silent running

My 98 WRX wagon is running at 1.4 bar on a Power FC with all supporting mods - mapped by Paul at Zen and no issues under full load and punishing driving. It's very fast. However, the Commander handset has shown very occasionally a 'spike' of very high level knock - an instantaneous peak of between 30 and 90 which only occurs under two conditions; on a steady light throttle cruise at motorway speeds, OR after a through-the-gears full-throttle blast up to 100 mph but then the knock spike happens on the overrun in 5th when I'm engine braking. Neither of these are times when you'd normally expect knock at all and mapping and remapping with lambda monitoring and det cans has revealed that there are no noticeable problems that would make this knock spike occur. What's more, it's not the sort of repeatable problem that's easy to track down, as even on a whole day's long distance driving it happens only once or twice, and often I can go a few days of all kinds of different driving without it happening at all. Then I can be taking it steady through town at 30mph in 3rd and suddenly there it is - a 50 knock reading, but no change in any other of the readings at all.

Paul's suggestion was that perhaps the standard PFR6B plugs might be the problem - a particularly hot plug setting off a bit of det every now and again - and that perhaps I should try a colder 7B plug considering the tune of the engine now. I had also recently changed the plugs myself soon after the original mapping and it had never done it before I don't think.

Anyway, changed the plugs over for 7B's no problem and they are running fine so far. No knock problems at all, although it's early days. A closer inspection of the 6B's that I'd pulled out showed a perfect mix, but one of the plugs had not had its crush washer flattened down, which would have effectively pulled it 2mm out of the combustion chamber. My fault as I didn't want to overtighten anything when I fitted them and I'd thought that it was as tight as I dared go! Anyway, as well as the uncrushed washer, it was also the only plug with a matt tan discolouration on the white of the top of the plug which goes under the boot of the HT lead. The colour was down at the hot end, the middle of the plug in other words. Also this plug was the only one whose top conductor had some dark grey coating, rather than shiny metal.

Any opinions? My first thought on seeing this is that by it not being fully down in the chamber, is that this would affect it's heat rating, but is it A: less plug length in the chamber makes it run colder because it's less exposed to combustion? B: Less plug showing makes it run hotter because there's less plug area to draw in and conduct heat away to the head?

Help!

silent running

Anyone?

jaytc2003

iTrader: (1)

there is a thread on 22b with a similar problem, they seem to reckon it is down to a possible sticking valve. The engine in question was stripped, nothing was found wrong, they re assembled it and the noise had gone!!

silent running

Bollocks! Thanks for the input though. I'll have a look over there and see what's what. The bummer is that even with the 7B's in place, it still did it again last night.

silent running

Just spent the evening rerouting all the breathers through a catch tank then back to the inlet (TMIC was full of oil btw) but no luck; it STILL gives me a big knock spike on the overrun in 5th gear, after hard acceleration through the gears. I got a picture of the Commander trace -it showed full vacuum and a very low level of knock bubbling away, then I must have given it a little tickle of throttle and it caused two big knock spikes in very quick succession, the second one was at 51, then it went back to full vacuum and very low level of knock.

buzzard

if all else is well,a sticky valve could do it. might be worth checking plug colours again. if one plug is still different to the rest a compression and leakdown test might give some clues. but why only in 5th? any strange noises from the box and being picked up by the knock sensor?

an intermittent injector or its connection could be to blame but if its only happening in 5th gear then seems unlikely. id wash the intercooler while i was trying to think of something else.

silent running

Cheers Buzzard. The nearside head has always been a little clicky, but nothing too noticeable. Problem with the plugs apart from that they're a ******* to get to, is that from what I've heard you could only be absolutely sure if you switched the engine off at the moment it knocked and then pulled the plugs straight away?

A compression and leakdown test might be the way forward though. As far as the 5th gear thing goes, it's just that that's the most reliable way I have of replicating the knock spike - overrun in 5th at high speed. But it has also done it on very light load, hardly any throttle (low vacuum), totally steady on the pedal, no acceleration or deceleration, just cruising in town traffic in 3rd. As usual, absolutely no reason for it to happen, nothing untoward on any of the sensors or gauges. Then it doesn't happen again until a day or two afterwards. Sometimes I can go on a gentle drive to work never even going over 4000 revs and half a bar of boost and the first I know of it is when I glance at the Commander once I'm at work and see that it hit a 60!

The injectors were just changed and the problem has remained so I'm ruling out that for the time being. Could be the gearbox maybe? But how to tell. The one thing I am thinking is that how can it possibly be detonation if I'm on the overrun with the injectors shut down? That idea of the sticky valve is interesting. Can you explain exactly how that might work? Would it be sticking open and knocking the piston or sticking shut and suddenly opening with a clunk?

buzzard

you're right about pulling the plugs that very second but if there is a recurrence of some fault it might indicate a slightly diff colour to the rest. if an inlet valve was slow to close a leaner mixture could result giving you a high knock level. piston to valve contact would produce a healthy noise level for the knock sensor. since you have a lot of oil in the pipework its well worth giving your boost sensor a flush.

silent running

Yes good point. What would cause a slow valve - carbon/oil build-up? A weak valve spring? I've given the boost solenoid plenty of cleaning in the past but never seem to have done the boost sensor. It's the other widget next to the solenoid isn't it?

buzzard

yup. the upside down one