MS code error?
Turbo Mini Forums

I am using the fast idle output port to regulate my electric fan. I have set the idle control algorithm to none. In the extended menu I have set the PM2 port to switch to 1 when coolant temp > 88C.
Everything works fine and the engine warms up nicely with correct AFR’s. When the fan kicks in for the first time everything is still normal but after it switches of all cylinders are running lean until the fan kicks in again and this keeps repeating. It looks as if something in the code is triggered on the falling edge of PM2.
Does anybody know how to fix this? I have added my msq file.

Regards Will van Gemert

There is no msq attached. You should also post a log of this. And posting on the msextra forum may be the best way to get more responses.

Jean

http://www.jbperf.com/

Sorry forgot to attach the MSQ file. here it is. I wil make a log also later today and shall also post this on the MSextra forum.
Thanks

Attachments:

10.36kb

My guess is that it is something to do with the fan dragging the volts down and affecting the injectors.

Particularly with the low pulse widths that you are using.

I would try increasing the "Battery Voltage Correction"

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Thought of that also but did not check it the strange thing is that it only first occurs after the fan switches of again. Also the higher voltage with the fan of should make it run richer instead of leaner. Will do some measurments and post a log.

Your VE table is a little hilly in the idle region.



I've drawn a line around the area. You have some big changes in fueling over some small steps. This will cause the changes in AFR due to the alternator load alone.

I would suggest making all those cells the same value.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

I agree with Paul.

Fan load on the alternator affects idle speed quite a bit.

This is what it says in the UK MOT test manual for testing emmissions,

"Many modern vehicles are fitted with electric engine cooling fans which can cut in during an emission test. The extra load on the alternator reduces the idle speed which causes the engine management system to react. This gives rise to highly variable readings.
If this happens during a test, wait until the fan switches off and the readings stabilise before continuing."

And that is meant for OEM systems where the designer will have gone to a lot of trouble to try and avoid it.

Do you have closed loop control on your IACV ??? That might help.

But as Paul says, you have some significant steps (steep slopes once the CPU processes it) in that VE table especially in the 55-60 fuel load area. What fuel load are you running at idle ???

Schrödinger's cat - so which one am I ???

Thanks for your respons.
Keep in mind that I am runing a 286 cam. at idle (1500rpm) my map is around 60.
Just did some more measurements and saw a voltage drop at the injectors of around 0.45V when the fan comes in. My previous note that problems start after the fan switches of the first time is not correct it starts wen it switches on (more logical so probablky not a ms problem). I also found out that my voltage measurement in the MS is not accurate it is 2V to high I will look into this right away. You might be both right that its just the bumby VE table together with the alternator load.
I only had the engine running at idle speeds up to now so there will be a lot to improve.
I will take a good look at the VE table around idle and flatten things out. Is it save to start ve analyze live at this point or should I first have a more or less stable setup?
I added a log of the 'problem' it starts around second 550 when coolant temp goes above 89

Attachments:

193.54kb

16.31kb

On that log you have a very high idle speed. As you say, about 1500RPM.
It drops as the fan cuts in (expected) but recovers quickly.

Personally I would try to get idle lower (even with a lumpy cam).

But at 1500RPM and 60 fuel load, your VE table isn't too bad (it would be a lot worse if your idle was 1000 or less) so battery voltage compensation might be more likely the cause.

Battery voltage measurement is performed by R3 and R6, a potential divider that drops actual voltage down to a level the CPU can read so check their actual resistance values with the CPU unplugged. But the compensation is based on the change in voltage rather than absolute values so being 2V out on absolute voltage won't make much difference to how MS2 measures the 0.45V change.

Apart from experimenting with different battery voltage compensation values, from your msq it looks like you are using low z injectors. You may have to alter the PWM settings or injector openning time value to get an accurate actual openning time at different battery voltages.

Finally - and this may be the most important - your msq shows you have the default settings in "accelleration wizard" of MAP vs TPS of 100% and the default values of mS per %/sec (starting at 15mS added) which are all far too high for siamese application as our pulse widths are nowhere near 15mS in the first place, let alone needing 15mS adding.

Read this thread, especially the second page. Very small voltage changes affecting the TPS signal to the CPU were causing massive changes to the fuelling.

http://www.turbominis.co.uk/forums/index.p...tid=378757&fr=0

Schrödinger's cat - so which one am I ???

Right, I'll go back to my original theory.

Your voltage correction is actually too high. As the voltage goes up when the fan stops, the pulsewidths drop and it goes lean.

You have two problems that I can see: Your voltage is far too high at 15-16v and your pulsewidths are too low at 1.0 to 1.2 mSec.

Sort the voltage problem first. then see what you can do with the pulsewidths. i would actually reduce the Battery Voltage Compensation.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Resistor R6 was 12 instead of 10K just changed and and wil give it a new try.
Its correct that I still have the default settings in the acceleratiopn wizzard. as I only was trying to get a good idle at this moment. I will read thru the thread on this subject before trying again. I indeed have low z injectors with the settings as recommended in the manual. What should I look for when altering the PWM settings?

Too much battery voltage compensation combined with a wrong opening time and large injectors will have a massive fueling impact at idle.

Jean

http://www.jbperf.com/

Schrödinger's cat - so which one am I ???

I don't think tht the problem is AE in this case. if you look at the log, at no time do the Accel or Deccel indicators light up.

The TPS signal only changes from 7 to 6% throughout.

My guess is that if you zero out the Battery Voltage Compensation, then the idle will steady.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Decreasing the voltage correction to half the original value seem to have cured the problem decreasing it further resulted in running a bit richer when the fan turned on. So I am a little step further now. Pulse width at idle is now around 1 msec I to aprox what values should I set the acceleration wizard?

Are you still running without the vacuum referenced fuel presure regulator?

This will mean a high differential presure across the injector at idle, meaning lower pulse widths to get the AFRs right.

Here are my AE settings for the 998Ti.

Edited by Paul S on 3rd Apr, 2011.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Yes I am still running without vacuum control on the fuel presure regulator. I wil by one next week. Just changed the injector opening time to 0.75msec this improved the controlability at idle. Wil try to change it a bit more to see what happens is has a big impact on the ve table values. I am running low z injectors from delphi do you perhaps now at what the correct parameters for these are?