[Diy_efi] Speed-density vs. MAF/MAP...

[Adam Wade]

--- Marcell Gal <cell at x-dsl.hu> wrote:

> (I reroute the discussion to the list if you don't
> mind)

Not at all.  I actually thought my original reply to
you was to the list.  ;)

>> In addition, you can read MAP before or after the
>> throttle.  That would give two very different
>> numbers.

> you're probably right.

Even if you're at full throttle, you have a
restriction in the throttle body itself, plus the
plenum acts as a capacitor for the air...  So unless
you're at a sufficient rpm, steady-state, to get
relatively smooth airflow through the throttled volume
(taking into account the intake valve opening and
closing), you should have pretty notable differences
between PAP (as you call it below) and true MAP.

> However since MAP is Manifold Air Pressure I would
> not call pre-throttle Air Pressure as MAP.

Well, there are speed density systems out there using
both techniques, so obviously there are ways to make
both workable, although I imagine they differ.  I'd
definitely like to experiment with that, or hear from
someone who knows the differences from experience.

>> If you read MAP from the throttled volume

> that's what I call MAP

Yes, I agree that that is truly MAP.

>> right before the intake valve closes

> I agree with that, good point.

I think I meant right before it OPENS.  If there is
some supercharging going on from tuned intake length
and reflected pressure pulses, that is where you would
"see" it with a MAP sensor.

Of course, VE cannot be inferred from that alone, as
you also have valve overlap, engine rpm, and exhaust
length as factors in the mix.  This is one area where
I believe mass air has the advantage when reversion is
not an issue; mass air automatically takes into
account any changes in VE as long as there is no
backflow through the MAF.  More VE means more air
ingested, and you can read that directly.  In fact,
you should be able to very easily compute VE with a
solid MAF signal and rpm.

Speed density CAN take into account changes in VE, but
it depends on how and when the MAP is measured.  If
MAP is relatively steady (high engine speed), then
it's trivial, and speed density should be as accurate
as mass air.  It's when the signal is not steady
(closer to idle) that issues might arise from how the
MAP sensor is sampled.

>> If you have a higher throttled volume MAP before
>> the intake valve opens, at a given RPM and throttle
>> opening, then you're going to get better filling.
>> If the engine itself suddenly becomes more
>> efficient (tuning the exhaust to that particular
>> RPM, for example), then the MAP should drop
>> slightly as more air moves through.

> OK. So we can confirm the appr. 

> VE= Constant * (pAmbient - MAP)/RPM
> relationship, ( only applies, when MAP(RPM) is
> measured at WOT, seems similar to an airbox pressure
> sensor config)

Fair enough.

I think the big advantage of the airbox MAP, then, is
that you can close the throttle(s) and still use the
same formula as above.  While MAP drops like a stone
at idle with the throttle(s) closed, that gives the
opposite of the above formula...  The MORE difference
bewteen ambient and MAP at a given rpm, the LESS power
the engine is making.

> which is quite inverse of what could be thought
> or some previous email suggested (though not stated
> explicitely).

I think I am seeing now where the confusion arose in
our communications.

>> I'm led to wonder, since the most obviously useful
>> MAP numbers seem to come from the throttled volume,
>> what the real-world difference is in plenum MAP and
>> throttled volume MAP.  

> Is there a better name for Plenum Air Pressure, like
> PAP?

Not yet.  Maybe I should invent a new standard in my
book.  ;)

I think I answered most of my own question in this
email, as well.

>> Does the plenum and throttle(s) act as a filter for
>> low-rpm pulsation to some degree?

In answer to my own question:  More than that, it
allows you to see the ACTUAL pressure drop from
ingested air at ANY throttle opening, while you would
need to know the throttle opening with a true MAP
sensor to know what the pressure reading was telling
you.

> I think so. The further away from the valves, the
> smoother the airflow should be -- generally.

And you have the big capacitor of the plenum.

> Why don't we take 32..64 MAP samples per revolution?
> (at least to get datalogs to support the 
> pre-intakevalve-opening theory).

I am working with someone who is going to be doing
some significant datalogging, and we are going to
outright test a lot of the ideas that have come up in
this threat to find strengths and weaknesses, useful
and less useful methods of doing things.  I've become
immensely curious, and it's time to find out the
correct answers.

> Even processing that realtime should not be too
> prohibitive with fixed point math.

I'd even be interested in taking oscilloscope traces,
synced to crank timing, and exploring different ideas
for sampling times in crank degrees to again find what
would be more or less useful in a practical sense.
 
> (definitely less resource than the pure wav data
> from a knock-sensing mic)

Heh, I almost think it makes more sense to look at
synced analog initially to get some ideas, and then
explore the actual application of those ideas in a
digital realm, once we know what our sampling
parameters need to be.

=====
| Adam Wade                       1990 Kwak Zephyr 550 (Daphne) |
|   http://y42.photos.yahoo.com/bc/espresso_doppio/lst?.dir=/   |
| "It was like an emergency ward after a great catastrophe; it  |
|   didn't matter what race or class the victims belonged to.   |
|  They were all given the same miracle drug, which was coffee. |
|   The catastrophe in this case, of course, was that the sun   |
|     had come up again."                    -Kurt Vonnegut     |

__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com

_______________________________________________
Diy_efi mailing list
Diy_efi at diy-efi.org
http://www.diy-efi.org/mailman/listinfo/diy_efi