Chuck Dunlap wrote:

What waste of power? The resistors do not dissipate much - the originals are just 5 Watts resistors, and I do not think they even dissipate that much. I think the reasons there have been failures in aircraft is from high engine compartment heat. The resistors are wire wound and I think the high peak pulse current fatigues the wire. The 25 watt resistors use a larger diameter wire, not as prone to failure. My original units are still working fine after 175 hours with cool air blast. I once flew my Mazda Rv6 with EFI for over one hour with the alternator turned off (within gliding distance of the airport) and the voltage still did not drop below 12 volts. This was with a standard 33 amp hour battery.

Paul Lamar replied:

OK Chuck. So you don't think the higher average current is causing the
failure? First I have heard that high peak pulse current can fatigue wire
at about 80-100 pulses per second. Light bulbs last about 1000 to 1500 hours at
120pulses per second and the wire is incandescently hot. Of course that
is a sin wave and ignition module works off a pulse. Do you think this a thermal
fatigue phenomena? Very interesting. Got any references on this subject
of wire wound resistor failure modes?

Paul,

I haven't read all of my email, so perhaps what I am about to say doesn't apply.

We here at Texas Instruments have alot of experience with "interesting" failure modes of high current devices (we're currently the #1 supplier of analog devices in the world).  What Chuck may be referring to is a phenomenon called "electro migration" failure.  It happens in small conductors at high current.  A small pre-existing defect in the conductor causes a narrowing of the conductor at some point. As the electrons go whizzing through, that small narrowing gets made smaller as atoms are slowly shifted "downstream".  This, of course, only makes the narrow part narrower, so you can see that a failure is in the making.  I've got some photomicrographs of this failure mode.  Here is one.  The official failure mode coclusion for this device is:

     The unit failed opens testing due to metal migration caused by high current flow while the bump was at an elevated temperature. This caused high resistance in the bump and cracks.

The "bumps" are contacts from the silicon device to the chip carrier.

[Image]

That happens enough in power semiconductors that we have special chip design rules to try to prevent it.  I don't know if this phenomenon is the cause of the failures in the relatively macro scale of the wire wound resistor.

This is a SEM (scanning electron microscope) picture of what happenes when one thermally shocks a device from -65C to +150C.  The leafy stuff is silver that has migrated on the semiconductor and evetually caused one of the electrical wires to lift off the contact.

[Image]

Russell Kent

The AirCraft Rotary Engine NewsLetter.              Powered by Linux.
ACRE NL web site.
http://home.earthlink.net/~rotaryeng/            http://www.linux.org