Milo dropped off a complete running RX8 engine
mounted on trailer with Tracy's EFI. It is on loan for two years.
First thing we will do is reverse engineer Tracy's
EFI system by measuring WHEN the fuel is injected relative
to the e-shaft angle.

Paul,

Do you plan on testing the ignition system that you designed as well
as the
555 EFI?  Any time frame?

Joe Berki
Limo EZ

Yes. I have everything I need to test that. I am not sure when I'll
get to it. Here is the idea for the new guys.

These are inductive coils and they normally require a computer.
Inductive coils are turned on for a period of time and then
they fire when the current is suddenly turned off. This off time
is called dwell. If the dwell is too long the coils over heat.
If too short you don't get full spark energy.

The breakthrough on how to make this stand alone and work without
a computer just occurred to me. The trick is to slot the trigger
wheel and use a  IR photo transistor, an IR emitter LED and FET switch
instead of a magnetic pickup or Hall Effect sensor. These devices
put out a single narrow pulse. Not what we need.

When the IR photo transistor sees the infra red emissions from the
IR LED it turns the coil current on using a high voltage FET solid
state switch. In the case of the GM coil the FET is built in. I used
something like this for years on my computerized road test equipment
and it is very reliable and durable. Also tolerant of dirt and dust.

It will work with just about any stock inductive plug on coil system
including the Corvette GM coils. The length of the slot controls the
dwell time. The end of the slot controls when the coil fires.

I have been thinking about this problem ever since the stand alone
coil on plug systems first appeared. How do we get rid of
the computer?

If you want to fire the leading and trailing plugs at different
times you can use four slots with four photo TX's and four
LED's other wise each photo TX fires both leading and trailing
plug coils at the same time.

Now some of you electronic types might say "why not just
use a 555 one shot with a fixed pulse width to turn on
the current in the coil for a fixed period of time triggered
by a magnetic sensor or Hall Effect "?

The problem with that is the spark timing will vary with RPM.
You want the plugs to fire at 22 degrees TDC over a wide range of
RPM's so that approach will not work.

Basically this thing works like distributer points.

Paul Lamar

The only problem with this approach is starting.
GM uses a 5 millisecond pulse, but good spark can be obtained with 3ms.
The coil will fire 100ms after pulse leading edge if the pulse does not
end before that.  If you use a 180 degree slot (5ms @ 6000rpm), powering
up the ignition below 250 rpm will result in a prefire as the coil fires
before reaching 22 BTDC.
Once above 250 rpm, all is ok

Victor Roberts

I assume that logic is built into the coil.
180 degree slot would be 4 ms at 7500 RPM which is the
max RPM we can reach with a 2.85:1 gear box and a 72 inch
prop. So we can cut down the slot length to 135 degrees
cutting the dwell  time to 3 ms at 7500 RPM easing
the starting problem. I think.  Does that sound about right?

I will also look into some simple 555 one shot circuits that
will limit the max dwell time. Perhaps just a capacitor that
charges????

Paul Lamar

The logic is built into the coil, to protect against damage from a
faulty trigger providing a continuous"on".

Once the GM coils are "set" by a positive edge, they will fire within
the next 100ms either upon the falling edge of the trigger, the
expiration of the 100ms, or the removal of the 12 volt supply.

The solution is to inhibit the initiation of the trigger by monitoring
the cranking speed  and not allowing the trigger to reach the coil until
the rpm has reached the minimum required.

GM latches the starter relay to keep it engaged for a period of time
after the key is rotated to the START position, to make sure all coils
have fired before releasing the starter solenoid.. On my Tahoe, the
briefest turn to START and release will result in the engine cranking
for almost two seconds.

A 555  rigged as a retriggerable one-shot can be used to determin the
minimum rpm has been reached before enabling spark initiation.
Allow me to correct a mental error in the previous post, 180 degree slot
requires 300 rpm, 135 degrees requires 240 rpm.  The 250 rpm number came
from my previous work on this problem, where I had chose that as the
minimum practical rpm for a system such as you proposed.

Another fix would be to have a mechanical vane cover part of the slot
that would be moved by centrifugal force to open the full length of the
slot. If the vane masked all but 30 degrees of a 180 degree slot, it
would work down to 50 rpm.  If the mass of the vane were about the same
as the mass of the material removed to create the slot, it would serve
to balance the disk.

Instead of a slot, invert the logic and use a lobe to cover the
photocell when you want an "on", the lobe could be hinged and grow from
centrifugal force to provide a greater angular duration at higher rpm.

Victor Roberts

Interesting idea Victor. I'll draw something up in 3D.

One of the things I will do shortly is record the dwell time
of Tracy's computer through the starting process. Perhaps we can learn
something from that.

----------------------------------------------------

Here are the new 3D's.

Brilliant idea Victor. Real easy to implement. One piece
of curved stainless steel and two small SS springs and two small roll
pins
(not shown) inside the springs to act as guides. You should apply for
a patent.
We can drill large holes to balance it. The plate is .25 inches thick.

The slot changes length from 135 degrees during starting to 180
degrees at high RPM.

Anybody can understand this and fix it if necessary. No degree
in computer science required :) Cheap to build with an NC mill
or a rotary table and a manual mill.

BTW there are probably some food containers out there that will
act as a housing. I need to go shopping in the super market :)

Paul Lamar

I think the vane needs to be longer, the trigger needs to shrink to 30
degrees for cranking at 50 rpm.
Perhaps pivoting the vane on one end would work.
Victor Roberts

I can make it longer with out a pivot.

Paul Lamar