Subject: pport tuning peculiarity
From: Rotary Engine
Date: 3/17/2009, 9:12 PM
To: AAA Put this in the To box
CC: Tracy Crook <tcrook@rotaryaviation.com>


 Hello Paul-

 Sent this to Tracy earlier today; thought I would pass it along in case
 someone on the list has seen this before...

 Had some peculiar behaviors in that latest tuning attempts.Hopefully
 you'll have some suggestions.

 Background-

 - 13b, p-ported, with slide throttle design & "tuned" intakes
 - LS1 injectors
 - EC2 with latest p-port firmware
 - EM2
 - 6800 field altitude
 - 55F OAT
 - Known to be under propped (has a 72x72 160hp 2-bladed prop.Needs a
 72x109 230hp 3-bladed prop)

 Symptoms-

 If we run the engine with the manifold pressure lines *DISCONNECTED *from
 the EC2:

 - Engine starts very easily
 - The engine runs quite smoothly across a pretty wide rpm band after
 tweaking the manual mixture control.
 - Reasonably smooth running as low as 1500rpm engine (526rpm prop!)
 - Very smooth running 3000 - 4800 RPM.
 - Above 4800 rpm we seem to "run out of enrichen" on the manual mixture
 control & the engine stops.
 - MAP table view on the EM2 real-time shows the EC2 using map location
 59
 or 91 (MP constant matching atmospheric pressure) depending on high/low
 throttle (as expected)


 If we run the engine with the manifold pressure lines *CONNECTED *to the
 EC2:

 - Engine noticeably harder to start
 - Engine will not run smoothly at any rpm
 - Significant surging pattern -- rpm increases rapidly, followed by
 engine stumble to low idle rpm, followed by next surge.
 - Engine often stops during the stumble / hard to restart
 - MP pressure varies rapidly based on engine rpm
 - Significant manual mixture changes can help ease the surges somewhat,
 but never enough to get the engine to smooth out enough to let us see
 if MAP
 values could be tweaked.


 Looking at the MAP table with engine off, it seems like all map cells are
 at
 0 -- as expected.I did not check all 127 cells, only the top cells (say
 100 - 127).

 Mark (Rocky Mtn High)

 That is good evidence the 555 system will work fine :)

 We noticed this phenomena on the dyno test of your engine.

 I am not privy to Tracy source code but the last time I talked to him
 about it was a couple of years ago and he was taking only one sample
 of the intake manifold pressure per e-shaft revolution.

 Using a computer this takes only about 10 usec.
 http://en.wikipedia.org/wiki/Microsecond
 Light travels only one mile in FIVE microseconds.

 That is 10 millionths of a second. 100 usec is one msec or one thousands
 of a second. This is a very brief period of time so the manifold pressure
 is jumping around a lot in a p-port tuned manifold. It could be anything
 or even above atmospheric pressure at some times. That is what p-port
 tuned
 manifolds do. See the enclosed chart. A lot depends on WHEN you sample
 the pressure. It changes with RPM in unpredictable and mysterious ways.

 IMHO this is confusing the heck out of Tracy's computer. IMHO you need
 to feed the manifold pressurefrom both intake tubes past the slide
 throttle
 using very small hoses into a beer can and then tap out of the beer can
 with
 one normal size hose going to the pressure sensor in the computer.
 Hopefully
 this will damp the pulsations and average the intake manifold pressure.

 Al Gietzen tried this as I recall and it did not seem
 to change anything. I don't know what size hose Al used
 but I think very small hose is essential. On the order of .030
 inside diameter or smaller.

 Paul Lamar


Mark,
You don't need the beer can. If you use 1/8 od tubing, make a coil of
tubing
using at least 20 feet of tubing. 30 feet may be better. The length of
tubing works to damp the intake signal to something that the computer
makes
sense with. My experience is the tubing can work better than a reservoir
as
it is using the friction of air molecules on the surface of the tubing as
the damper, and will respond faster than a reservoir, but still be damped.
That doesn't mean the maps are correct, as it sounds like the manifold
pressure side of the map is way off.
To trouble shoot being rich or lean, have a squirt bottle of fuel handy
and
give the runners a quick squirt as the rpms are declining (while surging).
If this makes them instantly recover, the system is going lean. if it
forces
it to die, the system is going rich. that will tell you which way to go in
the tuning. My guess is it is very lean with the sensor connected. The
sensor disconnected simulates wide open throttle with no manifold vaccum
with high fuel demand, so the ecu increases the pulse width to the
injectors. When the sensor is connected, and the ecu sees manifold pressue
decreased, it reduces the pulse width.
make sure you are careful when performing this test...... and of course
you
have a fire bottle handy...
hope this helps,
Larry

Hi Larry,

20 feet of tubing is a fair size expense and bundle.
My lathe tail stock chuck will not hold a tiny .030 dia. drill or I would
make some orifice up and send them to Mark.
Perhaps you could make up some orifice that will plug into
red silicone vacuum plumbing tubing. OD of the orifice would be .125 to
fit
the tubing. 0.030, .025 and .020 ought to cover it.

The rest of the world uses mass air flow sensors for a good reason :)

Paul Lamar


Paul,
plastic tubing from McMaster Carr #5345k515 is 15 cents per foot. vac rated
polyethylene.
Larry

To each his own. I like the high temp red silicone stuff. In case
of a fire it withstands 500 F. Also stretches nicely to grab on to barbs.
Looks more professional :)

BTW a couple of things. If one is tuning the p-port using Tracy's system
don't assume the manifold pressure on average is below atmospheric. That
is not the case. The P-port engine is at least 120% volumetric efficiency
so the average manifold pressure is probably above atmospheric. This is why
the tuned p-port generates so much power. The poor man's
turbo charger. If you make this assumption the engine will run lean
on the top end. Also the pressure sensors need to be chosen to reflect
this fact. In other words if the pressure sensor is scaled so atmospheric
manifold pressure is five volts it will never work with a p-port engine.
It needs to be scaled plus or minus 15 psi around atmospheric.

The pressure sensor Tracy uses (I think) is a Free Scale MPX4250AP.
That puts out 5 volts at 36 psi. So it should be a minor software
problem to adapt that to a p-port or turbo engine. Here is a pdf.

Also lacking a data acquisition system a quick and easy check
on the manifold pressure wave form is use an oscilloscope and look
at the output wave form of the pressure sensors. This will also
reflect the response time of the pressure sensors.
Here are pictures of Tracy's computer that Mark owns. Here
is where to connect the O-scope.

Mark, if I were not so busy I would fly out bring my scope and help tune.
I can't get away until the end of April.

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

Also I notice the response time of the pressure sensor is 1 ms
from reading the pdf.
If you sample if for 10 usec no telling what is going on.
Is the sensor in it's rise time phase, its peak phase or is it
in it's decay phase?

Another reason for taking many samples and averaging them.

Paul Lamar




Oversampling is always preferable to undersampling..If the response
time of the sensor is 1ms and you can't sample at least every 500us,
you may get unpredictable results.If you can't sample at least at 2x
the response time, you need some analog signal conditioning in the
form of a low pass filter before the digital sampling.

It seems likely that the response time for the sensor is driven by the
mechanical behavior of the transducer - mass, spring constant, and
damping.I doubt the response time is related to un-damped (resonant,
oscillating) behavior.


Regards,

Matt-

Every 500 usec is a piece of cake :)


Paul Lamar

Paul

It sounds like Mark has many variables that are creating non consistent
responses in his MAP sensing configuration.Adding more tubes, beer
cans - really now?- only add another item that can cause non
consistenacy.

Mass air flow sensors aren't without warts either, as they have to be
sized and calibrated for the expected flows. It isn't a
slap-it-on-and-go technology unless you are using the exact manifold and
fueling set-up for the engine it was designed to go with.MAF sensors
are sensitive to particles, MAP to location.Map sensors however are
fairly simple and a well baked technology. Basically a simple strain
gauge device.The information from either sensor, whether measuring
mass flow, or mass pressure is going to have to be manipulated to arrive
at a useful value. Both methods haveto account for temperture
differences, to calculate the amount of mass flowing through a
particular sized intake pipe. Neither sensor will operate in a perfect
linear response manner, over their intended range, in response to the
gas mass they encounter.In other words some error. Hence Tracy's
and MegaSquirt's approach to engine control. Add turbos and
increasing altitude to the mix and there is something good to be said
for look-up tables to tweak the fuel flow.

However your premise on increasing the amount of samples and averaging
them is correct.
 Increasing the sampling rate is a good thing but not by trying to do it
several times over the period of one intake cycle.The computer can
handle it, yes, but to no advantage with the relatively slow chemical
reactions happening in the engine.

 Why not takepeak MAP readings for at least ten intake cycles,
average them, and send that value to the CPU? In software flow - and
please correct any wrong terms- the data is buffered in a a first in -
first out sample file. On each cycleanother new value is sampled and
added to the file and the oldest one discarded. If each sample is of a
long enough duration to capture the peak signal, then some type of
conformity is happening. Peak MAP may not be an accurate depiction of
the actualMAP but again what we are looking for is sampling
consistentacy across the engine power band.

In response ;) to this issue, in the latest version of MegaSquirt,the
map sensing device is separated from the mother board. The small MAP
sensor with its signal processing PCB is installed right on the engine
near the sampling port. Power wires and signal wires travel back to
the remote CPU which has become palm size.

One point I would like clarification on is why do people tee the Map
sensor tubes together? Don't we only need to sample what is happening
in one rotor?

Doug in Japan

Generally speaking computers don't sit there and sample over and over
again. They do other things in the mean time. The process is called
polling. One could make a more sophisticated sensor that would send out a
pulse when the peak was reached. This pulse could interrupt the processor
so it could go out and read the peak value but that is not how it is being
done now.  I'll take another look at the Megasquirt but as I recall they use
throttle position, map and O2 sensor. Overkill for aircraft use.

Paul Lamar

Paul,

Megasquirt as you stated does look at throttle position, MAP and the O2
sensor as well as intake temperature.   O2 is handy for tuning and
possibly feedback  in LOP operations.  Throttle position sensor is not
needed.


Doug in Japan

What is near mandatory for tuning Tracy's system is a wide band mixture
monitor. The rotary will run from 10:1 to 20:1 so Marks engine mixture must
be all over the range. We saw that on the dyno with Tracy's system.
Mazdatrix
does not leave home without a wide band mixture monitor. I gave up trying
to tune it and stuck a Weber on there. Dave Lemon had it running fine
in ten minutes with one or two jet changes in the Weber.

My contention is almost any mass airflow sensor will keep the engine
in the range of 10:1 to 20:1 so manual leaning is all you need
as the engine runs at a constant speed and load for hours at
a time. Mass airflow sensors average the airflow and automatically
compensate for air temp so you don't get the wild variations in intake
manifold pressure characteristic of any organ pipe tuned system.

In the aircraft case there is no load on the engine up until
about 4500 and there after the load increases with the square or
the cube of RPM. Not at all like a car. A car is much more difficult
to tune properly hence the computer.

Paul Lamar


Paul/Doug

Dampening the pulses a MAP sensor sees with one reference w/ oriface
per rotor combined at a surge tank is common practice in racing EFI
systems. To my taste it is simple enough and reasonably effective at
getting a representative manifold pressure.

Doing this electronically is more complex, IMO.

-Chris
-- 
The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
Youtube key word UTUBPLEASE
Copyright 1998-2008 All world wide rights reserved.