Cosworth BOA swap start but backfire

[Baltazard]

Hello, I come from France I made a swap of a Cosworth 2.9 V6 24v BOA in a Ford Capri 3
and I have backfire problem, I will explain my problem in the 2nd position
Here is a summary of what I've done

My capri was of purchase


Bodywork







Motor receiving



Parts mounting












Wiring the ECU




Mounting

[Baltazard]

Engine starts very well, but I have backfire intermittently, I changed the fuel pressure regulator, I have NGK IX iridium spark plugs,
Cable plug and coil is new
I have the exhaust tube 2 and 6 very hot after starting 5 minutes. Spark plugs are black, I install the lambda sensor Bosch but no change
I put the edis unit next to the eec iv is there a risk of interference
Had you ideas where that might come I'm lost

[davew]

Now that's a lot of work and a lot of modifications etc.... like the 'customised headers' in particular, even the clutch servo...

Such overfuelling is probably an injector/wiring issue though... Check with Noids but also get a FCR and post up the Codes too !

Both EDIS and ECU are screened and so interference is possible but very unlikely

[Baltazard]

Thank you for your answer
Before motor fitting I disassemble the injectors and clean with WD40 and compressed air
I have drive with the intake pipe removed all the injector sprayed, one wire drive 1 2 3 and the other 4 5 6
It may be worn but it's weird for gasoline injector, no ?
What is Noids and FRC

Which is weird too, before I installed the steering pump, alternator and brake servo I had no backfire
I used pipe of EGR to the aspiration of the brake servo
https://drive.google.com/file/d/0BxLfeM ... JEb00/view

[davew]

Interesting... by using the vacuum of the EGR circuit like that you may simply have upset the ECU. If you just block off the vac. feed to the servo what happens to the backfire ? Or Is it possible that you just disturbed something else whilst fitting the servo ? As the backfire is intermittent this could just be bad wiring/continuity, too...(?)

- Injectors often 'gum up' over time but it needs better solvents than WD40 to remove these... (NOIDS are just LEDs you connect to injectors to see if they are working OK); The problem however is that clearly the injectors are probably staying open too long because the ECU is telling them to, based primarily on inputs from the MAF, TPS and the Lambdas . An (exhaust) manifold leak can suck in too much air and so the Lambdas see this as a weak mixture and over-compensate too. Vacuum leaks can do the same...

An FCR is s Fault Code Reader for the ECU; Based on your EGR mod. you may now get some Codes that are not easy to interpret but it is the bast way forward now !

[davew]

Incidentally do you have a video of the engine running properly...? The posted video tends to point towards fuel starvation...
and/or an intermittent electrical issue with one of the two main injector feeds !

[Baltazard]

This is before having to change the pressure regulator
Is it possible that the injection pump does not arrive enough sucked gasoline as it is in the engine compartment
(This is a new scorpio pump) through the original hose ? which could make a bad spraying gasoline ?

Here is my mounting to retrieve the depression of valve egr

There was no electrical connection to operate the vane egr it worked with depression capsul
I deleted it when I installed the stainless steel manifold
I tested when I close with finger the vacuum is like that with the servo, In the video he is in the air

When I say intermittently there is approx all 30 seconds, less since
I have lambda sensors and both the misfire and both the backfire
if I take a limited rpm it do less but I take a lot of rpm say a machine gun

What's weird is that I have the exhaust pipes of the cylinder 2 and 6
that's really hot the other his warm after five minutes of start
and 2 corresponds to the middle of the coil I'm afraid I have a problem
with edis or eec IV You were right it is best to read the error code
But how does it work with eec IV, I have this diagram for the ECU
Gunson 77032 FRC is good ?

[davew]

As you have such a non-standard set-up it is quite difficult to diagnose for sure.... as again both fuel and electrical problems can do this ! However:

-The issue/test with the EGR was really to check if the vacuum was compromised by adding the servo, but from what you just said probably not

-If you are still concerned about the proximity of the EDIS and ECU then try separating them (particularly the cables as 'pick-up' between these cables might happen - but the engine already ran OK like this previously)

-If you are now using an 'in-line fuel' pump specified for the Scorpio it should be fine, but check the filter too, it is important to have one big/clean enough !

You should also make sure there is no debris in the fuel system too of course.

- Are all the plugs blackened now or are some worse than others ?

As the engine is running so badly you will get MANY Error Codes though ! Yes the Gunson 77032 is fine for reading these (3-wire DLC)

[Baltazard]

I will do another test tomorrow but there was who was darker than on other
The filter and pump are for scorpio and new, It's an 8mm fuel line
with an adapter 13 mm to the underside of the pump and the filter before the pump
I do not know if the original scorpio a pump on the fuel line or directly into the tank
The spec of the scorpio pump are 4bar but the Capri 2.8i pump is 6,9bar
I asked myself if the pump scorpio have enough suction force for Capri line
Is what I will see with Gunson or I have to test with a manometer

[davew]

As previously stated it is unlikely to be the fuel pump.

A Fault Code Reader will not give you a single/definitive reason for the problem (such as Fuel Pressure), but should give some 'good clues' !
Unfortunately when an engine is running this badly the ECU will probably just be in 'Open Loop' anyway...

[Baltazard]

Here are spark plugs
3 2 1

4 5 6


I just noticed :
coil for 4-3 is really black
coil for 6-2 is black
coil for 5-1 is a little dark but smells gasoline


After cleaning spark plug I have a little less backfire but it comes back quickly as before
The small metallic sound just fan of the alternator rubbing on the pulley of the steering pump
I have to grind it slightly, before he did not touch but it rubs
https://drive.google.com/file/d/0BxLfeM ... pkd1k/view?

Here are without the brake servo
https://drive.google.com/file/d/0BxLfeM ... FpSEU/view?

[davew]

Ouch ! Now that we would say is not really 'backfiring' but in fact 'misfiring' - but anyway it is much more likely to be caused by electrical/timing issues.....

As for the coil problems the only way it can be 'blackened' is if there electrical arcing (sparking) is happening between the coil outlets, can you see any of this if you observe in the dark ?? (Watch out for the dangerous fan etc when doing this, obviously). Do you have another coil to try instead ?

Incidentally the engine does not really sound 'quite right' before you added the servo, either !

[Baltazard]

Desolate I misspoke this not be the coil that is blackened but spark plugs are in pairs on the same coil in coil assembly
spark plugs 4-3 is really black
spark plugs 6-2 is black
spark plugs 5-1 is a little dark but smells gasoline
4-3 are on the same coil, 6-2 are on the same coil and 5-1 are on the same coil
If that makes me feel that the coil is not similar controlling by edis module?

[davew]

Ah yes - that makes more sense now, the plugs not the coils are blackened !

Yes, that blackening does tend to indicate that the coil/plugs are firing intermittently/spuriously, and it is a combination of the ECU and EDIS that decide this....
Again separating these two modules (physically) would make some sense. Is the screening on the cables also earthed ?

It may be that you will have to use an oscilloscope to monitor and compare the signal feeds to the coil too in order to diagnose this further...

[Baltazard]

I need to test to separate modules, the shield of wires are connected to ground as in this schematic
By cons I just thought of a trick is that the shield of the wires of cps sensor may touch the ground or
only pin 7 edis because it is possible that it may touch the ground in the path

[davew]

You may be onto something there, the CPS on these engines rarely fails, but if the signal from it was grounded it could cause this kind of problem, yes !

[Baltazard]

I will try to isolate the shield of the wires of cps sensor with electric sellotape to see if that changes anything
When I receive the error code reader I'll post the results

[davew]

Well expect a lot of Codes ! Because of this the Gunson unit may not give valid Codes with the engine running either ('KOER' Codes)
In the meantime check that wires to MAF / TPS / ECT are intact too !! You may still need a 'scope analysis at some point though too...

[mk2 cossie]

Sounds like the same symptoms mine had, and that was down to the ht leads being on the wrong way around

[davew]

Yes I wondered about that myself to begin with Mk2 Cossie (due to the '5/6' quirk with these coils) but the engine ran relatively OK before the servo was fitted....!
That could be coincidental of course but 'something' has been disturbed to produce such over-fuelling (and it may be more fundamental than ECT and/or EDIS wiring etc, too...?) You are absolutely correct though that if it suddenly runs this bad it is usually 'something we did ourselves' !