Alex Pogossov wrote:
"Alejandro Lieber" wrote in message
...
I am thinking in building an AM transmitter for 7.1 Mhz (40 meters) using a
EL34 in class C as final.
The idea is to modulate it by varying the negative polarity of the
suppressor grid (g3).
Suppose I load the EL34 plate with 500 volts DC and a current of 150 ma
with 0 volts suppressor, does anyone have any idea what negative voltage
in g3 is needed to reduce the anode current by half, needed for screen AM
modulation ?.
Alejandro Lieber LU1FCR
Rosario Argentina
Does not seem like a good choice. EL34 is not for RF. It has high Cag
capacitance = 1pF. You are looking at interdependance of grid and antenna LC
tanks or parasitic self-excitation at worst. And 75W DC input is too much --
unreliable.
I have seen the EL34 used as exciter for 200 / 800 watts transmitters.
A 75W DC input in class C with 66% efficiency will dissipate 25W in the EL34
anode, still below it's capabilities.
Suppressor modulation is not a good choice -- requires about -150V dc and
the same AC component. Nonlinear. Negative g3 voltage repulses the electron
flow to the screen grid. It is easy to burn, unless a series resistor is
used to limit Pg2. (Ask Patrick what happens if you run a push-pull audio
amp without negative feedback at full power and then disconnect its load!)
If you (for whatever strange reason) decide to build AM transmitter (in the
21-st century when even SSB starts to look old fashioned), then use 807 tube
or any other specially designed for RF.
The best modulation is the cathode modulation. For example with 807:
1) get about 35Vm of excitation, connect g1 to ground via a 100K resistor,
apply AC excitation through a cap to g1.
2) Use Vg2= 250...300V
3) Between cathode and GND connect a darlington of a MOSFET with a 100R
resistor from emitter (source) to GND. Bias base (gate) accordingly to get
about 40...50mA DC on idle. Apply AF to base (gate) via a 1uF cap or so. You
will need about 5VmAC. Such transistor, as you can see, works as a
controlled current source. Thus cathode current of your 807 will be
perfectly following the AF input. Cathode of the 807 shall be decoupled to
GND via say 4700pF. Since it is the current that is controlled, this scheme
does not depend on the non-linearity of the curves of the tube. Cathode
potential sets by itself.
Very interesting. Do you know if it has been done in practice ?
Will the final valve still work in class C all the modulation cycle ?
I also remember seeing a rig using 2 6W6 (or 6Y6) in parallel in the cathode
of the RF output valve.
--
Alejandro Lieber LU1FCR
Rosario Argentina
Real-Time F2-Layer Critical Frequency Map:
http://1fcr.com.ar
---
news://freenews.netfront.net/ - complaints:
---