This is a read-only archive of the old Madabout Kitcars forum. New posts and replies are no longer possible, but every thread is still here to browse and read.
Well, for the first time in nearly forty-odd years of motoring, yesterday I broke down. After queuing through some road works for about 15 minutes the Marlin started to misfire. It would idle OK - but as soon as I tried to apply the power it would misfire and grind to a halt. Eventually I pulled in to the verge and called the RAC as I didn't fancy diagnosing the problem at the side of a very busy and narrow A road.
The symptoms were quite odd. It would start and idle fine but as soon as I tried to pull away it would cough and die. Although I had been in a traffic queue the engine temperature had barely rose above normal. Of course by the time the RAC turned up the fault had completely disappeared and the RAC man was as baffled as I was.
Any clues anyone? I wondered about a vapor lock in the fuel system. Is that possible with fuel injection? Or perhaps the air flow meter?
I once had a problem with similar symptoms (but before setting off...) and it turned out to be the connector on the AFM was a bit corroded and dry. A bit of contact cleaner/lubricant and it was fine.
If it's not that I think I'd be looking for a camshaft or crankshaft sensor that is going open circuit when it gets a bit warm but difficult to diagnose until it fails completely.
The engine should run fine without the AFM plugged in or at least this is true with MAF based engines. It will just sit on a default worst case map and not perform as well.
It sounds more like a lack of fuel getting through, probably the high pressure pump on its way out. When you open the throttle plate to pull away it's just taken a big gulp of air and died but there was enough to keep it idling. Unlikely to be cam/crank sensors I'd have thought as it wouldn't start properly if the DME can't work out the engine position. Modern cars will refuse to even try and start under this condition but things may have been different with the M20. I've had coils go before and the car idles fine but when you try to pull away it rattles you to bits but doesn't stall. I think your M20 has a distributor anyway so...can't be that.
Thinking about it this sounds just like my mate's 330 that refused to pull away the other day. It idled fine but as soon as you lifted the clutch it died. He's fallen foul of the common problem on BMWs where the siphon between the two sides of the tank stops working and you can't let the tank drop below a 1/4 without getting stranded. It's especially bad on a hill start where the fuel runs to the back of the tank.
Well I replaced the hp fuel pump and all seemed ok - until today when the same problem re-occoured. I am now wondering about the fuel pressure regulator. Does anyone have a spare I could borrow to try?
Also - can anyone suggest a source for a fault code reader. I still have the diagnostic socket so it might be worth trying that.
Have you checked all your engine electrical connections really carefully?
I don't know the M20 engine, but does it have an Idle Control Valve like the M50? If it idles OK the ICV is working, but when you accelerate you then rely on the MAF and TPS to tell the ECU what fuel to add. These, or the connection to these are therefore the likely culprits?
I had the same problem as yourself with my M50 going to Stoneleigh this year - I'd driven 80-90 miles, half of it on the motorway, then when I got onto some A roads it cut out above idle. I fiddled with the TPS and thought I'd cured it only for the problem to reappear 2 miles later.
Mine turned out to be a connection pushing through on the MAF connector (I'd obviously touched the MAF wires while fiddling with the TPS) - I know Ian says it should make no difference, but I am certain that was my issue as it has not reappeared since finding the problem and making a proper connection: a female in the MAF connector socket pushed out when the male was pushed in creating a tip to tip dodgy intermittent connection fault.
Out of interest I've read that a failed TPS should not cause the issue for the same reason Ian describes - it should revert back to closed loop and only result in less efficient fuel consumption but not a misfire.
Whatever your problem turns out to be, my money is on a dodgy electrical connection/component/wire somewhere.
I've just read the previous posts and I've obviously experienced the same problem as Peter having a dodgy MAF connection causing the problem - you might want to start there checking each wire very carefully - even for a broken wire?
I've been over the wiring very carefully - I don't think it's that.
I think the clue is that it seems to happen once the engine is warm, and then when it has been idling for several minutes. (The engine temperature is rock steady at about 90 according to the gauge.)
I would still like to track down a tool for the diagnostic port to see if that tells me anything. I have also ordered a cheepo fuel pressure gauge. If the pressure regulator is on the blink that should show something.
The only other thing I can think of is the low pressure pump - but that is by the tank so I don't see why that would be affected by a period of idling.
I think the clue is that it seems to happen once the engine is warm, and then when it has been idling for several minutes. (The engine temperature is rock steady at about 90 according to the gauge.)
I would still like to track down a tool for the diagnostic port to see if that tells me anything. I have also ordered a cheepo fuel pressure gauge. If the pressure regulator is on the blink that should show something.
The only other thing I can think of is the low pressure pump - but that is by the tank so I don't see why that would be affected by a period of idling.
Cheers, Robin
Hi Robin,
I've been pondering your engine problem and the symptoms do suggest 'heat soak' causing an intermittent connection or component failure, which 'fixes' itself when it cools down.
I agree the coil could be a candidate as it sits behind the radiator in the Sportster. (I know I had a spare but I haven't seen it since moving house so it may be in my lock up?)
Other possibilities are the crankshaft position sensor (I think I sold my spare, but I'll have a look?), Ignition lead sensor(I've definitely got one of those that you can have), Fuel pump relay?
You could try the old 'hairdryer' trick on the smaller components to try and induce the fault or maybe a can of freezer spray once you've created the fault.(This might be worth trying on the coil).
On fault code readers, the only one I am aware of is by Peake Research. But the early Motronic DME's had a very limited list of fault codes and even the manufacturers warn that the fault codes can be misleading.
I'm probably going over to my lock-up sometime next week so i'll see if I can find anything useful.
Thanks Peter. Any spares I can borrow will be usefull. I have already replaced the hp pump and filter and it's starting to become an expensive problem....
I have another theory. The fuelling arrangement as suggested by the manufacturer of my swirl pot is a bit odd in that the low pressure overflow from the swirl pot is teed into the hp return from the fuel pressure regulator. This means the same fuel could be recirculating through the hp circuit via the swirl pot. At idle when you are not using much fuel it could mean that the fuel is overheating. So this week I am going to try a one way valve in the lp circuit ensuring that once fuel has been through the hp pump once it returns to the tank.
I fitted a valve to prevent the fuel constantly being recirculated through the high pressure pump. Now it has to return to the tank. It runs fine, at full temperature, without any sign of misfire - in the garage that is. Tomorrow afternoon I will take it for a run and give it a proper test.
One slightly worrying other development. I noticed that the rubber O-ring that acts as a seal on the fuel filler cap seems to have swollen making it a pain to refit the cap after filling. It's a recent problem - it's been fine up until now. I wonder if the petrol companies have been fiddling with the recipe...
hi robin, I have a spare fpr you are welcome to have if it suits, it is a 3 bar pressure regulator for a 2.5 m50 engine, part number 0280160503. I have a thread on this forum about something very similar happened to me ....I replaced maf idle control valve tps, fuel pressure regulator, injectors cleaned... the problem was in the end so simple, a faulty connector on the alternator, dropping voltage to ecu, and causing every fault code in the book, but essentially gud idle and spluttering under fueling was intermittent and frustrating.
hi robin, I have a spare fpr you are welcome to have if it suits, it is a 3 bar pressure regulator for a 2.5 m50 engine, part number 0280160503. I have a thread on this forum about something very similar happened to me ....I replaced maf idle control valve tps, fuel pressure regulator, injectors cleaned... the problem was in the end so simple, a faulty connector on the alternator, dropping voltage to ecu, and causing every fault code in the book, but essentially gud idle and spluttering under fueling was intermittent and frustrating.
Thanks Jeremy. FPR seems to be the same part as for the M20. I will see what my "test" tomorrow brings and, if I still have the problem, I would like to try your spare.
Like you I have already replaced a few bits. My first thought was the HP pump as that was also making a strange noise. They ain't cheep and it didn't solve the problem Grrr. I can see now why some people still prefer carbs!
I managed to borrow a fault code reader and... there weren't any.
And I have now added a one way valve to make sure the fuel goes back to the tank after its been through the HP pump. Previously it could circulate through the HP pump and swirl pot giving it some chance of overheating.
I had similar a while back and it was poor connection to relay holder for the injectors. I had looked at all manner of things prior and really only found fault by chance . The cable at fault into the relay holder hadn't 'locked' into place so installing the relay pushed the female in the holder back making for intermittent contact.
I have at one time replaced the HP pump as it was making a hell of a racket and on its last legs but the engine still ran ok. so doubt its fuel related especially as it ran OK on its original set up.
Well - the first sign of warm weather yesterday and the misfiring has re-appeared. So far I have:
Checked all the connections
Changed the fuel filter (I had a spare so that was free)
Changed the HP fuel pump (£££ doh but at least I have a spare now)
Changed the fuel delivery so that fuel is always returned to the tank after it has been through the fuel pressure regulator (no chance that it can re-circulate and overheat).
Improved the tank venting so there is no chance of any restriction.
None of these has had any effect whatsoever except on my wallet.
My current thinking is if it were the coil or other spark problem then there would be unburnt fuel exiting from the exhaust which you would smell (you can't) and also would cause a backfire (it doesn't).
So I am now thinking it's a sensor problem.
The TPS is a pretty simple, mainly mechanical, device and I can't imagine that being much affected by temperature.
The oxygen sensor is only used to fine tune the mixture (as I understand it) so I can't imagine it being that either.
So that only leaves the engine (camshaft) position sensor.- I have ordered a new one from GSF. They seem to be on special offer at the moment so the price wasn't too bad (££).
I am beggining to hate fuel injection...
Now I just have to wait and see....
As usual - any constructive suggestions welcome!
- compression test on each cylinder
- spraying carb cleaner for air leaks as a lean burn would cause this
- each plug out in turn still connected to the HT lead but grounded and turned the engine over to make sure you're getting a good spark on all cylinders (fuel pump fuse removed).
- pulled the lead on each plug in turn with the engine running to check it bogs on all cylinders (not sure this is a good plan H&S wise with HT leads mind you)
- just pulled each plug to check whether any are wet with fuel, oily or damaged etc
- checked each injector is firing (screw driver on each with your ear against the end)
All my previous misfire issues on various cars have come down to a bad pencil coil. In all cases the car would idle fine and even pull away gently but as soon as you tried to accelerate the misfire would start. I've always had an OBD2 scanner to hand with makes life easier on modern engines as it will tell you which cylinder and often that it is actually a coil but the symptoms seem very familiar.
You won't necessarily get back fires or smell petrol either if just one cylinder is not firing 100%, just a serious shudder.
I wouldn't suspect the cam sensor as that would either prevent the car starting or see it revert to a batch injection mode which would just be less efficient (not sure if m20 was fully sequential though anyway). Most other sensors such as coolant, air and O2 should all have a similar effect and you'd notice a lack of power and high fuel consumption rather than complete breakdown.
my money's on either an air leak or something in the ignition chain between the coil and the plugs
The problem only occurs when the engine is thouroughly warm. After 10 minutes rest the fault disapears until the next time.
The m20 only has the single sensor for engine position which I assume is used to determine when to fire the injectors and spark. Without a signal the injection will stop as its the only indication that its running. Or am I wrong?
There does seem to be some history (google) of hall effect sensors becoming fa ulty when they get hot -which is why I decided to change it. I think the sensor has some simple electronics as well as the hall effect device. A single transistor I understand which can be affected by heat.
Anyway we will see... (Assuming last wednesday isnt the only warm day of the year.....)
no I think you're right. If the engine only has a crank position sensor rather than cam too, it would almost certainly stop totally (and not start) if it was completely dead as it's the only thing available to time injection (with a few other things such as temp thrown in). I guess the distributor is attached directly to the cam??? so that gives you your spark timing. But i think you said it idles with the fault but won't rev? I would expect a bad crank sensor to not even manage an idle. It sounds like the M20 will be using wasted spark and batch injection as normal so ignore what I said earlier about it reverting to that.
If the TPS were bad you'd definitely have drive-ability problems especially on wide open throttle as you'd not get the necessary enrichment, but the engine should still run without too many problems under normal driving. It is basically just a potentiometer but I've had 2 die on much younger engines. In that case the car just kept going into limp mode every few days, but still worked fine the rest of the time. Had several occasions where I had to quickly drop the clutch, stop and restart the engine on the motorway or going through bends at speed which isn't fun. On modern cars with drive by wire, the TPS has 3 tracks and the outputs are compared by the ECU for consistency. As a safety feature, if one starts acting up it will put the car in limp mode to prevent a dangerous condition where the throttle is left wide open by a fault that it then can't detect in the feedback loop. On a cable driven throttle this isn't necessary as the springs will always return the throttle plate so it may only have a single track. I can see how you'd be skeptical on how it would fail but the tracks can burn out. I don't think replacement TPS of this simple type are normally that expensive so it may be worth swapping that next if the CPS doesn't sort the problem. On a drive by wire car it normally means a replacement throttle body so £££.
After that I think you need to start looking at more general issues you'd get with EFI or Carbs.
[EDIT] the other thing to say is I'm pretty sure the M20 uses a VR sensor rather than Hall effect. Even the early non-Vanos M50s were still using VR.
Well the sensor has arrived so I can't put the job off any longer. It has a 3 wire connection and so I think it is probably hall effect.
I was looking at the job last night. The sensor is easy to get to but the wiring is a nightmare as it passes through a purpose made trunk in front of the engine and behind one of the pulleys. I will have to take the cowl off and the radiator out to access it which is a bit of a pain. It think it is routed this way on purpose to avoid proximity to the HT circuits so there is no cheating taking an easier route unfortunately. It's about time I changed the coolant anyway....
Well I managed to change the sensor last night without pulling the whole car to bits. I have fitted it temporarily with the wiring running over the top of the engine. I left the old sensor in place just tied up out of harms way. If that proves to have done the trick I'll do it properly later on in the year.
Well... it's not the crank position sensor! The problem occured again today. Now, as soon as the engine is up to temperature, it starts to missfire gradually getting worse until it stops.
The "good" news is that it is getting worse so at least it should be easier to find!
I havenow changed the coil. The saga continues......
