Re: Ancient Rectifiers

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Dr. Duncan Cadd by way of Terry Fritz <twftesla-at-uswest-dot-net>" <dunckx-at-freeuk-dot-com>

Hi Matt, All!

Date: 02 February 2001 02:15
Subject: Re: Ancient Rectifiers


>> Does anyone have knowledge of/remember/ever seen large
rectifier tubes
>> (~12in. high) with screw bases like a lightbulb? I know
they are rated >100
>> kV with 6.3 v / 6 amp filaments. Two are VR-17C, one is a
VR-6-CFR and
>one is
>> a Machlett ML-46-E.


Well, I've done a bit of digging.  No direct refs to the
above valve types.  However, I do have a variety of old
valve data and some refers to EHT rectifiers of this general
type.  For those used for voltages from 16kV to 150kV,
without exception they all use pure tungsten filaments, and
in all cases the anode current is proportional to the
heating power.  Extrapolating from this, I would conclude
that the rectified current for a 6A 6,3V pure tungsten
filament in EHT service is going to be of the order of 40mA,
i.e. around 1mA per watt of heating power.  If your tubes
are indeed pure tungsten filament, they will light up with
the brilliance of a light bulb when powered, i.e. white hot.
Thoriated tungsten is not quite as bright, a bit yellowish
because it runs at a lower temperature, though you may need
a clear lightbulb beside you to compare.

Now, the nice thing about pure tungsten filaments is that,
short of applying too high filament volts, there really is
nothing you can do which is going to hurt them, which is of
course why they are used in EHT rectifier tubes - no danger
of emission loss by back-bombardment of high energy ions
resulting from residual gas, and no emissive surface layer
which can be damaged or depleted.  So if you really want,
you can test them without fear of harming the emission
characteristics.  Simply stick a variac-controlled
transformer across a tube in series with a milliammeter,
with the tube filament suitably powered (see below) and wind
up the volts whilst noting the current.  It will "plateau"
out as the rated current is approached, and my guess is you
will only need a hundred volts or two before you see the
plateau, perhaps not even that.  With a pure tungsten
filament, this cannot harm the valve, though it can and will
with other filament types.  If at a plate voltage of 50V you
measure a rectified current of 200mA you know I'm wrong:-)
This test needs to be made as rapidly as possible as the
valve anode will be dissipating all the power.  It needn't
take more than a few seconds if your eye and meter are quick
(analogue meters are IMHO preferable to digital for this
one.)

The filament needs to be at ground potential to be used with
a conventional heater transformer, and you earth the heater
transformer winding preferably at a centre tap; this
unfortunately means that without special filament
transformers insulated for say 30kV you are restricted to
half-wave rectification using only a single tube in the
grounded line.  More tubes for full wave or bridge
rectification would require the other filaments to be at
different potentials above ground, hence more than one
filament transformer/filament winding with EHT insulation
between the mains and the filament winding and between
filament windings.  One of my projects on the back burner is
a 2kV psu using type 866A mercury rectifiers in
centre-tapped full wave configuration.  For this project I
have a "Thermador" filament transformer for the 866As.  They
only need 2,5V to light them up - but the windings have
massive ceramic insulators on them because the tube
filaments and their transformer windings will be at 2kV
above earth and the centre tap of these filament windings is
the positive HT hot terminal, so the tube sockets need good
insulation and clearance above grounded objects too.  The
transformer nameplate states that the windings are tested to
7,5kV above ground.  In the case of a pole pig, the LT
filament transformer windings are going to need at least
30kV and probably 40kV insulation from the mains primaries,
despite only putting out 6,3V!  Ouch!

There are also possible problems with dc saturation/heating
of the EHT transformer core with half wave rectification -
you will in any case have low efficiency.  Sorry, but it
doesn't sound too encouraging, though it would be a neat
project if you had suitably-rated filament transformers (and
tubes).  I hope at least that it helped.

Dunckx