I laid in a small supply of PIN diodes from Richardson after Paul posted
earlier about the availability. Thanks!
The PIN diodes got a bad rap when they failed in early production models,
but my original pair lasted almost 20 years. When one of them failed, it was
because the rubber insulator under L1 on the T/R board had disintegrated,
allowing the choke to arc to the PCB. Evidently the arc killed one of the
PIN diodes, but the other one was OK. I replaced the bad PIN diode, rewound
the choke and replaced the rubber pad. Haven't had a problem since. If you
an older 87A, I'd recommend carefully inspecting L1 and the area around it
for any signs of arcing, and check the integrity of the rubber pad. This can
be done by simply removing the top cover and the cover on the T/R
compartment (always unplug first and allow the interlocks to do their job!)
I haven't had a RAM/ROM chip failure, though I've heard that AlphaMax ROMs
are in such short supply that Alpha is saving them for repairs and won't
sell them to owners wishing to upgrade or repair their own amps. Is this
Not sure about availability of the CPU, but it can't be replaced in the
field. First, it has to be loaded with the 87A firmware. In theory owners
with flash equipment could do that, but the CPU also has to be programmed
with various amp-specific parameters, like the sensor levels for fault
detection and band switch positions. This requires a special test setup and
software, so the amp has to go back to the factory (FWIW, due to these
factors I don't think Array Solutions would be able to do this repair.) As
originally designed, this wouldn't have been necessary. When the earlier
amps were made, Alpha recorded the firmware settings for each amp in a file
so they could be reloaded without going through the calibration procedure.
That would make it possible for owners to obtain a CPU from the factory that
was pre-loaded with their specific amp parameters and do the replacement in
the field. But evidently all the 87A parameter files were lost at some
point, so this is no longer possible. I don't know if Alpha has parameter
files for later CPUs or 87As that it repaired, but it's worth asking.
I found all this this out when my 87A took a lightning hit that came in on
the RS232 line. It killed the MAX232 level converter and the serial I/O pin
on the CPU. Amazingly, the amp still worked because the I/O pins on the CPU
are internally fused. Only the RS232 function was killed. But I needed the
RS232 capability for remote control, so the CPU had to be replaced. Luckily
my contest club, YCCC, had previously arranged a Alpha Day in which an Alpha
tech (Brad) came to New England and fixed many Alpha amps owned by club
members, including mine. This happened shortly after the lightning hit and I
was able to drive the amp to get fixed on the same day. I wonder if Alpha
still has the program that exports the parameters from the CPU to a file and
imports/writes them to the CPU (and whether Brad recorded the parameters for
my amp.) If the program still exists, it would be nice if Alpha made it
available to 87A owners so we can record the parameters in case a CPU
replacement is needed.
Note: I'm not 100% sure where the amp firmware and parameters reside. The
AlphaMax upgrade didn't require replacing the CPU, so much of the code must
be in those ROMs. I remember seeing a post here from an owner who had a
corrupted ROM that had incorrect parameters in it and was somehow able to
reflash it. But AFAIK, the RAM/ROM chips in my amp were OK and only the CPU
had to be replaced, and I was told that this required loading it with
73, Dick WC1M
From: Paul Christensen [mailto:firstname.lastname@example.org]
Sent: Thursday, October 5, 2017 7:50 AM
Subject: Re: [Amps] Alpha
At some point in the future, 87A prices may reach a point where it's
worthwhile to own a donor unit, especially as 3CX800A7 tubes become more
expensive to purchase. I see the most problematic replacement parts being
the ROM/RAM chip sets, and Tx PIN diodes.
For those like me who will own their 87A amps until kaput, consider
purchasing an extra set of TX PIN diodes before they're gone as that part is
now obsolete by Microsemi. The original MA/COM MA4P4006D PIN diode was
subsequently renamed as the UM4006D when Microsemi took over the product
line. The "D" suffix signifies an isolated #4 stud mount. Although other
high-power PINs can be used, it's nice to take advantage of the 87A's
"thru-panel" heat sink.
Richardson RF has the PIN diodes on close-out. This spring they had a
supply of 275+ diodes. Today, stock level is 50 units and I doubt more will
be manufactured in the isolated stud-mount version. Price is only USD
$24.75 in small quantities. That's much less than original pricing. Two
UM4006D diodes are used in the 87A/89.
From: Amps [mailto:email@example.com] On Behalf Of Dick Green WC1M
Sent: Thursday, October 5, 2017 12:29 AM
To: firstname.lastname@example.org; email@example.com
Subject: Re: [Amps] Alpha
> I figured out the life of a vacuum relay (RJ1, HC1, GH1, all rated for 2
> million cycles) in QSK contesting in CQing and also S&P. When running
> (holding a freq and calling CQ) you get about 20,000 QSOs for the
> rated 2 million cycles, in S&P it's about 33,000 Qs, using my call as
example, with the CQWW contest exchange.
At first blush that number looks very low to me. After all, there are quite
a few very active contest stations here in New England that do 10,000-20,000
QSOs per year and you don't hear lots of stories about vacuum relays needing
to be changed frequently.
But I think that's because the vast majority of highly active CW contesters
don't use QSK for running. Too much noise, which gets really fatiguing when
you do 3,000-4,000 QSOs in a weekend. Besides, many contesters grew up with
amps that couldn't do QSK (meaning frame relays), so they've always used
semi break-in. Some use QSK for S&P, but even then it's not all that
necessary (if you're spending so much time in pileups that you need QSK,
The number of relay closures for semi break-in is dramatically lower -- off
the top of my head I think it must be 1/100th as many as QSK closures, maybe
Of course, relay closures have never been an issue for me because for the
past 22 years I've done virtually all of my contest running on an Alpha 87A.
And in semi break-in at that. :-) I did have to change out the PIN diodes
once, but that had little to do with cycles -- just a L1 arcing on the T/R
board. I had to change out the relay on my Acom 2000A shortly after I got it
because the factory had a bad run of HC1 vacuum relays. It's been going
about 17 years now, though mostly for S&P duty.
I've been lucky -- I've been able to do numerous repairs to my 87A without
having to send it to the factory. But I do worry about parts availability.
That problem started a long time ago, and it will eventually force me to
replace the 87A. On paper the 9500 would be a worthy successor, but honestly
I would want to make absolutely sure the amp doesn't still suffer from the
problems it had early on. And it may be tough to justify investing in a tube
amp with legal limit solid state amps getting better.
73, Dick WC1M
Amps mailing list
Amps mailing list