It is pretty easy to do your own EFI with a Maximite. The easiest
hardware control computer in the world to program.

Not a very steep learning curve. I can help with out a lot of time
or trouble and you would be doing all 1000 people on here  a big
favour as it would be open source. You could use all six injectors
for smother running.

This is basic'ly all there is to the program. Pardon the pun. A '
before the text is a comment and does not enter into the
functioning. EFI-program.jpg attached.


Paul Lamar



I will definitely have a lot of catching up to do!  So the
Maximite will handle both the EFI and ignition coils?  The
programming is beyond me, but my brother might also help.  I assume
I could test the Maximite on my RX-8?

I will look for a source unless you have one handy.

Scott Gettings.

CircuitGizmo sells them in the states.

http://circuitgizmos.com/

Version 2 is now out with higher performance.

The web site does not do it justice. Some times I think CG doesn’t
understand just exactly what it does. It is like a complete DOS
computer running QB with a solid state 64Gig hard disk.

"The CGMMSTICK and CGCOLORMAX have the Maximite BASIC (MMBasic)
interpreter programmed into the PIC32 microcontroller. "

Not only is the BASIC interpreter programmed into the PIC32
microcontroller the programming editor and a disk operating system
is also built in to the microcontroller.

The Maxm+doc jpg shows a Maximite set up next to another computer
running the PIC BASIC pdf doc for rapid access to any BASIC command
info. This is a major breakthrough in the ease of programming micro
controllers.


For $57 it is an extreme bargain.

BTW

Harbor Freight sells a low cost one cylinder piston engine. Perfect
for getting your feet wet in EFI.

Once you get that to work expanding the number of injectors is a
piece of cake.

Paul Lamar


I forgot to mention something about the Maximite. When I mention
operating system I take for granted what that means. That could be
a mistake.

What the built in OS in the PIC chip does  is store hundreds of
different programs on the PC board in the 64 Gig  SD card and
thousands of mixture maps that can be accessed in micro seconds.

That makes the Maximite ideal for EFI. One can share mixture maps
down loaded from here and try dozens in one's EFI program.

If a bug is discovered in a new EFI program  another EFI program
can be loaded and ran in usecs.

This is totally unique to hardware control micro computers.

This is far far superior to any EFI system. Even Motec let alone
Tracy's.

Amazing !!!!!!!

Paul Lamar


"If a bug is discovered in a new EFI program  another EFI program
can be loaded and ran in usecs."


I hope that I am not the only one to see a flaw in this approach.
My airplane has a best glide speed of 90 knots. It also has a
vertical speed of -700 feet per minute in this glide. So, 1.5 miles
per minute forward and 700 feet down. So, if my engine stops at
10,000 feet I have at most 14 minutes and 21 miles to find a
landing spot. If it happens at 2000 feet I have less than 3
minutes. If my engine quits I will not be ciphering through
multiple EFI programs looking for that one that might fix my
predicament, I will be looking for a good place to land, one that
I hopefully can walk away from.

My approach is different from this one- regardless of the language
you use to write the program you need redundancy. My approach at
the moment is 3 ARM boards/processors and dual sensors. Each of 2
boards reads both sets of sensors and compares the readings. First
check. Then these 2 boards calculate the outputs. The third board
reads this information and compares the info and will reject
anything that is way off from what is expected based on averaging
totals and tolerances that are kept in memory. Second check.

The third board outputs to the drivers, while the other boards
check this output and compare it to what they calculated. Third
check. If the third board isn't producing an output within the
parameters of each of the first 2 boards a notification is sent and
the pilot can then select which computer to land with. The notion
that in the event of an engine issue related to the EFI the pilot
would be changing programs or adjusting code is nonsense. All of
this should be worked out long before you leave the ground, and all
that would be left to chance would be an actual failure, not an
adjustment. Kevin Alderman

Of course one has to be prepared with a program name that is known
to work.

That should take all of 30 seconds. Of course one would not do that
at low altitude.

Take off with the known good old program test the new one at 10,000
feet.

Mikro Electronica told me they are implementing the same thing in
their very powerful ARM microcomputers. A sea change in programming
and ease of use. Talk about I/O. They are swimming in I/O :)



Paul Lamar



Paul, how long does it take from power on until the program starts
working- meaning does it "boot up" before it works.

Kevin Alderman

The same time it takes to get a command prompt on windows 7.
In a blink of the eye. Compared to Windows loading off a hard
disk it happens instantaneously. As I recall DOS is only several K.

Same for the EFI programs once DOS is loaded in the Maximite.

Any multi tasking OS is bloat ware compared to DOS.

Years ago I had the second computer store in the world.
We started with Apple II's. My first Apple II was a mother board
only. No case, no power supply, no key board let alone
a monitor. The OS executed out  of room instantaneously.

That was back in the days of one megahertz processors.
Now the preprocessors one or more  GIGI hertz.
A thousand time faster.

The Apple II OS was written in assembly language.
The interpreted Maxmite executes interpreted BASIC  instruction
in about the same time as a 6502 does in assembly language.

Read this IEEE pdf. I wrote a multitasking OS in 6502 asm with
22 different programs in 4K or ram.

The Super KIM  mother board I designed in the early 70s was the father
 of the ARM processor. It was  built with discrete IC's.

The Phd's at Los Almos wrote the IEEE paper but did not do any hardware
design or programming. I did all of that.  All they did was convert the hexadecimal
source code I provided  to 6502 assembly language that ran on the Cray 1 6502 cross
assembler  they had. .

As far as I know this was the very first embedded processor system documented.
It took me a year to write it in hexadecimal

http://www.rotaryeng.net/RT1.pdf

Paul Lamar


The Rotary Engine News Letter. Powered by Linux. If you want off
the list PLEASE let me know and I will remove  you.  ACRE NL web
site. http://www.rotaryeng.net You Tube http://tinyurl.com/beqqxas
Copyright 1998-2015 All world wide rights reserved.