UEGO IDEA

[Ludis Langens]

Stuart Bunning <stuart at kenelec.com.au> wrote:
> I have removed the carby and
> installed the TBI and now (havn't done this yet) are going to keep altering
> the VE table until the base pulse width for the injectors will supply the
> same amount of fuel as my carby did at idle over the same time period.
> Hopefully I will be able to start the car and it will idle. The problem is
> if I change RPM or load of the engine ( touch the trottle ) I would
> instantly go lean. If I was to set the entire VE map the same maybe your
> approach will enable to use BLM numbers to adjust the rest of my map.  

You won't go as lean as you think.  The MAP sensor is used to index into
the VE map.  But it normally is also multiplied into the fuel
calculation.  If you had a totally flat VE map, an increase in MAP would
still increase the fuel delivered.  (Further, an increase in RPM would
cause fuel to be injected more often - also increasing the total fuel
delivery over a unit of time.)

> For example if at 3000RPM I have maximum torque at WOT 100% efficient could
> I assume
> that at 50KPA it would we only 50% efficient,25KPA 25% etc...

I checked the VE table in a production PROM.  At most RPM values, there
was almost exactly a factor of two increase in VE from 20 KPa to 100
KPa.  (Once the MAP gets multiplied in, that means a factor of ten in
fuel from 20 KPa to 100 KPa.)  Across the RPM range, the highest and
lowest VEs varied by about a factor of 1.5.

The only exception was the idle "corner" of the table (low rpm, low MAP
/ high vacuum).  In this portion of the table, the VE dropped below what
would be predicted by the previous ratios.  Perhaps you should do your
carb measurement at a load other than idle - and use the results to set
an initial (flat) VE table.


Regarding tuning with a UEGO:  A wide range sensor allows the BLMs to be
determined almost instantly.  If the sensor indicated a ratio of 12.3
when the ECM wanted 14.6, a little bit of math can jump the BLM to very
close to the correct value for the next cylinder cycle.  There is no
need to bump the BLM (or INT) by one count, wait a while, and perhaps
repeat.  Of course these smart BLM changes would require a code change
in an ECM's program.  With appropriate software, the VE table could be
tuned fairly well on a dyno with just a quick sweep through the RPM
ranges at 10 or 20 different throttle positions.

-- 
Ludis Langens                               ludis (at) cruzers (dot) com
Mac, Fiero, & engine controller goodies:  http://www.cruzers.com/~ludis/