Subject: Patent is now pending on new rotary ignition
From: rotaryeng
Date: 4/29/2014, 7:46 PM
To: AAA-rotaryeng



   As of Friday afternoon.

   I am buying parts for 100 coil on plug units to be used for testing
   by Mazda (hopefully) and others.

   Will fire six times or more around TDC at 11,000 RPM.

   If somebody wants one or more for testing let me know.

   Paul Lamar.... Still in Virgina

   If all goes as planned be back in SOCAL Sunday night.

   Fantastic Paul, Milo



   Can I trigger them with the older crank angle sensor?

   Lewis Bjorkacro


   You can but you will only get two sparks close together on the front
   rotor. I take it this is the one you are talking about. A simple
   circuit could be built to use the lower part and count pulses.

   Best to take it out and save the weight. ABS wheel speed sensors are
   cheap and two are required for a two rotor. Or you can use 3rd gen
   or RX8 sensors. You can also make your own as all they are is a bar
   magnet with a coil of wire wound around it.

   This ignition system can also be triggered directly by a computer
   but it must be reprogrammed to eliminate the dwell time and enable
   multiple sparks.

   Paul Lamar   Back in SZP


   I'd like two for my single rotor.  Tell me how much $ and
   when/where.

   Lewis.


   In about 2 weeks. Introductory price is $175 each plus shipping.

   Paul Lamar

   Paul, one per plug, or one per rotor?

   Kevin Alderman

   One per plug. That is a standard in the auto industry now.

   Also the rotary racers use one MSD system per plug but one would also
   have to buy and old style coil one per plug for another $50 or so.

   The MSD stops multi-firing at 4500 RPM. The multi-firing it does do
   at low RPM is all grouped around 20 deg BTDC within a few usec.


   Paul Lamar


   I suppose if the leading and trailing plugs are both fired at 20
   BTC, you could get one waste spark coil per rotor.

   Dale Davies

   That is a feature I have not yet implemented but it is possible with
   my system.


   Paul Lamar


   Kevin,

   Forgive me if this is a little simple minded and has already been
   covered, but if you used a stock trigger wheel or something like a
   60-x  wheel or similar, couldn't you program your ECU to fire as many
   pulses as you want at any given advance?  This assumes of course you
   have enough outputs to fire all four ignitions.

   It seems to me that fixing the advance on an ignition is a step
   backwards, especially considering the potential of this new ignition
   system.

   Paul - I know it's possible to delay an analog signal with some clever
   circuit design... how difficult would it be to put together
   'controller' board that can sense RPM and delay your pulses
   appropriately for the load condition?  This way you set the triggers
   to max advance, and delay it with the board.  This would work more or
   less like a mechanical advance.  If you wanted to get really clever,
   you could incorporate something to sense vacuum and boost, but having
   a two variable advance would likely be much more difficult to do with
   hardware alone.

   -Josh Jacobs


   IMHO Kevin is programming the Ardunio as a full blown EFI. No need for a separate
   PC board. The Ardunio can do it all.

   For a car like the RX8 one would need a computer to take into
   consideration the 2 ms dwell and then fire the plugs six times
   in 10 degree increments at the proper E-shaft angle. I will eventually get
   around to doing that. One needs a variable delay depending  on RPM.
    My computer of choice would be the:

   http://www.mikroe.com/mini/stellaris/ <http://www.mikroe.com/mini/stellaris/>

   ARM Mini-M4 available from Mouser for $30 or $40.

   A simple tach circuit is a 555  and for MAP  a Free Scale pressure
   sensor sending analog voltage to the Mini-M4 A to D
   input. Then you can get fancy with advance inversely proportional to MAP.

   My language of choice is Mikroe compiled BASIC.

   EFI then becomes cheap :)

   Paul Lamar



Thanks for the input guys. My system isn't the Arduino anymore, I went with a Freescale chip with X-Gate. The X-Gate chip reads all of the inputs and does the calcs offloading the interrupt duties from the main processor. My code has grown quite a bit, and with testing will likely be cleaned up and shrink a bit. Yes I can use a table for spark advance and I might do that- set the pickup at the max advance and then wait to fire the plugs. This can be handled with math or a lookup table, no biggie. As far as the stock CAS, I have one of the 2nd gen units and actually have it installed in my engine. However, I can use lightweight HE sensors strategically placed and save weight as well as get rid of something that rotates at 3000 RPM that could fail. I have several extra I/Os, so other than writing more code that isn't a problem. I can even write in code that checks for the HE signal at a specified time at any particular RPM as a failsafe, and if it isn't there use a table to fire the plugs while it sends an error message.

My system is a little different than most- I plan to have my system be more like a FADEC- push throttle forward to detent 1 for max power consisting of 8500 RPM, 8psi boost and a limit on temps, in flight cruise will be detent 2 set at 6000 RPM whether in climb or descent, detent 3 for decending to land, flight idle and ground idle with full manual control as well. I know I know but if you met me you would understand, its just how I am.

At the moment I have 2 planes almost completed which is taking up a lot of my time. One is a Sonex for a friend and the other is a Wheeler Express just like my rotary powered one except it had a Lyc IO360 on it when I bought it. Almost done with both of these which frees up a hangar for the other one :).


Overboard as usual,
Kevin Alderman

-

The Rotary Engine News Letter. Powered by Linux.
ACRE NL web site. http://www.rotaryeng.net
You Tube http://tinyurl.com/beqqxas
Copyright 1998-2014 All world wide rights reserved.