Got the Aussie system set up on the bench. The power   supply
does
not appear to be oscillating so far. No  current drain  :(

Paul Lamar


Paul

On the Aussie  ign there is  probably a 2nd connection
for +12V switched that needs to be  powered  in addition to the main +
input.

regards Robert

I had  the  trigger signal connected to the wrong terminals.
The PS osc does not   run when the engine is stopped.

This thing is AWESOME!!!!. My wimpy  12  volt power supply dropped
off to seven volts at 3 amps and still  the spark  is jumping
1/4 of an inch at 250 sparks per second or  15,000 RPM for
a  one rotor. 7500 RPM for a two rotor It starts to  miss
about 16,200 RPM for  a one rotor but that could be
the power  supply. Only 21 watts input power.  The stock 2nd
gen coil is running  cool as a cucumber.

I have a USB  router with a 2 foot cable on my  notebook
computer in close proximity   and Windows is going crazy  thinking
I am plugging and un plugging  USB  devices. Linux is  too smart
for that :)

I think Aden Lindsay sent  this to me  from Perth years ago
and I am just getting around to testing it.   Thank you Aden.
This is the system we will copy. Only two for each  spark  plug
might work instead of four I had first  thought.

Next thing is  to get the bigger power supply hooked  up.

Paul  Lamar


----------------------------------------------------
Nice  photos!

I have to caution you that operating the CDS system on low  voltage is
risky.  The power MOSFETs that drive the primary of the  push-pull DC-DC
converter's transformer need to have a Vgs (voltage on the  gate with
respect to
the source) of 10v in order to turn fully  ON.  As the input  voltage drops,
the power dissipation in the  MOSFETs will increase, due to the  increased
Rds(on). (Rds=resistance  from the drain to the source  terminals.)  If the
 power
dissipation becomes excessive, the MOSFETs will   self-destruct.

You will need to look at the datasheet for the  particular MOSFETs in
question to find the Id (drain current) vs Vgs  plots.

While you are testing it, I suggest that you ensure that the  input supply
does not drop below 10v.  Even if you don't fry the  MOSFETs during testing,
you may subject them to significant and abnormal  thermal stress which will
have  an adverse effect on their service  life.

Let's talk for a moment about the gap you're using for the   spark.

It is generally accepted that to ionize a 1 meter air gap at sea  level, 3
million volts are required.
You can extrapolate downwards, and  say that it requires 3,000 volts to jump
a 1mm gap at sea level; 1mm  being roughly 39.4 mils (thousands of an
inch).

However, your engine  has a compression ratio.  I don't know what your
engine's compression  ratio is offhand, but let's just assume 12:1 for the
moment.  Since  you're testing in open air at (roughly) sea level, you will
 need
to  multiply the distance of the gap you will be using by the   compression
ratio of the engine, plus some if you wish to see what it'll do  with  worn
spark plugs.

So, if your standard spark plug gap is  39.4 mils, and you have a 12:1
compression ratio, you will need a gap of  473 mils (0.473") in order to
simulate
the spark gap while at 1  standard atmosphere.  It will require 36kV to
ionize the air in that  gap.  If you wish to test under more adverse
conditions (say, worn  spark plugs) then increase the gap by 5% to  10%.

Note that the  shape of the electrodes used for the spark gap will have an
effect.   If you use electrodes with conical points, the air will ionize
more  quickly.  If you use a cylindrical rod (like brazing rod or
brass/copper
wire) with the end chopped off cleanly, you will more  accurately  simulate
a
fresh spark plug.  If the shape is rounded  (half-spherical) you  will
better simulate a worn plug.

I  suggest that using a chuck key in your spark tower for a test is not
doing  it any favors.  Even though the test was of relatively short
duration,
you may have caused some arcing/pitting in the coil tower,  which  will
accelerate wear.  For best results, you should use an  ignition  wire of
the type
you will be using on the aircraft to  minimize wear and tear on  the coil.

Regards,
Steve  Wilke
----------------------------------------------------------------

You  must be clairvoyant Steve. I blew it last night when I was fooling with
the  power supply current limiter. Do you think it is the FET's?
I have it apart  to fix it.

The  MTP3055E TO-220, N-channel MOSFETS, (Q1,Q2), 14A  at 60V was
used and may no longer be available.
Can you suggest a better  one?

Paul Lamar



----------------------------------------------------------------
Paul,
I'm very sorry to hear that my prediction came true.  The bright side  is
that you won't be tempted to use the ignition with it's pre-stressed MOSFETs;
 simply replace both of them.

Digikey stocks those MOSFETs:
_http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=MTP3055VLF
S-ND_
(http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=MTP3055VLFS-ND)
$1.01 each; and they will ship SMALL orders via USPS 1st Class to  save you
on shipping.  Don't be surprised if they charge you sales  tax.

Rather than re-engineering the CDI at the moment (which would require a
good bit of testing), just go ahead and install the same MOSFETs.  The  3055's
aren't terrible, but there are certainly more modern MOSFETs  available.  I
just have a bit much on my plate to take on yet  another project; it would
be nice to actually finish a few of them.

Consider replacing that 1N914 with a more robust diode.  A fast 100v  1A
diode would be a good choice.  Otherwise, the 1N914 is a weak  point.

The circuit could definitely use an under volt lockout.  That kind of  thing
is normally taken care of with a supervisor IC.

Steve


I do biz with Digi Key all the time. Great service.

I had no plans on using that particular system but I will use it for
a bit of lab R & D. After I looked at the bottom it was obvious the MOSFET's
had failed before as the single side boards traces had come adrift
from the glass epoxy when somebody replaced them.

I'll look into that supervisor IC as I was thinking the same thing.
It does need to start at 8 volts or so however.

I have already redesigned the power section for a double sided PC board
as I am thinking of building my own power supply if I can find the right
transformer. There were a lot of dumb things done on the original board
including the single side. A dumb way to save money. Still a work in
progress. Not quite finished.

Paul Lamar