My old beast has a bit of a misfire and is down on power. The plugs are getting a bit old but I recently checked and set the gaps down to 0.7mm and I stuck on some new Intermotor plug leads (the Magnecores it had were falling apart).
Is there any way to check if the coil itself is failing, or do you just test by substitution? Also, is there anything that actually fails over time in spark plugs apart from the gap opening up?
It may be something other than a HT issue of course - any Diagnostics info ?
Yes, substitution is the best/quickest way to check coils; Have you looked at the coil physically (for cracks)
and/or observed in a very dark environment to check for leakage (sparks) yet ?
There aren't any fault codes stored. It's particularly stumbly and rough on petrol, but mostly runs okay on gas. On petrol it's often pretty lumpy accelerating away - on something with a carb I'd say it felt like the accelerator pump jet was blocked, it feels like it leans out massively as you go from idle to about quarter throttle.
It's single-point gas, which is set up pretty much correctly (it's maybe a little rich).
If running OK on gas, then that kindof indicates ignition side is OK. Generally a weak ignition will be worse on gas cos it needs those healthy volts even more than petrol.
I'm thinking that there's some coil testing data in ECU SID, but I don't have it for GEMS on this PC.
Maybe trims have drifted a long way off? I'd do an adaptive reset if it were me, but that could make things worse of course
Have you reset the adaptives? If it's running rich on gas then it will have set the trims on petrol to attempt to lean it off, so it will run rough on petrol. The longer it has run on gas the worse it will get. For some reason when mine is run on gas one bank shows permanently rich while the other shows a correct mixture with the lambda flip flopping as it should. That makes the other bank adjust the trims to max lean and it will run rough when switched to petrol. Resetting the adaptives cures it and it runs fine until it has been run on gas again for a while.
I haven't reset the adaptives, I'll need to get a shot of some diagnostic equipment to do that.
Surely the trims being wonky wouldn't make it misfire on gas?
No they won't, I thought the problem was only on petrol. If it's on both then it does point towards either an ignition problem but, as OB has already said, that would usually be worse on gas or a leaky injector. When you had the plugs out were any of them particularly blacker than the others? That would signify a rich mixture on that pot so could be down to one or two dual fuelling. As your gas system is only taking a lambda reading from one bank, if you've got a pot or two on the other bank running way out all will appear to be normal. I actually suspect that's what I've got. Mine takes the lambda signal from bank 2 and I'm pretty sure at least one injector on bank 1 is intermittently leaking petrol in there too so that bank always shows as running rich and at odd times it feels slightly down on power.
A certain amount of lumpyness has been cured by reprogramming the gas ECU so its default stepper point isn't 255.
This has also brought the fuel consumption back to about normal ;-)
Sounds like you have the same quirk as I found on the Ascot (HSE, silver one, whatever). That is also using an OMVL Millennium with R90E and taking it's lambda reading from one of the 5-0V sensors. Got it all set up and running fine but then left it idling for a long time and got too rich and too lean errors. At idle the sensor stops giving an output but the GEMS petrol system runs open loop at idle so it doesn't matter. However, the Millennium sees a lean output so starts to move the Default to compensate until it reaches 255 and can't go any further. That flags the lambda too lean error. Then when you drive it, it's running rich so flags a too rich error. What I did was adjust the R90E to get the default the same at idle and out of idle then used the default lock to hold it there. Set the actuator limits to +-40 out of idle and +10, -30 at idle so it will keep the emissions down at MoT time. I also turned off the too rich and too lean diagnostics but left the lambda fail dagnostic enabled. I suspect this is why some people have said that they don't work too well using the original 5-0V sensors and why mine has a separate 0-1V sensor fitted in one downpipe solely to drive the LPG system.