I have thought about it and thought about it and I have come to the
conclusion the
best way to do it is in hardware. The MC68HC908 does not have the
additional timer
needed to do it in software.
This uses a CMOS 555 pulse width generator to generate a pules width that
will delay
a 3 ms one shot to fire LS1 coils and RX8 coils. These coils will overheat
and fail
if the dwell is much longer than 4 or 5 ms. I have tested all these
circuits and the
CMOS 555 pulse width generator works extremely well. Much better than my
wildest
expectations. It works the same way the 555 EFI system works.
The RPM parameter is in the form of an analog voltage. In other word an
analog tach.
It is a simple circuit that generates a DC voltages when fed with a
constant pulse
width depending on the RPM. Basically a resister and capacitor. High RPM
more
voltage low RPM less voltage.
This is fed into the CMOS 555 to generate a varying pulse width delay. Low
RPM long
delay. High RPM short delay. Therefor the signal is reversed by an
inverting op amp.
The whole rig is triggered about 50 degrees BTDC.
Here is the preliminary circuit.
I'll do a bunch of tiny PC boards and sell them. You can buy the parts
from digikey.
Paul Lamar
Will this work with the self exciting coils on the older RX7s? The engines
I bought came with several of the older self exciting coils. Do you have a
wiring diagram for these coils so that I can test them? Or is it better to
replace them?
I am assuming that the boards are one per rotor like the MS EFI- so include
me for 2. If there is anything I can do to help the MS EFI project, let me
know.
Kevin Alderman
I think you are talking about these coils. We call them 2nd gen. They are somewhat
more robust except the ballast resister was failing in the early use and Tracys
suggested increasing the wattage. That could of been because Tracy was not dwelling
them correctly or the higher continuous RPM the aircraft engines were seeing. That
is why I recommend heat sinks for the RX8 coils.
I am not sure how Tracy handled the dwell and timing issue for 2nd gen coils. I did
take them out of the igniter chassis and tested them stand alone as CD coils and they
worked great. I am sure they would work great with a MSD CD system. The CD testing
was done with a high buck, 300 volt, off the shelf potted power supply. As I recall I
was using 10 amp bench 12 volt supply and it was not up to the 300 Volt power supply
surge current requirements. Always use a lead acid battery for testing as they will
put out a 300 amp surge current. Even the motor cycle type.
Both lower plugs were fired with one coil and that had two towers. Obviously waste
spark. Somewhere in the thousands of words of megasquirt ignition the 2nd gen rotary
coils may have been discussed. So far I have not seen any but then again I am not
looking as I am focused on getting one coil per to work plug right now.
Paul Lamar
I've been watching all this development with great interest, as I had
recently concluded that I would be foolish not to design my project
plane around a rotary. One of my biggest worries was engine management,
as the primary problem I'd been noticing in doing my research was the
difficulties people seemed to be having with this (the other biggie
being cooling). I very much agree with your KISS approach, and I'm glad
to see how fast you're making progress. However, now that you're
getting into the ignition side, I was wondering how difficult it would
be to include a knock-sensor retard "loop" in your setup. Obviously no
one wants to add unnecessary complexity, but rotaries are so smooth I
would think a knock sensor would have an easy time of detecting onset
early, and with aircraft engines tending to run lean and heavily loaded,
it would seem like some sort of rudimentary retard could be a "lots of
bang for the buck" sort of addition (and almost required for turbo guys,
which I am leaning toward). What are your thoughts on that, Paul? And
thanks again for all your exhaustive work on this project...I'll
definitely be one of your customers once I'm at the engine stage.
Greg Gibson
Knock is not generally a problem with the rotary unless you are looking at an ungodly
amount of boost. One of the reasons is the NA compression ratio is very low by todays
standards at only 10:1 in the RX8. Also there are off the shelf knock warning subsystems.
Mistral did everything in their power to create knock and the knock, when it
occurred, was unable to damaging the Iannetti ceramic apex seals :) The cast iron
seals have been known to fracture.
Paul Lamar
--
The Rotary Engine News Letter. Powered by Linux.
ACRE NL web site.
http://www.rotaryeng.net
You Tube key word PaulLamar2
Copyright 1998-2012 All world wide rights reserved.