[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: My First Toob Coil
Original poster: "by way of Terry Fritz <twftesla-at-qwest-dot-net>" <cwillis-at-guilford.edu>
Hi Shad,
Tube coils are very entertaining, and I'm sure you'll enjoy the experience
with those 833s.
>I haven't wound the primary yet, as I don't have a good value for the tank
>cap or primary yet. I was planning on using around a 5nf tank cap,
and however
>many turns it took me to get it resonant at ~300khz (i can get there
roughly by
>calcs and scope, and then get it close to dead-on with the scope) The
>bottom-most turns I'll leave tappable so I can get the adjustment I want.
The 5 nF capacitor sounds like it will work. I've used L / C ratios
(henries per farad) of 50,000 up to 100,000 in tube coils.
Changing the L / C ratio in such an oscillator changes the primary /
secondary coupling, the circuit Q, and the impedance
that the tube sees, so it's hard to know how to design the resonant
circuit. Playing around a lot seems the only sure way
to get the best results.
>plate tranny - MOT with the voltage doubler in place - 2 MOT caps in
>parallel
Here's something I learned from experience: don't use the oven
rectifier. It doesn't handle high peak currents and
chances are, you'll be looking to replace it soon. Make a string of 20 or
so 1000 PIV / 6 A diodes, each bypassed with
10 megohms and at least a 0.001 microfarad ceramic cap. This setup has yet
to cause trouble for me.
I'll pass on some advice given to me by John Freau and Dave Sharpe a while
back that may save you some of the problems
I ran into. Make sure to bypass the transformer with about 0.5 microfarads
or so- RF will cause burning and internal shorts
in the secondary otherwise. A nice added safety bonus would be a plate
protection resistor (high-power wirewound) in
series with the plate lead to limit the energy of accidental plate flashover.
>the schematics James sent me calls for Mica caps, but I'm sadly lacking those,
>so I'll be stringing up MMC caps to fill this duty. I have enough
strings to
>get my to the necessary capacitancies. Though the voltage ratings
will be far
>(10-15x) overkill, I don't know about the current ratings. I'm not
worried
>about toasting caps though, I'm more concerned about the tubes.
I've had great success with MMCs in tube coil service, and I cannot tell
the difference in performance between them and
the high power mica caps from an induction heater. The mica caps are very
hard to find at a good price now. You should
give a thought to current handling, since the caps in tube coils do heat up
some. At least make sure you have two
MMC cap strings in parallel.
>How critical is the value of the tank cap and grid leakage cap? The grid
>leakage R determines the "breakrate" of the toob coil. I'm
assuming you'd want
>it to run the tube as hard as you can for the least plate redness while
>charging the tank cap up as to full.
Tuning is real critical in all the tube coils I have made. The effects of
capacitor heating on detuning the coil are
notable. I recommend a high-voltage air variable cap immersed in mineral
oil for fine tuning, placed in parallel with the
main cap. Otherwise, you can fine-tune the secondary to the primary's
frequency by adding more or less topload. If you
do use a topload, you can affect a bit of tuning by moving it up or down
above the secondary. You should always have a
breakout point if you use a toroid though; my tube coils seem to run into
flashover problems if they aren't discharging.
Regarding the gridleak, the tubes are most efficient running with class-C
bias (that's the only way you'll get the mentioned
1600 W per tube.) The RCA handbook method for calculating the grid resistor
is good enough. This component can be a
rheostat of the required power rating in order to play around with the
value. If you have two tubes in parallel, take the
value of the leak for one tube and cut it in half (since twice the current
will be flowing) to get the right grid bias on both
grids. If the gridleak RC (the product of the two) is too high, the
oscillator blocks or "motorboats"; and thus R will determine
the pulse rate. This mode of operation with high grid RC has been called
"sputter mode." It's useful for limiting tube
dissipation and increasing the spark-length-per-power efficiency. It is one
type of pulsed mode. "Staccato" is another
type of pulsed mode that only applies to AC coils and level-shifted
(doubler) coils. The advantage of staccato is that
the coil is turned on and off in syncronization with the zero crossings of
the AC plate voltage, reducing kickback
problems.
>do I have to ground *everything* around the coil that I don't want to get
>shocked by?
Non-grounded conductors near the tube coil will pick up RF power, which can
cause minor burns.
The RF burn problem is more substantial with tube coils than spark gap
coils. The biggest
safety issue for tube coils especially is the plate supply, with high
current capability. Unlike with an NST, you
cannot expect to survive MOT current. The same safety considerations that
apply to "regular"
coils also apply to the tube variety- be extremely careful with allowing
direct discharges to yourself because
through arcing to the primary or a grounding mistake, the plate supply can
show up at the secondary.
Have fun! Definitely mention your progress and results.
-Carl