The V8 Owners Forum -

I have started fine tuning my new engine. I have noticed a discrepancy in lambda corrections across the banks of about 12%. So either there's a fuel problem or an air problem.

I have checked the airflow into both banks, and this appears even to within a couple of percent.

So I therefore turn my attention to the fuelling. I have a standard fuel rail fed with a Bosch 044 pump, matched injectors and an aftermarket 1:1 pressure regulator. The lean bank is at the regulator end of the fuel rail, meaning there might be a pressure drop across the rail.

Has anyone seen this sort of behaviour and how was it dealt with? Thinking a parallel feed to two separate rails might be an answer.

Thanks

I don't know the answer but I will be interested to know the outcome.

I think I have a similar setup 044, standard rail, and Bosch 155 green injectors.

What fuel pressures/duty cycles are you running? Are you running boost? What power are you making/aiming for?

I have seen someone use the later square fuel rail (from Thor equipped engines) I seem to remember the larger cross sectional area reduced the fluctuations in fuel flow.

Tom.

My thinking is that if there was a pressure drop between banks, there would also be a pressure difference between injectors on each bank?

There might be a pressure drop all the way across the rail. I can only measure the average lambda per bank so would not see this.

Further thinking about it says the discrepancies are too big for pressure drop unless the fuel pump is on its way out. My alternative theory is there is an air leak somewhere.

Rover fuel rail is has a pretty reasonable inside diameter and of course an 044 pump is pretty handy too. So it wouldn't be at the top of my list of things to check. Might be worth monitoring fuel pressure at WOT, either electronically via a pressure transducer or a pressure gauge in the car.

Have you had the injectors flowtested (or swap them with the other bank)

Could be a tired Lambda. Try swapping side to side.

Regards Denis

I thought this too and put in 2 new lambdas. They are much cheaper in th US...

I checked the injectors a while ago using the fuel system and the ECU to deliver a known amount of fuel into a set of jars. They were tolerably even.

Bizzarely, applying the fuelling offsets I estimated from the assumption of the fuel rail pressure drop problem leads to a completely happy stable smooth engine.

It looks like I have the right cure for the wrong problem. The fact that the fuelling corrections fix the problem tends to reduce the chance of it being a misfire, along with the sound the engine makes...

dnb wrote:I have started fine tuning my new engine. I have noticed a discrepancy in lambda corrections across the banks of about 12%. So either there's a fuel problem or an air problem.

I have checked the airflow into both banks, and this appears even to within a couple of percent.

So I therefore turn my attention to the fuelling. I have a standard fuel rail fed with a Bosch 044 pump, matched injectors and an aftermarket 1:1 pressure regulator. The lean bank is at the regulator end of the fuel rail, meaning there might be a pressure drop across the rail.

Has anyone seen this sort of behaviour and how was it dealt with? Thinking a parallel feed to two separate rails might be an answer.

Thanks

There are many things that can effect distribution, whether air, exhaust, injectors etc
The chances of it being a fuel rail issue is almost zero unless it's an incredibly small diameter rail.

The lean reading could even be valve clearances, slight misfire, wonky sensor, air leak in exhaust etc

Yes parallel will always be better when lots of fuel flow is required, but the rails will not be the problem.

12% sounds like a lot though

How were the injectors tested and listed as flow matched ? At a single 100% duty, or at a few different duties ? Are they new ?


The injectors I'm using at present were supposed to be matched...but when I had them tested a few years ago independently, there were variations of around 5%

I tend to agree with your analysis Stevie. I was beginning to clutch at straws because all the other options seemed exhausted.

It's a normal early EFI fuel rail. The only difference is that I have reversed it so the fuel flows round it the other way to standard for ease of connections. (This should make no practical difference)

The puzzle is that I doubt I could conspire to put all of the under-flowing injectors on one bank. The air leak is a possibility, but the discrepancy seems too big for what is essentially a relatively well behaved engine.

The injectors were flow matched (to 1% allegedly) but this is several years ago so they're probably a bit off now. It might be time to get them tested.

Is it 12% under all operating conditions ?

Same type of wideband on each bank ?

All operating conditions I can reasonably get at on a car with no MOT - I have access to a small amount of private road so I can test with a little bit of load.

Yes, the same type of lambda on both banks, and I swapped them to see if the problem would move. It wouldn't.

I think there must be a mechanical issue.

I cant see an injector problem causing a 10% difference unless it was major...although a single injector not firing or almost not firing at all might just cause something like that, but I'd expect more than a 10% difference in readings if one injector stopped working

it cant be really an ecu or tuning problem.

So it must either be cam/intake/exhaust/valvetrain related problem

A 4 gas reading might help pinpoint. But even a relatively small air leak pre sensor whether a hole, gasket, misfire etc could easily cause the dodgy reading.

Where are the O2 sensors located?

stevieturbo wrote:I think there must be a mechanical issue...

I fear you are right. I am resigned to taking the inlet manifold off tomorrow afternoon to see if I can pressurise it and see some leaks. I can deal with the valley gasket oil seal at the same time and have a look at the exhaust.

The cam and valve train are newly assembled, so I would like to believe they aren't causing problems, at least until I exhaust all the other possibilities.

The sensors are a couple of inches after the 4-into-1 collectors.