[AMRadio] Re: Gates BC1-T operational again

[D. Chester]

What I did to make the BC1-T operational on CW would apply to many other 
broadcast transmitters.

My main keyed stage is the parallel 807's.  However, I had some backwave 
evident, so I now key the 2nd 12BY7 stage as well.  The total cathode 
current is in excess of 200 ma, so I wouldn't recommend keying directly with 
a bug or electronic keyer.  One of those antique straight keys with the 
quarter-inch contacts, used to key spark transmitters, might be ok, but I 
use a TV sweep transistor as an electronic relay to cathode key the stages 
with negligible voltage drop.  I use the same keyer with my homebrew 
transmitters as well, just plugging it into the appropriate key jack.  I use 
an inductor and capacitor between the transistor and the cathode circuit of 
the tubes to shape the waveform and avoid key clicks.  For the values of 
inductance and capacitance, I use  the standard info from the older RADIO 
handbooks or ARRL handbooks, plus some trial and error.  The TV  sweep 
transistor circuit is described in the 1980 ARRL handbook.  I used the 
highest voltage-rated transistor I could find.  Mine is  rated to switch at 
least 1400 volts at several amps.  The extra voltage capability is necessary 
to avoid blowing the transistor with transient spikes from the shaping 
circuit components, especially the inductive kick from the inductance.  I 
built mine probably 15-20 years ago and have never had to replace the 
transistor.

The main problem  with operating a typical broadcast transmitter on CW, is 
the power supply filter.  Most BC transmitters use a common power supply for 
modulator and final.  The voltage for the class-B modulator is held nearly 
constant by the 100% duty cycle load of the class C final, which acts like a 
4-6 kilohms, 1 kw+ bleeder resistor.  When operating CW, the modualtor is 
turned off, but the class C final is keyed intermittently.  The power supply 
is designed for a  constant heavy load, so the power supply filter choke 
seldom has enough inductance to maintain optimum or even critical inductance 
under key-up conditions.  The result is that under key-up conditions, the 
power supply filter will begin to act more like capacitor-input than choke 
input, and the voltage will most likely soar to almost 50% above the nominal 
value.  That may be enough to damage the power supply filter caps and other 
components, and the keyed waveform will be horrible.  You need to do 
something to maintain better voltage regulation.  One solution would be to 
substitute something like a 5-40 Henry swinging choke in the power supply, 
and appropriately reduce the bleeder  resistance.  Such a filter choke will 
be hard to find, and probably expensive. The stock filter choke is probably 
something less than 10 Henries, which is sufficient for AM with the class C 
load on all the time, but not for the intermittent load presented by a keyed 
final.

But you do have the needed additional inductance available - in the 
modulatior reactor.  The trick is to rearrange the circuit in the CW mode so 
that the modulation reactor and power supply filter choke are wired in 
series, giving a total inductance typically of  40-60 Henries under full 
load.  It is very easy to do this.  You will need a reliable high voltage 
double-throw switch. For CW, switch the power supply filter capacitor and 
bleeder  resistor from the bottom end of the modulation reactor to the top 
end, where it feeds the class-C final.  This will leave the two chokes in 
series, with the output capacitor and bleeder at the top end of the mod 
reactor, so that the series-wired chokes serve as the power supply filter 
choke.

You will also need to lower the bleeder resistance in ohms from the typical 
100-150K to a value no greater than about 900-1000 times the total 
inductance (mod reactor plus filter reactor) in Henries.  Critical 
inductance L (Henries) =  R (load resistance in ohms)/1000.   The total 
inductance is now enough to maintain optimum inductance even under key up 
conditions, and the power supply filter remains effectively choke input at 
all times. This holds the STATIC voltage regulation to less than 10%,  but 
chances are, the DYNAMIC regulation is still unsatisfactory, and so will be 
the CW waveform.  Put a scope at the power supply output and you will most 
likely see that the instantaneous voltage jumps all over the place, almost 
as much as the original power supply, except that the voltage variations are 
of such short duration that an analogue meter will indicate excellent 
voltage regulation.  The solution to poor dynamic regulation is to change 
the output capacitance from the 8-10 mfd found in a typical 50's-60's 
vintage AM broadcast transmitter to at least 25 mfd.  The more capacitance 
the better, but too much capacitance will require step-starting the power 
supply.  I have found that with the BC1-T,  25 mfd gives a  satisfactory 
compromise between dynamic regulation (and keying waveform) and the 
capacitor charging current surge.

Don k4kyv