I'll throw a few things out there based on my experience with v-belts. We
have literally thousands of v-belts on a wide variety of equipment. Our
mechanics have become very picky about which belts they use. Sometimes we
run into a change that the manufacturer has made that causes the same brand
and part numbered belt to do strange things. We've seen things such as a
change in the composition of the rubber of the belting material causing
additional friction and "pulley climb" and changes in the number of layers
of reinforcing fabric and changes to the hardness of the belts causing
problems because it is either more or less flexible than the previous belt.

Do you still have the old belt(s)? Are the belts that were being thrown off
worn, chaffed, or stretched unusually? Polished or shiny spots on the sides
of the belt will indicate slipping. Cracks and fraying will indicate that
the belt was either overtight and/or old when installed. Do they lay flat
when you simply lay them on the floor? Twisted belts can indicate
misalignment. I'm just wondering if there was something wrong with those
specific belts. How old where they when installed? Brand and part number
identical to the latest 3 hour belt? Can you measure them to check them
against the original length and what the belt number states?

Could either pulley have slipped on the shafts? Do you have room to run a
straight edge across and between the faces of each pulley to check for
alignment?

Too tight? Too loose? Rule of thumb tension is a deflection mid way
between the pulleys of 1/2 the thickness of the belt, but check your MM.

Jim

"Viperdoc" wrote in message
...
Threw the belt again after twenty minutes in IMC on the way to the
avionics shop. Destination was VMC, so elected to continue rather than
turn around or land, and the other alternator was handling the load
easily.

Avionics guy fixed the radar (took three guys around 30 minutes). Belt
replaced.

Flew 1.5 hours back- no problems. After return I tightened belt again, and
then flew another 1.6 hours doing LNAV/VNAV approaches in IMC (much
smoother than ILS).

Checked belt again, which appeared tight. So, last belt has now gone over
three hours, where the others have all thrown in less than 30 minutes.

What gives- can a new belt stretch so much that it gets thrown that
easily, or should I still suspect that the engine is detuned and has a
sticking counterweight?

The engine shop quoted a cost of over $3,000 as a minimum to change the
counterweight bearings, plus any costs associated with finding something
wrong with the cylinders.