I am highly encouraged to say the least.
 This was done in 4 milliseconds. To put that
 into perspective that is two times every
 revolution at 7500 RPMs. A millisecond is
 1/1000 seconds or one one thousandth of a
 second. Amazing!
The first picture to look at is the scope output.
 The top trace is a function generator out putting
 a saw tooth wave form. This is fed into pin 82
 which is an A to D input channel. This will become
 the manifold pressure sensor.
The bottom trace shows two simulated CAS pulses
 about 4 micro seconds wide each with 4 milliseconds
 between them. The scope amps were set to two volts
 per vertical division.
The second picture shows the test set up.
 The CAS simulator I built is on the right.
 The function generator is in the back left.
Here is the successful extremely simple ZBasic program.
 The injector pulse width output is still to be added
 but that takes only Usecs as does the rest of it.
************************************************************
 Sub Main ()
 Call PutPin(12, zxOutputLow) ' pin 12 CAS 1 will be at logic zero
 Call PutPin(13, zxOutputLow) ' pin 13 CAS 2 will be at logic zero
 dim x as integer
 'check and wait for CAS
 'do
 for x = 1 to 10000 'develop only
 If (GetPin(12)=1) then
 Call rotor1
 end if
if (GetPin(13) = 1) then
 Call rotor2
 end if
 'loop
next x'develop only
 end sub 'main
sub rotor1() 'it is 22 degrees BTDC
 dim add1 as integer
 add1 = GetADC(82)'MAP sensor val must be an integer
 Debug.Print "rotor 0ne A to D = " & CStr(add1)
 end sub 'rotor1
sub rotor2()'it is 158 degrees ATDC
 Debug.Print "rotor 2"
 End Sub  'rotor2
************************************************************************
Ultimately the rotor 2 program will be identical to the
 rotor 1 program. It is not filled in yet to keep the size
 down.
Now here is the kicker. Reading the A to D takes the most time.
 If one cannot do that in 4 ms and all the other stuff that
 needs doing game over. That statement applies to this timed injection
 approach only. There are other ways of doing EFI.
 If we had to add a hardware high speed A to D it would
 not cost much.  This is the log of what happened.
Downloading file "E:\3rd-efi.zxb":
 .............................
 Download complete.
 Verifying download:
 .............................
 Verification complete.
 ZBasic v3.0.3
 rotor 0ne A to D = 761
 rotor 0ne A to D = 586
 rotor 0ne A to D = 372
 rotor 0ne A to D = 544
 rotor 0ne A to D = 918
 rotor 0ne A to D = 257
 rotor 0ne A to D = 63
 rotor 0ne A to D = 771
 rotor 0ne A to D = 605
 rotor 0ne A to D = 64
 rotor 0ne A to D = 533
 rotor 0ne A to D = 439
 rotor 0ne A to D = 53
 rotor 0ne A to D = 696
 rotor 0ne A to D = 943
 rotor 0ne A to D = 625
 rotor 0ne A to D = 48
 rotor 0ne A to D = 690
 rotor 0ne A to D = 455
 rotor 0ne A to D = 289
 rotor 0ne A to D = 678
 rotor 0ne A to D = 460
Paul Lamar

Paul,

Will this be fast enough for a 3-rotor?  After all, we all aren't
flying 13B's.  ;-)

Mark S.

I don't know yet but I think it will be.
I may have a better answer today.
However it is just one approach. It is timed
injection which you don't really need in a
and aircraft engine.

Another approach is to add a high speed A to D like this
chip. We have plenty of ports to interface it to.
It does an A to D in a couple of usec. 3 rotor no problem.

Paul Lamar

Hi Paul,

It looks to me like this chip could control the fuel injection and
the ignition. It looks like with the software, you can set it up to
control as pulses go hi and low in comparison to the crank pulse
wheel, AKA RX8. Also, I will bite on what you are using the
accelerator pump for?

Dale Davies

That is what we are planning on doing. I'll add the ignition
triggers today. Real easy to do. No accelerator pump plans yet.
I don't think we need it in an aircraft engine. Tracy does not
have one.

Yes it is easy to change the CAS from positive going to negative
going. That will be done in hardware. I picked the hard parts
first to see if the overall scheme is doable.

Paul Lamar


Hi Paul,
My question about the accelerator pump was in reference to one of the photos.
It looked like a wire/cable was attached to the "lower" end of one.

Dale Davies