[Collins] [collins_radios] Re: help needed

[Dr. Gerald N. Johnson]

On 1/31/2012 12:14 PM, Richard Knoppow wrote:
>
><SNIP>

>>
> In the mid-1940's RCA published a graph showing
> filament life vs voltage. The chart was meant for
> transmitting tubes with tungsten or thoriated tungsten
> filaments or for lamps. I don't know how valid it is for
> indirectly heated tubes like most receiving tubes. With this
> in mind the chart shows that a 5% reduction in voltage about
> doubles the life of the filament. The chart goes only to a
> 5% increase but shows a significant decrease in life
> although not as extreme as the increase from reducing
> voltage.

A similar chart is in some GE tungsten lamp catalogs. Along with that 5% 
reduction doubling life it cuts light (and electron) emission by 10%. At 
the same time a 5% increase increases output by 10% at the cost of 
cutting life in half. Hence projection lamps are relatively efficient 
but have a lifetime rating as short as 25 hours. NIST transmitters for 
WWVB and WWV have taken this to an extreme operating transmitters at 
much lower than rated power and cooling filaments to achieve phenomenal 
tube lifetimes more than ten times the expected lifetimes.

There's a lamp in a San Francisco fire house that's been running for 
about a century, but its a 240 lamp on 120 volts judging by the dim 
glow. If one works the geometric progression of subsequent 5% reductions 
in voltage and doubling of life, its within its expected lifetime at a 
century of operation compared to 1000 hours expected life at rated voltage.

> Currently (no pun intended) mains power in the USA
> seems to be higher than the design value for most older
> equipment. A few decades ago AC line voltage in most places
> was standardized at about 115 or 117 V, it is now more
> nearly 120V and in some places up to perhaps 125V. The
> reason is probably that line losses are lower as voltage is
> increased so the increase probably puts off having to rewire
> the grid for higher capacity. Of course distribution
> transformers could be used to keep the voltage to individual
> homes constant but there are millions of them and they would
> have to be replaced.

I think much of the raised voltage comes from regulatory agencies 
demanding minimum loaded voltages and power systems typically having 
voltage regulators at sub transmission, but above distribution voltages, 
e.g. 35Kv circuits with two transformers between that sub distribution 
voltage and the consumer 120/240 service entrance. Historically such 
transformers have had a typical impedance of 4 to 5%, so the 120/240 
voltage will drop 8 to 10% going from no load to full load. With the 
regulating agencies setting a minimum voltage (so as to keep resistance 
heating appliances working well, and preventing excessive currents to 
induction motors) the utilities have had to raise the no load voltage to 
not violate that voltage minimum.

> Reducing voltage to equipment meant for a lower voltage
> is not as simple as it might seem. Variable transformers
> like the Variac are expensive. Installing a bucking
> transformer requires either modifiying the equipment or
> constructing an external housing. Using a line dropping
> resistor is wasteful of power, creates heat, and the drop is
> dependant on the amount of current drawn. This last is
> important for many pieces of equipment which have variable
> power demand. Among these are transmitters and tranceivers,
> I think a simple dropping resistor is not acceptable for
> either.

The bucking transformer is the best of the options. I know one old 
electrician who used one for his whole house. The variac works decently 
but is subject to unauthorized adjustment unless its totally hidden. 
There are numerous suitable buck/boost transformers in the internet 
market places at reasonable prices, compared to purchasing a new high 
end solid state rig.

> In the same way a dropping resistor in the B+ supply
> can cause problems with regulation. As pointed out by David,
> a dropping resistor in the B+ does nothing for the filament
> voltage but such resistors are often applied where a tube
> rectifier is replaced with a solid state one. In this case
> the loss of regulation may not be so significant simply
> because vacuum tube rectifiers have an internal resistance
> so that the voltage drop is dependent on the current draw.
> The drop will depend on the particular kind of rectifier and
> on the type of filter. However it can be on the order of 50V
> for HV rectifiers. In comparison a silicon diode has about
> 0.7V drop which is fairly constant with current (but can increase to
> 1.5V for some diodes under load). In most cases a
> simple dropping resistor will do to maintain the output at
> the right value. since the vacuum tube diode also looks like
> a resistor the regulation will be no worse. Where mercury
> vapor rectifiers are used, as was very common for high
> current supplies in transmitters and similar equipment a
> simple resistor is not suitable because mercury vapor diodes
> also have a constant drop, about 15 volts. Probably the best
> way of compensating is to stack silicon diodes so that the
> drop of each adds to the right value. Since one is usually
> dealing with a high voltage supply this may be a way of
> solving two problems at the same time.

The power in a series resistor is the same as that of the vacuum 
rectifier with the same drop, but the solid state rectifier still saves 
the filament power heating the power supply. Often that filament power 
provides half the heat from the tube.

> One can also use diodes to drop the voltage when
> replacing vacuum tubes with diodes by using a regulator
> diode with suitable voltage drop and current capacity. This
> will also improve the regulation of the supply.
> An expensive but satisfactory solution to the high line
> voltage problem is the use of voltage regulating
> transformer, like a Sola constant voltage transformer. These
> work quite well but can be mechanically noisy and should be
> operated at near full capacity for reasonable efficiency.
> Where line voltage varies a lot the Sola transformer is
> quite valuable.

Sola have always been quite expensive, and often the output voltage is 
very distorted, and they put out a strong magnetic field that can affect 
audio equipment like dynamic microphones and low impedance transformers. 
In one test setup at Collins where I had a large Sola to stabilize 
signal generator and RF VTVM, I had to fight with the engineer across 
the bench working on updating a Collins audio console because the 
magnetic field coupled to the audio input transformers injecting so much 
hum he couldn't work. I couldn't work with line voltage variations so we 
had to alternate working at the benches.

The Sola works by saturating part of the core (distorting the line 
current) and resonating the secondary or an auxiliary winding a few Hz 
above the standard line frequency so that increased line voltage 
increases the saturation of the core and shifts that resonance higher in 
frequency lowering the output voltage. Its not appropriate for operation 
on a portable generator and its not practical to convert a 50 Hz Sola to 
60 Hz operation and still achieve regulation.
>
>
> --
> Richard Knoppow
> Los Angeles
> WB6KBL
> dickburk at ix.netcom.com
>
>
73, Jerry, K0CQ, Technical Adviser to the Collins Radio Association.