Nearly stalling when starting from cold

[Tony C]

Not really a problem, but something i have been working on for many months now,
When starting from cold the engine fires ok BUT then the revs drop to near stalling then pick up again,
does not matter what temp the engine is when first started it does this, it is ok when been running, if you switch off after starting from cold and start straight away its also fine,

any ideas??

[rob@vems.co.uk]

What sort of pulsewidths do you see on the main screen?
You're probably looking to make a change in the Priming/Cranking/Afterstart settings, once the engine is in afterstart you are either going too rich or too lean.  The Pulsewidth gauge and/or datalogs will tell you by how much.

Rob

[Tony C]

i will do a datalog and check the PW gauge in the morning when i do another cold start mate,
ta very much.

[cliffb75]

Based on the fact that you've said a re-start is fine, then it will be lean in the afterstart. The reason it re-starts better is that the additional fuel from the first start is left hung-up in the ports, which then gets added to to the normal afterstart fuel which is applied during the re-start.

For what its worth, the way I think about fuelling is this. Crank fuelling sets up the initial wall film, and supplies enough fuel for the first round of combustion. Afterstart fuelling then maintains this wall film, and adds fuel for poor vapourisation, which quickly reduces as the combustion chamber warms up. However, at this point the ports are still cold as the coolant temp has not increased. The warmup fuelling deals with this aspect, and supplies the additional fuel required to maintain the long term wall film requirement.

If you plot a curve of the combined effect of the terms against time (or combustions) you get a dogleg shape, where the afterstart fuelling decreases very rapidly (since the chamber comes up to temp quickly) then a dogleg and the warmup fuelling decreases more slowly

Try experimenting with more afterstart fuel but over less combustions. Also preferably log the lambda during start with a wideband gauge, along with the pulse width, engine speed and MAP to see exactly where the problem lies.

Good luck.

[Tony C]

Please find below the data file for my cold start along with recent .msq file, i changed the after start enrichment to 60 and it didnt die down to a near stall, not sure if that is a fluke or not, but take a look at the two files and see what you think please.

http://www.capriracing.co.uk/vems/vemsdata.zip

[rob@vems.co.uk]

Using the Datalog viewer we get a good idea of what's going on, I've drawn the ASE start and stops points.

It looks as if the 13ms pulse of fuel spins the engine up nicely, then fades out as we'd expect it to.
You may get less of a peak in RPM if you lessen the fuel pulse, but then its working so you may not want to spend too much time messing around chasing perfection for an event that happens at best 7 times a week!

[Tony C]

be gentle im still learning mate   
does every thing look ok now that i have changed the ASE % to 60? was at 40.
or is there room for more improvement,

[rob@vems.co.uk]

From my point of view thats looking pretty good, it pulls the engine up to its idling speed and everything is fine and dandy.  I'd wait for Cliff's verdict as he knows far more about the fuelling requirements and what its supposed to look like

Rob

[cliffb75]

Hmmm, OK here we go then.

The run up is OK, but not perfect. There is a 'stumble' (i.e. the engine speed drops momentarily) which could be caused wither by too lean or too rich fuelling. For the moment, it looks like the crank fuelling is OK for your needs, but you could experiment with it a bit to see if you can get the run up perfect.

The flare is not as high as I would have expected. Not sure what your target idle speed is, but I'm assuming this is a start and idle, in which case I would normally aim for a flare peak of around 100 to 200rpm over target idle speed, which would seem to be about 1100rpm, so you should flare to about 1200. You can do this aither by increasing air (need to balance the cold requirement against the hot requirement bearing in mind the range of the bypass valve) or sometimes spark advance, and possibly adding a bit more fuel if you are currently on the lean side.

On the spark note, is that actual spark timing thats logged i.e. degrees BTDC? If so you seem to be running very retarded. I would expect to see between 10 and 15 deg BTDC at idle, and much more advanced then that when the throttle is pressed. However, that is not fixed in stome, and epends heavily on the idle air settings.

On to the bit we had concerns with, the area immediely after start. I'm a bit surprised by how active the pulsewidth is - its up and down all over the place! But looking closer thats because its following MAP. Whats going on here? Throttle press? Doesn't seem like idle speed bypass activity as that seems to be flat out until 510s. However, with the MAP unstable it's difficult to draw conclusions about the fuelling.

Lambda trace - not sure whats going on there. Why does it suddenly jump up from 0.6 to 0.8 at 500s?

So in summary, it would be good to see a log where lambda is working right from the beginning, the load is stable (should be atmospheric as start of cranking and drop quickly to the pressure for idle). However, from the log we have it looks like crank fuelling is OK, maybe a touch lean, afterstart fuelling has the right decay time, but not sure about the quantity, and warmup fuel is about right. However the spark and air side of things seems a bit screwy, and everything changes at around 500s for some reason, where lambda starts reading, IAC valve ramps down (to control the increase in speed that occurred at 492s) and spark strts to be advanced of 0deg again.

[Tony C]

I aint a clue whats going on thus why i asked
im not using any ignition control yet mate,
still running on a dizzy,

[cliffb75]

Ahh, maybe I've presumed too much?

So, tell me:

Was this a simple start and idle (i.e. you didn't touch the throttle at any point)
Are you using an idle air valve (or similar) or just a fixed throttle stop?
Are you using closed loop wideband lambda control?
What is the normal idle speed cold and warm?
Anything else you think is worth mentioning?

I'd certainly advise going to 'proper' spark control as soon as possible. You'll find the gains in driveability and probably performance are well worth it 

[Tony C]

well went round to the car tonight, and it was the same as the other day, starts up runs for few seconds at around 1100-1000 rpm then dies down to around 300 then picks up again and then idles,

have changed a few things in the afterstart table, but im clutching at straws
The settings below is what i have changed tonight to try after work tomorrow.

changed primepulse[ms] from 2.0 to 3.0
changed Hot start temp scaling [%] from 9.0 to 14.5
changed crank cold PW from 7.5 to 9.0
changed Crank warm PW from 2.0 to 3.5
changed Cranking threshold from 300 to 400

Ahh, maybe I've presumed too much?

So, tell me:

Was this a simple start and idle (i.e. you didn't touch the throttle at any point)
yes, the only time i touched the throttle was at the end when i reved it to about 1900rpm

Are you using an idle air valve (or similar) or just a fixed throttle stop?
Yes using the rover idle control valve

Are you using closed loop wideband lambda control?
Yes closed loop with wideband lambda control

What is the normal idle speed cold and warm?
1100 cold 950-1000 when warm

Anything else you think is worth mentioning?
not off hand 

I'd certainly advise going to 'proper' spark control as soon as possible. You'll find the gains in driveability and probably performance are well worth it 
this will be once all the fueling issues are sorted

[cliffb75]

Make sure you send us the log to look at

For the purposes of this testing, it would be useful to see lambda from the start, so make sure you don't have any delay set, and turn the ignition on and leave it for about 30 seconds for the lambda sensor to start reading so that we get a log of the whole start. Once you've got the starts sorted you can put the delay back in to protect the sensor.

Rob, am I looking at the right things on the log? Iac DC is the idle air valve duty cycle right? Have a look at the coldstart07 log Tony sent the other day. What do you think is happening at 492 to 510 secs? MAP doesn't seem to tally with the IacDC to me? Whats your experience with the Rover valve - could there be a problem with it?

[dnb]

I had this with the TVR - same parts.  The rover valve is a stepper and it can be a pain to get working...  It has a very large range so it's easy for it to mess around without obvious sign.  It sounds like the valve is about right when cranking, but is too closed during after start so is starving the engine of air.  If you press the throttle a bit (only 2% or so - enough to not be in idle mode in the ECU) does the problem persist?   

[Tony C]

i will try next time i can get round he garage Dave,
if i open up the stepper, wont i have a problem of idling to high while cold??