Rovergauge help with MoT fail

[Milstad]

Having read a lot of previous threads regarding MoT failures/ rich mixtures and Rovergauge, I have just installed RG and found it a great insight into what's going on.

It was really disappointing after a 3 year chassis-up rebuild to be told by my local MoT station that I'm best off abandoning the test, as I have some work ahead of me to sort out some very high overfueling issues before it would pass. At least they didn't charge me.

On RG I'm seeing the even bank of cylinders functioning correctly, cycling each side of zero on the fuel trim, but odd side is showing constant -100% trim.

I've checked the MAF and see 0.3v at ignition on and 1.83v with engine running. I think this is as expected from other posts. For the lambda sensors I see; odd - good ground on white lead, 13v when running (white/ red), 9.6 ohms across white/ red, and 1.4v drifting to 1.8v on white/black, with no cycling. Even bank, I see same, but with 0.2-1.4v cycling on white/ black.

I wonder if I need to replace odd bank lambda?

I have also noticed some missing, so will need to investigate plugs/ lead/ dizzy cap/ rotor.

Anyone have any advice on the lambda they could share please.

[blitzracing]

The lambda output is stuck at really rich on the faulty bank which tally's with the MOT reading. Start by removing all the plugs on that bank and check the colour. Misfires cause a fake lean signal to the lambda probe, so it will overfuel to get the probe to switch but yours stuck and not cyclic. Possibly an injector stuck open?

[ITVRI]

I would recommend the use a infra red thermometer to quickly check each exhaust pipe and diagnose which cylinder(s) are not working.

[Penelope Stopit]

ITVRI said:

I would recommend the use a infra red thermometer to quickly check each exhaust pipe and diagnose which cylinder(s) are not working.

Clever diagnostics, like it

[blitzracing]

Have a look at this article- good breakdown of how the probes switch.


http://www.g33.co.uk/img/fuel/14cux-fault-finding....

[Milstad]

Thanks for the suggestions guys. I had already run the infrared thermometer and didn't see any cold exhaust flanges, but there was a difference - some at 320ºC others more like 280ºC after cold start up.

Plugs out next for inspection. Distributor cap, rotor, leads and plugs were all replaced as part of the rebuild, and injectors were cleaned and serviced.

[spitfire4v8]

Milstad said:

Distributor cap, rotor, leads and plugs were all replaced as part of the rebuild, and injectors were cleaned and serviced.

Probably your first mistake then ..

I had a chim 4 litre on the dyno yesterday .. the owner was a lovely chap and had replaced his cap, rotor, ht leads, lambdas all with new parts.

On the dyno it made a misfiring 165bhp .. put some good secondhand items on that I had and in the end it made 210bhp, bang on for a 4 litre.

Just because it's new doesn't mean it's any good.

We did a couple of tests with new airflow meters that the guy had spent hundreds of pounds on (reputable sources) .. the difference in mixture between them was over 10 percent at full throttle with one being so poor in output at small throttles that it maxed out the lambda trims and still didn't achieve lambda=1. Again, a good used genuine AFM sorted it.

Unless you have a VERY good reason for putting new parts on, my advice is DON'T.

[ITVRI]

Milstad said:

I had already run the infrared thermometer and didn't see any cold exhaust flanges, but there was a difference - some at 320ºC others more like 280ºC after cold start up.

If there is no large temperature changes try physically swapping the lambda sensors over. If rovergauge then shows the problem on the same bank you can rule a faulty probe out.

[blitzracing]

it would be an odd fault for a probe, as he has the voltage readings for an over rich mixture, just as the MOT main said. If it where something like fuel pressure it would affect both banks, but it is really odd its just one side. A misfire will make the cylinders that are firing run rich to compensate for the one that's not, but firing the ECU will need to add fuel to do this to get the probe to switch, and this is the opposite with the probe saying its rich and the ECU trying to remove fuel. It should have thrown a fault code as well if the probe has stopped switching for a period. Im thinking plug colours should tell you a lot, as Id expect them to be jet back if its genuine rich.

[Colin RedGriff]

[quote=Milstad]

It was really disappointing after a 3 year chassis-up rebuild to be told by my local MoT station that I'm best off abandoning the test, as I have some work ahead of me to sort out some very high overfueling issues before it would pass. At least they didn't charge me.
/quote]

Is this the first time it's been tested since the rebuild?

What work was done on the engine during the rebuild?

Where any ignition or fuel system components changed?

Just trying to understand if this is something that has come up, possibly as a result of the rebuild, or if it has developed after the rebuild.

[Milstad]

I should get a chance to pull the plugs tomorrow, so will have a look and see. Since rebuild I've started the engine in the garage and run it a few times and then driven to the MoT, which is five miles away. Not sure if that will be enough to blacken the plugs? It was very noticeably rich from the strong smell of unburnt fuel though.

Like many, I'm sure, I had the 'whilst the engine is out of the chassis, I may as well....' moment. Block went to Powers and was skimmed on both decks, adding head porting and enlarged trumpets. Crank was also reground, with new rods, clutch, fully balanced, etc. New piston rings, new cam. On top of this, new ignition parts as I said above. All the injectors were serviced by Powers too. No other components changed, but new fuel lines everywhere.

I did notice that the offending lambda sensor had a broken wire as part of the rebuild, so it must have been running for sometime before the rebuild not operating properly. That has been repaired, heatshrinked etc and the circuits are functioning, as per voltages in post above.

[blitzracing]

Don't forget the ECU has to re learn the long term fuel trim. Get the engine hot and let it idle for about 2.5 mins and see if the long trim moves. It will start at 0 after an ecu reset, and should move slowly untill the short term trim cycles around its mid point. It may be that simple.

[Milstad]

Hmmm... I swapped over both lambda sensors and the -100% trim remained on the odd bank, so I guess they are both OK.

I checked the plugs and they all seemed to be a bit sooty - no single plug was worse than the others. Then reassembled without the plug extenders.

I've dug out the old distributor cap/ rotor leads to give that a try.

I do seem to recall that the engine has always been a bit lumpy on tickover, which could be me just seeing an old problem.

ECU reset, and still running too rich on odd bank.

I wonder where to next? There must be somewhere in the system that issues different commands to the left and right bank. I'll try the old ignition parts just to check, but it feels like a fuelling problem on the odd bank?

[blaze_away]

Are you certain you have correct firing order ?

Just my own personal observation...Testing exhaust temps I've always found a bit of a challenge as heat soak happens quite quickly and the exact place you measure on each port has a huge effect.

[blitzracing]

Check the long term trim. If the fuelling is miles out for one bank, the long term will have reached 100% on that bank and run out of adjustment. If its less than 100%, and the short term is still not cycling, its not completed setting the long term up yet- so more heat and more idle time.

[Milstad]

Thanks for the help on this one. In the end it exceeded my capabilities and I gave it to Dom at Powers to have a go - it turned out to be a faulty ECU, so thanks to Dom and I'm back on the road after a long rebuild

I mention it as I would not have thought to check the ECU itself, as it was working OK before strip down, so this may help someone else with these issues in future?

[TarquinMX5]

Glad it's sorted . Thanks for updating with the result; it's not too bad on the TVR section but it's becoming more noticeable on other sections/fora where requests for advice aren't acknowledged and/or results aren't posted which, if they were, could help others in the future.