Re: 833A's plate color

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: "D.C. Cox" <resonance@xxxxxxxxxx>


Dull red is the normal color for 833s.  Bright 
orange means its running too hot.  Fab a chimney 
with forced air cooling (ref: older ARRL Radio 
Amateur's Handbooks for articles and 
photos).  PVC is acceptable with a good blower, ie, 200-300 CFM.

Dr. Resonance




Hi Frank, and anyone contemplating a VTTC:

    If you possibly can, avail yourself of a
reprint of RCA manual TT-5. I believe that
"Antique Radio Supply of Tempe AZ still sells the
reprints. There is a wealth of theory and
measurement, application and circuit diagrams in this book. (<$15)

Matt D.


In a message dated 3/22/06 1:41:36 P.M. Eastern
Standard Time, tesla@xxxxxxxxxx writes:
Original poster: "Philip Chalk" <philip@xxxxxxxxxxxxxxxxx>



Subject: 833A's plate color

Original poster: "Angelini, Frank J" <FrankJAngelini@xxxxxxxxx>

>   I am running an 833A based coil.  The tube is NOS RCA from a now
>defunct AM radio station.  I have approx. 3000 vac on the
>plates.  I'm getting a nice 18 - inch brush discharge, but the plate
>color is bright orange.  I have been told that I am pushing this
>tube.  Is this correct?  What should my plate /current/current
>be?  How can this be measured?
>
>Thanks


Hi Frank,

Sorry, this turned out very long - hope it helps  :-)

I've long wanted to play with 833a's, & never met one up close, but am
reasonably familiar with it - "bright orange" does sound like pushing it
a bit, depends on your subjective perception, description etc of the
term.

According to RCA's TT-5 Transmitting Tubes manual :-

"Plate shows an orange-red colour when tube is operated at its maximum
CCS or ICAS ratings."

3000v rms is about 4,250v peaks, so it's a little above the ICAS max
rating of 4000v **with forced air cooling**  With natural cooling, it's
3,300v max (ICAS. 3000v max CCS)

As a class C RF Osc/Amp the tube's max ICAS ratings include also :-
Plate current = 500ma, Plate input = 2000W, Plate dissipation 450W. Here
the max values co-incide, which isn't always so. At these values, you're
getting 1,550 Watts output from the tube.

Your input may well be above 2000W rms, and your plate dissipation sure
sounds like its greater than its rated. As plate dissipation is simply
the difference between the power input to the tube & the power output or
extracted from it, to reduce plate dissipation you must either reduce
input or increase output power. And power of course is the product of
voltage & current. So reduce input power by reducing plate voltage,
current or both & increase output by better impedance-matching from the
tube to the load (coil system).

Btw, CCS = 'Continuous Commercial Service' & ICAS = 'Intermittent
Commercial & Amateur service'.  Given that RCA say, even for CCS, which
is like an A.M. broadcast station transmitting 24/7, that the plate
'should be' running orange-red, you probably don't have to worry too
much if you're only doing short runs, with intervening cool-down periods
- depends on just how bright "bright orange" is  :-)

If you haven't got any forced-air cooling on the tube, you should
definitely add some. A heat-dissipating connector on the plate terminal
could also help a little. RCA TT-5 again:-

"With forced-air cooling, an air flow of 40 cubic feet per minute from a
2-inch-diameter nozzle directed vertically on the bulb between grid and
plate seals is required to limit the temperature between these seals to
145°C."  Do you have equipment that  you can measure this temperature
with ?  (BE CAREFUL of the HV)

Definitely measure the anode current if you can. Especially as you're
feeding it with ac, this is not a completely trivial exercise.  You
really need a true RMS meter to get useful measurements, due to the
'spikey' nature of the anode current waveform. If you have virtually any
kind of true RMS meter with about a 1, 2 or even 5 amp FSD scale you
could use that. ( AC Current meter goes IN SERIES with the HV supply
lead to the anode )

BE EXTREMELY CAREFUL to not touch the meter or any associated wiring
while the coil is running.  Set up the meter, making sure it & it's
leads are well-insulated & kept well away from anything else, then hands
off, coil on & note reading.  A 1/2 or 1A fuse in series with the HV
supply lead if paranoid.

Do you have an oscilloscope ?  This certainly gives you more options.

Sorry to preach again something that is constantly reiterated here, but
at least for my peace of mind:- With ANY kind of High Voltage
measurement you need to be very sure you know exactly what you're doing
& TO BE EXTREMELY CAREFUL. Any such measurements around this sort of HV
carry very real risk of serious injury or death at your very first
mishap or careless slip-up.

Some kind of protection ACROSS the current meter is worthwhile if you
can arrange it - several 1kV diode strings back-to-back, or a tiny
spark-gap of about 1/32 inch or less, etc.  If the meter goes
open-circuit for any reason, it'll have the full 3kV ac across its
terminals & is very likely to blow-up, possibly literally, in protest.

Reducing the plate voltage some will certainly give the tube a happier &
longer life, as will reducing the filament supply a little - down to
about 9.5v, or current-limiting it.

Depending on the circuit, increasing -ve bias on the grid might also
help, possibly simply by using a cathode resistor, bypassed; or reducing
the grid current somehow - reducing coupling of feedback coil, or
whatever - all depends on the circuit. Possibly you need to load the
tube more heavily by increasing overall coupling from the tube to the
secondary. Again - what's the circuit?

Certainly taking any measurements you can will help getting a feel for
the conditions, safety etc of the tube.

If you'd like to email me more details, type of transformer & the
circuit, or a link, or something, I might be able to offer you more
specific suggestions.  Haven't built a tube coil yet, but keen & getting
close. Got a 4-250A lying around just asking for it. vk2poc@xxxxxxxxxxx