So, Ive been having issues with my ignition system, well, for as long as I can remember. More detailed discussion here: https://www.thirdgen.org/forums/elec...-ignition.html

The current one is a vortec ignition with the crab distributor cap and PCM controlled remote coil. The thing loves to eat caps, and blows holes through the plastic in between the terminals. The coil also looks like a plasma globe at night with all the corona discharge.

So, I made a simple circiut to measure the peak primary voltage. Basically two 1kV diodes in series and a high voltage mylar 0.1 µF cap to act as a rectifier. Tapped the "AC" side into the coil primary terminals, and the "DC" side to my DVM. With the car running, the primary voltage never goes below 400V, and at high RPMs, it goes up beyond 1000V, maxing out my meter.

I also tested the coil (stock A/C Delco) with a function generator, and on simple low voltage sine AC, it gives the advertised 100:1 step-up ratio. If the same applies here, then my ignition voltage is going up to beyond 100,000V. Seems a bit excessive, to say the least.

First question is: does the coil act as a simple auto-transformer during inductive flyback? Second: if yes, then why is the voltage sailing out past the clouds, and into the stratosphere?

 

I'm not really picturing how you're measuring the primary voltage.

 

Hopefully this quick schematic simplifies things. Basically its a half-wave rectifier that allows the capacitor to charge to the peak coil voltage on the primary side. The excess voltage leaks back through the volt meter (which is 1 MΩ of resistance) as well as the diode, so its not perfect, but it should capture the peak voltage on the primary side.

Edit: The diodes pointing the wrong way in the doodle, but you get the idea of what its supposed to do. Charge to the max voltage thats developed during inductive flyback.

 

Well it's a crude peak detector so it'll be a crude measurement. Is your diode fast enough? How good is your DMM?

I'd just use a scope.

 

Buh, buh, buh..... so you're measuring basically the peak self-inductance inside the coil's primary windings? I'm no electronic's expert but I don't know if that translates directly to output voltage. The fact that it's doing a moon-shot as you increase RPMs is weird.

I wouldn't know anything about diode speed or anything like that. I'll defer to others more knowlegable than myself.

Is there any chance that the built-in variable dwell circuitry is messing with your readings? It limits out around 2500 RPMs and stays constant above that point. Not sure if your readings are showing a similar curve or if they just keep going up and up.

 

This is a quote from the other thread:

To me that is a lot of gap. And can explain why the caps aren't lasting as that spark is also heating up the area of the cap between the posts.

I would check that the star wheel on the shaft is properly located, it may have slipped or has been assembled incorrectly. On the earlier small cap distributors the reluctor has an index pin to the body. Something else to check.

RBob.

 

Yes, I agree that it probably doesnt directly translate to output voltage as the core saturation threshold and inductance both probably reduce the ideal step-up ratio to something less than 100:1. How much less, Im not sure. But the peak voltage is important here, as that will be what causes the insulation to break down. Once theres a path, then the higher current can leak through and burn a carbon trace. On two of the caps, it burned a dead short between two terminals as soon as I turned the key.

Its a crude "order of magnitude" reading only using a standard switching diode and capacitor to get and "hold" the peak the voltage, but it does indicate somethings up.

A scope would be better, but I wanted to get a feel for whats there before sticking some expensive piece of equipment there and unkowingly burning it out.

 

This is a whole 'nother animal . Its a 96-01 vortec ignition system. Theres nothing in the distributor but a hall effect and a shutter to act as a cam position sensor. The cap also has a HUGE terminal base circle to accomodate the horizontal posts. Its this one here:



The other thing is that these ignition modules arent modules, theyre amplifiers. The old HEI modules had zener diodes in them to suppress the maximum voltage that could be developed by the coil AFAIK. The vortec module just has a pre-amp, and a large high current transistor. No ballast or anything like that, so it can run wide open. The intent was to have the computer limit the voltage by limiting the spark advance. The computer does clamp down on the advance, but obviously not enough.

There is one nice thing, the PCM calculates the position of the distributor, and allows it to be located within ± 22.5º of #1 TDC. My thought is to simply advance it 20º. That will close the gap quite a bit, and maybe limit the runaway voltage.

I did check to be sure nothings slipped, and it hasnt as far as I can tell, but to be sure I can get the actual position of the rotor from the PCMs datastream since it tracks its position relative to the crank using the cam sensor.

 

Yes of course, forgot all about how the Vortec's are set up. Has the crank trigger under the cover instead of the wheel/reluctor set up in the distributor.

I agree, in this case it would be that the base timing of the distributor is the key. It needs to be set to 'line up' with the crank trigger. Otherwise the rotor is too far from the posts.

Usually, the zener in the ICM is to protect the drive transistor on the primary side. Otherwise they don't last, even with a 1.5 KV transistor (BTDT). There should be current limiting, which in the small cap pre-Vortec ICMs is about 5 amps.

RBob.

 

I always wondered if the constant failures had with those early HEI ones where due to the zeners failing first?

One thing with these is that so far, Ive never had one fail, even in this situation of way too much voltage.

I would assume that the max current in this case is no more than 7 or 8 amps. The coil primary resistance is very low, on the order of 0.5Ω, so its limited by the module.

 

I gave some more thought to what the primary reading actually represents. Since the primary resistance (~.5Ω) and the secondary resistance (5.5kΩ) are so far apart, the coil probably behaves more like an iron core inductor rather than an auto transformer on the primary side. So even a dead short on the secondary would only reduce the flyback voltage on the primary modestly. I think what a 1 kV voltage reading means is that its not making 100,000 volts, but its capable of making 100,000 volts. In reality, the voltage is limited by the break-downs in the gaps along the way.

One way to test it is to lock the timing out at 0º. If it changes a lot, then the two are related. If not much change, then no real conclusions can be drawn from the primary voltage, other than what the max output can be.

 

I ran it again today with no advance whatsoever, so the rotor and terminal where lined up when the coil was firing, and as I suspected in my last post, no change in primary voltage, so during inductive flyback it behaves entirely as an inductor, and the secondary is just along for the ride. So, the primary and secondary voltages are only related insofar as the primary dictates what the max output of the secondary can be. IOW, pull the coil wire, and it will make close to 100kV.

Hmmm...

I suppose its possible to measure the secondary voltage directly with a divider circuit and a scope but thats kind of dangerous. Not quite ready for that last ride in a hearse yet.

 

I don't think your peak detector is working too well. Use a dual-trace scope and check both the primary and secondary. Use a non-contact probe on the secondary instead of a voltage divider unless you like buying new scopes.

 

Would a non-contact type of probe give anything other than a qualitative plot of the voltage? I dont think you can make any definite measurements of the actual voltage, can you?

Fluke makes 40kV AC/DC probes. Im thinking of trying one of those with a somewhat safe tap-point on a rigged coil wire and measure with a cheap chinese DVM to see what happens. If it lives, then maybe cautiously approach with a scope and see what it shows.

I was kind of hoping to be able to pull off something quick and dirty to see whats going on. The peak detector does work, but understates the voltage as the capacitor is small, and the diodes leak a bit in the reverse direction. For example, plug it into a wall socket, and it shows 112 volts rather than the 168 peak to peak or whatever it ought to show. Theres probably an expensive way to actually measure it, but at that point, 8 LS1 coils and a wire harness start to look pretty attractive.

 

You can get at least well within an order of magnitude, closer if you calibrate it against a gap.

A DMM isn't a good choice for measuring something transient, they sample too slow and filter too much.

 

I totally agree. Hook it up to a scope.

I also agree that you'll get secondary voltage rise with a large gap in your distributor; rotor>post. You need to "time" that type distributor such that the spark is coming off the trailing edge of the rotor tip at full retard, and off the leading edge at full advance. That way your rotor-tip gap is basically ABOUT the same all the time, and should be, what, <1/8" or so?

 

Oh, I know it wont work well, and at best will just capture the mean, but if it dies immediately, then its way higher than 40kV. It would be nice if they made ones that go higher, maybe they do.

With this, though, the voltage across the gap in the cap, which I know with a fairly high degree of certainty with the measurements of the dimensions and the advance, are very close to the max rating of the plastic (150 mils of Rynite, which is 50kV), so its a real tight margin. Utlimately, I need a way to measure it to within 10 percent if Im to draw any real conclusions.

The indirect way is a good idea, but that needs to be well isolated too if its to be hooked up to a scope. Even the wires themselves give off some corona discharge. Not sure how much that amounts to, but that could easily take things out, too.

Edit: as far as the divider goes, it is a practical way to measure the voltage if its done right, has very good isolation, and has some means of limiting the max voltage at the output. Another stringent requirement is a very good ground to ensure it works properly.

 

Thats what I plan to do, move the distributor to about 20 deg BDTC, which will split the difference between the timing at light load, and the retard during DFCO.

The peak detector does detect the peak voltage, or at least somewhat close, but thats all it does. The scope would be good for using the dwell tables to limit the voltage, though. Thats another way to control it. If it only has 500-600V to work with then at least it will let me know its got a problem by some other means than just blowing out the cap. The dwell tables are more or less the same as the ones in the old HEI computers. I dont think they put too much thought into them.

 

Wow TGO top techs while u guys are on the subjects not to be rude if I am very sorry just need help very bad on my ignition system. I have a vid too. I have to advance it like 20* to get it to do this (counter clock wize)
http://static.photobucket.com/player...8-15-30-37.mp4

 

I think Im going to make a divider to measure the actual voltage. Basically a box with two well isolated pass-thru HV terminals, and something like two to three HUNDRED resistors in series embedded in epoxy to get a divider network that wont break down at any point. The only good thing about that many is that I can precisely choose the tap point that gives nearly the exact division ratio Im looking for.

Even if it doesnt help fix it, it will still be a neat experiment...

 

Hey- a thought here- check your rotor phasing. In fact, I think this might be the whole issue. You know how to do that, right? I mean you have enough caps laying around at this point you can cut a hole in one to shoot your timing gun through to check rotor position when the spark actually fires.

Here's my thinking...... even since the days of the old HEI the rotor isn't directly in MECHANICAL phase with the plug terminals when the mag pickup points are lined up. There is a slight offset for a number of reasons including partial compensation for vacuum advance base timing orientation.

Now you've got a distributor that runs 100% on computer adjustment. Effectively the timing changes are all done in a way that GUARANTEES rotor position will be different at every different amount of advance.

You see where I'm going with this. The combination of that timing adjustment method and the fact you never really know EXACTLY what offsets the engineers are making at the mechanical level means you can't just look at how the parts line up and know what your rotor phasing is on a running engine. You need to check phasing the old fashioned way- with a timing gun and a cap you can see inside of at the rotor.

And way-off rotor phasing would DEFINITELY explain all symptoms you are experiencing.

 

Yes, I certainly agree that theres all sorts of backlash and position tolerances at work here, as well as any odd design decisions at GM when they cooked this thing up. Worth checking on for sure.

I did actually measure the timing with the PCM locked out at 0 advance, and it was pretty close to TDC (relative to the balancer/timing set marks, at any rate). At least to within a degree or two. This was done a while back, and is probably worth checking again, but the PCM can get it pretty close at a variety of RPMs as the mark barely moved. This, however does not give any indication about where the rotor is, as you say. I guess drilling a hole in front of the #1 is worth doing with the timing locked out at no advance and see where it actually is.

Another problem here is that there are so many variables. It could be too much dwell, it could be excessive inductance in these MSD wires with suppresion coils around the core of the wires, or it could just be too much gap in the distributor, mechanical issues like too much backlash in the gears, offset cap, etc...

 

Yeah, it could be a lot of things, I agree.

It would be interesting to know where the rotor is pointing when it's firing at true #1 TDC. I would THINK that would be at the bottom end of it's electroic adjustmet range (-22.5* from the zero point, according to your research on how it functions). And the rotor should be well past (CW) the plug post at that point.

But when the engine is running under typical/normal conditions the timing probably spends most of it's life in the 25-35* BTDC range which SHOULD mean the rotor will phase very close to the plug post at that amount of timing advance. Should.

 

Its certainly possible that its designed with that intent in mind. Its interesting to note, though, that the PCM limits the max advance from 30º to about 33º, depending on temperature and baro pressure. I remember seeing a large, convoluted routine in the PCMs logic that acted on the final spark advance at the dist., but could not figure it out at the time what the intent was. At this time, though, it looks like the intent was to limit how far the rotor could be from the terminal. Its interesting to note that the trucks also had very little spark advance, especially under load, and large portions of the spark table where either single digits, or negative advance. If advancing the distributor "removes" the limit, then it indicates that the code is indeed to limit the max rotor/terminal gap.

I do remember checking where the rotor pointed to with respect to the cap when I bought the dist. Theres a "8" and "6" arrow on the base pointing to the #1 cylinder for 6 and 8 cylinder engines. I recall it being more or less in line with the start of the hall effect shutter and #1 post in the cap. Also keep in mind that those values I stated are the tolerance limits, or the points beyond which the PCM throws a code.

But, punching a hole like you said and observing where "0" really is will certianly be worthwile, just to make sure. If it is as you say, then its not very promising as it means that the voltage isnt due to the excessive gap.

 

I checked it, no phasing. The rotor points directly to the #1 terminal at all times with no advance on the #1 cyl. firing. When it does have advance, the spark to the terminal is long as hell.

 

Decided to just go ahead and jump it a tooth and advance it 20 degrees. Runs WAY better now, and doesnt misfire under light load or run pukey from the timing being retarded. I have full advance now (yay!). The distributor sits cockeyed now so its a headache to re-route everything, but if it works, Im happy.

 

Cool. I think your next set of ignition parts will last a good long time now.

 

Oh, and by the way, if you want to "fine tune it" you can always rotate the distributor gear 180* on the shaft. That will change things up top by half a tooth. There are 13 teeth on the distributor gear so a 180* flip puts you exactly half way between 2 teeth. You get the idea.

 

Thanks. Hopefully they last so I can put this thing to bed already. The cap on there is one of the ones that was misfiring. I figure if it runs good on one thats compromised, then it shouldn't have any issues on a fresh one.

Thats an interesting idea about the dist. gear, wouldnt have thought of that. That might help get it a little more "stright" than it is now at its funky angle.