Hello, since I got my 02 Justa about a year ago, it has always had intermittent difficulty to start when cold, "cold" being defined as "not being since at least 2/3h", not a temperature issue (I'll get back to temperature).

When I go to start the car normally when it's cold, one of three thing happens:

• it starts right up (1 to 2 cranks), which is what I expect and what I want to happen every time
• it struggles to start (3 to 5 cranks, never more), which I qualify as a "hard start" and not something I'd expect on a well maintained car (which it now is, after everything I've done on it)
• it's a crank no start (I stop cranking after 10 or so cranks, I don't want to fry something), then it starts right up after I shut down and try again.

Important note: I always wait for the fuel pump to stop spinning after putting the key to position 2 and before actually starting the car.

Now if it only struggled to start, I would have left it alone, but last winter I had a few crank no start, and it started again this winter.
I do not know if it is a coincidence or not, but all the times it happened I was at my in-laws place (not just in the neighbourhood, specifically at their place). It generally happens around 5°C to 10°C. Not lower (it started right up when it was snowing) and not higher (it never showed trouble during summer). The only common factor I remember is rain: it was already raining, or had rained not long ago, so moisture might be an aggravating factor.
I do not want to accept that and I've started digging hard to solve that. That includes reading the 32 pages of the sticky https://www.northamericanmotoring.co...-starting.html and seeing that there is no accepted fix, just a lot of different parts that have been changed, sometime successfully, sometime not, lots of people blaming software and lots of people blaming gas.

For those 2 last parts:

• I do not want to go to a dealership, ever, so the firmware version that's on it stays on. If that's the issue, it'll have to stay like that and be called a quirk
• I live in France, and we got RON95 and RON98 gas, equivalent roughly to 91 and 93 octane gas in the US. I use exclusively RON95, as per owner manual. Apparently Mini USA recommended to go lower octane and that solved issues for people. There is no possibility to do so here. And I don't really want to put higher octane gas either.

So hopefully it's a part misbehaving, and I want to troubleshoot as far as I can without throwing parts at it and hoping it fixes it.
Luckily my Mini still starts, so I'm determined to see it through the end, however many months it will take. It's already been 1 month and a half since I really started digging into it, I don't have any deadline and lots of time and patience on my hands (though I cannot work on the Mini has much as I'd like, so things can take a while between each operation).

Here's what I have done so far that's even remotely related, including maintenance done since I own it (so everything is less than 1 year old), and none of it made a difference to that issue:

• oil change (of course). Nothing weird in the oil, though it was running very low on oil when I got it.
• spark plugs and wires. Again nothing weird there, I replaced them because they were of unknown mileage and a wear part.
• coolant change. Again, unknown mileage/age, wear part, nothing strange.
• injector cleaner bottle in the tank when I got it. It's been a year, I'll put in another one soon, but don't expect to see a difference.
• replaced the gearbox (Midlands with destroyed bearings). New one works great. Didn't touch the clutch/flywheel because I have documentation it was done recently.
• changed the crankshaft sensor o-ring, and obviously cleaned it.
• changed the fuel filter. It was completely black, so probably overdue. No change afterwards.
• tested the battery. It is now 4 years old, but both voltage and cold cranking amps are good and well above the minimum.
• got the starter out, did a ghetto bench test, greased the bendix. Also tested when starting with the battery tester (which measure voltage drop), and voltage stays above 10V, so everything's good
• cleaned the coils plugs. They were already clean with no sign of corrosion, but I sanded them a bit.
• fuel pressure tested (gauge at the fuel rail) the system. The time I tested it, it puts a strong 48psi in less than one second when putting key to position 2, then is about 42 psi at idle, it spikes up to 50psi when you stab the gas pedal, but is a bit lower to 38/40psi when you keep the gas pedal a bit down. After shutting down the engine, it keeps a steady 43psi pressure after 30 minutes. All in all, it seems within spec.

I tested various ways to start the car:

• holding the clutch pedal. No difference
• pressing the gas pedal before starting it (to get the throttle plate unstuck if it was). No difference.
• instead of starting the car after the fuel pump stop spinning, shut off and do another fuel pump cycle. That seems to help, the car mostly start right up when doing that, I actually can't remember a hard start that way, and no "crank no start" one for sure. I'll do that all the time from now on when not pressure testing fuel, and see if it ever fails to start.

The thing is, the time I tested the fuel pressure was at a time it was not exhibiting issues. I'm left wondering what it would show when it's exhibiting the issue. Due to the intermittent nature of the issue, it's really hard to catch it into the act, especially as I need to connect the fuel pressure gauge beforehand (because I need to relieve pressure on the fuel rail).

Yesterday though, my Mini finally gifted me with a new clue! And I even recorded it on video, after 20 or so uneventful start sequences and missing the only crank no start. See video below.
What happened is that I had tried started it 5h before (so there was still pressure on the fuel rail), and didn't do the "shut off and 2nd fuel pump cycle" dance. It started right up, but struggled to idle and stalled a few seconds after.

My theory for now is that the fuel still in the rail allowed to start and run a bit, but something sticky in the fuel system prevented from sending fuel. Unfortunately for now it's only a theory, I'd need to catch it into the act with the pressure gauge connected to confirm it. Because there are still other valid theories around air/fuel mixture.

What's your thought with everything laid out so far? What would be your next steps?

Right now, my only planned next steps are:

• change the camshaft sensor o-ring, and clean and properly reseat the connector. Someone mentioned that fixed his issue. That seems far-fetched for my situation, but I'll try regardless, gaskets are cheap enough.
• pull and clean the MAP sensor, throttle body and change related gaskets. My father in law (before yesterday development) was convinced it's an air issue, so it makes sense to rule that out. Again, gaskets are cheap, and I just installed INPA to reset throttle body adaptation.
• carry my fuel pressure gauge around, especially at my in-laws and try to connect it at every chance to try to catch it into the act. I'm only there about once a month, so it might take a lot of time, months, perhaps even years if I'm unlucky/situation doesn't allow enough tests with fuel pressure gauge connected.

Thanks for any help!

Here's the video. At the beginning, I try pressing the gas pedal. Right before it stalled, I tried giving it some gas to no avail. It started right up afterwards.

 

Have you checked your engine ground strap?

 

I take it there is no CEL? No pending codes?

Second what NC TrackRat has to offer.

I'd add check all battery cable connections not only at the battery but where the cabling connects. It is not a bad idea to clean the battery posts and cable connectors to ensure a good connection. At the other end disconnecting the cables/wiring and cleaning the contacts would also be a good idea. A 2002 is going to suffer from age/exposure and this certainly includes electrical connections.

An intermittently bad fuel pump (or if present a bad fuel pressure regulator) has to be suspected. Unfortunately an intermittent fuel pressure problem is hard to catch. You might find an error code related to this in the proprietary error codes. To access these you will need a MINI specific diagnostics tool. A fuel pump replacement with no high confidence diagnostic is not something I would advise. But with a 2002 and you don't provide the miles a fuel pump can be the problem. (My 2002 Boxster's fuel pump quit at around 200K miles. I don't recall how old the car was when this happened. But I sold the car in Dec. 2017 with 317K miles on it.)

There can be other explanations. The ignition switch is suspect. When starting the engine wiggle the key, put pressure on the key in one direction or the other to see if you can affect the starting behavior, or effect other things like the dash lights.

I'm not a fan of throwing parts/repairs at the engine as you suggest you are going to do. I'm talking about a camshaft sensor o-ring (checking the connection is ok but be careful That wiring is old...) and cleaning the MAP sensor, throttle body and changing gaskets. (Sure if you pull these replace the gaskets).

The risk is with engine manifesting untoward behavior for as yet unknown reasons if you mess up this additional work then you can introduce a new problem which makes diagnosis even harder.

If you want to check sensor connections check the crankshaft position sensor connection. If the crankshaft sensor signal is not getting through the engine controller will not know when to trigger the fuel injectors and spark. The engine cranks but that's all. One possible clue the crankshaft sensor is acting up is if the tach during cranking doesn't "move". The connection could be ok but the sensor is acting up intermittently.

Also, speaking of sensors if you can monitor engine air intake temperature and coolant temperature.(I used an OBD2 scan tool for this on a number of cars.) A sensor that reads too high can result in engine start issues as the engine controller will not fuel the engine correctly for the actual (and colder) temperature. Actually found a coolant temperature sensor acting up that accounted for my hard to start from cold VW Golf TDi.

Don't know about the fuel. All I can offer is use what grade of octane MINI recommends. Buy from a busy station to ensure fresh fuel. If you can switch brands to see if the different fuel has any effect on the behavior.

 

Thanks for the answers!

Not yet. I need to find that Bentley manual picture someone posted (Daftlad if I recall correctly) with all the 22 or so ground straps and work my way through it.

Indeed, no CEL nor pending code. I check from time to time, the CEL has never come on since I own the car (apart from when I inverted spark plugs wires, but that's a mistake on my part that was quickly fixed).

I'll do that. Overall I don't think it's going to be an issue as the car is nearly 100% rust-free, but it doesn't hurt to check. Posts were already cleaned and there is a good connection there, and no electrical gremlins anywhere else.

Thanks, that's actually an answer to something I was wondering (if the fuel pump could fail that way). I recently installed INPA, I'll need to find out how to check proprietary error codes. If someone could point me the right way, that would be super helpful.

Oops, my bad, I forgot that. 130km miles, age of the fuel pump is unknown (probably original), and overall really rust-free. I don't have the detailed story of the car, but I would bet mostly short city trips from the brief speech I got from the previous owner.

You are right, I didn't think much about it, but the EWS do cut fuel supply. I'll keep that in mind. Do you know at what point the fuel supply is cut? Because on the video I posted, I did hear the fuel pump coming on when putting key to position 2. If the immobilizer is not detecting, would I hear the fuel pump?

I'm not either, but those o-rings/gaskets are cheap. I actually wanted to clean the throttle body anyway (as part of preventive maintenance) and the camshaft sensor o-ring is probably dead too (the crankshaft sensor o-ring certainly was). I'll keep that for later now that I have more leads to follow.

Thanks for the explanation. My knowledge of how camshaft/crankshaft sensor work and affect how the ECU act is very partial. I've actually changed the crankshaft sensor o-ring somewhat recently (it was leaking slightly) so the connection should be good.
How would you test this sensor signal? (apart from the tach not moving when cranking, I'll pay attention to that)

I forgot in the list, but I did check those. Though, like for the fuel pressure, I didn't check at the specific moment the car was acting up, so there could be an intermittent error here too. One more thing to keep an eye on. It might justify the purchase of an Ultragauge.


Thanks again for this detailed answer @RockC , it provided me new perspective and insights.

 

My experience with some cars is (was) I hear the fuel pump run when I turn the key to position 1 (which I think is engine run) and before I turn the key to start. Obviously the pump should run remain energized at least from whenever it is first energized to when the engine is shut off. You'll have to pay attention to the fuel pump, listen for it, and note its behavior when things work right, then note any difference when things don't work right.

With more modern cars -- no key push button start -- I don't think I hear any fuel pump.

130K miles is either too many miles or not enough miles. That is a fuel pump can quit at any time. It is great when a fuel pump fails outright. Not so great when it fails intermittently.

You replaced the crankshaft position sensor o-ring? If you tell me the starting problem appeared after that you need to check the connection. This falls under last thing touched first thing suspect. You also need to check the CPS is secure and has not moved closer to or further away from the flywheel and the toothed ring that generates the voltage signal

I've never check a CPS sensor signal. My understanding it is a voltage signal a "square wave" that I'm sure an oscilloscope would be able to display. I used a 'scope at work but I never owned one. They can be expensive. This includes automotive 'scopes. I can't recommend one.

Can't offer any help regarding how to access proprietary error codes. All my OBD2 work involved accessing OBD2 codes/data/telemetry.

 

I've been listening for it since I learned what the sound was. I can always hear it prime for a few seconds when putting the key to position 2, including the times when there were issues starting. Unfortunately I can't hear if it's on or not when cranking/with the engine running, because the pump sound is faint and drowned by the engine.
Maybe I'll run without backseat nor cover when testing things to hear it better.

I agree. I do not have a comprehensive history of the car. The pump could be newer, or it could be original. I fully expect to be on the side of the road waiting for a tow truck one day, it comes with the age of the car and all the delayed maintenance of previous owners. Hopefully when it fails I'll be able to kick it back to life to get home, or it'll happen close from home.

The problem started way before I replaced it, and the car has been trouble-free the few months after, as it was during summer and the issue didn't manifest between last winter and current winter (apart from being slightly hard to start, which it has always been).

Thanks for the help so far!

I've had new developments today: while driving, shortly after leaving home (engine was not up to operating temperature, though coolant was), there was two times where there was a rough idle, going below the 700/800rpm it normally does. The first time it went slightly below 500rpm and nearly stalled. The second time not as much, but still below normal idle rpm. Both time were recovered by giving it some gas pedal action, and it didn't do it just right after. That's the first time I see the car doing that (though I generally don't let it idle a lot because I'm mostly on highway).
Not sure what to make out of this yet, though since I could give it some gas, it feels more like a sensor, throttle body or ECM issue, not like a fuel issue.

Of note: it was raining, which is again a lead that something doesn't like humidity.

I'll be able to do some testing this week-end (and start checking ground nuts), what values should I watch for with OBD2 live data? I was thinking short-term fuel trim (is the ECM pulling or adding fuel?), intake air temp (does it seem to match real life?), intake manifold pressure (not sure what I should be seeing, but I could graph it against engine RPM afterwards to see if there is something strange when there is a rough idle, if it reproduces), timing advance (if something's strange there, it could mean ... maybe an issue with crankshaft/camshaft sensor?).

 

@Daftlad I found the thread where you posted all the grounds straps from the Bentley manual here https://www.northamericanmotoring.co...ml#post4662771 Thanks a lot for posting that.
Except you have put page ECL-23 twice, and it's missing ECL-22. Any chance you could post that missing page? Also if there is a R50 engine compartment picture, it would be great.

Thanks!
I'll be printing that and ticking them off.

 

^^^ I'm online now and have my Bentley manual (soft copy pdf) open to that page. See attachment...

 

Nice, thanks!

 

Small update:

• I checked all the easily accessible grounds in the engine bay (both sides of the engine ground straps, 2 on the picture, ground bus on each side, 1 and 17 on the picture, and the DME ground, 18 on the picture). None were loose and there is no sign of corrosion. It's too cold to really start disassembling things, both for me and because it increases the odds of breaking plastic clips, so "going though all the grounds" is relegated to an unknown future date
• I configured the Torque Pro app display to show intake temp, intake pressue, short term fuel trim, timing advance, coolant temp and revs. I looked at it during errands, but so far it has not reproduced any issue, so no update on that side
• gaskets will arrive tomorrow, but I'll hold off on those for now.

I'll keep the thread updated when there will be something worth posting about. Could be a week, could be several months, could be next year.