If it happens again, I'll have a cardboard tube ready to localize the snapping sound to either the plug or the sensor. Thanks for the info.

I remember this oscillation mode has been mentioned before but it has been a few years and I cannot remember. I have not personally experienced this.

Because the ground runs along the high tension lead the chances of the spark traveling down the sensor leads are very slim. It might be possible if you removed the ground from the high tension lead inside the ignition box and ran it to the engine case. I chased bad spark plug boot problems all last summer and the hall sensor never failed. This was an RCEXL ignition. The trouble is probably a bad boot. I had very similar symptoms. The engine would start, idle for a few seconds, and quit. Then it would be impossible to start. Pulling the plug would show that there was spark, but under compression spark would actually occur in the boot. This was with a boot I made myself. The first one I made ran for 2-3 hours. Two subsequent preplacements worked for minutes each. I replaced the ignition with one that had the 1/4-32 boot already.

The reason you didn't blow the sensor was that the spark was happening inside the boot and then to ground at the base of the plug. If there had been no path at the plug end of of the high tension lead, then the spark would jump at the sensor and take it out.

My point is that with the ground lead surrounding the high tension lead the chances of it taking the path of the sensor are pretty small. The path is right there all the time. Maybe if you peeled the shielding back far enough from the core wire you'd have problems. but you shouldn't be powering up the ignition in that condition.

TRUE

But people do all kinds of things at the flying field or at home

Typical, they pull the plug out to see if it is firing. They may or may not ground it to the crankcase. A lack of understanding is detrimental.

Greg I know you have a good understanding of the system and would never make the "no-ground" mistake but others do it all the time.

I am kinda curious how the high voltage electricity is getting from the coil back into the small low voltage wires of the Hall Effect sensor. I've had mine apart and can't see the electrical path it would take in doing this. What I have found, in most cases of failing to properly provide a path to ground at the end of the plug lead, is the spark will jump inside the coil burning the thin insulation on the windings making them fail and then arcing will begin happening inside the coil even with the plug connected. In the case of the RcExl, it's ruined, in the case of the CH it can be taken apart and the spark coil can be replaced. Been there done that. If someone can help me understand the path back to the Hall Effect it would be appreciated.

The spontaneous and continuous snapping sound occured again this morning. I used a tube to localize the sound and it is coming from the plug area. Regardless, the engine seems to run. I have in the past been guilty of sparking the plug without it being screwed into an engine, but that plane continues to run and fly well. Could someone explain how grounding the plug to the engine instead of just the plug's own metal housing is any safer? The engine isn't grounded any better than anything else. It just has more mass. The return path is still through the hi tension shroud, right??

Ran 1/2 a tank thru the engine...at idle. It quit once, but restarted easily. RPM was around 1800, cylinder head temperature right after stopping was around 260 degs F on the intake side of the head, and around 310 degs F close to the exhaust stack.

That's the thing, you never need to ground it to the case if it has the supplied grounded boot on it. If you take the grounded boot off then it's up to you to ground the braided cable to the engine case and the same when testing plugs. If you are removing this you should have some idea what you're doing. The manual says nothing about it, but everyone knows this.

I agree with Jezmo. But, I only have what I've read here. They also say never to power up an ignition without the sensor connected. Been there, done that, nothing happened.

I should be safe then, the way I was testing the ignition. I have the braided shield soldered to the metal shroud. The sparkplug is installed in the shroud, contacting both the hot lead and the hex ground. The engine is not part of this setup. Still should be ok though, right?

I wonder if I damaged the RCEXL (or the sensor, as suggested by W8ye) when the connection was intermittent in the engine. Remember, I said I had to add a second spring to make the hot lead contact the plug reliably? That means sometimes the rig was firing with essentially no plug installed. I don't know if the spark was finding another path to ground instead of the plug. Maybe I damaged the electronics, and now it fires repeadedly, even with no sensor trigger. If this is the case, it would be too bad because I'm sure that 1/4-32 shroud comes off a lot of guy's engines out there. It doesn't lock on very well.

To check your hall effect sensor on the RCEXL type ignition

If you are in a quiet area, - If you hear the ignition spark when you turn it on, and no spark any other time, the sensor is shorted.

In this test, you can also remove the plug from the engine and connect it properly in the spark plug boot with a ground wire from the threads of the spark plug to tio the engine. In this case you should see the plug spark once when you first turn the ignition on even without turning the engine over. The plug will not fire as a result of bringing the magnet by the sensor. Then your sensor is shorted.

Jim, mine sparks one time when I first turn it on and then afterwards only when the magnet passes "out" from under the Hall Effect sensor. I was told this is the norm. My CH Syncrosparks do it as well. All of my ignitions run fine.

From the way he describes it there does sound like something may be wrong with his Hall Effect Sensor. It certainly shouldn't continue to spark with no rotation.

It must not be shorted then, because despite the "oscillating" (sparking repeatedly upon switching on RCEXL) the engine does run. Have you ever used a grill with a battery powered starter? Not the piezo type. Hold the button down and you get a series of sparks until you let off the button. THAT is what this sounds like, at the plug. I have another sensor I could plug in and see if the same thing happens, after making sure the plug is contacted.

In your "out of the engine" version of the test, I'm still confused by the need for the external added wire you mentioned, going from the plug threads to the engine. Isn't the shroud-shield-ground touching the plug hex sufficient? The engine is floating electrically, so what does it contribute? The sensor isn't grounded to the engine either, so that can't be it. Is that wire just extra "insurance"?

Is this a quote from the manual? This is a procedure to test for a failed sensor, nothing else.

That is correct

The CH 1/4-32 plug boots have not come loose in my use. What kind of battery are you using? Are you using an electronic ignition kill switch? It's rare to see an intermittent hall sensor. They either work or they don't. Unfortunately, I don't know about the self sparking.

If the base of the plug is not electrically connect to the engine, you are going to have trouble.

You can have the plug tightly installed in the head and run the engine with the spark plug wire properlay attached to the spark plug.

But if you are going run the ignition, the ignition spark wire must be attached properly to the spark plug and the spark plug must be either screwed tightly into the cylinder head or you must have a ground wire attached from the engine to the base of the spark plug.

Otherwise, do not have electrical power going to your ignition

The spark plug wire of an RCEXL or CH consists of two wires in reality. The ground is the braided wire around the high tension lead. This ground is supplied connected to the spark plug boot. The crankcase does not come into play unless you disconnect the metal boot from the braid.

That is indeed how it works Greg. I use my plug inserted into the metal "boot" to check and set my timing. I don't "ground" the plug to the engine "block". The is "no" connection between the engine and CDI box on any of my systems other than the one created by the shield around the high tension lead. The voltage goes out through the plug wire "core", through the plug, across the gap and returns back to the CDI through the outer metal braid or "shield".

There are several different ignition designs floating around out there and the designs vary considerably

On my CH and RCEXL ignitions, the braid on the plug wire connects at the engine end of the high tension lead only.

It is against RF shielding convention to ground a shielding braid at both ends least it become a transmitting antenna

6 volt supply, as specified on the outside of the ignition box. No kill switch. To stop the engine, I just unplug the battery.

Jim, I can't speak for sure about CH because I don't have one to look at. However, I have three ignition system boxes right here. They are Yy Ti (Victor engines), CRRC, and XYZ. On all three, a continuity check between the braid and the battery ground reads zero ohmsgrounded. Thus, as I have said, the return path should be thru the shroud-to-the-braid-to the ignition box. No engine should be necessary to test the sensor as long as the plug is correctly stuffed into the receptacle (with a braid). Many old time ignition sites sell systems with a single wire and clip going to the plug. In such a case, obviously a seperate ground wire would be necessary.

Look at the ignition housings for all these systems. Mostly, they're plastic. And if the braid were not connected at the box end, then the mfg'er would provide a ground lug, or wire to go to the engine. I don't see that on any of my systems. The plug wouldn't fire if the braid wasn't connected at the box.

As for the RF shielding theory, I've heard it both ways. Ground one end, ground both. But most antenna aren't connected to the same conductor source at both ends, right?

Now, an update on the Saito's misbehaving ignition box. More careful observation on my part has revealed that when I plug in the 6V battery to the ign box, clicking (sparking) comes from BOTH the plug and the ign box too. They come rapidly at first, and then the sparking slows down until it stops completely after about, say, 1 minute max. It reminds me of a camera flash strobe whining as it's capacitor charges. Then stops when fully charged. Same kind of rate. After the sparking stops, running the magnet by the sensor causes the plug to spark and also gives that same click from the ign. box. I've never heard ANYTHING like that from any other ign box before. This seems odd, though the engine seems to run.

I swapped this Saito's sensor for a known good one, from a Victor engine (a Yi Ta ign box). The misbehaving ign box from the Saito behaves the same way with this new good sensor. So I preliminarily conclude the sensor is good, and the ign box is at least suspect. Perhaps, as I said earlier, damaged from having been run when the plug wasn't connected all the time. We ALL agree this is bad I think.

Here's a picture of an RCEXL box, with the same electrical leads as all the other kinds have. I don't see any independend means of connecting a ground to an engine.

If you have an CH or CHEXL ignition with a metal sparkplug cap then it is self grounding when a spark plug is correctly inserted. If you have a CH unit with a rubber plug cap then you must connect the ground braid to the plug to test it.
If you fire a CH unit with out having a grounded plug in the cap then you have a 50% chance of knocking out the sensor. Lets see 50% then 50% of that and then 50% of that , 3 sparks and you have a 12.5% chance of still having a good sensor.
When you fire an ungrounded ignition the spark has to go some where, the sensor is only 1/8" wide the spark will seek the ground path through the sensor. Most transistors/IC chips do not take kindly to have 25,000v ran through them. If the spark can find another way to ground because of the internal resistance of the chip is high it will, but if it can't then its through the sensor and POOF.
Also IF your CH or CHEXL is constantly buzzing a spark when battery voltage is applied, then you have blown the SCR and your ignition is toast. Although is some rare cases to much voltage will cause the SCR to oscillate, but that usually just before it dies.

What kind of 6V supply?

You will find that battery negative is the same as the braided wire on high tension lead. Check with an ohmmeter. They are system ground. The braid is there to return the spark current. The sleeved high tension wire works like a coax cable and this keeps noise inside the sleeve.

I guess that will depend on how the components are laid out on the board and how good the insulation is in the coil. I had my RCEXL jumping a 1/4" gap in the boot without a plug while I was trying to sort my boot problems.