749 O/L idle help needed.....

[David Papworth]

Regarding IPW vs. RPM:

 OOOpss, yes, you are right. I forgot that the amount
of fuel injected is proportional to both the pulse width and the number of
pulses
per unit time

The correct equation should be

Fuel = (magic constant) * displacement * RPM * VE * MAP

Where Fuel = IPW * Pulses per second * (Magic constant 2)

and Pulses per second is determined from the crank sensor
or ignition system and thus itself a function of RPM

After cancellation and factoring constants

IPW = (constant) * displacement * VE * MAP

Which is the same thing as what you just said below.

But observe that IPW DOES track MAP, so variations in MAP
will affect IPW even if all the VE entries are set the same
(as you describe).

With a fixed throttle and IAC opening and constant fuel,
the MAP will pretty much track RPM.

Thus the IPW will seem to follow RPM as the MAP keeps
changing.

I'm assuming you've checked the TPS to make sure the ECM knows that
you are at idle and thus want it to set up a steady idle.

If the ECM is going to insist on leaning out as RPM drops, you could fight
it by adjusting the timing tables to deliberately advancing timing a degree
or two
as idle RPM drops. A little bit of timing change at idle has a much stronger
effect on idle
speed than a small change in mixture. You could thus set up a negative
feedback "bucket" at your desired idle speed that the fuel control would
be hard pressed to counteract.

What is the period of the RPM fluctuation? What is the phase
relationship between RPM and IPW?

It may be that the lag filter on MAP is causing the ECM to induce
the oscillations due to too much phase shift in the control circuit.

Something like, RPM drops momentarily due to imperfectly
atomized fuel or whatever, then returns to normal. MAP drops with
RPM, ECM reacts to lower MAP by backing out IPW, but RPM and
MAP have already returned. RPM drops again due to lowered IPW,
ECM reacts to prior RISE in MAP by adding IPW, which raises RPM again.
If you can figure out how much phase shift there is between
MAP changes and IPW changes, you could determine whether
the ECM is inducing the fluctuations or tracking them.You would need
to use an oscilloscope on the bench to figure this out.

Alternatively, study the disassembly of the 749 ECM and try to figure
out its idle strategy.

You could try to play with the VE tables to force effectively a fixed
IPW at low RPMs, assuming the 749 is in fact using the tables at idle
and not following some other strategy.

This would be done by increasing VE at lower MAP values and reducing
it at higher values (which is what you must do to hit your goal of constant
IPW). This would cause VE and MAP to compensate for each other
in the "improved" equation above, and the ECM would just sit there
at a constant IPW. You probably only need to play with a couple of
entries right around MAP at idle at your normal altitude.
Whether a constant IPW would or would not result in a steady
idle is hard to say, but at least you could feel like you accomplished
something!

One more thing,
Some fuel injection systems switch to a different strategy at low/high RPMS.
Does the Syclone setup continue to set the rate of injector pulses per
second
proportional to RPM all the way down to idle, or does it switch to a fixed
frequency mode with a variable pulse width that tracks RPM as well as MAP?

This would make more sense with TBI, where one might get better control
of fuel with higher frequency injector pulsing that would happen if it just
fired on a per-revolution basis. It appears the Syclone is not a TBI system
(from looking at the syty.org website),
which would make such a strategy rather strange. Still, you can verify that
the injection frequency tracks RPM on the bench easily enough.

Good luck, and sorry for the previous bad information.

----- Original Message -----
From: "Barry Tisdale" <btisdale at cybersol.com>
To: <gmecm at diy-efi.org>
Sent: Sunday, October 29, 2000 6:25 PM
Subject: Re: 749 O/L idle help needed.....


> Thanks, David -
>
> I don't understand why the RPM makes any difference.  My understanding is
> that RPM does not enter into the BPW calculation; only things that affect
> the amount of air drawn in per cylinder "suck" (air density, basically -
> secondarily injector voltage, etc.) will change the BPW.  I see how the
> total fuel needs of the engine are reflected in RPM, but as RPM
corresponds
> 1:1 to injector events, there shouldn't be any variation in pulse widths
> depending on RPMs.
>
> On my benchtop ECU tests of the chip, varying the RPMs thru a normal idle
> range of 500-1000RPMs, the BPW does not change, as the VE tables are set
> alike over this range.
>
> However, varying the MAP, even just a bit, will change the BPW - again,
> even though the table is set to zero change between cells.  Everything
from
> 20kPa to 100kPa and 400-1000RPMs is set to the same VE.
>
> This fluctuation in BPW is seen in the vehicle as fluctuating RPMs @ idle.
> - +/- 25, but I'm after zero.
>
> Something is causing an increase in BPW with an increase in MAP - oxy
> sensor's out of the picture (O/L), so I'm missing something somewhere.
>
> Thanks again - Barry
>
>
>
> At 05:04 PM 10/29/2000 -0800, you wrote:
> >There are a ton of adjustment factors, but
> >the basic fuel injection control equation is:
> >
> >IPW = (magic unit resolving Constant) * Displacement * RPM * MAP * VE
> >
> >As your RPM fluctuates, the IPW will track it in direct proportion.
> >
> >Thus if you get 1.25 ms at 750 RPM, at 650 we would
> >expect 650/750 * 1.25 = 1.083 ms, near enough to what
> >you are seeing.
> >
> >Setting your VE table "flat" means that you are asking for the mixture
> >to stay constant, which means the PCM is doing exactly
> >what you programmed it to do by varying IPW in proportion to RPM.
> >
> >The problem is, the VE table does nothing to establish a consistent idle
> >speed.
> >(You can de-stabilize it by causing the idle mixture to be too lean
across
> >the board).
>
> --------------------------------------------------------------------------
--
> 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
>

----------------------------------------------------------------------------
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