Missfire/backfire under track conditions

[dapinky]

Don't forget to do that in service mode, and for good measure I would do it when the engine is warm.

I never seen anyone recommending to do with a warm engine in this forum, but I recently learned that for my Toyota the reading is more accurate when hot, I guess it won't do any harm to do that with this car either

This bit of your answer awoke memories of times past, which have been completely lost within my memory banks until this morning!!!!!

Back when I first picked up a spanner and wondered what all those bits under the bonnet actually did, it was well before electronics got involved (in everyday cars, although I accept they had a part in exotics and racing cars). The mantra then was always set static timing on a cold engine (well, I suppose some vehicles which I didn't work on may have specified a hot one, but all mine were cold!).

As electronics started to creep in, setting the ignition timing was somewhat left alone as (although it had some effect on things) it had far less effect than when the only things which changed stuff was plugs, points gap, timing and valve clearances....

.... I am a bit of a dinosaur, and (once set) I rarely even check my timing on anything these days - except the Triumph, where it matters - and it is to the detriment of the cars' efficiency. I really ought to check mine more often - it really can make a big difference when it gets to the stage that the ECU cannot (easily) compensate under every condition encountered.

[RonR]

It might be worthwhile checking the knock sensor electrically, details are in the closed loop manual section EMJ.4 under "Code 43", page 62-63.

[peteoddcar]

I think you need to put it onto a rolling road dyno for a full load run up with you own diagnostics tools running and try to recreate the symptoms. ( shouldn’t be too costly if you need just 30 mins or an hour and can be flexible with when you attend.

[MPx]

Many thanks for the ideas/suggestions so far. This morning I tackled the base timing as best I could - followed the WikiLEC sheet - so thanks for whoever put that up. Its not mentioned, but I decided to also take the wheel arch cover out as I couldn't really see the timing marks properly otherwise. Being billy no mates, I couldn't quite do it as Joao suggested but it wasn't too hard to do it on my own. But I've not done my own mechanicals for many years and no longer relish it, so well out of my comfort zone this morning. I reckon its the best part of 40 years since I last wielded my timing light!

Some interesting finds - not least of which was that it was not set at 16 degrees. Anyone got any ideas why there would be a very clear white mark (and only the one white mark) at 0 degrees? That someone felt they needed to mark TDC, but not 16 degrees - seems odd? Anyway, I tried to mark 16 degrees but found it hard to get at and rather ham fistedly marked both the 16 and 18 markers with some white. But at least I knew I wanted the right hand one! I also marked the nick as that was very difficult to see - as you can tell from the photo taken before I marked it.

So this is how it was set

You can only just about see a slight mark on the pully and its between 6 and 8 degrees.

And this is how it was set when I'd finished


I'd set my switcher to program 1 for the stock S2 program while I was working on it. I then went for a 15 min drive and all seemed well at moderate road speeds. Then switched back to pragram 6 (Everest) and did another 20 mins. Slowish due to traffic but gave it full beans up a long hill and it boosted to 0.9 and felt very strong. All good, but prooves nothing as it would have done the same before I'd started. But at least it showed I hadn't messed anything else up.

FWIW I'm not (yet) convinced by this. I know Iain (RIP) had to set his timing way off Lotus Spec to get his to run at its best and SWLC have done so much to the car in trying to eliminate misfires they may have reasons for it to be set at 6/8 degrees. But I'll certainly be asking Phil about the TDC marker and if they would have set it "wrong" on purpose for any specific reason. Obvs it could have moved off on its own anyway - it doesn't do many miles, but the vast majority it does do are well above 3000revs so it is worked and if they do "go off", I guess mine would just like any other.

Ron, I did look at the EMJ.4 Manual and printed the first 10 pages of the section and the Code 43 pages (actually pages 88-99 and 149/50 in the version of the manual I found in the manuals section). Was troublingly close to gobbledegook to me. Also as I haven't got a Tech 1 I didn't think I'd be able to follow the instructions very easily - and not sure how all that ignition off/ignition on stuff works with the S2 alarm which shuts stuff down after 20 secs if the engine hasn't started. I guess I could have tried to find the connector cavity A12 to see what the voltage was - but since I wouldn't have done test 1 first...not clear what it should be. So I've stopped again for now. Will await the arrival of your data-logger once available and see if this base timing change has done anything about the more general knocks issue. Cheers all.

[dapinky]

Mike,

The reason that the TDC is often marked is because a "professional" timing light has a dial on it which you select the number of degrees you require, and the machine flashes the strobe at the right time to hit the TDC mark.

The reason being, that on some cars it is easy to mark TDC, but not necessarily the number you want..... or even have a scale fitted.

As the scale is on aplastic moulding, it may or may not be exactly in the right place anyway..... but finding an accurate TDC is easy by whipping out spark plug #1 and using a dial gauge to see when the piston is at the top of its travel (or even just drop a pencil into the hole and watch it!).

With the piston at TDC, you can then put a mark onto any (easily seen) part of the pulley and a corresponding pointer on the block..... then just dial in 16degrees BTDC into the strobe light and watch the marks line up at the right time.

[MPx]

Ahhh. Thanks Dave. So the likelihood is I've just undone some professionally sorted timing by lining up on the plastic marks. Oh dear! Will be interesting to see what the next scan has to say.

[dapinky]

Not necessarily, Mike... It may be just the particular angle that the marks are viewed from..... where you see it lining up with one, if you move your head round the wheel it may look different.

Any mechanic will set things up in the way that works for them at the time.... only time will tell if what you've done has improved things or not.... The easiest way to judge timing is to see how many (or how frequent) the 'knock' sensor has to make an adjustment on Elanscan.

In previous times I have just used a bit of deserted road and timed my acceleration between 50-80 in 4th gear and adjusted the timing until it was the 'best' without running awfully..... and that is where mine is still set to (I'm no expert!).

[MPx]

That sounds just like what Iain posted about the way he did it. Would be interesting to know what yours is reading according to the marks, given its what's "best" on your car.

[dapinky]

I've booked a day in the garage for tomorrow, so I'll check (Mrs P is allowing me a sunny day to play cars!).

[MPx]

So following my change of the base timing I had a trip down to SWLC Performance on Saturday and finally plugged in my laptop for most of the drive down and got this Elanscan file

EleanorSWLC.zip

covering the first hour and a bit.

I was driving quickly where I could but obviously not track day quickly. The car didn't miss a beat all day and was pleasingly quick. I don't really know where this leaves me. A couple of graphs I've picked out:

Many, and particularly Simon, were very keen on me checking the base timing (again!) because of the knock count on track. So I checked and thought I needed to change it. Its now set different to how it was and is perfect on the marks, but are the marks right? This is the knock count graph - the pink is the track session the blue the trip to SWLC after the timing change. I guess its better to have less knocks in over and hour on the road than in 10 mins on track...but I'm still getting knocks. Interesting to me that its not recording any retards to any of the knocks. Any further thoughts on this aspect??


So I looked at the advance graph to see if that showed any retard but its a bit too busy to tell really.


And just for completeness on an earlier theme....looking at coolant temp suggests all is well in that department now


If anyone has the time, I'd be very grateful if anything else in the scan stands out. I'm guessing I wont get any real progress on this until Combe next year!

[Tuga2112]

I'm not following the thread very carefully. So this is something someone may have mentioned already.

But...
Have you checked if the knock sensor is properly threaded in and doesn't have any play?
And there's nothing loose near it that may hit it while driving which may cause false positive readings.

I know it's a very basic check, but sometimes stupid simple is the solution

[RonR]

Not sure if it's a red herring, but have a look at this graph of wastegate operation vs knocks:

Wastegate Vs Knocks.jpg

Looks like the knocks coincide with the wastegate closing.

[MPx]

Have you checked if the knock sensor is properly threaded in and doesn't have any play?

Ha! I'm guessing this was a test for a bit of a laugh Joao! Took me a while to look up where the knock sensor is. Back of the engine of course makes it almost impossible to get at. Did my best and I'm guessing its this bit arrowed in the mirror image??? If so its all set nice and tight.


While I was back there, I did come across this connector which is always slightly worrying when not connected to anything! As my symptoms are all very sporadic/occasional its almost certainly not relelvant but would be interested to know what its for?

[MPx]

Looks like the knocks coincide with the wastegate closing.

Thanks for this Ron - an excellent spot. It certainly coincides a number of times, but also plenty of times showing one without the other both ways around so its not strictly in step. I don't think I've got a sticky wastegate (checked a few times) ...but is there anything else that the wastegate might be doing that would cause a knock?

On a slightly different tack, some posts back, Simon suggested that here shouldn't be any knocks, but in looking up the knock sensor in the engine manual it gives a write up of the operation and its certainly written as though knocks are expected and then describes how they're handled. So maybe I shouldn't be so worried about their presence, but more why there's no retards recorded when they occur?

[RonR]

Hi Mike,

You mostly use program 6 on the switcher, which is Everest Version 5. Alan says this on his web site:

V5 Chip:
The V5 was developed by created 16 power enrichment maps that varied smoothly between the stock Elan table and a considerably leaner (10.2:1 compared to 12.3:1 at the richest point) candidate table.
......

Map 8 was selected as the default map for the V5, it offers the best power boost with no measured knock.

As Version 5 runs the car leaner than standard, could this be the source of your knocks? Alan doesn't say if the V5 was tested on track, just on Dynos.

It would be interesting to see if you could run at next year's CC using program 3 which is Everest 4.5, to check against V5.

ALSO - Have you disconnected your battery since CC this year? If so the "Prom Error" code should have been reset, but it's still there in your latest scan.

[MPx]

Thanks again Ron. You're right that I run prog 6 all the time except for the MoT garage visit. I hadn't realised I had more than one version of Everest I just thought they were repeats of the one and only "Everest". If its on Prog3, then I'd be very happy to try 4.5 and see what the knocks might be on that version. I assume the fuel map is less lean on 4.5?

ALSO - Have you disconnected your battery since CC this year? If so the "Prom Error" code should have been reset, but it's still there in your latest scan.

No - the battery is never disconnected, indeed its on an Optimate to keep the battery properly conditioned. Is that another thing worth doing though? If I disconnect the battery do all the flags automatically reset?

[Saltire]

If I disconnect the battery do all the flags automatically reset?

According to the WSM, yes. I needed to do this a lot when doing my original electrical work on the car

However, everything else that requires a constant 12v feed (like the radio presets) also resets

I’ve not needed to reset again, so I’ve no direct experience, but I wonder whether unplugging the ECU (ignition off) and then replugging would rest the flags without losing other stored information. I’ll have to check the wiring diagrams at some stage and refresh my memory.

[MPx]

Thanks Jonathan. I'm not worried about the radio (never switch it on!) but I am worried about the Alarm as its the original which currently works fine, but I'm aware that they can be flakey. Don't know how sensitive it might be to a disconnect/reconnect although I do remember from changing the battery that if you do do that then the alarm triggers/sounds on reconnection.

[Tuga2112]

like Jonathan, i would have to confirm reading the books.

but im pretty sure you can unplug the ECU by removing a fuse. which would leave your alarm still connected

[Saltire]

My original Gemini alarm never sounded when I was disconnecting/reconnecting the battery, but I know others have different experiences. (I now have a non-OEM alarm where the issue doesn’t arise)