[TMC] Rectifier replacements

[Roy Morgan]

At 08:33 PM 12/6/2005, w3jn wrote:
>Use the solid state replacements.    Mercury vapor rectifiers are cranky 
>and insolent, and WHEN (not IF) they flash over they can wreck a power 
>tranny.... never had anything that used 872s but I would expect the same 
>problems in.re. flashovers, having to warm them up for some time after 
>they haven't been used in a while, etc.
>>The SBT-1K uses 872 rectifiers.

John,

Your mileage may vary.  There are some things that can lead to happy 
relationships with mercury vapor rectifiers:

Larry is working on a TMC-built transmitter. That means that the power 
transformer is likely way oversized for the job, and that the B+ system is 
well-protected for overloads above a very conservative level.  TMC stuff 
was built to run 24/7/365, and, if something did mis-behave, to come back 
on line quickly

My fairly limited experience with a couple of TMC GPT-750 transmitters was 
good. These, by the way, are earlier than the one he's working on, I think, 
and reportedly built on purpose to be "sailor proof" in all respects.  In 
one case the 872A's flashed over repeatedly, each time popping the B+ or 
main circuit breaker.  Apparently we finally got enough time on the things 
so they settled down and worked just fine after that.   Note that the 
GPT-750 has LOTS of fan power, and that may have kept the tubes from 
warming properly.  In the second 750 I worked on, the rectifiers worked 
just fine and never gave any trouble.

I've recently been working on a Johnson Desk Kilowatt which has recently 
been the prime subject of restoration and repair at W0YVA.  The 872A's in 
that have given no trouble at all.  They have been run in room temperature 
conditions at least once a week for some time now.  They have also 
withstood the insults of a faulty HV choke and modulation transformer.

I have made up in the true tradition of early ham radio a breadboard 872A 
"Cooker."  It is a piece of board from a shipping pallet equipped with two 
sockets, a 10 volt 5 amp filament transformer and a line cord. (the tube 
filaments are in series)  I can set a pair of 872A's to warm up and clear 
the mercury as I work on the transmitter.  The filament heat alone is 
plenty in room temperatures to evaporate the mercury well down to the 
bottom of the envelope.  I have a similar cooker for the 866 types.

If you are like the guy we heard about in Montana whose shack is truly an 
unheated metal hut out in the back yard, I would not try to run a Valiant 
with 866A's in it, or any other MV rectifier for that matter.  (The 
unfortunate fellow has orders from the wife unit to keep the house 
completely clear of all radios.)

Note that the ratings of the 872A are way above the demands of a KW-1, 
GPT-750, or Desk Kilowatt, so in all these cases there is plenty of reserve 
capacity in the tubes (and everything else about the high voltage parts, 
too).  Very likely not so with the 866A's you have used.  The 866A may be 
pretty much loafing in the Valiant, but in other transmitters they probably 
are worked much closer to their limits. Also, the 866A is smaller than the 
872A by a lot - and the leakage paths that are likely culprits in 
flashovers are much shorter.  The reputation the 866A has for being 
difficult may be well deserved. Fortunately, the 3B28 Xenon tubes are 
fairly easy to find, unlike the 4B32 upgrade to the 872A.  (I'm looking for 
some at less than commercial prices.)

I have a small number of the predecessors to the 866A. These are type 866 
or a similar earlier tube that seems to have gone out of production at the 
time the 866A was brought out. Most of these old tubes are globe shaped, 
and likely date from the 1930's.  They all show evidence of disaster: 
clouds of white filament coating flakes inside, bits of filaments rattling 
around, severely discolored envelope inside surfaces, bent and discolored 
plates.  I shudder to think what happened to the chokes and HV transformers 
during the significant events.

I read in an old tube manual that to diagnose difficulties in mercury vapor 
rectifiers that are flashing over, wrap each plate lead around an iron 
nail. After a flashover, the nail will be magnetized in one direction for a 
working tube due to the net DC current, and not for the one that flashed 
over (or something like that).  Do not try this at home, please.

Happy rectifying to all,

Roy


- Roy Morgan, K1LKY since 1959 - Keep 'em Glowing!
7130 Panorama Drive, Derwood MD 20855
Home: 301-330-8828 Cell 301-928-7794
Work: Voice: 301-975-3254,  Fax: 301-948-6213
roy.morgan at nist.gov --