While working on the EDIS logic diagram, it occurred to me that the duty
cycle on the coils might be too high for use with the rotary engine. My
reasoning is as follows. A 4 cyl, 4 stroke engine has one ignition event
per rev. Auto engines seldom stay at max torque or HP for very long. I
don't know what RPM that would be in one of the engines the Ford EDIS
coils were intended for, but I suppose it would be in the 3,000 to 4,000
range. This would equate to 3,000 to 4,000 spark events per minute. I
expect most of these engines would operate at 2,500 RPM or so most of
the time, except in city driving. I think this results in a relatively
low duty cycle.

The rotary can be expected to run most of the time around 6,000 RPM,
perhaps more for take off and climb out. And I understand there are 3
spark events per rotor per revolution or 6 events per revolution in the
case of the 13B because the two rotors fire at different eshaft points
(degrees of rev.). I seem to recall that the rotor RPM is different than
the eshaft. I believe that there are only 3 spark events per eshaft
rotation or six in two rotations. If my understanding is correct this
would equate to 18,000 spark events per minute. This is a duty cycle of
more than 4.5 times the original design. NOT GOOD. This is so far out of
the ballpark that testing is probably a high risk endeavor.

Even if you used two coils, one for each rotor, the duty cycle on the
coils would exceed the one the designer intended for the Ford. Using two
coils would introduce a complication in the eshaft sensor position
output, as you would need some way of generating two separate pulse
trains; if you are going to that much trouble, why not use a more robust
coil? It is also possible that the Ford EDIS coils cannot operate at the
frequency needed by the 13B, e.g., electrical time constants.

I think it might be better to massage the output of the stock CAS with a
small electronics package to produce the correct signal for more robust
coils, perhaps the LS1s Tracy uses. What I envision is a small can
containing the electronics that would screw on the top of the CAS in
place of the flat plate. Four wires would then go to control the coils.
Just off the top of my head, it seems to me that a Schmidt trigger and a
phase locked loop with maybe some driver transistors would do the job.
If you wanted spark advance, add a PICAXE; possibly the PICAXE could
also do the job of the PLL and might be a better overall solution. I
expect the cost of the parts would not exceed $100. I would expect
to publish the schematics and construction directions and let everyone
who wanted to roll their own. For those who would prefer a wired and
tested unit, that might be possible also if there was enough interest.

If there are errors in the above discussion, please point them out to
me. Thanks.

Now for the requests.

1) Are the LS1 coils the best ones for the job? I assume so because that
is what Tracy recommends. Suggestions are welcome.

2) I need some design information on the chosen coils. What I mean by
this is, what kind of signal does the coil require, and what kind of
load does it present to the driver? I would be grateful for this
information or a pointer to where I could obtain it.

3) Can anyone help with the mechanical packaging? For example, a custom
can to fit on top of the CAS might cost more than all the electronic
parts put together! It might turn out that a remote, i.e., not on top of
the CAS, electronics package might be more economical if packaged in an
off-the-shelf box.

Activities update:

We are presently wiring the hangar, or will as soon as DAC and Ken get
back from vacation in MN Saturday. Right now I am working on the
physical design of the intake and exhaust systems. (I am so old and
decrepit that I can not do much phyically.)  DAC and Ken are in the
middle of fabricating the motor mount and custom oil pan (preempted by
the hangar); unfortunately they both have full time jobs so it is a time
available situation. I expect they will get back to it after the hangar
is done. During slack periods (HAH) I will look deeper into the ignition
problem. I have NOT given up on the PICAXE controlled super simple fuel
injection (SSFI) system that Paul suggested many months ago. It is just
that I have more pressing needs. If we can get the SSFI working, it
should be much more economical than a carb. I talked to Paul about it
when we met at Mark's hangar last Saturday :) (PIX attached),
specifically for the need of a throttle position sensor. Which brings me
to the next request:

4) The throttle position sensor in the slide throttle needs to pass a
rod through the spacer. I need to seal the rod against leakage. (I don't
like gasoline vapor under the cowl, especially when trying to start the
engine.) I am requesting suggestions on the best way to do this.
Currently my thoughts are an O-ring held in place with some kind of
screw-in device. Custom parts are too expensive unless absolutely
necessary, so I would appreciate a pointer to where I could get an
off-the-shelf part. I have seen them, I just don't know where to buy
one. And I do not know what the proper name would be, so searching is
difficult. (Thingies doesn't work very well). :)

Congratulations on the bell housing to all involved.

Doc Custer
Retired Software Engineer and sometime Audiologist


Rotor rotates at 1/3 of e-shaft rotation.

One spark event per e-shaft revolution  per rotor.
Both plugs firing at the same time. Front and rear rotors
are 180 degrees out.

I am not sure LS1 Corvette coils are the best. The coils used in
the RX8 are used to firing up to 9000 RPM which is a lot higher
than the LS1 engine can go.

Paul Lamar

Doc,


Just now I found out about E Mags.  Sort of a hybrid magneto that uses
electronics to collapse the field and produce the electrical surge.

http://www.emagair.com/NextGen.htm

As Paul said each rotor has one spark event per e-shaft revolution. You can
see this on the Pineapple Racing website.

I can't say at the moment the E Mag will work for us but the possibility is
intriguing  because it is a wasted spark system designed for four banger
aircraft engines.  The mag wastes one spark on one of the cylinder pair
exhaust stroke for each rpm.  We could put that wasted spark to use and fire
both plugs of  rotor #1 at the same time.    It would be nice if we mount
this E-mag in the CAS location because then it would run at a more
reasonable speed.  For example 3,000 rpm instead of 6,000 e-shaft rpm. at
cruise.  After 180 degree rev. of the rotary's  eshaft (90 degrees of the
CAS shaft) there would be another firing event for both plugs on rotor #2.
Obviously we would need four pick-ups.  0 and 180 for rotor #1 and 90 and
270 for rotor #2.

One of these models has a alternator built-in so if battery power goes south
it automatically starts generating the power it needs to run.

As for coils there are lots of types and output voltages up to 60K volts,
and firing durations of 200 micro secs to 500 micro secs.

Anybody know the specs. on the standard Mazda coils?
Included is a  small PDF file that lists Mallory coil types and specs.


How this helps a bit.


Cheers

Doug in Japan


The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
Copyright 1998-2006 All world wide rights reserved.