Magneto problem?

[N2709P]

You most likely have a Bendix ignition switch that has a recurring AD. the AD requires at every 100 hour or annual, the ignition switch it be checked for grounding out the magnetos. The best way to check this is with the engine at normal operation temperature during idle, momentary select off on the ignition switch and make sure the engine quits firing. Immediately go back to both. If you wait too long in the off position, you will get a pretty good pop. Not good for the induction system. If you wait too long in off, just let it die. Also, when checking the mags during run up, as you select “L”, “R” and both, attempt to pull the key out. If the key can be removed in any position other than in off, the ignition switch needs to be replaced. The “P” lead wire is normally a single braided wire in a shield. The shield is to help eliminate electrical noise. Current thinking with most shielded wire is to only ground one end. I grounded my shielding at the ignition switch and the shield at the terminal on the magneto is not grounded. Your loose shielding at the magneto could be that shrink tubing was not placed over the end of the shielding so that it did not fray. The shielding should not affect your ability to turn off the magneto. 500 hours or engine overhaul is about a normal time for magneto overhaul. If you can keep clean fuel and good spark to these engines they are pretty hard to stop.
N2709P

[azevedoflyer]

Thanks N2709P.
Will check as told.
Surprised by your "current thinking..." part in the text!
azevedoflyer

[N2709P]

As explained to me by my favorite electrical engineer (it takes 2 E’s to spell geek), old school thinking was to ground both ends of the shielding to build a Faraday or “noise shield”. Further experience found that in most cases , especially with GPS equipment, this is detrimental to the shielding’s purpose, as grounding both ends provides a continuous loop for “noise” to enter the system. Leaving one end of the shield not grounded still protects the shielded wire from radiation patterns and limits these radiation patterns from having a path into the system.
N2709P

[Gilbert Pierce]

All modern avionics manufacturers installation instructions that I have seen require only one end of the shield to be grounded.

[JPerkins]

And I was told to ground one end of the shielding on circuits carrying audio signals and ground both ends on circuits carrying digital data signals. Hmm, there must be a paper somewhere that explains all that and sets it strait.

[Steve Pierce]

Has anyone found a bad Bendix mag switch in recent years, I have not. Have had to clean the contacts on the old AN switches. Have found P-leads where the primary wire was contacting the shield on very old p-leads. I do quite a few 500 hour mag inspections and rarely find anything, mainly adjusting the e-gap to make up for the point wear. Most problems I see were created by Champion when they bought Slick and started changing things. I love the SureFly SIM, starts 3rd blade every time and no maintenance, look for leaks, wires aren't rubbing and all is good. No timing adjustments or 500 inspections.

[azevedoflyer]

To all that contributed, many thanks.
I stand educated about magneto shielding.
About points, I found that both are from Kelly Aerospace and not, as stated, one of them from Tempest.
Additionally, under a magnifying glass both have the movable arm (the one pushed by the cam shoe) contact crowned. The perceived angle comes from this feature.
However, while one contact is centered, the other is not. On this non-centered, the crown contacts the fixed part towards the edge, explaining the visually "greater" angle.
Difficult to take a picture! To complicate things, I do not know how to post pictures on threads!
Cheers,
Azevedoflyer

[Vagabondblues]

Has anyone found a bad Bendix mag switch in recent years, I have not. Have had to clean the contacts on the old AN switches.

Probably because the AD is only applicable to the “twist to start” switches made over 50 years ago. Most short wings still have the big starter contractor and don’t have airframe voltage going through it for the starter solenoid. Some late model Tripacers do use it however.

Bendix now has a coaxial condenser for their mags as a retrofit. The coaxial capacitor which was unique to Slick used an antenna style shielding. Designed by a former Bendix engineer while employed at Slick. The new Piper Tomahawks then had points that would fail in 10 to 15 hours. It was found that Piper installed noise filters inline with the P leads. Electrically speaking capacitors in series, double the capacitance. That is two 10 mf capacitors in series provide 20 mf. Coaxial circuits are parallel. So if you have two capacitors in parallel the capacitor is cut in half. Two 10mf is now 5 mf. The primary purpose of the capacitor is to prevent arcing of the points as they open collapsing the field at E Gap.