How I fixed bank not firing
Discussion
Hi all - you might recall a few weeks ago, some postings from me about my '98 4.2 who's odd bank of cylinders weren't firing.
Well I thought I'd recount the tale of how I got it fixed, without the aid of a dealer, in case it helps out anyone else.
This story includes some lessons I learnt too!
Cyliners 1,3,5 and 7 - those on the passenger side - weren't running.
I first suspected the coil pack, and it did look a bit iffy, a bit puffy and split.
So I got a new one and put it in. Actually I got two new ones - one from Offords and one from the local Citroen dealer as the same part is used in a 1996 ZX 1.6i !
But no change.
Then I decided to get technical, and got out the oscilloscope and checked whether the coil pack was getting power and signals from the ECU.
Pins 3 and 4 on the coil pack input connector should have 12v; Pins 1 and 2 get pulses to ground (negative) from the ECU. Pin 1 goes to ECU pin 40, pin 2 to 55.
(on the other coil pack pin 1 goes to 19, pin 2 to 20)
Using the scope I was able to compare the signals arriving at the dead side coil pack with those arriving at the working one, and look at the signals actually at the ECU connector.
All those signals looked fine.
Then I did what I should have done first, removed the air inlet, removed the spark plugs and checked for a spark.
There were sparks!
So it wasn't an ignition problem after all. Neither was it a lambda sensor thing, the ECU software showed they were working.
So the only other thing that would stop the cylinders firing is no fuel.
Each injector has two connections - 12v and a signal from the ECU. The oscilloscope told me that the red 12v wire did indeed have 12v on it, but the yellow signal wires had nothing.
On the working side though, the yellow wires had nice negative-going pulses.
I started to fear for the health of the ECU at this point.
So I looked at the pins on the ECU connector which feed the injectors - strangely all four injectors are fed from the same pin - 38 on the odd side and 53 on the even (so all four injectors squirtnat the same time!) Both mine were OK! So the signals were there at the ECU but not there at the other end of the wire, at the injectors.
A ohm-meter showed a very high resistance between the injectors and the ECU - somehow that one yellow wire in the loom had gone high resistance. I'd suspect a splice joint in the loom somewhere.
(not in the loom connector - as my car doesn't have one!)
I identified some spare wires in the loom, there's about 5 that in the engine bay are labelled 'stepper' but not connected to anything. So I used one of those to replace the faulty wire, soldered it to pin 38 at the ECU and to the nearest splice to the injectors in the engine bay (which necessitated pulling back some of the black sheath to find the first splice).
All fixed and running lovely!
My thanks to Jerry Plummer for his suggestions and especially to Dave Connolly in Belfast (a great TVR mechanic) for his telephone advice and help with wiring connections and understanding the way the Cerbera's ignition and fuel system works.
Well I thought I'd recount the tale of how I got it fixed, without the aid of a dealer, in case it helps out anyone else.
This story includes some lessons I learnt too!
Cyliners 1,3,5 and 7 - those on the passenger side - weren't running.
I first suspected the coil pack, and it did look a bit iffy, a bit puffy and split.
So I got a new one and put it in. Actually I got two new ones - one from Offords and one from the local Citroen dealer as the same part is used in a 1996 ZX 1.6i !
But no change.
Then I decided to get technical, and got out the oscilloscope and checked whether the coil pack was getting power and signals from the ECU.
Pins 3 and 4 on the coil pack input connector should have 12v; Pins 1 and 2 get pulses to ground (negative) from the ECU. Pin 1 goes to ECU pin 40, pin 2 to 55.
(on the other coil pack pin 1 goes to 19, pin 2 to 20)
Using the scope I was able to compare the signals arriving at the dead side coil pack with those arriving at the working one, and look at the signals actually at the ECU connector.
All those signals looked fine.
Then I did what I should have done first, removed the air inlet, removed the spark plugs and checked for a spark.
There were sparks!
So it wasn't an ignition problem after all. Neither was it a lambda sensor thing, the ECU software showed they were working.
So the only other thing that would stop the cylinders firing is no fuel.
Each injector has two connections - 12v and a signal from the ECU. The oscilloscope told me that the red 12v wire did indeed have 12v on it, but the yellow signal wires had nothing.
On the working side though, the yellow wires had nice negative-going pulses.
I started to fear for the health of the ECU at this point.
So I looked at the pins on the ECU connector which feed the injectors - strangely all four injectors are fed from the same pin - 38 on the odd side and 53 on the even (so all four injectors squirtnat the same time!) Both mine were OK! So the signals were there at the ECU but not there at the other end of the wire, at the injectors.
A ohm-meter showed a very high resistance between the injectors and the ECU - somehow that one yellow wire in the loom had gone high resistance. I'd suspect a splice joint in the loom somewhere.
(not in the loom connector - as my car doesn't have one!)
I identified some spare wires in the loom, there's about 5 that in the engine bay are labelled 'stepper' but not connected to anything. So I used one of those to replace the faulty wire, soldered it to pin 38 at the ECU and to the nearest splice to the injectors in the engine bay (which necessitated pulling back some of the black sheath to find the first splice).
All fixed and running lovely!
My thanks to Jerry Plummer for his suggestions and especially to Dave Connolly in Belfast (a great TVR mechanic) for his telephone advice and help with wiring connections and understanding the way the Cerbera's ignition and fuel system works.
. 
Roll on the summer evenings, 
The coil packs were roughly the same price actually, maybe £10 cheaper from Citroen.
Now, though, I have another problem with the car! The driver's side bank is acting up, but I think its the lambda sensor, as it's not reading correctly. May be knackered Cat, so I'm going to take both cats off and have a look through them before I change the sensors.
At 36,000 miles the sensors are due changing anyway, so I'm told.
Now, though, I have another problem with the car! The driver's side bank is acting up, but I think its the lambda sensor, as it's not reading correctly. May be knackered Cat, so I'm going to take both cats off and have a look through them before I change the sensors.
At 36,000 miles the sensors are due changing anyway, so I'm told.
If you buy Magnecor leads for a cerbera be careful! There are two types. The ones sold to me for my 98 4.5 were actually for the 4.2 cerbera. Although I was unhappy with them and phoned the TVR dealer before continuing to fit them, the dealer assured me they were correct. I expect a main dealer to know better than me so I believed him. Subsequently, I mentioned my concern to Joospeed because I was unhappy with the way the leads fitted. Joospeed discovered the main agent's mistake and supplied the correct leads. However I do now have a little used set of leads for a 4.2 cerbera which are surplus to my requirements. Perhaps a PHer would be interested?
Gassing Station | Cerbera | Top of Page | What's New | My Stuff