engine missfire under load

[simondosborne]

I have a missfire problem with my 1999 500 Griff with 15,500 miles on the clock. Under heavy load the engine appears to drop a number of cylinders to such an extent that the car slows rather than accelerates when you give it the beans. NO PROBLEM WITH STARTING OR UNDER LIGHT LOAD OR TICKOVER ENGINE SPEEDS. i HAVE CHANGED THE PLUG LEADS, IGNITION AMPLIFIER, COIL, FUEL PUMP AND ALL EARTH CONNECTIONS HAVE BEEN MADE GOOD. Any ideas as to what it can be? the problem improved for a short time after coil and ignition amplifier were changed BUT AFTER 10 MILES OR SO BACK TO SQUARE 1.

[carsy]

Plugs themselves.

[43scotty]

Do you have the original induction hose fitted ? its worth disconnecting the hose at the air flow meter between the AFM and the air filter, mine looked fine but because its double walled the inside was collapsing under heavy acceleration and restricting the air flow.

[blitzracing]

Ive seen the above as well- but why do you assume a misfire is ignition related when it could be lack of fuel just as easily? (heavy load- heavy fuel requirement)- it needs some step by step diagnostics. You can check for ECU fault codes, and fueling with plug in readers like ECUmate, or RoverGauge. This costs much less than blindly throwing bits at it. You can also do fuel tests under load with a simple test meter on the lambda probes, but you have to extend the test wires so a passenger can read the meter when you drive the car. You should see at least 1 volt (1.2 preferable) at over 3400 rpm between the white and black wires. Below this the voltgae should cycle 0 to 1 volt at .5 to 1 second intervals.

Edited by blitzracing on Wednesday 16th January 20:46

[simondosborne]

Thanks for that Mark, I used the roverguage software i believe supplied by you, fuel systems & voltage all worked fine when checked by TVR engineer in the car whilst i was driving

[blitzracing]

So no error codes at all? AFM voltages OK? If you have a misfire you will get very high levels of short term negative fuel trim as the ECU sees unburnt mixture as a lean mixture, so removes fuel trim to richen the mixture- which is the wrong thing to do- result is black plugs on the ones that are firing as it over fuels, so plug colours are worth a check on both bamks. Mind you the ECU is only in closed loop below 3400 rpm or 2/3 throttle, so if the fault is not there in closed loop, you wont see the ECU trying to correct it, as the lambda's are ignored in open loop above 3400 rpm. I have not checked how RoverGauge shows this, but I suspect it will show 0 short term trim. (not that its easy to read a laptop under load tests anyway) A simple test meter is a better bet on the lambda outputs for the passenger to read in this specific example. You will quite clearly see the voltage drop to zero volts as the revs rise if its genuinely running lean. On the ignition side, check you have a genuine Lucas Rotor arm, as there are some nasty pattern ones that short out.

In terms of HT diagnostics its harder- an old school garage with something like a Crypton tune can check HT voltages and dwell / ignition primary switching on a scope- it will tell you a lot. Gunson used to do a simple hand held unit that clamped around the plug leads and showed peak HT voltages on an LED display for the home mechanic- a very useful item, but no longer in production I think.

Edited by blitzracing on Thursday 17th January 18:52

[simondosborne]

Thanks Mark, we will try the tests you suggest, but no fault codes at all came up on the tests so far

[ruaricoles]

I had a strange issue with my Griff earlier this year, when it ran fine all the time except under full throttle in 1st, 2nd or 3rd; all it was was the air intake elbow pipe becoming detached from the plenum under bigger engine movements and letting air leak in. Someone else has already suggested the pipe collapsing, so worth checking I think.