 02 error - Live feed to ECU has been interrupted. Which clears automatically.
I went out later that night and i got 21 error - Tune resistor out of range.
I tried to clear this code, but after 10-15 attempts it would not clear.
So i had to un-plug the ECU again. This fault code has not returned.
It cut out again on the left bank on monday lunch time
and going home in the evening, but still did not generate a fault code.

I was hoping to get a 34 - Fuelling fault in nearside injector bank,
and praying not to get a 29 - Checksum error.

So when i got home i decided to check the 21 error - Tune resistor out of range.
I un-plugged the ECU and tested the loom connector that plugs into the ECU.
According ot the "Bible" to test the "tune resistor"
ECU disconnected, Ignition OFF,
test pins 5 & 27 (cat models - 3,700 - 4,100 ohms)
Using my multi-meter at 20,000 ohms
(the next setting down is 2,000 ohms, below the expected test results)
the result was 1, which is the reading when the probes are not touching. when I touch the probes together it goes down to zero.
this seemed a bit odd, so i tried some of the other tests in the "Bible"
I can only try the "ECU un-plugged" tests as i can't get to the connections when the ECU is plugged in.
I tried the Tune resistor test, The Throttle Potentiometer test,
The Stepper Motor test and The Lambda Sensor test
But i got the same "1" as above on the test meter.

Ok, I've just gone out and tested the tune resistor by disconnecting it.
It came back at 3,880 ohms (within the Bibles limits)
So i checked the wiring (see picture)
Pin 27 to one of the tune resistor connectors - this was fine.
Pin 5 (according to the Bible) to the other tune resistor connector - nothing.
After trying most of the other pins i found it on pin 32

Has anybody got this sort of connector, or am i miss understanding the diagram ??

and why wont the ECU re-set when i do the re-set procedure ?

Steve

Edited by Bensonv8 on Monday 4th June 21:34

[That Daddy]

Steve its just a thought,right the injectors are switched by a negative - feed why dont you rig up a 12v LED lamp to your steering column(so you can see when driving)now run a positive+ feed to one side of the LED to the ignition switch + live (fused of course) and the other to the injector negative- switched side from the ECU plug directly(of the faulty bank)now drive the car(the light should pulse faster as revs rise or slower as revs drop)drive untill fault occurs,if the lamp stops flashing then you have lost the switched - output from the ECU,which would indicate the ECU as gone AWOL,you can either replace or get your one repaired,its not a bullet proof way of testing but it is a sure way of finding out if you have a duff ECU,you can also use a dwell meter but this will still tell you what you need to know,good luck

[Bensonv8]

That Daddy said:

Steve its just a thought,right the injectors are switched by a negative - feed why dont you rig up a 12v LED lamp

That Daddy,
That's the sort of thinking i like, i have a dwell meter, so i'll give that a go.
and i think i've got my head around the ECU connector, the diagram in the Bible is of the ECU, not the cable connector. - - - - - - -

[That Daddy]

Bensonv8 said:

That Daddy said:

Steve its just a thought,right the injectors are switched by a negative - feed why dont you rig up a 12v LED lamp

That Daddy,
That's the sort of thinking i like, i have a dwell meter, so i'll give that a go.
and i think i've got my head around the ECU connector, the diagram in the Bible is of the ECU, not the cable connector. - - - - - - -

Steve you will get that fixed,i had an intermittant dying at speed like you switched the ignition off(done the 2 relays etc)all seemed fine for 10 months guess what its returned and in aint the relays this time,taking into the workshop tommo, dont you just love faults like this.

[spend]

The diagram shows what you plug the connector into... Aways a pain with MMMMultiway connectors visualising which is the male to female mapping (well it is for me anyway)

Dave

[GreenV8S]

I always had a bit of a problem with those diagrams in Steve's book, as far as I can see they're the identical connector just drawn upside down?

I have a set of test leads consisting of a m/f mini-timer and 12v LED which plug in between the injectors and the loom, the leads are long enough to bring the LED back to the bulkhead so I can see all eight of them from the driver's seat. Knowing the approximate injector duty cycle, and the fuel pressure, and the manifold depression, is enough to get a pretty darned good idea what the ECU is trying to do.

[griffter]

1) There is no fuel pressure sensor - so you won't get any fault codes relating to fuel pressure.
2) The diagram in the bible is off the BACK of the ecu plug (ie as you describe, it in fact correlates to the socket on the ecu rather than the connecting face of the ecu plug). Unfortunately the bible doesn't make this clear.
3) You can access the back of the ecu plug - as per bible diags - while the plug is connnected, by (with the plug disconnected) removing the rubber seal (use a small screwdriver to gently lever it out) then undoing the two phillips screws which are underneath it. Now you can gently draw the plug out of the plastic shield into which the cables run. This is really useful because you can now stick wire probes carefully into the backs of the terminals and as GreenV8S suggests, monitor injector pulses in real time with 12v LEDs. If you're losing a whole bank at a time there's no need to monitor all 8 as they're fired in banks of 4 - so a left bank monitor and a right bank monitor (which is all you can access from the ecu plug anyway) will be fine. However see (4).
4) Are you sure the left bank is CUTTING OUT? I've had a problem where the left bank goes to maximum fuelling and floods the cylinders, thus the left bank cylinders do not fire. You can obviously see if this is the case with the 12v LEDS (the injector pulse duration is so long that the LEDS will appear to be on constantly at higher revs), but an easier (although less conclusive) diagnosis is to floor it when the 'cut-out' occurs. If at at about the 4k rpm the car suddenly appears to come back to life and takes off like a scalded cat, it's because the bank is not cutting out, but over-fuelling (as with mine). Suddenly there's not too much fuel, but just enough (for wide open throttle and >4k rpm, so the engine runs happliy under full load over 4k rpm). The problem then is identifying what the cause is. It could be a duff ecu, but I changed mine and this didn't fix it. Top culprit (following a helpful discussion with Mark Adams) is an intermittent break in the ecu 12v supply. Not enough to cause the engine to cut out completely, but enough to cause the ecu processor to corrupt. Such an intermittent break could be almost anywhere - metal 'Bosch' relays, loom, fuse box, behind dash wiring, immobiliser, ignition switch etc... I've been over all those with fine-toothed comb, and the ignition switch seems to have done the trick (fingers corssed...).
5) Good luck!

Edited by griffter on Tuesday 5th June 08:09

[GreenV8S]

Although it's certainly *possible* to tap into the wiring at the ECU to monitor the injector pulses, the way I prefer to do it is to insert the test light between the loom and the injector. This means that you will pick up any wiring faults, it is also less intrusive than opening up the ECU plug and a lot easier to do. In general I'd be very cautious about powering up the ECU with the plug dismantled because of the danger of damaging the back of the plug and blowing something up. I do general sensor testing from the *front* of the plug i.e. just disconnect it from the ECU and put the meter on the exposed connectors.