I have three 87As that I have used fairly hard in the 1990s.
Two of the amps are about ten+ years old and one is five+
years old. I have never had a PIN diode failure in any of them!
From the failures that I have seen in other amps, it appears
that the PIN failures most often come from system (station)
design. Not always, of course, but most often.
There are two areas that need to be studied to determine
station problems. I think the most common issue has
to do with antennas that are not DC grounded. If you take
an ohmmeter and test from the center pin to the shield of
the coax there should be a DC ground. The problem (based
on my experience with the locations of the failures in areas
of high lightning activity and most often with dipoles and other
non-DC-shorted antennas) seems to be static buildup on
an ungrounded antenna--so that when the antenna is selected
there is a large static charge that is discharged into the amp.
Of course the amp has an RF choke that drains the DC charge
on an antenna--but, if the antenna is left open-circuited and is
then switched onto the amp, the static spikes due to the time
constant of the RF choke. First thing I would do is to check
all of your antennas with an ohmmeter to make sure they
are all DC-shorts. If they are not then I would add a shorted
1/4 wave line to each--if they are mono-banders. If they are
multi-band antennas then an RF choke on each one should be
installed directly on the feedline--this prevents the charge from
ever building up--even when the antenna is not being selected.
The effect of this static discharge over time can even eat away
the insulation on a PL-259 coax connector if left ungrounded.
I have seen this on my old phased array of 160M wire
verticals that was suspended between two towers. The insides
of the connectors just "disappeared" over time. So a solid state
device like a PIN diode is really subjected to abuse when a
large static charge is introduced into the amp when an ungrounded
antenna is selected.
The second most common thing I have seen is in multiple transmitter
situations where close antennas allow an 87A in the RX mode to be
subjected to large amounts of RF from a second transmitter. This RF,
if large enough, can damage the PIN diodes (or possibly even confuse the
microprocessor to sequence improperly). Additional filtering to reduce
the coupling of RF into the 87A while in the receive mode can help
here. Again, shorted 1/4 wave stubs can provide some protection
depending on the specific relationship of the two bands in question.
Easiest thing to check in a multi-TX station is to see if you see
reflected RF on an 87A when it is not transmitting, while you are
transmitting on another band with another amp. If you see reflected
RF of 20 or 30 watts or more getting into the amp when it is in the
RX mode there is little protection to the PIN diodes. Station (system)
design is again the key element to work on.
Obviously I don't know what your station setup is--so I don't know
if these issues apply. But I have seen this same sort of problem in
a commercial station that is using multiple 87As and was having
occasional PIN diode failures and once these two issues were
addressed the failures ceased.
And my own experience with multiple 87A amps in contest service
shows (at least to me!) that the amps can perform for many years
with NO failures. I have not experienced any failures of any sort
with my three 87As--except for one instance when I did something
really STUPID! ;-)
To demonstrate the ruggedness of the PIN diodes I have pulled off
the dummy load while running 1500 watts and transmitted into an
open circuit and never seen a PIN failure. The protection circuitry
in the 87A is fast enough and good enough to protect the amp
when transmitting into an infinite VSWR. So that is why I feel that
the first two scenarios are the most likely for scenario for those
experiencing repeated PIN failures.
I'll be happy to continue the discussion either on-line or off-line if
you think I can help.
73 es gl John W0UN
BTW You don't need lightning for a large charge to build up on an
antenna. In dry climates you can get a very large charge built up on
large antennas even during clear skies. I once disconnected the
grounded balun from a large log-periodic to do some maintenance
in it. I reminded myself that the antenna was ungrounded and
to be careful. The day was cloudless and there was virtually no
wind. But in a relatively short time that log periodic built up enough
static charge to jump out and knock me to the ground even though
I was wearing rubber-soled shoes and hadn't touched the antenna.