Re: Turbo MAP & MAT interaction

["nacelp" <nacelp@bright.net>]

I'll take you word about all you say, but, there are some critical areas
that need addressed in my opinion.  The stock senders are so slow in
resonding to temp changes, that while the ecm might be making changes based
on what the sensors tell the ecm are, they change so much from a heat sink
at idle staging to 4-5 PSI boost (Brakes Locked, car still), to 20 PSI at
5,500 rpm, and this can happen in 5-10 sec, the ecm can't forcast what the
operating conditions will be, so it has to have at least quick responding
sensors, again in my opinion.
   Personally, I want full time baro, and instant CTS, and MAT corrections.
   Why they have such good MAP sensors, and then drop the ball on the temp
stuff makes no real sense to me, other then it's not an EPA requirement, so
they got it close, and it feels good seat of the pants and they just left it
there.
   I just dug it out but I'll be ordering a 12160855  transmisssion temp
sensor to see how it compares to a normal MAT/IAT/CTS sender.  This one is a
pipe thread,  there is one other one I got in my notes to try.
   Other thing is that you can fail safe the calibration much better.  ie,
sudden spike in CTS, and kill the timing, and you reduce the power and the
damage from say a blown headgasket.
Grumpy



> > I have been monitoring the ALDL data on my car for over a week now, and
> > have noticed that the Manifold Air Pressure (MAP) sensor reading is
> > varying from about 1 bar with the engine off, to 0.25 bar at idle &
> > overrun to about 1.7 bar at full whack. This seems correct from the
> > documentation I have. However the Manifold Air Temp (MAT) sensor output
> > hardly changes, only rising by a few degrees in a few minutes, and that
> > appears to be more to do with the engine warming up than anything else.
> > There is no short term, MAP related temperature change.
>
> You are inquiring about monitoring MAT temp changes as far as the direct
> effects of air temp increase related to boost pressure increase? OK.  But
> your MAT does elevate in temps as the engine warms to normal temps right?
> Assuming these then yes (and you have an intercooler I assume?), you would
> expect the temp to rise accordingly as boost pressure was raised but not
as
> much as we can expect with these sensors and the layout of your system.
> This is not uncommon and efforts (I say efforts, not absolute resolution)
> are addressed in the ECM by code working with tables to give a more
> meaningful and usable strategy to provide good values for the SD equation.
> In the chip there are calc variables such as tables that provide simulated
> linearized MAT that take the raw A/D MAT reading using a 1K ohm pull-up
> resistor to provide more true values for those MAT sensors.  Then there
are
> also offsets that compare the CTS to the MAT and scale this out to
> represent a more true reading.  Then you can consider that a fully warmed
> engine will already have hot intake runners warming the incoming air, and
> now these calcs make more sense.  There is some more massaging but I think
I
> explained it enough here for it to not look so off to you.   Now, I would
> hesitate to relocate the MAT to try to avoid the heat soaked effected MAT
> temp as you would then need to completely rework the above mentioned calcs
> and, would no longer have proper readings of the air temps as they ARE
> heated by the engine, and this would then make for a cold (perceived) MAT
> reading making your calc offset towards the rich side of fueling, and then
> your oxygen sensor would wonder why the mix is so rich and would be
working
> to try to resolve this in closed loop.  There are so many details worked
out
> in the chip that one would spend a lot of time trying to come up with a
much
> better approach, but for those who want the challenge, and to what
> gains???? go for it.  With this type of setup as it is, I have yet to see
> one that is THAT far off to warrant a total revamp of the system, though
as
> with this list, we know we can always tune it a little better and look for
> those challenges ;-).  Now if someone really wanted to do this, there are
a
> host of sensors that are faster acting and accurate, but all the sections
in
> the chip would need to be told this to work properly.
>
> Jeff M
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> --------------------------------------------------------------------------
--
> To unsubscribe from gmecm, send "unsubscribe gmecm" (without the quotes)
> in the body of a message (not the subject) to majordomo@lists.diy-efi.org
>

----------------------------------------------------------------------------
To unsubscribe from gmecm, send "unsubscribe gmecm" (without the quotes)
in the body of a message (not the subject) to majordomo@lists.diy-efi.org