The ultimate ignition system

Some have asked so I will say that I do have enough aluminum to make 6 or 7 heat sinks.  I will get them done this week.

Those interested can PM me.

I'm thinking maybe $50-60 for the heat sink, new hardware, heat sink compound and a paper template to show you where to drill the holes.

I'll install it for you for an additional $20.

DG

Well, I have driven the snawt out or it.  Yes a fair bit of 5000-6500 shifting and no problem other than some pinging.  Need to see John and get a chip done for this camshaft. 

I routinely pop the hood and feel the heat sink.  It's always about the same temperature no matter how hot the engine is. 

Our ignition problems are over, simple as that.

Capt Fiero wrote:    SO how is the test mule doing in the wifes car, been able to drive it much?    I know she probably won't rev it that high, but have you had a chance to scream the RPMS up for any prolonged sessions?  Any chance to get it hot enough for the rad fan to kick on in Traffic?  I am sure a lot of interest will be there once people know this is a bonified save your butt item.

Here's a picture of the latest batch of heat sinks in case anyone is interested.





Got a few left, don't know when I'll be making more.

Thanks.

Damien

How is this probe hooked up?...bottom of the coil to common(Ground)? It looks like that's it.

I would like to see ch1 across the coil, and ch2 hooked to a current probe (.01ohm maybe, 5A = 50mV signal at max) in the primary to show primary charging current...that would nail it.

 

I think this module is suspect.

To define what I'm talking about, the period where the waveform is high is when the ig module is open circuit. When the voltage is low, but not 0 the current is building in the coil and the module is conducting. The small voltage is due to the saturation voltage of the transistors in the module. Then, the negative going spike is when the energy in the coil primary is transfered to the secondary and a spark occurs.

 

Now, look at the timing...When it's cold, is it open circuit for 4ms, then charging for 5ms or so (This is the dwell on a points system, btw), then it discharges to the plug over 1-2ms period. Good stuff.

 

Now look at the last one, the ugly. It's o/c for about the same period, then charging for a very short period of time...1ms or so. This causes a dramatic reduction in spark energy, because the coil only has 1/5 the time to build up energy. The current builds linearly, so 1/5 the current, but the enery is proportionl to the square of the current, so something like 4% of the previous energy is delivered. Not good, and it shows in the discharge waveform, which is very low amplitude and I suspect the spikes are due to multiple crappy arcs.

 

The dwell is supposed to be fixed in this system...ie, the module is shot.

 

It probably is not strong enough to deliver a signal to a timing light.

 

Chay

I read some more via megasquirt, and their typical values for dwell are 2-4ms, depending on the L of the coil used.

 

For us, if we assume a spark discharge time of 3ms, and dwell of 4ms (Each of these is conservative) we have a total system time of 7ms for each spark. We should be able to maintain this until:

7ms*3sparks/rev = 21ms, 21ms^-1*60 = 2857rpm, at which point the ecm would lower the dwell and the resulting spark time in order to keep it firing on time.

So, we get a really good spark at up to 2800rpm, then by 3000rpm, the spark energy is about 1/4 of what it was at 2800 (Because the dwell is /2).

 

Chay

Oh, yeah, one MORE thing. The amplitude of the reverse voltage on the coil primary will be equal to the spark plug breakdown voltage*ratio of the ignition coil. Essentially, the ignition coil becomes a transformer, with the primary and secondary "disconnected". Therefore the voltage will build on the primary and secondary until one or the other exceeds the voltage level required to drive current...it lets go on the secondary when the spark plug fires.

 

This means that there is no good way of reducing this spike, because in doing so you'll be absorbing energy which ought to go to the spark plug. The focus should be on attaining a good dwell time, so the spark plug fires once strongly rather than the weak restrike I see in the last of your scope shots.

Comments?

Chay

I have to try this myself. I thought I had it all figured out, and the change in probe position has thrown me into confusion. I don't get how it works, but I'll try to figure it out with the new info.

Chay

I have a new thought which seems to fit. You can confirm it with the current probe.

The high voltage corresponds to the dwell. The dwell is long enough that the maximum peak current is reached. This is a module maximum current and protects the module and stops the coil from overheating. The value increased through the years. It started at 5 amps and unded up at 8 amps. I'm not sure where ours fit in.

Anyway, when the max current is reached, the voltage across the coil drops to the value required to maintain the current flow which is quite low because the only factor is the parisitic resistance in the coil primary. This drop from high voltage to low slightly positive voltage should be coincident with the ramping current leveling off. Then sometime later, the module interrupts the current flow and the large reverse voltage is the reflected voltage required to fire the plug on the secondary.

As the frequency increases, the amount of 'max current' time should decrease until it dissapears. Then the voltage should go from the positive directly to the negative with no intermediate level.

I'd be interested in the results of the I probe. It should be quite reveling.

I haven't had a chance to read the link. I'll do it when I get a chance.


C

The clamps might work if it was the right type. All the current to voltage types I've used add delay, so they would show the waveform correctly, but it would not allign to make sense with the voltage waveform and be conclusive. The best way to go is a good old resistor...the trouble is breaking it into the circuit of course.

Chay

Well the exact same setup is in Michelle's daily driver.  Other than that connector issue, it's been working perfectly.  If a problem of any kind pops up I will report it.

As far as testing goes.  It's going to take a while before I know exactly what good and bad really are.

If anyone wants to bring me their suspected dead or old modules so I can beat on them that would be good.  Eventually, I will learn how to tell what shape they're in.

Capt Fiero wrote:    Damian, when you do sell one, make sure to get the buyer to keep a report on how well it works.   Another interesting test would be to take some "bad" modules or pickup coils, and test them in your fancy device and see if the bad parts come back to life.