[Gmecm] Strange effect while trying to tune my GM-EFIed beast

[David Allen]

  Tomas, since the effect is so profound, probably the error is something obvious if you just happen to focus on it.  For instance, I had a car that needed a little more canking fuel to get it started when cold.  I intended to modify the cold-engine cranking fuel pulse-width.  The car started perfect cold and hot, but there was a temperature range when it would NOT start.  My mother discovered this (it was her car) and I had to drive into town to get it started.  Using clear-flood mode got it going in a cloud of black smoke!  
  After farting around with it, I noticed 2 cells in the table (which I had not intended to modify) were off by a factor of 10!  I must have "bumped the mouse" and accidentally edited these cells.  Correcting this fixed it.
  Look at your tables in "graph view" and see if there are any strange peaks, valleys or cutoffs. Anything major wrong should show up.
  Do you save each old calibration?   So you can therefore "go back" and undo the changes?  I've got about 7 calibrations where I tweaked on different aspects of my turbo motor. This has saved me hours of work on more than one occation!
  Which model Bimmer have you got?  (used to work at an independant BMW / Mercedes shop)
Later,
David

----- Original Message ----- 
From: "Tomas J. Sokorai Sch." <tsokorai at minimania.org>
To: <gmecm at diy-efi.org>
Sent: Friday, August 25, 2006 5:56 PM
Subject: Re: [Gmecm] Strange effect while trying to tune my GM-EFIed beast


> On Friday 25 August 2006 18:17, David Allen wrote:
>>   What was the last set of modifications you mention?
> 
> Uhh ... hard to remember... I'm a bit "non-methodic" when doing table & 
> consts. changes: sometimes I get into a "change everything" frenzy  :)
> The main were a bit more of fuel (VE lower and upper tables) and a bit more of 
> advance. I also played a bit with the DFCO parameters, but I only increased 
> the RPM threshold, because before that I was getting engine stalls on 
> downshifts on long decels (manual tranny).
> The strange thing is that before my last modification session, the engine ran 
> at any RPM, but very anemic at high RPMs.
> 
>>   Have you scoped the "reference" signal from the DIS module to the ECM? Is
>> it faltering away over 2000 RPM? What kind of crank trigger are you using? 
> 
> I'm using a "frankenstein-sensor" ;) it is a VR sensor from the TDC diagnostic 
> setup BMW put on these old k-jetronic engines, but the signal was too low, so 
> I modified the sensor a bit stacking three NdBFe magnets on the back to get a 
> higher signal level.
> 
>> Could it be gaining resistance as it warms up?  If it's a VR sensor
>> (magnetic pin with a coil winding) then the signal is usually dirtier at
>> high-revs. Anything interfering with it would be worse when hot (more coil
>> resistance) and at high-revs (dirtier signal). Have you verified the
>> polarity of the crank trigger coil with relation to the DIS module? To
> 
> I checked the polarity very early on my adaption phase: with the switched 
> polarity, it barely runs at any RPM :)
> 
>> isolate tuning versus hardware causes, you could replace the temp sensor
>> with a fixed resistor.  This resistor would give a low "temperature"
>> reading which corresponds to the actual engine temperature reange where the
>> engine performs properly.  If the engine runs right when fully warned up so
>> long as the ECM is being told it is not warmed up, then it is a tuning or
>> code issue.  If  the problem persists, then it is a hardware issue and not
>> a tuning issue. Hope this helps!
> 
> Thanks!.
> I'm getting the feeling it is a too big advance, but the strange thing is that 
> instead of knock or backfire, I get a complete missfire and even an exhaust 
> bang! 
> I don't think it is the crank sensor because before my last changes it was 
> running more or less smooth at any RPM. 
> Now I need to narrow down exactly *what* in my modification-fest gave me that 
> result.
> 
> Well... it was fun trying to drive the car across the whole city with a 2000 
> RPM limit ;) 
> 
> -- 
> Tomas J. Sokorai Sch.
> _______________________________________________
> Gmecm mailing list
> Gmecm at diy-efi.org
> Subscribe: http://lists.diy-efi.org/mailman/listinfo/gmecm
> Main WWW page: http://www.diy-efi.org/gmecm
>