Puzzle: alternator vs. vibration

I am a former SR22 owner (N22GX), and now, since April, I have a Bonanza with the Atlantic Aero IO-550-R conversion, an engine that is somewhat similar to the SR22’s engine. I would appreciate any thoughts on a problem I am having now with breaking alternators, as descibed in the email from my expert A&P / IA, Dave.


A quick summary:

We have seen three alternator problems;

First: The alternator that was on the plane when you bought it had a broken inboard casting at one of the four attach lugs (lower fwd). When we removed the alternator, I found the (elastomeric) drive coupling loose
on the alternator shaft and the keyway slot worn in the coupling to such an extent that it allowed the coupling to rotate approximately 45 degrees from one end of the “slot” to the other.

Second: An overhauled alternator (from Kelly Aerospace) and an overhauled coupling (from RAM aircraft -AEC646655) was installed. After 12 hours of operation, the rotor (armature) shaft failed internal to the alternator between the fan and the inboard bearing.

Third: A replacement overhauled alternator was installed (re-using the RAM coupling). This experienced the same failure after 18 hrs

Kelly has inspected both failed units and determined that the shafts failed in a torsional mode. The rotors/shafts of the two units were not from the same manufacturing batch, and they have not seen any other similar failures recently. I discussed the installation procedure and coupling torque with Kelly and verified that I was installing it properly. (the original alternator did not have the drive coupling installed correctly, as they left the shipping washer in place between the coupling and alternator spacer/seal)

We conducted a ground vibration/balance run with the following results:

Alternator not loaded: 1/2 order .663 IPS, 1st order, .241 ips, 2nd order .306 ips, 3rd order 1.03 ips, subsequent orders all less than .020 ips.

Alternator fully loaded: 1/2 order .668 IPS, 1st order, .043 ips, 2nd order …065 ips, 3rd order 1.575 ips, subsequent orders all less than .180 ips

Bob @ TCM tech support did not seem to be too concerned with the vibration data. I also spoke with Mike from Hartzell tech support and they feel the 3rd order spike is an anomaly from blade passage during ground runs, and would not show up in flight.

So far, both Hartzell and the prop balancer manufacturer seem to feel that this is caused by dynamic counterweight pin/bushing wear (or sticking pins). Hartzell seemed to be fairly convinced that this is the likely cause. I have not received a possible cause from either Kelly or TCM (I have been playing phone tag with the TCM rep Thursday and Friday).

I am certainly open for ideas,


Thanks in advance for any suggestions.


Interesting data Arlen.

plenty of questions in my mind off the bat…

give me a call.

715 4911 303