Interpreting Scope Waveform

[lothian]

2000 Honda Accord EX 3.0L V6; 110k miles

Troubleshooting P03xx codes, the technician scoped the voltage of each OEM ignition coil (Honda p/n:30520-P8E-S01) in sequence. Strictly for comparative purposes, the technician replaced the original coils with compatible known-good coils and re-scoped. Result: an obvious difference in the scope output between the six original coils and the “known-good” (used) coils. The technician cannot explain the significance of the information, and that’s why I’m posting here. I also saw the output. The scope reads as follows–hopefully my description of the waveform output makes sense:

Scope of OEM coils (active codes P1399, P0300, xx301, 3, 5 and 6):
Square waveform reading on scope. All coils show charge/discharge cycle is within mfgr. spec. voltage and millisec intervals: coil reads at zero voltage (horizontal baseline), charges to nominal voltage immediately at computer signal (left-most vertical line of square wave), holds voltage for nominal time in milliseconds (horizontal “top of square” line), then discharges to random negative value (right-most vertical line of square wave). This waveform is consistent at each original coil.

Scope of ‘known-good’ coils (after code reset):
Square waveform reading on scope. All coils show charge/discharge cycle is within mfgr. spec. voltage and millisec intervals: coil reads at zero voltage (horizontal baseline), charges to nominal voltage immediately at computer signal (left-most vertical line of square wave), holds voltage for nominal time in milliseconds (horizontal “top of square” line), then discharges to zero (right-most vertical line of square wave). This waveform is consistent at each known-good coil. No P-codes result.

The technician cannot explain:
– the significance coils that appear to discharge to negative voltage (DCNV);
– whether or not DCNV is normal behavior;
– why DCNV behavior would apparently cause P1399/P03xx codes;
– why every OEM coil shows the same waveform pattern on the scope.
Current theory: failing capacitor in each coil.

The technician left the known-good coils in for a two-week evaluation. Day three as of this post–so far, so good.

I welcome anyone with knowledge in this area to post their thoughts and opinions…

Topic

[Gumpy Guss]

The way a coil ( actually, a transformer ) works is that the voltage turns on, that starts a magnetic field building in the coil primary circuit. When the driving voltage stops, the current and magnetic field drop to zero. The rate of change in the magnetic field is what induces a voltage in the secondary coil. The secondary coil has lots of turns, and you get a certain numbre of volts per turn, which adds up to many thousands of volts.

Now there’s a complication– a coil cannot in the long run supply DC, so any positive spike to say 20,000 volts has to be balanced out by an equal and opposite amount of power on the negative side. Actually, it’s the area under the curve that has to be the same, so say a 1 millisecond spike to +20,000 volts can be balanced by a 20 millisecond wave to -1,000 volts.

Now some driving circuits can tolerate this backswing voltage, but some cannot, so some ignition coils have an internal diode or on older ones, a capacitor. The diode is used to conduct the backswing to ground so you don’t get a largish negative backswing going back to the ignition module. A capacitor is a somewhat lower-tech solution, it flattens the backswing out in time, which tamps down its voltage to keep the area under the curve the same. These capacitors are under a lot of stress from these strong pulses, so they have been known to fail open.

So without knowing what the ignition modules are designed to tolerate, we don’t know if the coils with the negative backswing are working as designed, or have failing capacitors, or failed diodes.

[Kevin Criswell]

Is there anyway you can get a capture of the waveforms and upload them here?

[twiggy]

[quote=”grg88″ post=75267]The way a coil ( actually, a transformer ) works is that the voltage turns on, that starts a magnetic field building in the coil primary circuit. When the driving voltage stops, the current and magnetic field drop to zero. The rate of change in the magnetic field is what induces a voltage in the secondary coil. The secondary coil has lots of turns, and you get a certain numbre of volts per turn, which adds up to many thousands of volts.

Now there’s a complication– a coil cannot in the long run supply DC, so any positive spike to say 20,000 volts has to be balanced out by an equal and opposite amount of power on the negative side. Actually, it’s the area under the curve that has to be the same, so say a 1 millisecond spike to +20,000 volts can be balanced by a 20 millisecond wave to -1,000 volts.

Now some driving circuits can tolerate this backswing voltage, but some cannot, so some ignition coils have an internal diode or on older ones, a capacitor. The diode is used to conduct the backswing to ground so you don’t get a largish negative backswing going back to the ignition module. A capacitor is a somewhat lower-tech solution, it flattens the backswing out in time, which tamps down its voltage to keep the area under the curve the same. These capacitors are under a lot of stress from these strong pulses, so they have been known to fail open.

So without knowing what the ignition modules are designed to tolerate, we don’t know if the coils with the negative backswing are working as designed, or have failing capacitors, or failed diodes.[/quote]

Your explanation is spot on..

[lothian]

Thank you for the explanation of ignition coil theory of operation. I actually intended to specifically request that very thing in my original post, but I plain forgot.

Your explanation, combined with scope output I cited, I should infer 1) that the original coils (circa 2000) contain caps; AND, 2) each cap within each ignition coil may (coincidentally) be failing (by spiking to negative). Am I correct?

I’m tempted to replace this set with new; however, I’m not convinced the problem is isolated to aging ignition coils. I’m in the midst of the technician’s test using a set of his ‘known-good’ coils. After three days and ~100 miles, I get the same series of P-codes. (We scoped his coils for comparative purposes; the square waveform showed a discharge to zero–never into negative voltage).

I don’t feel like I’m making progress finding this gremlin.

[Gumpy Guss]

If you are getting the same error codes, that tells you that the coils are not the problem.

The negative backswing you see on the scope is probably not significant. Coils will have different backswings depending on whether they have capacitors or damper diodes. The negative backswing isn’t important, the important part is to have the strong positive spark pulse.

I would start looking at other causes for the random misfire codes, like old plugs, air leaks, and the others.

[lothian]

“The negative backswing isn’t important, the important part is to have the strong positive spark pulse.”

Exactly the information I was looking for. Thank you for validating my suspicion.

When the mechanic showed me the disparate waveforms between the original and “known-good” coils, he focused on the negative backswing as a possible cause of the misfire codes. When I pressed him about the significance of that negative backswing, he cited the “failing cap” theory. However, he would not commit to the coils being the definitive source of the misfire codes–and so, he moved to the next t-shooting step of swapping coil sets for comparative performance; iow, see if the car runs ok without throwing any misfire codes. So, from my perspective, the mechanic seems to be on track with his step-by-step process of eliminating variables.

Your comment concerning “…old plugs, air leaks, and the others” aligns with t-shooting work I performed months before I presented the car to the mechanic.

In Jan. 2013, I replaced the plugs and the injectors; I [color=#0000ff]carb-cleaner method[/color]‘. I must mention at this point that I did this work as part of t-shooting a different issue (read my thread on it in this forum by clicking [color=#0000ff”>here) The stalling problem went dormant for a few months, then returned accompanied by the P-codes mentioned previously–which was new–and I decided it was time to let a Honda tech investigate the problem… which brings us to this point.

Your recommendation to “…start looking at other causes for the random misfire codes…” aligns with the Honda tech’s opinion. He suggests that, after eliminating plugs, coils, etc., as the source of the misfire codes, we should next check th’ valves–and adjust, if needed. I am reluctant to go this route, for apart from the expense, I cannot understand a direct connection between the valve play and misfire codes… yet.

My goal is to ferret out and resolve this problem with the misfire codes, and then resume the search for a solution to the stalling problem. While the results of the “known-good” test seems to rule out the coils as the source of the misfire, it sure doesn’t seem like I’m making much progress.

[Author]

Replies