Missfire/backfire under track conditions

[MPx]

I'm starting a thread here to bring together various posts around subtle issues with my car that I've been chasing for a couple of years now. Circs have meant the posts have all been made in threads on related topics but they weren't supposed to be about my car!

This has included various misfires as well as a performance issue. Much work has been done by Phil Webb at SWLC and last year I think he finally fixed the sporadic misfire - I benefitted from him swapping out all sorts of components and then replaceing the originals when the fault resurfaced. So I've been pleased to have had no stutters for the last few '00 miles. Sadly there is still an issue, but one very hard to reproduce, particularly on the open road. On the road - if I can find the right place/conditions to put the car under strain for a while, I can feel a loss of top end shove at high revs. The car is still quick, but just not as quick as I know it should be. There are no fault codes produced. At Castle Combe this year, the track conditions meant I could drive the the car into a fault condition. The car would be perfect after the warm up lap for a couple at full chat. But as it got further into the session I lost some top end - fast to 80-90 then just wouldn't go higher despite boost still at around .9 (Everest programme). Then on lap 5 or 6 I'd get a big pop, sometimes two, (like a backfire) and the CEL would come on. I'd obviously lift and within a few seconds the CEL would go off....but if I resumed it would happen again.

For the last run of the day RonR leant me his data logger and later (below) posted the file. On the run in the attached file the first few laps were on 6 Everest (no pops but I lost top end) - then switched to 5 (Kilimanjaro?) and got the pops on the second (which was my last) lap. Grateful for any pointers on what might be happening.

Well despite polishing up my credentials as a complete knob...not least by running over the great and good by starting the car while stood beside it - forgetting it was in gear - Doh! I had a great day!! Only possible of course by the company of so many of LEC (and indeed other clubs) and the tremendous help everyone gave me. Special thanks as ever to Pinky for actually working on the car to replace the wastegate acutator pipe and Ron for fitting his data logger for my final run and then studying the Elanscan file - just a shame the computer basically appeared to have recorded that the car was working OK. Well it was most of the time and gave me some great laps but just not as many as I'm used to! Many thanks one and all - greatly enjoyed.

Here's the ElanScan file for further diagnosis. There's a climbing Knock count, an Overboost later on in the run, and a Prom error which could be the mountain switcher doing something untoward. The PromID appears to be 599?

230105-104503.zip

A chance duscussion with Ray at Silverstone led me to question the CTS

I realise that this is happening in the wrong thread, but its where it is so I'm continuing unless it can be stripped out somehow. I'm still getting top end issues once the car is hot, and my instinct is that it may be overfuelling. Spoke to Ray about it at Silverstone yesterday and he suggested Coolant Temp Sensor having seen other posts on here. I've searched them now too and it appears to be a possibility. Before I ask SWLC to change the sensor, I thought I'd check here. Am I right in thinking this Graph from the run that RonR captured for me looks like the sensor isn't reading right - it covers 6 laps of Combe? tia

Joao wasn't convinced...

certainly others will know better than me. however i shall be dropping my ignorant opinion regarless.

if i recall correctly, the thermostat opens at 93 but thats a mechanical thing. so. not necessarily the most accurate. BUT
your temperature ranges from 89 to 109 and this according to your message is within running laps in a circuit.

for a pressurised system. i would assume the readings your getting are reasonble and if i had no input from anyone else. i would be looking elsewhere for the problem your describing...

but lucky for you... you dont need to rely on my ignorance. and you can wait for some real knowledge on the matter

id be curious to know what is happening to your machine.

Interesting thoughts Joao. It seems very odd to me that I go on track and do a warm up lap - maybe 120sec then do a couple of hards which for me at Combe in the M100 are about 95secs, and then I start to loose power and can no longer maintain the pace. I'd have thought the temp would climb (maybe higher than normal) and then stay there. What would cause it to drop at all if I'm still trying?

if i recall correctly, the thermostat opens at 93 but thats a mechanical thing. so. not necessarily the most accurate.

i would assume the termostat in your car is opening a little higher than the intended temperture. we know the elan was tested in death valey desert, although i dont know details about the weather there, im pretty sure a car tested in those conditions will have perfectly adequate cooling design to handle a trackday in british weather.

but ignoring that (again based on ignorance) opinion. have you ever noticed how slowly you need to be driving after being standing still to lower the temperature of your coolant from the fans being on to normal operating temperature ? Ive been in queues in motorway when the fans turn on and as soon as i get the chance to drive at 20mph the needle drops to normal operating tempoerature in less than a mile.

althought i already said it. im just going to repeat this to be sure you take my words with a grain of salt. but this is just ignorant opinions on the topic based on my understanding and experiences with the car.

the most i can help you at this point is to highlight what i would be doing if i was in your position. the figures from that graph dont appear to me unusually high readings from the sensor. BUT you can check the engine manual to see if theres any information there about what the ECU expects the Temperature sensor to report within limits. and maybe even see if it enters some kind of limp mode due to it.

just to give you an idea of the type of idiot your talking to, the following is just an extract of what i was typing and the thoughts in my head

"one thing that came to mind just now is that the secondary throtle may be one of those things that only operates at higher revs ? i think its operated by vacuum, but maybe if the boost pressure is too high it could work the line free? .... but then again, that would induce a vaccum leak which would be always present afterwards..... so i probably should delete this sentence now "

whats the readings of the MAP over that run ? a bad reading from that would make the ECU think theres not enough or too much boost and over/under deliver fuel.

also, this is a issue i had very recently on my 4runner... the TPS was not tight and after a certain amount of throtle it moved with the throthle... which made the car perform in a very weird manner, but certainly very underpowered.. just something worth checking i guess.

Then Phil chipped in on another thread....

I had problems with a CTS many years ago and been looking through my notes I made, before there was the useful page on WikiLEC. http://wikilec.com/view/CTS
It is a thermistor, which is 'electrical engineer speak' for 'thermal resistor' meaning a resistance that changes with temperature.
If you have a simple multi-meter with ohms scale the key information, easy to check, is : in iced water it should be about 8k ohms, in boiling water 185 ohms.
They are fairly robust and not prone to changing values, just stop working. Mine went open circuit, so very high resistance (infinity is high ) so ECU thought engine was cold. It was fine on first start up but soon too rich as the engine warmed up. (I found an old volume control 'pot' that covered the approximate range so 'lashed up' a manual adjustment until a new CTS could be purchased and fitted.)

I'd be very interested to see the graph from MPX with his new CTS fitted. On track, a full throttle blast down the straight will generate a lot of heat in the head so CTS will then read high, braking and part throttle will allow it to reduce, then high again for blast down the next straight. The graph covers 1000 seconds so 16 mins 40 seconds and range approaching 110°C doesn't surprise me - that's why the system is pressurised, to elevate boiling point ... 15 psi elevates it to 125°C. You get more power at higher temperature, particularly with cold air, which is more dense. Variation for 110° to 90° is within my expected range. BTW I run on an Everest chip and a diesel Astra thermostat which opens about 5°higher than the OE 'stat.

If you do the boiling water / iced water check and figures are about right I'd try a new thermostat before replacing the CTS. Thermostats can get a bit sluggish and not fully open - I usually check mine boiling it up in a saucepan - but only when my wife is out ..

I'd be very interested to see the graph from MPX with his new CTS fitted.

Thanks Phil, and so would I. Two issues are 1: I can only generate the symptoms in track conditions not the public roads - I will do a public road run to see what happens but it wont be conclusive whatever the result. 2: I'm currently in a Covid household and went positive myself on Sunday so I'm grounded anyway!

On track, a full throttle blast down the straight will generate a lot of heat in the head so CTS will then read high, braking and part throttle will allow it to reduce, then high again for blast down the next straight. The graph covers 1000 seconds so 16 mins 40 seconds and range approaching 110°C doesn't surprise me - that's why the system is pressurised, to elevate boiling point ... 15 psi elevates it to 125°C. You get more power at higher temperature, particularly with cold air, which is more dense. Variation for 110° to 90° is within my expected range.

And thanks again and very much chimes with Joao's expected variation as he told me on the Castle Combe thread. On the other hand...That graph isn't going up and down between corners and straights. Its slightly under a 100s lap for me if I'm trying - clearly not when I'm coasting if its gone bang - but anyway the temp warms up over the first lap or two then drops like a stone for a couple of laps then reverses again so these are more than whole lap swings. The symptoms start on about lap 4 and terminal on 5 and 6. Also very recent advice from Phil Webb at SWLC was that once the car is properly warm and thermostat open then temp should be stable and vary by very few degrees no matter how hard its used (obvs excluding traffic issues etc - and there weren't any!).

Anyway, I'll get my new graph and post it in my own new thread, and migrate all the other posts there too so that I stop hyjacking other's threads.....sorry!

[MPx]

And today I went for a run and produced this coolant graph as promised


In a sense this is good as its much more what I would expect - going from home(cold) to the nearest petrol station - temp climbs to "normal" and pretty much stays steady until I'm sat next to the pump pffaffing about with the laptop to stop logging and save the file while the engine is idling and getting hotter. So the new CTS appears to be working well. Trouble is the 6-7 miles were on a country lane with traffic and at 30 to 50 mph speeds....so not really much of a test!

I did give it some beans on a slightly longer circuit home, but didn't get to any loss of power situation.

[lotusflasherman]

I have some old Elanscan files on this desktop computer but most are on an old laptop I keep to connect to ECU with USB cable.

I have attached a couple of images from my file created when I drove out of the village (30mph) onto a bendy country road with some 90° bends, before turning round and driving home to park on my drive .. temperature rise at the end is after stopping ..

data cts1.JPG

This shows a steady rise to 86°C at which point the thermostat opens .. and cool water from the rad' enters the engine so it drops a bit before climbing to a fairly constant operating temperature of 84°C . The rise at the end is after parking on the drive but did not switch on rad fans - I usually 'trundle' home at below 30 mph in a high gear, usually 5th (which is around 1500 rpm at 30 mph) so I don't have to do the usual 'cool the turbo' procedure.

data cts 2.JPG

This one has the road speed plotted against the CTS temperature graph shown above which gives a bit more information ... such as I turned around and headed homewards about 415 seconds into the trip and was stationary on the drive around 580 seconds when coolant temperature rose a bit with no airflow.

Interesting point maybe is that my CT plot is dead straight and horizontal at 84°C while travelling back ..

[MPx]

As you say Phil, interesting that your plot is dead straight on 84 degrees, but I think with my new CTS I'm happy enough that mine is now flat enough around "normal" temps...but would be interested to see if it saw-tooths a bit more when pressing on.

So I'm not really sure where this leaves me. I was working on the possibility that the old CTS was signalling cold when it wasn't and the ECU was then over fuelling resulting in loss of top end performance and then backfire. That would fit my symptoms. But it really only manifests properly on track. If I have a particularly energetic blast about, I was sometimes aware it lost top end, but that's very hard to reproduce at will. I guess I'll have to wait for Combe again next May to give it a proper test.

[MPx]

So finally got back to Combe for a real test. Car behaved perfectly on the Motorways and A/B roads getting there (as normal). The temp sensor change has certainly helped - no backfires or ECL...but also not cured the problem. Car goes like a train for 2 or 3 laps but then just loses power at the top end. Sounds the same, all instruments look right, but just won't pull at the top end, pace goes, and cars I was faster than glide past. Arrrghh! Doesn't happen at road speeds in road conditions so obvs easy to live with, but it certainly takes the shine off an outing to Combe. As last year, Ron was a star and lent me his new, even neater, data logger - but I was only out on the parade and the car was fine, so not expecting much from that file when it gets loaded. Not sure where to go with this now. Can't reproduce the issue off track, and these days I only track once a year. Seems to be one of those where I'll end up taking a punt on something random and hoping. My current thoughts are potential fuel issues - maybe injectors - just because that side of the car is still original (so 30 years old, - but it only clocked up 60k miles on the trip up to today), whereas electrical/exhaust/etc has all been renewed more than once.

Interested in any ideas people may have of how this might happen. Beats me how the car's just brilliant, but then goes off. Will be brilliant again next time....until is not. All the stuff I can think of would either work or not - and not come and go.

[Saltire]

Mike, my only idle and probably mad thought is this when I used to track cars (many years ago now, and not a Lotus) I found I was consistently getting underbonnet temps much hotter than I could “achieve” on the road. Given that the Elan is pretty packed under the bonnet, and has a turbo as well, I just wonder if that could be the root cause. On the road, your temps aren’t high enough, but on track, extra thermal expansion takes place and is affecting either the mixture or an electrical circuit in some way.

I’m probably talking rubbish - as usual - but it might account for your symptoms.

[MPx]

Thanks Jon its a good thought - must have merit in helping identify what's "unique" about more than 3 laps on track, and might revert to normal later! Its just where to look for something affected by the additional temps. We know its not general internal engine temps which are monitored and "normal" in terms of coolant. But maybe a sensor or something electronic misbehaving when hot, or something deforming, or maybe fuel vapourisation. Hard to think of anything electrical rather than electronic that might be affected.

Is there anything carrying high pressure around the engine bay? Turbo boost? Fuel? Vacuum? I'm guessing a very hot pipe might expand more than it should and drop some pressure? Probably a numpty thought...but I'm really struggling to rationalise what I experience!

Anyway, thanks again for the steer.

[RonR]

Mike,
I've attached your files below, it seems the file headers weren't corrupted after all - for some reason the logger didn't rename them from ".ec_" to ".ecu" after closing them.
Did you change the switcher map at any time while the car was running? You have an error 51 logged, which can sometimes happen if you switch maps at just the right moment. I've not looked at them in detail, just enough to make sure the logger was working properly and to notice you were seeing some knocks on the parade lap.

mikeMPX-240525.zip

[MPx]

That's really helpful, thanks again Ron. The car was (as always) in programme 6 - Everest - on the switcher. I only ever change it to 1 (Std S2 map) for the service/MoT each year. So no it wasn't switched at all. But I did press that red button on your logger a couple of times while we were in the queue waiting when I wanted to reset before we set off - didn't touch anything thereafter until back parked up - can't remember the sequence/order of engine off, switch logger off, disconnect connector so I may have caused an issue there somehow?

Pleased to see the coolant looks more believable than last year. Coolant temp rises while idling/waiting. Dropped for the time of the laps on track. Rose again once coming off track until shutdown. I'm hoping that's what we would expect - reacting to the lack of cooling air flow through the car at low speed/stop? Seems believeable to me anyway despite being slightly counter intuitive that it cools slightly when under max load!

What is Coolant raw? Its graph shape is the opposite of Coolant temp.

There were a lot of knocks (60) recorded in the 10 mins recording. As far as I could see, no retards - why no reaction? Or am I looking in the wrong place?

As a short roadspeed drive, we had no issues so I'm not surprised it doesn't appear to show much - but then I'm a numpty at this. As ever grateful for other's thoughts.

[Tuga2112]

following on Jonathan's idea.
is your intercooler piping all connected right ? is the air intake temperature recorded by any chance ? if the air going in is considerably hotter, theres nothing that the computer can do to compensate, it wil lose power.

[MPx]

Thanks Joao. Honestly dont know about the pipework as not sure I'd recognise if it was right or not.

Is the MAT temperature relevant to what you're asking? If so here's the graph from this year


Looks fine to me (but it would since we didn't suffer the issue!) - higher on idle while queuing, then oscilating lower while on track, rising again when we came off before switch off. Comparing to last year where the problem was dire...its all much hotter, but also looks OK to me - but maybe that's too hot???

[Tuga2112]

yes Mike, that is the sensor that would report what i was thinking.

now, this is all ignorant statement. and im sure others will know better than me. BUT i looked up air dencity tables based on temperature and found this page.
https://www.engineeringtoolbox.com/air- ... d_600.html

if you scroll halfway down, you find a table (or just ctrl+f "Density and specific weight of air at 1 atmosphere pressure, at temperatures given as °C:")

and there i found these figures first number is temperature and second is kg/m3
60 1.060
80 1.000

now, if i interpret this correctly (and crudely) at 0 bar, you have 6% less air for combustion, if your pushing an everest chip, that be 0.92 bar, which is nearly 2 atmosferic pressures.
im making the crude assumption that would be close to a 10% less air going into combustion in between the 2 readings.

i mean, this is all armchair ignorance. but in my head. i would expect at least 10% less power between runs. which i guess could feel like a huge difference when on a track ?

i just wanna make you aware, i have no engineering background, so this is all assumptions but im hoping letting this thought out there can open up (or dismiss) other avenues of investigation.

[Simon_P]

My guess would be ignition timing. Either base timing or CAS sensor fault.

I'll guess that when you're tracking it you get enough errors for the ignition timing to be retarded enough that you notice it.

Interesting that you get Knock on the parade lap - there shouldn't be any.

[MPx]

Joao, thanks for the analysis, I can see your logic. However, I dont think your line of thought on the graphs reflects the symptoms I get. On last year's graph, the car was really excellent up to around 600s, and only then did it loose power and eventually backfire and CEL. I'm not seeing a change in the MAT graph that would trigger/result from that albeit there is a slightly less concise movement - but of course I'd be reacting on track (backing off!) which may itself have had some effect.

Thanks Simon. I too am concerned about the knocks - particularly as this year there was no retard reaction recorded at all despite the 100 count. What might cause no reaction to be recorded? Last year there were retards



There were around 100 knocks and plenty of retard over the first 600s thereafter although knocks climb another 100, much less retard action. I dont "know" for certain about my base timing as I dont typically work on the car myself. I've been trying to get to the bottom of this with help from SWLC for more than a couple of years. We've spoken about base timing and I've been assured its been checked and rechecked and is spot on. I believe them - they do this for a living! As part of the elimination of a more prevelent stutter they tried a different dis unit (no change) and then I had a new (s/h) CAS and Hall Sensor assembly which appeared to do the trick. This was in Feb 2023 before last years Combe. Clearly it didn't fix the issue in extremis - or quite possibly there was more than one issue. I did later fit a new Temp sensor myself and that also appeared to help, giving a more believable graphical output. Since then I've not been able to generate the symptoms on roads, just on track, and even there I no longer get a backfire/CEL just a loss of top end power.

[Simon_P]

Base timing must be checked in service mode - anyone with a strobe can do it - take it to a friendly garage just for a second opinion. You need to know that it is correct or you are wasting your time.

There should be 0 knock - something is clearly not right.

[RonR]

Mike, I found something in last year's file which occurs at 506.4 seconds:

Flag ALDL 59 Bit 1 is set - Boost Spark Bias Active
Thereafter ALDL 38 Bit 3 toggles on/off - PE_TIMER_FLAG Exceeded MAX

Image8.jpg

It'll need some more digging to see exactly what this means, but at least there's something to investigate!

[MPx]

Wow, that's a great spot Ron - many thanks. Certainly very interesting as an event change happening at about the right time.

The sad part is that I've absolutely no idea what most of the words mean let alone where to start digging! I'm guessing these are just two of the many unnamed flags in Elanscan?

So once a flag goes "Active", the PE timer starts to exceed its max. Any idea what a PE timer is? Or the Boost Spark Bias come to that?

I'll look forward to getting one of your loggers and seeing if I can capture a similar event on the road....

[RonR]

Looking at the spark advance graph, the baseline seems to be 20 degrees, and after the Boost Spark Bias Active flag is set there is more retardation of the spark going on than before.

Image10.jpg

The flag is set as a result of calculations on the number and rate of knocks, so I think we're getting closer.....

[MPx]

Ah OK. So we're heading back towards a timing issue and a problem caused by the number (and rate) of knocks detected. So I need to take on Simon's mantra - base timing is all - and try to either do it myself of find someone happy to do it while I watch. I hope to get some time to have a go at that later this week. Appreciate the help - thanks again.

[Tuga2112]

Ah OK. So we're heading back towards a timing issue and a problem caused by the number (and rate) of knocks detected. So I need to take on Simon's mantra - base timing is all - and try to either do it myself of find someone happy to do it while I watch. I hope to get some time to have a go at that later this week. Appreciate the help - thanks again.

Setting the timing on this car is fairly straightforward but easier if you have 2 people.
The most practical way I know of is to lift the front drivers side corner and take the wheel off,
Have your helper describe where the notch is while you adjust the sensor (distributor cap on the na)
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