I've Megasquirted my L69, as you may have heard in a previous thread, and I've been having mis-fire issues like mad. Won't rev past 2000rpm. I've sent the fueling all over the place, and I know where it should be, so it's not the issue. It's definetly the ignition/timing. Now, at idle, my coil-in-cap ignition requires 4.5ms of dwell to maintain a happy motor. Any less and I get mis-fires. At about 2000rpm or so, my dwell drops(expectedly) to about 3.2ms. Unfortunatly, this causes mis-fires like mad. I've retarted the timing by 30 degrees across the board(in order to increse dwell), and it can reach 4000rpm(barley, but it gets there), however with about 10 degrees of advance at 4000(not good). Is this a coil issue?? Should I need less dwell to fire the plugs? My plug gap is set to .040", and well, I'm out of ideas. Would carbon-fouled plugs be a possible issue. I don't think they are that bad, but I'm open to ideas. I'm about to cry at this point, as it's been a month since I've started tuning this thing.

Ruling out any ignition issues, not enough dwell will cause misfires. Its the dwell that determines the spark energy from the coil. Not enough, and the pressure in the cylinder as well as the gap between the rotor and the terminal in the cap will cause a misfire. Id use whatever GM used on stock apps. It seems to work fairly well. Ill see if I can dig the tables out of one of my hacks...

Ok, here it is...

GM also has dynamic dwell. Sudden changes in RPM will cause dwell to be added in equal to the change in reference period. The reference period is the time between reference pulses. Large MAP changes will cause the dynamic dwell to be set to a maximum 3.9 msecs. This is also limited to 11.25 degrees of dwell to ensure that there isnt too much dwell when the static is added in. This provides extra spark for transitions

As for the static dwell, from what I can tell, the stock static dwell curves in an ECM are generally like whats below:

DRP > 25 msecs, dwell = .25 x DRP - .41 msec

7 <= DRP < 25, dwell = .0625 DRP + 4.7 msec

DRP < 7 msec, dwell = .5 x DRP + 1.2 msec

The DRP, for a V8, is equal to the 1/[(RPMs x 4) / 60]
----------
There are probably dwell curves for old school distributers floating around as well that are in degrees. Those can be converted to milliseconds by the following formula, again for a V8: msec dwell = DRP (in msecs) x (deg. of dwell at crank)/90

Also, how is the megasquirt set up? In the GM ECMs, the dwell is independant of the desired timing and is handled entirely within the computer itself.

The right way to do it (I know you won't be able to do this, unless you have access to some good stuff), is use a current probe and an oscilloscope, and set the dwell (tested at each voltage level), to whatever is required to get the specified current in the coil. Specs are typically in the 5.5-7.5 Amps range (remember ballast resistors? they completely limit max current to about 6A, like for points cars with tons of dwell at idle). That's if the ECU is driving the coil directly. If you're using a module that has current limiting built in, you can err on the long side, but it's not ideal to be making the module limiting too much, or it may overheat. Dwell is VERY dependant on operating voltage though (like roughly 40% more needed at 12V vs 14V, and 65% more at 10V vs 14V). Low voltages are encountered at cranking, mostly, and that's if your ECU is controlling the coils.

Coil charge time is not linear, and depends on the resistance and inductance. If you know the specs like those:
L in Henry
R in Ohm
V in Volts
t in seconds
Tau=L/R
Amps = Volts/R * (1-e^(t/Tau))

or
t = -ln(1-targetAmps/maxAmps@V)*Tau

The amount of time available for dwell is not affected by SA timing. It is affected by the number of cylinders and the engine RPM. Basically the time between cylinder firings. I'm not sure how the MS gets calibrated for dwell, but it may be that you are not supplying a minimum off time before starting the next dwell cycle. GM uses a minimum of 600 usec's of spark plug firing time.

At 4,000 RPM there is 3.75 msec between spark plug firings. This leaves 3.15 msec available for dwell.

At lower RPM's GM limits to dwell time to reduce coil & module heating and reduce energy consumption (fuel mileage).

RBob.

Yes I understand how to find the proper dwell for the coil and whatnot. Those equations are covered on the MS forums. Your right in that I don't have the proper equipment.(yet, I plan on using a voltage divider to feed the signal into the sound card on my laptop and use free 'scope software)

Off the top of my head, the "time before next charge" is 2.0 msec. It is a constant, and never modified. I'll try to decrease this. Maybe to .8 msec(800 usec)
That leaves me with 2.95 msec at 4000rpm. Still not quite enough I don't think. The coil likes 3.5 minimum, barley works at all below that.

Dwell in MS is set with a "Max Dwell" settings. There is a table of Voltage/Dwell amount modifier, but its only useful for getting good spark when starting. The code uses the "Max Dwell" setting unless less time is available(i.e. 2000rpm and up). It then shortens dwell and motor gets very mad.

At idle, I can't go less than 4.5ms otherwise it mis-fires. This is the big clue I'm finding. What I don't understand is why my coil takes so long to charge? Is it bad? Maybe a bad alternator? I haven't really peeked at it yet, but the battery is still charged after 2 tanks of gas. Plug gap is .040", but could they be fouled? Even a bad ground could probably cause it. I'll go through this stuff this weekend and report back. Thanks for your thoughts.

FWIW:
I'm not familiar with MS, but it seems to me you may want to verify the distrib connections
again before tearing entirely into the S/W.


If you accidentally reverse the coil pick-up wires where they attach to the 7-pin HEI module
it causes the timing reference point to drift around at higher engine RPM's. This causes the
engine to run like crap in general.


Also, IIRC the GM HEI module GROUNDS the reference signal at the timing point (e.g. 6*BTDC).
Depending on which edge of the reference signal your S/W is set up to read (e.g. it wants a
rising edge instead of a falling edge), you can throw your base timing off by ~30* AND the
computer will not see enough time between DRP's to charge the coil sufficiently.

Was there not a settting in MS to pick either a rising or falling edge signal? Seems to me I
read that somewhere...


Just a couple of hardware things to look for. HTH

Yup, there is a setting for that. I'm using the "rising edge" scheme. I'm not entirly sure if the MS inverts the signal in the VR circuitry or not. Currently asking in a thread on their board. I know it inverts it with the Hall circuitry.

Haven't tried the other way, though. I'll add that to the list.

Just wanted to credit Kaiser for catching the fact I was triggering on the wrong edge of the tach signal.

After correcting this, it worked much better, only to discover I had a MASSIZE vaccum leak, thanks to the aluminum manifold and some stripped threads. A heli-coil should be all I need to get this thing going.

Thanks again guys, you've never failed to pull my butt out of the fryer.

Pizza guy what did you end up doing to fix this did you have to rewire sompthing or just change the spark inverted output? I had similar problems with mine untile I temporarly gave up and am now using a mechanical small cap dizzy and running fuel only.