ROTAX 912 "p-lead" shielding and grounding

[FredG]

Rotax documentation is a bit fuzzy on shielding procedures for the leads from the ignition modules to the ignition switches that, when grounded, will shut the ignition and stop the engine (called "p-leads" on conventional magneto systems).

On page 9 of Section 24-00-00 of the most current Installation Manual, Rotax says to use shielded wire and to ground the shield on both ends ("shielding braid on both ends grounded to prevent EMI").  Figure 5 on page 10 shows this grounding method.  However, the Wiring Diagram on page 19 of the same section clearly shows grounding at only the switch end and NOT on the engine end of the wire.  

Furthermore, review of the schematic in Rotax SI 912-013, "Standardization of the Ignition Unit" shows no shielding of this lead, at all.

Meanwhile, the "AeroElectric Connection" by Bob Nuckolls, has a diagram on which the lead is shielded and that the shield is grounded only on the engine side, and that to stop the engine, the ignition switch (when "off") will ground the ignition lead to the shield.  The shield is never actually grounded at the panel, only at the engine.  

It seems to me that grounding the shield at the panel and at the engine is a recipe for a ground loop exactly as Corey described in an earlier thread this week.  Shielding at only the engine side or only the ignition switch side would not result in a ground loop and would still hold the potential in the shield at ground voltage thereby controlling electromagnetic interference.  

So, does anybody know what the best way is to shield the "p-leads" for the 912 engine?

Thanks,

Fred

 

 

[Anticept]

It should only be grounded engine side. If it's going to be grounded inside as well, there probably should be a resistor placed in the connection. 100 ohms is probably sufficient.

The kind of voltage used going into the aircraft ignition switch is pretty small. It's grounding out the primary side.

Grounding at both ends is for high frequency applications. I don't think our application quite falls into high frequency. Gigabit ethernet, for example, would use shielding grounded at both ends if it's using any shielding at all.

I would say if one gives you noise and the other doesn't, then go with the second configuration.

EDIT: Fred: here's a good read on this. https://interferencetechnology.com/simple-method-for-predicting-a-cable-shielding-factor-based-on-transfer-impedance/

[FredG]

Corey, thanks for your guidance!

Fred