This looks very promising for our use.

   The $15,000 GM LS7 V8 is a high RPM racing version with Titanium rods.

   Apparently the GM part number "??" is one step up from
   the LS2 coils. Hopefully the dwell time will be more like
   the RX8 coil coil 3 ms dwell time.

   The coils can be mounted on a heat sink from the looks
   of them. That looks like a huge advantage to me.
   The LS2 has a small heat sink but it would be better
   if it could be mounted on a heat sink which it cannot.

   I'll have to get one to look at it and test it.

   Paul Lamar

   I just talked to MSD and they said the dwell time
   is 4 to 4.5 ms which is an improvement on the 5 ms LS2.
   Also the bottom is flat so it can be mounted on a heat
   sink with two bolts rather than the RX8 coil which only
   has one hold down bolt and would require an additional
   clamp.

   These points seem minor but believe me they
   are important. The LS2 cannot be mounted on a heat
   sink but it has a very small heat sink on it and
   we are attempting to fire it twice as often as it
   as it is designed to do and or see's in a 7500 RPM V8.
   That means it is drawing twice the current and generating
   twice the heat. The V8 cylinder fires every other rev rather
   than every rev as in a one rotor.

   The icing on the cake is these MSD coils have CD built
   in and fire multiple times at low RPM.

   Paul Lamar

   Paul, The MSD coil looks like a nice part but the price at
   performance parts.com <http://parts.com is $92.95 each or $699.95
   for a set of eight and the dwell time is still twice as long as
the RX8   coils.   I am not up to speed on the electronic circuitry
but a recent   post said you had eight RX8 coils on hand for testing
and then later   that you were down to two. I was wondering if it
the test rig which is   failing the coils. Possibly not firing the
plug under compression is   encouraging multiple sparks because the
spark energy needed is lower   just firing in air. If the intent is
for developing multiple spark   ignition then testing that should be
distinguished from standard systems   application. I have been under
the impression that Mazda coils cooled by   a blast tube are a
reliable component and that if one were to fail it is   unlikely
that the remaining three would fail on the same flight. The   coils
with an outside air blast tube would be significantly cooler than
 the underhood temperature when installed in the car.

   Have you tested a Ford waste spark coil which has no internal
   electronics since that function is in the driver (EDIS) module.
   It might be interesting to see how it stacks up with the
secondary for   the secong plug grounded firing only one plug. These
coils are reported   to work up to 8300 RPM firing once per
revolution (due to the waste   spark). If the coils are usable they
might be tested with the EDIS   module driving them first with waste
spark then with the second plug   lead grounded.

   george grimes

   If you own a $15,000 LS7 what is another $800 for coils? :)
   I am sure there is a huge retail markup. I notice Ebay has a
   lot of MSD stuff much cheaper.

   The dwell on the RX8 is 2.5 to 3 ms while the MSD is 4 to 4.5.
   Not double.

   Yes there is no question the test rig was failing the RX8 coils.

   They don't like an out in the open spark anywhere around them
   They continually re-trigger and the dwell becomes infinite.
   Then they soon over heat lickity split and fail permanently.

   I just got another 3 out of my RX8 and all were discolored.

   I am working on a heat sink solution for cheap RX8 coils
   if you don't want to use the MSD versions. Either will work
   on the heat sink. A small 1 inch square 12 volt fan
   is optional on the heat sink.

   The waste spark is problematic on the trailing coils
   of a rotary but will work on the leading coils. The 2nd gen
   RX7 engine was waste spark on the leading coils. I have not
   bothered to test the Ford stuff as I must be sure it is
   a long term solution and not based on obsolete parts if
   possible. The coil on plug seems to be the long term
   solution for the future.

   Paul Lamar

Umm, Paul, would a simple coil check, almost possibly part of a
preflight be to simply swap the plug leads between leading and trailing
plugs and run the engine up. Won't a failed coil  then reveal
itself? I guess it could be an in car test as well:)

John Evans

I think Tracy has something like that in his installation schematics.
I am still doing a heat sink mounting that will make the
coils more reliable. Silly Mazda mounted the coils on the hot
engine block with a steel plate. Merely making that plate out of
SS instead will reduce the heat transfer from block to coils.

Paul Lamar

Looks like the bean counters won that discussion, Eh? Ferrous bracket.

Dale Davies