This is an array of standard .2 inch diameter infrared
photo transistors and LEDs. There is a slot in the slide
that allows the light to travel from one LED to its corresponding
photo transistor. Very simple and straight forward. These parts
are about 80 cents each. This gives slide throttle position
to one part in 16 or 1/8th of an inch. Two $2 chips called a
8 bit priority encoders are used to out put a 4 bit
(1 part in 16) binary number to a parallel input port.

The advantage of this approach is you know
where the throttle is and you know if it has moved quickly
so an enrichment pulse can be sent to the injectors to
prevent stumble. It is possible one could do away with
the manifold pressure sensor and just use this information
along with RPM to adjust the mixture.

This is a piece of cake for a DOS-QB EFI system.
The parallel port is read by one statement called
a PEEK in QB. QB can tell how fast the throttle
is moving by recording the last time it was moved
and subtracting that time from the next time it moves.
If it is less than a .1 second than somebody jammed
the throttle open :)

Paul Lamar

Paul.

O.2 inch is about 5mm. I have seen 3mm ones and their might be smaller
ones available. You could also accomplish the same thing with one
optical reader like my torque sensor idea. Have it read dots, or
patterns that correspond to binary numbers.

Doug in Japan

I was just going on what DigiKey was offering. It would be nice if they
were smaller. I could not find any. Got any data on your optical reader?
I have had problems with optical readers in the past and I have
found the interrupter design to be more reliable. The one side optical
readers is sensitive to ambient light level. The other advantage to this
approach is the current slide throttle covers provide the package and
holders for LEDs, photo transistors and the PC boards. No device needs
to be mounted or attached to the throttle cable. I might offer a kit
way down the road if enough people are interested in the PC DOS-QB EFI.
It can be retrofitted to current slide throttles that are out there.

Paul Lamar


How about this:

4 photo diodes, 4 LED's. Mount them in a single line perpendicular to
the axis of the slide. Binary encode the position by drilling lines
of grey code holes across the array. Grey code so that transitioning
from one position to the next only changes one bit of the code -
eliminating spurious glitching. As the slide moves, a different count
is made present across the photo diodes - no other encoding required.
At one end of the slide, only one hole, at the other end of the slide,
4 holes (16 counts).

To deal with radiation contamination (that's DC offset/drift), the
emitters (LED's) should be modulated at something like 20kHtz, and the
photo diodes should have a simple high pass filter driving a follower
(buffer) driving a rectifier and low pass filter, driving a
comparator. Have to look at the datasheets to see the response curves
to pick a good frequency. It might be helpful to turn on only one
emitter at a time and read only that emitter's photo diode at that
time - to reduce the chance of contamination from one to the next.
The computer can easily switch between them fast enough to not miss
any movement.


Matt-

Draw it up Matt. I would like to look at it.
If there is any width to the place where the holes go that could
interfere with blocking the intake ports. I explored that and gave up
on it for that reason. I drew it out to scale. The current scheme
needs only one narrow slot in the slide.

Paul Lamar


Why not just build it up as a unit separate from the slide throttle and
place it in the cockpit and mechanically connect it to the throttle cable?

Joe Berki

More work. More time.

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

OK I found some smaller transistors. This reduced the
area needed on the slide throttle.

Here is a pdf on the transistors and a few 3Ds of the
slide throttle position sensor.

I can use the MTIME function in PB to tell when somebody has slammed
the throttle open.

Paul Lamar


Paul,

My drawing skills are quite rusty, but attached is a picture with the
basic idea.. The slots are the additional openings that would be
added to the existing throttle plate. The round objects are the holes
that would be added to the existing throttle housing - sensors in
holes on on one side, LED's on the other.

The slots could/should be narrower.

And the sensors/slots don't have to be right next to each other..
They could be arrayed along the top of the throttle plate, one sensor
for each series of slots. Just so long as the corresponding
sensor/LED and slot are moved together.


Matt-

Yes I see what your doing now. That will work. IMHO it has two drawbacks.
It needs to be done in a mill and it won't fit between the tubes with
out leaking air when the slide is wide open. It must be two inches
long so one would have to increase the length of the slide by one or
2 inches or so to fit it between the aft tube and the throttle cable.
I would like to be able to retrofit to current slide throttles.
The old scheme can be done in a drill press.

Paul Lamar


Concur on the need for milling. However, I believe my approach can be
applied to existing parts.  If the slide movement is two inches, and
the length is eight inches, there's room for four evenly spaced
sensors arranged in series along the top edge of the slide. Each of
the sensors will see the same slots as depicted in my drawing - just
arranged in series instead of parallel.


Matt-

Well that too will work. You will need a long skinny PC board
to feed power to the LED's on one side and the same on the other for
the photo TXs and amps. It does cut the need for the octal encoders.

Paul Lamar