BPW and Cylinder Vol.?

[rrauscher at nni.com]

Having been through this <g>, I hope I can add some insight.

The inital tuning in forced open loop is to get the tables
roughed in. If the INT is allowed to run as you drive through
different map/rpm ranges, the differences can be so great
that it can be dangerous. An example would be where the INT
is low (98, pulling fuel), then you hit a lean spot. Not only
can the engine die, you also don't know if it's lean due to
the INT, or the ve table.

(also, don't drive too far, I almost didn't get the car
back on my first trip out <g>).

Once the tables are roughed in, mostly on the rich side,
you can go closed loop for tuning. However, there are several
other variables that can skew your results. I've been known
to disable: BLM (set min/max to 128), PE, baro update, DFCO,
and reduce the proportional terms. All of these will effect
the final bpw.

Make sure the engine is up to temp before collecting the
data. Also may want to reduce the high map timing a bit,
to be on the safe side (for wot tuning). Can also do light
load tuning with the drive tires off the ground.

Now start driving around, trying to stay at a particular
map/rpm long enough for the INT to change and stabilize.
It's easier to hold the rpm at a table point than the map.
(IE: ve tbl rpm's are 2000, 2400, 2800. Hold at each of
those rpm's, not at points inbetween).

As you go through the data, look for areas that the rpm
and map are close to a table point. This will give the
best results.

As you tune the VE% tables, keep in mind that you are
doing it backwards. That is, you are using the final
fuel term to adjust the VE% of the engine. The VE% is
just one of many terms that contribute to the final bpw.

This is why it is important to 'lock' as many of the
other terms as possible.

Oh, on the INT, the update rate depends on the amount
of air the engine is moving. The speed to which the O2
sensor reacts is also affected by the distance of the
O2 sensor location from the exhaust port (transport delay).

At low air flow, with the O2 sensor in the header
collector, the transport delay can exceed one second.

BobR.

Who believes that a dyno and an eprom emulator could make
short work of this.

>I've been trying this method (14.7, CL, tune from int/blm).  It can be done
to rough in the ve
>tables, but it seems that the integrator is slower to respond than I first
thought.  I got my
>tables roughed in and disabled CL per suggestions.  After about one day of
driving, logging,
>burning, I had really good tables.
>
>I suggest the OL method after my experiences over the last couple of weeks.
 The difference to me
>is the speed of the feedback, the O2 sensor is instant reading of lean/rich.

>
>YMMV
>
>Squash
>
>--- Shannen Durphey <shannen at grolen.com> wrote:
>> Funny what you can find by reading through old mail...
>> So you're disabling CL and watching the O2 values to determine
>> rich/lean?
>> How about Roger Heflin's suggestion to set 14.7 AFR and leave CL
>> enabled, using BL and INT as guidelines?
>> Shannen
>> 
>> rrauscher at nni.com wrote:
>> > 
>> > To force open loop, set the enable temp to the max value. For
>> > tuning the VE tables, also set the open loop afr ratio to 14.7.
>> > 
>> > On the '747 there are two tables to change, one for the afr
>> > vs. temp., and the other is afr vs. map. In this manner, you
>> > can then 'discover' what the ve is at each point in the table.
>> > 
>> > This is really tuning in reverse, but hey, it works <g>.
>> > 
>> > BobR.
>> 
>
>
>=====
>Trains, Trucks, and EFI at http://www.trailrunners4x4.org/users/realsquash

>

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