Mysterious injector issue - the story so far...

[Mervyn44]

I have a 2007 C4 Picasso, 1.6 HDi.

I've really been through the wringer with this car, It has been to several mechanics and been off the road now for over two months. My Dad is terminally ill so I really need transport and have decided to take matters into my own hands now.

It is very low on power, it makes a clicking noise and runs quite rough at idle. Upon acceleration it is lumpy then gets into its stride and finally pulls when the revs get up and the turbo kicks in. I've transferred my insurance onto our old beat up Pug 306 (which has served the family well for many years, but is now on its last legs... don't tell it I said that!)
Injector No.3 was replaced by my mechanic, this made no difference. He also pointed out some DTC's relating to the CAN bus and said that injector No.2 was stuck at 6 Bar, which is unusual apparently!

Since then I've learned a lot about injection systems, I bought myself a Lexia/Diagbox interface and have been doing some tinkering. Here's the problem...

Diagbox tells me that there is an intermittent open circuit on injector No.2, it also flags up the same CAN bus codes on the BSI, BSM and instrument ECU, these are F00A and F07A. I'm not sure if these are relevant to my problem, everything seems to work OK and I can't find information about these code on the internet.

Getting back to the injector issue, Diagbox shows reasonably decent readings for injector adjustment for cylinders 1, 3 and 4 (all well under 2). The live data shows them moving and approaching zero as the revs increase but injector No.2 is flatlined at just over 6... all the time, regardless of engine speed.

At first I though injector No.2 must be duff. I considered moving it to another cylinder, I reasoned that this would move the problem as well and therefore indicate that the injector was definitely the problem... Then I had a brainwave... why swap the injectors when you can just swap the electrical terminals?

So I swapped the electrical terminals on injectors 1 and 2, expecting the issue to migrate to cylinder 1, indicating the bad injector on cylinder 2... but it didn't! It stayed on injector 2 and the car ran really rough... would not rev above 2500.

I then decided to swap the connections between cylinders 2 and 3, I couldn't get the car to even start in that mode and it then flagged up an intermittent open circuit error on cylinder 3... which suggests that the injector may not be the problem at all.

Is this a wiring loom issue? or is it a fault with the BSI? Maybe the CAN bus codes are related to this, perhaps a wire in the harness is broken?

One thing that is weird is that when I just disconnected injector 2 the car wouldn't even start, so if it was a break in the wiring (even intermittent) then this would prevent starting occasionally, but the car starts first time on the key every time!

Dying to hear your views on this, I'm pretty much at my wits end right now.

Many thanks in advance for your advice - Mervyn.

[Peter.N.]

If you swap the leads you will be injecting the fuel at the wrong time - thing of the effect of swapping spark plug leads on a good old fashioned petrol engine.

Have you tried doing a leak off test on them?

Peter

[Mervyn44]

Thanks for your reply Peter.

I know that swapping the connectors will mess up the injection timings but I was hoping to see the indicated fault move to another cylinder when I changed them around, I figured this would point to the injector being the problem instead of the wiring. But when I swapped 1 and 2, cylinder 2 still flatlined at 6 so this suggests an electrical problem... but when I swapped 2 and 3 I got an intermittent open circuit fault on cylinder 3 (now cleared again) which is what I expected when swapping 1 and 2, and suggests the injector is faulty! So I'm not sure what to think now.

Update:
I plugged in Diagbox again last night. I checked the listed injector codes against the codes I could read on the actual injectors themselves and nothing makes sense. Injector codes 1 and 2 seem to be the wrong way round and the newly replaced injectors code (No.3) is not listed, instead it has the code that appears to be on injector 4! Its almost impossible to read the codes as the numbers are really faded but either I'm way off the mark here or there is definitely something wrong.

I'll try swapping the codes around to what seems right and see how that goes. If that doesn't help then I'll conduct a leak off test and report back.

Thanks again - Mervyn.

[BX]

Swapping the leads will cause poor or no running. The non firing cylinders will show up as all sorts of problems. The ECU will probably try to trim the non firing injector and have things all over the place. The only way to achieve what you want is to swap injectors. It is quite possible that the fault is in the wiring or the ECU itself. Mechanics love taking out injector and cleaning them with a wire brush. Do not attempt this on any common rail engine. It can cause a serious drop in power and drivability which usually improves with speed.The biggest difference between French engines and all the others is that the cylinders are counted from the flywheel end not the timing end. After that the differences are in things that do not affect maintenance or repair such as valve timing and overlap, the fine details of the combustion chamber shape etc.
A second hand ECU if one is required can be cloned off the original leaving both units plug and play in a given car.

[Mervyn44]

BX, thanks for the feedback.

I can't tell you how thankful I am for some help as this car is putting years on me! As I said, my dad is terminally ill, we are now relying on our old 2000 pug 306 which scraped through the MOT recently but really is on its last legs (hate saying that because its been a faithful old workhorse for many a year).

I swapped injectors 2 and 3 today, hoping that the problem would migrate from cylinder 2 to 3. This would indicate the injector as the culprit as suggested by BX. Unfortunately when I connected my diagbox/Lexia interface on the laptop it still showed problems with cylinder 2. Al cylinders appear to be injecting normally but injector 2 remains as a flatline on the live data, at just over +6, which I believe is the maximum adjustment the BSI can make.

So it seems to be an electrical problem so I am going to rig up a home made NOID light to see if there are pulses going to injector 2. If not then I will check the injector wiring loom.

[BX]

Check the wiring loom. If the fault turns out to be in the ECU then a second hand one can be obtained and cloned off the original. This makes it plug and play with no need to fiddle with anything else on the car.

[Mervyn44]

Many thanks for that BX, I was under the impression the BSI, BSM and keys all had to be replaced as a set or they wouldn't work... so I was kinda dreading having problems in that area... that they are able to be cloned is a big relief.

I rigged up a noid light made from a 10w festoon bulb. At 12v this draws nearly 1Amp which is similar to what the injector would draw (according to what I've been reading). With the injector lead removed and the noid light inserted the car failed to start, it tried for an instant and then just churned over on the starter... although the light did blink once or twice as it tried then stayed out. I suspect that what is happening here is that the BSI is sending impulses to the injectors and then stops doing so as soon as it gets feedback telling it an injector has been disconnected. The funny thing is though, the car will start without the light when the lead is disconnected maybe the 1Amp load is wrong and is confusing the BSI causing it to shut down? Its exactly the same story when the same test is done on cylinder 3. I may conduct the test again using an LED, these draw next to no current. The reason I didn't do that to begin with was because I had read that the injector circuit had to be stressed (to mimic an injector being connected) because the problem might be a weak signal which might fire an LED but not an injector, so give a misleading result.

Either way, the light did briefly blink which suggests there is (at least initially) a signal getting to the injectors... leading me to suspect a problem with the BSI or its programming. I know a guy who specialises in this sort of thing so my last resort is to call him and get a diagnostic run on the BSI circuitry and software.

[RichardW]

I wonder if the exhaust cam chain has slipped and put the valve timing out; this may account for the clicking noise and other faults. Difficult to check though, without stripping half the engine down. How many miles has it done?

[BX]

A filament bulb that draws that much current totally wrong for testing injector pulses. A filament lamp is slow to heat up and emit light. The resistance of a filament lamp is initially low but increases as the filament heats due to the positive temperature of the filament. The final current is generally much lower than the initial current. Injector solenoid coils are inductive. An inductive load has very low dc resistance. However it greatly resists current change. This means that when a voltage is applied across an inductor the current rises gradually over time. If it is a dc voltage the current is eventually regulated by its actual resistance and can be very high. In the case of your injector the pulse duration is too short to allow this to happen.
When voltage is suddenly removed from an inductor its magnetic field collapses. This induces a voltage (back emf) in the coil which is often much higher than the original applied voltage.
We need the injector to snap open and to snap shut as well. The back emf is often stored in a capacitor by some types of ECU. This is then used to fire an injector next time. Often it provides an itial voltage of over 100V for firing the injector. This pulse rapidly tails off and only lasts for a few microseconds. For the remainder of the time (still only milliseconds) the injector is held open by 12V.
This is an extremely simplified run down on what happens.
There are only three components for a mechanic to check and repair / replace. The injector, the wiring loom and the ECU.
The BSI and all the other stuff is not involved other than to supply power and some information to the ECU.

Richard is probably correct about the chain causing the clicking etc. However I think it is unlikely to cause a fault to show up on a single injector.

[Mervyn44]

Thanks once again for your input chaps. I made another post after that last one, but it doesn't seem to have been displayed. It was an observation that when running the exhaust smoke is white and smells really rich (diesel fumes) suggesting over rich running. It is not big plumes of white but it is noticeable, mainly because of the smell! Perhaps this is usual when the engine has not yet reached running temperature? I've had compression checks done and they come back healthy, the car has just over 94K on the odometer.

When I swapped the injectors the car had not been running for 1-2 days but when I removed them they were soaked in diesel, other than that they seemed relatively clean, no sign of carbon or gunk. They came out easily, in fact some of the securing nuts were surprisingly loose... although having now swapped them and torqued them down the overall situation is still the same.

[Mervyn44]

BX, I know that a filament bulb is slow to react but suspected that it might at least strobe a little, indicating impulses to the injectors, which it did momentarily. Having read your post I now understand that this was maybe not a helpful test, so many thanks for that helpful info... even if I didn't understand all of it!

[Mervyn44]

BX... hate to be plaguing you with this but you said something that made me think.

Its either the injector, wiring harness or ECU, the BSI is not involved. I was under the impression that it was the BSI that controlled the injectors? So the ECU you mention, what is this? is it the engine relay? Earlier in the thread I noted that there was an "F" code DTC logged relating to various systems including the engine relay (whatever that is). According to what I have read, these codes are very normal/common, do not mean much and are maybe not related to my problem. They are F07A and F00A (if memory serves).

So my big question is... if the BSI is not responsible for the weird figures I get in relation to number 2 injector, then what is?

I have now removed the BSI, it is a bosch EDC16C34 unit. I was going to buy a new one and have it cloned... but if this is not the unit controlling the injectors then I am barking up the wrong tree here!

I really need some clarification on that, and very much appreciate the assistance you guys have given to this point as I'm very much out on a limb here!

Kindest of regards - Mervyn.

[RichardW]

Too rich on a diesel =black smoke (and that's difficult to achieve at anything other than WOT!). White smoke = incomplete combustion, possibly caused by a timing fault - hence my musings above regarding the cam timing....

[Mervyn44]

Wouldn't this flag up in Diagbox?

[Mervyn44]

Well yesterday I did something I probably should have done long ago. Instead of simply doing tests with the car idling in the driveway I actually took it for a test drive. Turns out the exhaust smoke was a red herring, I still think there is an incomplete combustion issue but the smoke died away when the engine got warmed up. The idling quality and power are also improved and I put this down to properly re-coding the injectors and maybe also torquing down loose injector bolts... but there is still an uneven idle and a lack of power until the revs get up...and then it pulls like a mule!

I cleared the fault codes for the injection system and none have returned although No.2 injector is still constantly off the scale reading around +6 or "invalid value" in Diagbox. No EML, no associated codes.

One very interesting thing I did find this time related to the EGR. This is a new unit fitted before I bought the car. My mechanic had told me that it was throwing up codes but appeared to be working fine. I have yet to see any EGR codes being logged although the live data did uncover a few anomalies which I don't know enough about to say if they are significant. These are...

Bypass butterfly selenoid valve OCR (if present) = Value invalid.
This never changes, maybe there is no bypass butterfly selenoid?

EGR throttle electrovalve OCR = Value invalid.
this persists 85 percent of the time, occasionally switching to a changing numeric value depending on accelerator position.

EGR valve electrovalve OCR = 13.
This changes with pedal and rev changes as you'd expect but goes to Value invalid about 15 percent of the time.

There are no associated fault codes nor is the EML illuminated.

It would seem that a faulty EGR would explain all my problems quite neatly, although this still wouldn't explain the flat lining injector No.2... unless there is more than one fault source, not impossible I suppose. RichardW mentioned that one of his injectors was also stuck at 6, maybe this is not as significant as I think?.... I'd really like to get to the bottom of that.

I will do an EGR actuator test and see how that goes, is there any other method of knowing whether the EGR is operating properly?

As always, all comments and advice warmly welcomed - Mervyn.