Kevin, your experiments with the 2nd gen igniter gave me another idea to
try
for a cheap CD system. The 2nd gen lower mag sensor puts out 12 trigger
pulse per rev.
Feed that to a FET that turns on and off the primary of a pulse
transformer.
At 400 RPM that is 80 PPS. A simple circuit can double that to 160 PPS.
at 8000 RPM it is 266 PPS.
A common low cost ($30) 220 V 60 cycle control transformer is CT 24 volt
to 220
volt. Use the 12V half to jack the voltage from 12 Volt square wave to
220 volts AC.
With a full bridge rectifier or voltage doubler circuit you might get
near 400 volts
DC. Use that to charge the CD capacitor.
The old Heath Kit CD ignition used a control transformer and a
mechanical
vibrator
to generate the high voltage back in the late 50's and early 60's.
You can find a manual on ebay. Heathkit Manual Capacitive Discharge
Ignition
If you want a cheap pulse transformer they can be had from salvaged PC
power
supplies. It takes a lot of fooling with them to find the windings that
work.
A CD systems trigger with no delay.
BTW with AVR development board are you using?
Paul Lamar
-------------------------------------------------------
I just had a better idea.
Use an ABS sensor on the flywheel teeth and get a much higher frequency.
4000 pps at 400 RPM and 8.3 Khz at 8000 RPM. Now we are in the range of
switching
power supplies found in PC. Much smaller transformers.
No 2nd gen CAS needed. Will work with any rotary 12A to RX8 that uses a
starter.
Here are some typical PC power supply transformers and CD caps.
Paul Lamar
I was fooling with a transformer out of a PC power supply. I found a
couple of
windings that seemed to work. 10 volts in pp gave 100 volt out pp.
This was at 20 Khz but it worked over a wide freq range.
Remove the primary, use larger wire and cut the number of windings by
about 1.3rd.
Paul Lamar
I think I found a source for transformers. If the price is right I will
get some and
build some CD's. These are so cheap we can use one per plug. I'll donate
a
set if you
are willing to test them. You must be in the U.S. so I can get there to
witness the test.
Kevin, if you don't get yours to work I will be in Florida about Oct 1st.
I might
have them done by then. Those 2nd gen Mazda coils make great CD coils
when
removed
from the igniter's.
In the past I tried to build CD systems but with out good results. I
thought the HV
had to regulated. That was the main problem. I now realize that is not
neccessary as
the outboard type are not regulated. Nor was the Heathkit for that
matter.
In other
words it is open loop. Here is an open loop CD system using a 555
oscillator at 68 Khz.
Paul Lamar
I am not sure I will be done by Oct 1, but I work every time I get a
chance. I have read the discussions about the CD ignition and although it
looks good on paper, I am skeptical. I am a boater (USCG Licensed Captain)
and have spent lots of time on the water in everything from small
outboards
to big tug boats and commercial vessels- My experience tells me that 1.
The
only boats that have left me limping home are outboards, and 2.
especially
before the past 5 years or so ago more often than not due to a bad power
pack that controls the ignition. While the reliability of outboards has
increased substantially in the past few years, I still do not trust it on
an airplane. I remember a particular boat that we had- a 2006 model 35
foot
center console with 3 outboards. I seldom ran out a tank of fuel that we
didn't come back with one engine kicked up and not running. We had 3
engines, and most of the time only were able to use 2 of them.
More recently, for outboards we have a 150 Yamaha 4 stroke EFI, and a
Mercury 25 four stroke EFI. Both have suffered fuel problems, electrical
issues, and the Merc- only 3 years old- needs both electrical work and EFI
work to even run again.
On the rotary, I am encouraged by my weekend experiments- with redundancy
and simplicity in mind, here is what I am planning:
I will initially use an Arduino Pro Mini (Atmega 328 clocked at 20 mhz at
5.5 volts) with a program just to fire the coils through a Darlington or
MOSFET transistor. A wasted spark setup would be easiest, but a coil
firing twice as many times as it should will generate excess heat, and
will
shorten the available charge time by half. My tests for this part will be
completed using carbs from a local outboard repair shop- again simplicity.
I am not looking for power yet, just the ability to test the engine at a
variety of RPM settings and put a few hours on it. The Arduino has 14 I/O
ports some analog and some digital. I can use the outputs to record RPM
and
other variables. I also have a similar Arduino already set up as a
datalogger. Right now it records movement, temps, etc, but I plan to
modify
it with different sensors to record the engine parameters. This unit will
likely be a permanent part of the plane.
http://arduino.cc/en/Main/ArduinoBoardUno
These boards are manufactured by lots of different people now, and the
ones
I have have interchangeable crystal oscillators. Clock speed now is 16mhz
at 5v, 20 at 5.5v. I think this will work especially for testing, but if I
need to push it up I will.
Back to the ignition- assuming satisfactory results, the final design is
anticipated to consist of twin ignition computers in separate cases
connected in an A/B configuration. I used the attached schematic as a
base, with a few changes- I removed the switch shown. In its place I
connected one of the leads of the CAS to the Base of the Tx. I connected
the other to a common ground. I connected the common ground to the frame
of
the coil. I fed 12v to the tan wire, and the trigger to the red wire. Just
for this test, I didnt use any resistors or diodes, capacitors, etc. I
just
wanted to see if it would work. The final circuit will have a few more
components I am sure. And, I used the Darlington Tx because that is what
is
what I had in the goody box.
That is the update for the moment-
Kevin Alderman
I assume the Arduino is programed in C. Strange for a boat captain to be
proficient
in C. :) I will spend two months in Orlando and take a few days off to
visit the keys
anyway. I think I have been on your island and seen your house at the
airport.
I have a real old friend that lives in south Florida by the name of Dave
Smith.
He is also a licensed boat captain and a road racer. I think he still
runs an RX7.
BTW look at FET's instead of darlingtons. They have much lower on
resistance and
therefor generate less heat which translates to more reliable.
Darlingtons are old
school :)
Paul Lamar
Thanks Paul;
I would have used FETs, but I didn't have any in the goody box. My
intention this weekend was to just see if I could get the coils to fire.
Now that phase 1 testing is complete, we move on to phase 2.
As far as proficient in C, that might be a bit of a stretch. I am 45,
straight out of high school I started an electronics engineering course. My
last semester the class canceled due to lack of interest- took 10 to keep
the course we only had 8 left. It was at this point that I started business
as a Contractor eventually having 74 employees and many headaches. Over the
years I would go back to the local community college for courses here and
there- programming, physics, lots of calculus, etc. Always learning. As you
know after all of your years, it is mostly how your brain works, not what
stuff you memorized along the way. So, I find a problem, learn as much as I
can about it, and then proceed to solve it one step at a time.
For the beacon on the plane I made a 555 timer connected to two red LEDs,
one on top and one underneath. Works well and I am positive it will meet
the lumen requirement, BUT, I plan to make a step up circuit to boost the
voltage to 36v to make them really bright. I took apart a power supply and
a couple of amplified car stereo units, so now my goody box has more stuff
in it.
Kevin Alderman
Check this out- one made to communicate with Android!
http://www.mouser.com/arduinoandroid/
Might make it easier to output the display info to an android on the panel-
Yippee, I get another project to test out!
Kevin Alderman
Kevin, this can be programed with arduino software. $27. If I don't use it I will
send it to you.
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.
