"Solved" - 1st start difficult, after its better

Hi everyone,
Car : citroen 1.2 puretech 82cv e85 biomotors (French part) 105'000kms - MAP car
Issue : Difficult 1st start - from 15° to more 25°
Battery ok, crank ok, no DTC, new fuel pump (contain fuel filter), new spark, new MAP, new oxygen sensor, exhaust good, no issue with air flow.... and still issues (i'm still a part changer when i dont find to test and learn). I don't have strong scan tool to deep dive into issue.
My question is because i don't understand behaviour of my LTFT. Before replacing Oxygen sensor (seems moving good from 0 to 0.9v) , LTFT was -7, so car was seems running in rich condition. With new Oxygen Sensor during 1day LTFT was 0. STFT runnning near +8. My first conclusion was bad old Oxygen sensor. Fuel trims move to opposite value..
1st start next morning in OL, 15°C. Car start and stall. Fuel pressure is good, but like not enough fuel.. 2d start OK. waiting CL, and in close loop, STFT around +20... but few minutes after, LTFT move to -3.2 and STFT around 0 ... wow.. I'm stuck right now. Was thinking coil issue and so positive STFT. But Why ECU correct to negative if STFT was so Positive ? Any idea for next troubleshooting ?
Thx for ur help,

 

Hi Ludo.

That could be different issues, but first you need to go in deep. For instance it would be great to know if ECM is receiving real information from different sensor involved in start stage (Map, ECT, IAT, CKP and CMP) if aplicable.

My first shoot would be fuel pressure. You can lose it through fuel pressure regulator, hoses (inside or outside of fuel tank), fuel pump and injectors). The best test i´ve found is by spraying starting fluid, gas o even two or three tablespoons of gasoline inside intake manifold downstream throttle body. If engine starts inmediately (even if it stalls afer few seconds), it´s a good indicator to me that starting issue is due to lack of fuel.

Another test is pump fuel system several times (i would say 6 or 7 times) by closing and opening ignition switch and hearing ECM turning pump On and Off every cycle (there are ECM´s programed to only turn fuel pump one time even if you closed ignition switch, so pay attention). That way you allow fuel system to rise pressure.

Let us know your report after these tests.

Hi,
Thx for your answer. I have already tried the second test. I heard fuel pump and run 4 or 5 times but was not helping. I will try second time with more times like 6/7. As fuel pump is new part, first part that I change and comes with fuel pressure directly attached inside and was same will last one, i was focusing on electrical/electronics issue. And as second start is always good I was really focusing on this but i will try to add fuel in intale maniflod to confirm lake of fuel, good idea. When i was young i added fuel directly into spark to test but your test is quickest and good way to start a clean troubleshoot. Your help is very appreciated.

About your question.

PCM is programed to make oxygen sensor bouncing between rich (0.9v) and lean (0.1v), by increasing or decreasing the amount of time the injector are open. Let say for instance at certain factory operating conditions, pulse width oscilation is between 3.3ms (rich) and 2.7ms (lean). In this case stoichiometric pulse is 3.0ms and no corretion is needed (STFT bouncing +/- 10% and LTFT 0%).

You must bear in mind that STFT and LTFT values changes through different conditions (load, temp, rpm), so maybe you are watching different memory positions values for different conditions. take care to compare the same condition. Also is important to know manufactures in general consider +/- 10% range is normal (even 15%).

In your case, you had LTFT -7% (oxygen sensor was telling PCM to run a little lean, but still in range). You didn´t mention STFT for old sensor, so I cannot conclude. Now you have LTFT 0% and STFT 8% (now oxygen sensor is telling PCM to run a little rich, but still in range). This could be a little difference in oxygen itself and I would say we cannot conclude yet because of LTFT.

On the other hand, +20% STFT when engine is cold, could be an indicator of engine normal worn out, but i wouldn´t worry about it, since after engine warm up, fuel trims become normal. That could be also the cause of hard start in the first try, but we need more tests.

 

I will try to give more information and hope i remember well and I don't know if I'm doing good tests in fact, so if you find clue it's great. I'm focusing at LTFT more than STFT because at the end for my point of view as i need to add both value and result near to 0, i think its good. I tried to find a clue with rich condition or lean for my starting issue.

Do ECU at 1st start in OL use LTFT? or it's only in CL? In other words, does LTFT -7 or -3.1 can reduce fuel at starting in OL?

- before changing oxygen sensor:
at idle, cold, hot: STFT move frequently around -3 +3 sometime more ; LTFT -7 (at idle and various rpm i didn't notice changes during drive test) -> spark was little brown and dry; Seems good for me.

- 1st day with new oxygen sensor:
at idle cold (but I changed oxygen sensor just after 2d start with old oxygen) to hot : LTFT 0, STFT moving like -3 + 8 will say (battery was removing because I have check many think with air flow. So probably reset LTFT ..)
drive test during more than 40 kms, all rpm; LTFT 0. STFT more in range of +3 +10, so like car need more fuel for me. You can tell me if I am wrong. So I was really thinking than LTFT will raise in future.

- 2d day (today) with new oxygen sensor:
at cold (iat 16°) idle after second start: LTFT 0; STFT move a lot but rise more +10 +15 till +20 and LTFT suddently move to -3.1 (surprised me) (iat 25°). Right now, at all rpm i have tested, LTFT is -3.1 and STFT moving frequently around -3 +6, at idle someting more but not often.

-> My Feeling during drive test is that with new oxygen, car react more quickly, more stable.

Tomorrow morning I will test with 7/8 fuel pump run before starting the car and I am quite sure that LTFT will be -3.1 and car will not start fist time but I will check that. Tomorrow morning i can't test with fuel in intake manifold, will try another day when im off, I want to follow your advice and troubleshoot.

My feeling but I can be completly wrong, coil are not working well at first start, ignition was bad, 2d start coil are charged (If they need to be charged... I suppose that and im not sure at all), ignition was better but not really perfect, so oxygen is not completly burning, so ECU add more fuel cause oxygen sensor detect oxygen and when coil are in good range (they have 100000kms), ignition was great and no more oxygen seen by ECU so rich condition trigguered by ECU and was reducing with LTFT ... without this I can't explain STFT +20 and in same time LTFT move from 0 to -3.1 and STFT around 0... like opposite LTFT reading...

Thank for reading my poor english

Long term fuel trim corrections will be made/present whenever the car starts. Short term fuel trims will be made once the car enters into closed loop. STFT corrections will be suspended each time the car gets out of closed loop.

ok guys/Ludo. I am going to try to catch your ideas:

My feeling but I can be completly wrong, coil are not working well at first start, ignition was bad, 2d start coil are charged (If they need to be charged... I suppose that and im not sure at all), ignition was better but not really perfect, so oxygen is not completly burning, so ECU add more fuel cause oxygen sensor detect oxygen and when coil are in good range (they have 100000kms), ignition was great and no more oxygen seen by ECU so rich condition trigguered by ECU and was reducing with LTFT ... without this I can't explain STFT +20 and in same time LTFT move from 0 to -3.1 and STFT around 0... like opposite LTFT reading...

About coils. They don´t need to be pre-charge or something before cranking. They are only charge few ms before they need to be fired (usually 3-5 ms). In fact, electrical issues are related to high resistance due to heat. So it doesn´t make to much sense you have a problem which dissapears with heat.

PCM is not able to watch O2 sensor, because it could be too cold (so it cannot respond to exhaust gases). PCM is programed to watch when starting RPM, ECT, IAT, TPS & MAP. Another story at CL, but your idea suggest me you are thinking about OL (starting issue)

I will try to give more information and hope i remember well and I don't know if I'm doing good tests in fact, so if you find clue it's great. I'm focusing at LTFT more than STFT because at the end for my point of view as i need to add both value and result near to 0, i think its good. I tried to find a clue with rich condition or lean for my starting issue.

As juergen.scholl said LTFT is used everytime and STFT on CL, so for starting issue stay focused on LTFT. But i don´t know the exact strategy used by manufacturer to detect whether engine was succesfully started. Neither maximum fuel trim percentage (which seems to be 20%). Let´s suppose PCM is detecting a failed starting and is adding fuel, but it cannot add more because is at it´s maximum. Then it seems is telling you a lack of fuel.

Due to PCM is not able to watch O2 sensor, It is trying to add fuel by letting injectors opened a little longer, but if we have a lower pressure, it is actually adding fuel and air bubbles until pressure rises.

I suggest you a better trick for starting troubleshoot. Turn fuel pump ON constantly for ten seconds before cranking, by jumping to ground fuel pump control line from PCM or bypassing fuel pump relay, so that you can likely push air bubbles through pressure regulator (only in returnless fuel system type)
 

Ok that make sense to me.

When i replaced fuel pump that integrate new filter and presure regulator, i notice only one pipe. So i ask myself for return pipe and it seems on this car there is no return. One other pipe on tank is connecting to a canister and a soleoid managed by ecu if im correct connected to intake manifold for gasoline vapor.

I not seen any purge till injector (indirect) rail but i will check again. If there is no purge, how air can be evacuated ? By injector ? I am thinking step after. Is your test is a procedure to evacuate bubble or to point issue with lake of fuel pression during long stop?

Will test asap and come back to u.

Thx for your help

Hi guys,
I did the test. Fuel pump run more than 10sec before starting and same issue happened. I let the starter for a while but the engine won't start. Second try, engine start direct without any issues. LTFT is -2 since yesterday. So in running condition all seems good.
If you have other tests in mind... i will try gasoline directly inside intake manifold and probably will buy coil tester to see if electricity is here during first start not due to coil issue but ECU wrong info..
I don't know how to check camshaft sensor and flywheel or pmh sensor. If they can be a source of issue for the 1st start.
Thx again.

Hi ludo. Wow really? I didn't expect that.

Anyway i'm still thinking is lack of fuel. Is curious always starts in second try (which means to me fuel priming)

Nothing changes mechanicaly that quick (no thermal expantion or something). Neither electrical. And after start everything look pretty normal.

If you want to check spark, you can let a spark plug wire in the air, very close to ground (some metallic part on engine which is connected to ground) so that you can see spark jumping on cranking.

You can also see injectors shooting at the air while cranking (just take of unplugging coil control wires and letting everything dry before trying to start again)

 

I'm lazy with age . I use to do that when was young.. now I'm waiting before disassemble injector. If dont find will do it, and ultra sound perhaps. And for spark.. quitte same, i'm alone currently, need to find a way to see spark... was saying lazy.. battery seems good 12.6v. Tomorrow morning i put booster just to be sure of that too (dont know this ecu. Key is good, i checked immobilier, seems good too) . at the end i'm sure it a stupid thing.. fuel was my first though , just ltft -2 with good oxygen sensor bring me away... but you still focus on fuel so will put diretcly into intake manifold double check that point (full pump test for before injector ; into intake manifold for after injector). Thx sir !

I know you can live with a second try start, but i'm guessing that like me, you enjoy to solve these kind of details. In my case, since long time ago, i've been developing and refining my own ecu, so i think i learned a lot at manufacturers level.

That remains me there is another detail we need to take care. It is less probable but who knows. You could be dealing with IAC out of step which in the fisrt try is very closed, restricting too much the air. After start somehow it reset to the right step.

Have you try to cranking engine with throttle on the floor on the first try?

 

I dit it with throttle on floor, my guess was to push fuel... but same and was difficult to start second time and third time i remember, and after i read that can be a mode on ecu but i don't remember, like a test and normal to avoid starting. You should know better than me and probably see that on SD video. Im waiting to received SD book i bought yesterday.. just to learn more.
IAC is with Map and new.
One thing, i can test injector with my scan tool and when i test it. Difficult to start after. Like to much fuel. Can indicate lake of injector ?

After injector test, you could have a flooded engine (wet sparks or very saturated mixture).

About IAC, i meant not a bad one, but not calibrated (pcm memory 'thinking' is at certain step, but it isn't.

 

Is spark timing controlled manually on that engune?

if you have an injector leaking and the cylinder is getting too much fuel, it would be harder to start the more you try
 

I said IAC with Map but its IAT with MAP. Don't know where iac is.. on TPS?

Spark timing could u precise?

My scan tool can relearn or adapt Throttle but Nevers test that.. will try.. if it is that... i send u flowers

I got it on fact. Is controled buy ecu. I cannot move by hand ignition advance. On scanner i remember -10° when i start...

IAC (idle air control) is on the throttle body. Is a bypass air duct which is not controlled by you (throttle pedal), but by pcm.

Ignition advance -10? It depends on load, rpm and IAT. Theory says if i remember well, that should be positive high, but i can be wrong (i need to review)

Will check value when i start i can make mistake.
On manufacturer doc don't see IAC but remember air pipe on air shutter.
Will do learning adaptation on this TPS to be sure that amount of air is correct for starting. If 1st start wrong is it possible that ECU/TPS close much to have rich mixture for 2d start ?