This new GM coil 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 MSD 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 an MSD 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 it 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.
snips...
Paul Lamar
snips.....
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.
Paul Lamar
All this time and work to find a solution to cool a defectively designed
coils; plus projecting that all coils have a cooling issue when it seems
to my simple mind; using the D580 coil which has had none of those
problems (at least that I've heard of) would solve it.
Brian
Well not yet. Remember the problem took 30,000 miles to show up
in a car that averages only 3,000 RPM. The fact of the matter is
the D580 was designed to fire half as often as we are asking it
to do. It cost more to design something that is over kill by
100%. There is no evidence the bean counters will let them get
away with it. As I said, do the numbers and you will find
the 580 is near full on current all the time. The RX8 coil
was designed to fire twice as often as a rotor fires once
per rev while a piston once every two revs.
The primary resistance is .57 ohms. If the current is on all
the time and E = I x R that is 21 amps. I don't know about
you but that is a scary number to me. :) The power is
E x I which is 252 watts. Next time you get a chance grab
a 250 watt light bulb when it is on :) I am building heat sinks
for mine.
Paul Lamar
But remember the D580 has an iron core extending outside; a natural heat
sink is it not?
Brian
To a certain extent but not 250 watts worth.
The LS2 coil does not do that. It has a small heat sink
but remember it fires only half as often in a piston engine
then it is being called on to fire in a rotary.
Paul Lamar
Paul,
The LS2 coils I have look like the heat sink is removable. If that is
the case, then we could remove the small heat sinks and replace them
with larger ones.
My LS1 coils installed on my 3-rotor worked fine for almost 200 hrs...
no failures. They were installed low in the cowl.
Mark
I think that small LS2 heat sink is in physical contact with the IGBT
with some sort of clip or screw. Attempting to remove it would
destroy that contact. Here is a 2009 data sheet on an IGBT intended
for ignitions thanks to Mr. csmal. Note max junction temp is 175 C or 375 F.
Also note max current is 21 amps which coincidentally I calculated
in a recent post above :) That translates to 250 watts.
I think there is a reason the small aluminum heat sink is connected to
the steel shield. I would not modify it.
It is interesting that GM is in the fourth generation of
transistor-in-coil design. IMHO they are still perfecting it :)
Mazda is falling behind IMHO. Only 3rd gen and up to 2007 or 2008.
The last gen GM (I will call the LS3) looks a lot like the RX8 coil :)
Paul Lamar
--
The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site.
http://www.rotaryeng.net
Youtube key word PaulLamar2
Copyright 1998-2010 All world wide rights reserved.