On 10 Mar 2005 16:55:54 -0800, John wrote:

It's a bit tough to infer anything while I did this, because the b40
reading will decay down towards zero fairly fast even if the ptt is
held closed. However, it doesn't seem that wiggling the coax cable
connector has any effect on the vario spike/behavior...

Could be that the bad connection is at the antenna end. A bad
connection to the load (antenna) will cause standing waves of RF on
the coaxial cable, including its outer conductor.
An SWR measurement will give you the answer here.
I hope the problem is not at the antenna end, as it could be rather
hard to access for repair.
Cheers, John G.