Stalling, HELP!!!, Im Not so Worried

[Sprocket]

Fueling has been mapped, however i have had a problem with stalling as i dip the clutch when aproching a junction, lol, its a little inconvenient The second problem is starting it again. it takes a couple of trys most time with a little throttle so i suspect the crank PW is not quite right.

What may not be helping is the fact that the idle PW is 1.4ms and the TB is giving TPS of 2%. TPS drops to 1%

I ill change out the injector for one that is 10% less and scale up the VE table by 10% and alter the crank PW, though im a little unsure what im looking at. So here are some logs i did yesterday, the first is the cold start. It started second try and was real nice other than the fast idle rising, i altered the stepper refference table and it sorted the problem, everything was fine from that point until it was warm and stalled at a junction. Starting after it stalled takes a couple of goes and with a bit of throttle.

Im after any advice if some one can help.



cold start log before mapping



Stalls after mapping










Settings

[rob@vems.co.uk]

It would be nice to get hold of the datalog files so that I can see the values where things go wrong.  It would be interesting to see what the ignition thinks it should be doing as well as the MAP value.

While we're here...

EGO settings:
Set Lean limit and rich limit to 20
Step size of 5
Engine cycles before changing correction to 3

Warm-up time to 60
Max TPS 95
min RPM 600 is okay
max RPM 7000
Min temperature 17

Idle settings:
I've never seen TPS for idle settings that high ( before, try taking it down to 2 or 3
Disable the ignition based idle control for the time being.

Looking at the PID settings...  That I value looks a little high and consiquently we see in the first shot what looks to be Integral wind-up.  If you are happy with your P value then try setting your I and D to 0 (as this turns them off) and then start tuning using this:
http://www.expertune.com/tutor.html
As a pretty good guide.

[dnb]

My old engine idled with a pulsewidth of 1.6ms, so I don't think 1.4 will be the problem.  I agree with Rob that the issue is with the IAC PID settings.

Rob - Isn't your suggested EGO rich limit a little high?  Mine's 10.  (I initially aim the base map to provide 102% of the required fuel and have asymmetric lean and rich limits)

[rob@vems.co.uk]

My standard method is to start high and tune the settings from there, I tend to use 10% for both, but this is a single point injector so I'm suggesting that the EGO net is cast wide!

Rob

[dnb]

I see your point!!

[Sprocket]

Here are the two logs, one from the morning with cold start, and the second from the afternoon on the way home from the dyno. The stalling is not as a result of the mapping, it has been there all the time. It can get interesting with heel and toe braking to keep the revs up as i dip the clutch. Once the revs drop gently it will settle at idle nicely.

Also any recomendations on the crank pulse widths, how much more should I go, i dont think its far off at cold, but it is further away when warm as it wont start without a little throttle.

Im going to have a play in a bit, thanks for the pointers, kep them coming. Logs at the bottom of the post.

One other thing I need work out whats going on is the coolant temp when driving drops to 80, yet the stat is an 88. on the dyno it never fell much below 94. The coolant sensor is in the bottom of the inlet manifold, im just wondering if the coolant flow is wrong, IE the sensor is in the leaving water rather than entering and the chilling effect of the charge is causing it, its an odd one that i did not see the last time.

http://freespace.virgin.net/colin.milne/Datalog.zip

[Sprocket]

I spent the last 3 hours driving round and tweeking things, trying everything that i could think would be causing the stalling, no joy

Idle quality has been improved as has the EGOc, so the only real issue now is the stalling, I cant work out whats going on, Logs again at the bottom

This is with a smaller injector and the smaller TB. The smaller TB requiring more TPS for idle, and hopefully not closing off entirely. I set the throttle stop with the stepper at zero, so that the TPS read 1% to make sure that if the stepper ever sees zero that the trottle is at least that little bit open. Idle threshold is now at 6% as idle requires 4 to 5%. with follow TPS enabled if the idle threshold is not set higher when the throttle is released, the stepper freezes and will not return the speed to idle, staying high..

It is now clear to me that this is not a problem with the throttle closing off. Looking at the logs from today, TPS stays around 4%. PW increases as does the ignition advance, yet Lambda still registers very lean. RPM falls to around idle, MAP increases to what you would expect at idle, but nothing happens and the engine stalls.

It only happens after over run from mid or higher rpm, if you bring it dowm gently its fine.

I made sure that the Main relay control was set for ON all the time and over run fuel cut was dissabled by setting them high. I also removed the low VE entries along the bottom of the table  and set the low KPa/RPM igntion teble points to 30 to try and give it a kick when the revs drop.

http://freespace.virgin.net/colin.milne/Datalog2.zip

[rob@vems.co.uk]

Unfortunately I've tried to download those new datalogs but winzip says the files are corrupted

Lets start by turning everything like overrun fuel cut and fuel cuts above and below KPa off (25500 for overrun values, 510 for cut above and 1kpa for cut below).

Is MegaTune reporting any trigger errors?

What version of firmware are you using?

Rob

[Sprocket]

Bugger I'll sort the logs even if I have to email them to those who want them, I just need some one else to look at them.

Yes trigger errors are reported, but this is only after it has stalled and i restarted by jumping it. I was generating the condition, rather than letting it happen at a junction. Driving allong with the revs at about 3500 i would lift off and go into over run, then just dip the clutch, the engine would stall and i would then slowly bring the clutch back in to jump it, this is when trigger errors are reported.

I'll check the air gap on the sensor but other than that its good, with the teeth passing the center of the sensor

I already turned off all the over run fuel cuts and removed the low VE entries at the very bottom of the table at the higher RPM. I had also set the fuel cut above to 255 and below to 0, this may be a problem??

I also made sure that the main relay supplying power to the injector and ignition is on all the time, though the Lamda sensor is fed off this relay and that still reads during the condition so I dont think its this.

The engine compretion is high and the inertia starter is finding it difficult to turn the engine over, it wont if you dip the clutch, or have any electrical loads on, its not the clutch dragging, its just that starter. I wish I had bought the pre engaged ring gear now, a whole lot less hassle

The car is going for its MOT tomorrow and then on friday Im off tho Long Marston, for the British Mini Showdown. Not so much bothered about this week end but we are intending on going on holiday it the Mini so would like to get it sorted soon as so it can give me some time If I need to make the decision and put the standard stuff back on, though i dont really want to.

Got stuff to do to get the car ready for the MOT then I'll have another play.

As ever, all the help is apreciated

[cliffb75]

Is this the engine with the single point injection? I assume that its well upstream of the inlet valves?

What I'm thinking is that it could be that you are simply removing all the wall film on the overrun and then not replacing it in time as you get to idle. With such a long distance from injector to valve, there is a lot of wall film to maintain.

Try putting some big nmbers in the low MAP VE boxes and see what happens - i.e. log of lambda again. Also try and do some overuns at different MAP values (i.e. dfferent pedal angles ) and see if there is a point at which it suddenly becomes a problem.

Also, does the car run OK with no load. i.e. with the car in neutral, can you hold the revs steady at any speed you want or does it go lean, start hunting etc?

[Sprocket]

I have re packed and uploaded the last zip file, I have checked it and it does work for me.

http://freespace.virgin.net/colin.milne/Datalog2.zip

I forgot to mention that im running 1.0.71 firmware.



Cliff, it is indeed the single point TBi, i have a feeling you are dead right, i just couldnt work out how to over come it, I'll up the  bottom line on the VE table at the lower RPM and see whats going on in the logs.

When I monitored the OEM fueling i noticed the narrow band lambda on over run went to 200mv then just before fuel resume where it rose to 205mv. Maybe just a little fuel to wet the manifold?

It will have to wait till tomorrow after the MOT, i spent this evening fixing those little things that are just silly fail items, and the steering column bushes were a bugger to change having to get the grinder on the shear bolts

Oh yeh and im going to see how i get on with the cat test, i have another option if it wont pass but we shall wait and see

Onwards and upwards

[Sprocket]

Things have improved. I swapped about the low kpa of the VE table so that the bottom line is 10kpa, the second line is 20kpa and the third line is the idle kpa. second and third lines with idle VE at the low rpm end and then the bottom line with a RICH VE up to the fuel cut rpm. I upped the fuel cut rpm to 2200 and the resume to 2000 as cruse speed at 30mph was right on the fuel cut.( I put fuel cut back in as i need to fix it with the fuel cut)

In the Spark table, the very bottom left corner set to 30 with neighbouring sites the same, then down to 20 with the idle advance at 15.

It only stalled at one point but has not otherwise. I may be onto a winner

I need to do a little more testing o see how it reacts

[rob@vems.co.uk]

Nice work... What is the lowest kpa your engine sees?

[Sprocket]

Rob the lowest kpa experienced is below 10 and above 5.

I have spent an hour looking through todays logs at over run points and it is clear now where the problem is. On long duration over run situations the stepper appears to close the throttle to 2%, dipping the clutch allows a sudden drop in rpm and stall. The increase in ignition and fuel most times alows it to recover, but there is always a dip in rpm below idle set point. long over run events cannot recover as stepper closes throttle to almost closed

Three possible ways round this i can see. Speed up the stepper duty cycle to react faster or, set the stepper to zero and then adjust the throttle to 3%, to allow for increas and decrease of steps without hitting zero, or adjust the throttle by pass screw.


The first i shall try, and then failing that onto the second, but i think the third is not such a good idea as the thottle will close off entirely.

Comparing whats happening with the stepper control on VEMS with the Rover MEMS, it may apear that VEMS is missing something to hold off idle control on over run with the throttle closed to its natural idle position, IE, current situation - over run event, foot off throttle, MAP at below10kpa, TPS at 4, idle threshold at 5, rpm high at say 1500, idle control kicks in to try and bring the speed down, closes stepper/ throttle to minimum, backs off the advance??, dip clutch, rpm overshoots and engine stalls.

Anyone got any ideas

[cliffb75]

It this because the ECU goes into idle control mode whenever you back off? Is there an idle entry speed that can be calibrated? (I couldn't find it)

Glad it looks like you found the problem