[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: VTTC - audio modulation



Original poster: "Dr. Duncan Cadd by way of Terry Fritz <twftesla-at-qwest-dot-net>" <dunckx-at-freeuk-dot-com>

Hi Herwig, David, All!

>In order to minimize distortion I moved the operating point of the
>Armstrong oscillator towards the class A operating point using an
>external control grid bias supply. Of course the exact class A
>operating point is not suitable for modulation, as a nonlinear part
of
>the VT characteristic needs to be used here.

Um, why?  I would expect any rf oscillator to be capable of being
modulated linearly, regardless of its class of bias.  More below.

>Despite I lowered the
>number of grid coil turns before, distortion was large unfortunately.
>Too much turns still?


I think the problem lies elsewhere, but unfortunately until I can play
with the Q&D VTTC a fair bit, it will have to be mere speculation :-(

>> What you want is linear modulation of that class C stage
>- That is what I am looking for indeed. I wonder whether it is
>possible in principle to amplitude modulate an *oscillator* because
>usual modulators require carrier signals with constant amplitude.
>Wouldn't modulation cause the oscillation to stop?


It all depends on the modulation depth, aka modulation percentage.  If
you look at the very first valve AM transmitters, a good 80 years ago,
stuff like Hugo Gernsback's publications etc, or the stuff offered on
parts of this website http://www.oldradios.de/ (you can order CD roms
of all manner of stuff, in particular I recall Rainer, DC7BJ, who runs
the site has a German radio textbook from around 1920 or so by Paul
Fischer on CD rom, which you might find very exciting!) you'll find
plate modulated oscillators are used.  Even 55 years ago, British
textbooks still illustrated these obsolete circuits, mainly because
thanks to WW2 a lot of our merchant navy were still saddled with the
wretched things!  Modulating an oscillator in the normal course of
events is considered bad practice, but only because when this is done
the frequency usually shifts a bit.  Not enough to bother a VTTC, but
plenty to bother the guardians of the airwaves - also, the harmonics
sweep up and down in sympathy, and all of this is of course avoided in
the more sophisticated transmitters.  A modulation depth of say 20%
(in which the plate voltage varies +/- 20%) isn't going to hurt your
average power oscillator.  You might get away with more - I don't
rightly know the modulation depth used in the old plate modulated
oscillator transmitters, and I can't easily decipher it from the
circuit diagrams.  As you correctly surmise when modulating an
oscillator, it all depends on how seriously the oscillations are
disturbed.

>> It could be that the mere act of connecting your modulator to the
>VTTC is causing the modulator itself to distort
>- I didn't try a multistage approach, consisting of oscillator,
>modulator and class C amplifier yet. The main reason is, that the
>Armstrong power oscillator I'm using is selfstarting and
>*selfregulating*. I'm afraid that a class C amplifier, driven by a
>separate and frequencystable oscillator, will behave unpredictably
>when varying load, temperature, natural frequency of the secondary
>coil etc. Of course this multistage solution would permit an optimal
>modulator design. However i shy away from the additional expense
>and the increased complexity as well.


That I can understand, but I wouldn't suggest you need to go that far.
In "modern" (OK it's 1936 technology, but it's GOOD 1936 technology)
AM transmitters, you have this multistage setup it is true, but if you
revert to the bad old days you'll find that class C oscillators do
modulate linearly to give good quality AM - spurious emissions and FM
excepted.  Without finding time to do the experiments I'm now left
waving my hands around, but the point I would make is that there is on
the face of it no good reason why a class C power oscillator with
suitable modulation should not produce good quality amplitude
modulated rf output.  This, of course, is exactly what Ulrich's
original circuit does, so it is possible!

>> because with the VTTC operating in class C, it will offer a
>> widely-differing load impedance to the modulator stage
>- If the tube was cut off, the modulator would see a high impedance
>and if the control grid was driven into grid current, the impedance
>would become quite low. Right? Therefore a cathode follower should
>be use to drive the toob.


Spot-on!  The requirements for say screen modulation are less severe,
because of course the impedance here is pretty constant.  But not much
use if you're using triodes!

>> Full plate HT on the grid is not recommended :-)
>- My tube survived a strike to the control grid, blowing a fast 200 A
>peak diode, which is connected from the control grid via a 800 V
>gas discharge voltage protector to GND.


Ouch!  But excellent idea regarding your protection scheme!  Great
idea!

>> You can locate high rf potential on such a plate choke with a
>small neon bulb held on an insulating rod
>- As my plate voltage is 10 kV DC I'm "a little bit" afraid from
>getting close to the system when it's in operation. But it's an
>interesting modification you have made. Way to go?


Er, with 10kV on the choke I think I might be a little uneasy too!

>> If you have a working Q&D VTTC
>- quick and dirty?  :-))


Yes!  It's Ulrich's circuit without the modulator!  So if it distorts
when I get that round tuit, there will be some explaining to do ;-)

Best of luck with it, I think that until I can do some experimenting
myself, I really haven't got much more I can say to help you.  It's
touch-and-go with the micro-magnifier, but I think I'll just make it
for Saturday's Teslathon :-)

Dunckx
Geek#1113 (G-1)