I recently fired up my old KnightKit AM Broadcaster, the three tube unit which was sold as a kit by Allied Radio Corp. in the 50s and 60s, to check some of the operating characteristics.
The measurements were made with a digital storage oscilloscope with the probe clipped to the antenna output while a 3 meter wire antenna was attached. Since the transmitter is line operated it was connected to the mains through an isolation transformer to prevent problems with the grounded scope probe reference lead.
The unmodulated carrier waveform looks very good on the display with a voltage reasonably comparable to that produced at the radiator by solid state transmitters using a loading coil which suggests that the range for this transmitter will be comparable.
A spectrum display shows the second harmonic is down 39 dB below the carrier which is good. The fourth harmonic is down almost 50 dBc and there appears to be no odd harmonics. This is also good.
A 1 kHz sine wave was applied to the audio input of the transmitter and the amplitude was adjusted for modulation without clipping. The modulated signal envelope shows that there is a problem. The envelope is not symmetrical about zero and there is a time shift in the positive and negative peaks which indicates distortion. A probe of the signal at the plate of the modulator tube shows a good audio sine wave but this produces a poor modulated waveform probably due to non-linearity in the modulation process at the oscillator/final tube. Despite this, when audio is monitored with a receiver it sounds good but a careful listen reveals some distortion. A bigger concern is that the distortion can cause some splatter onto adjacent channels but this was not assessed.
Though this transmitter is "certified" by Allied Radio Corp. to be compliant the DC input power to the final was previously measured at approximately 180 mW which could also be a problem if operation under Part15.219 is intended.
The next step is to put the transmitter back in storage but perhaps if the modulation problem can be solved this would be a nice transmitter to put on the air. This might be a good Winter project.
Neil
Keeping tube equipment in service is as meaningful as keeping classic cars running.
Regarding the 180 mW to the final... given the drift of line voltage over time what actual range of input powers might result as the AC floats from, say 108 VAC to 130 VAC ( or whatever a typical range might be)?
Does the isolation transformer change the AC line voltage seen by the transmitter compared to raw line voltage?
It could be expected that the DC input power would change with changing line voltage but I did not test for this. The final stage uses cathode resistor biasing which would tend to stabilize the input power.
The isolation transformer I used was a 1:1 100VA unit so the output, though not measured, should be close to the input which here is 120 Volts.
I agree that keeping tube equipment in service or at least operational is desirable. Tubes are still available and though the prices appear high they are in line with the original prices when adjusted for inflation. Here, the remaining tube equipment which still operates include this transmitter, a Crosley vintage 1936 radio, a 1943 vintage military communication receiver, a Zenith 1960 era TV, and my low band Drake ham transceiver. In addition to the warm reassuring glow of the tubes these devices produce the smell of burned house dust when operating.
Neil
Bias wrong somewhere?
I just thought I'd throw that out.
I suppose you could "starve"
the final tube a little bit so the DC
input goes down to 100mW. You
would change a resistor for this. (?)
Several years ago, there were some
really interesting articles in QST about
the aging of parts in vintage radio equipment.
As I'm sure you know, parts age and the values
of these parts age. Perhaps the transmitter was
once 100mW DC in - but that parameter went up
because the transmitter is now about 50 years old.
After reading the QST articles about component
values changing with age, it almost makes me wonder
how vintage equipment can work at all. But then again, a well
designed circuit will be tolerant of those changes.
In one of these QST articles, a Heathkit DX-20 was
stripped and completely rebuilt. I seem to remember
the owner of the DX-20 had the inner cabinet gold plated.
Most people would not bother with that, but the owner
was able to do the gold plating for either a small cost or
for free - and he either had help or was
familiar with that sort of thiing.
When the project was finished, the the DX-20 worked much better
than it had before the rebuild. I believe it even worked better than
what was stated in the original specs.
I'm sure you remember that MRAM (?) did a lot of mods on the
imfamous Lafayette KT-195 Part 15 AM transmitter. He made
it much better, but then again - that's an other animal entirely.
Whatever you do, I hope you have fun with that Knight Kit transmitter.
Bruce
TV was in its prime in 1960, but can a vintage set receive anything today?
In fact, is the NTSC television system now extinct?
I liked NTSC, because it was still around when I did pro-video production and enjoyed the time-base correctors and vector scopes.
Carl Blare wrote: I liked NTSC, because it was still around when I did pro-video production and enjoyed the time-base correctors and vector scopes.
NTSC was capable of producing acceptable-ish color video displays -- given the correct adjustment/matching of the video sources at the studio, a nearly perfect transmission path to TV receivers, and correct adjustment of those receivers (black level, white level, color "hue," and color "saturation/intensity").
Modern 8VSB digital r-f signals whether received over the air, or via a cable TV system produce video displays on consumer-level HDTV sets with far better resolution, less noise, over transmission paths containing multipath components, and with far fewer source/receiver adjustments to achieve and maintain that performance.
Plus, such over-the-air digital formats can carry up to three separate video programs having better end-end performance than NTSC could produce under the best conditions.
Rarely is NTSC transmitted in the U.S. these days. But some NTSC receivers are still used by consumers as displays for the output signals of the digital-to-analog r-f signal converters available during the digital transition, some of such were subsidized by the U.S. government.
At least some use still exists for the old horse (NTSC TV).
Guess there comes a day when we say goodbye to old chums.
A friend just went to the Fiji Islands and shot video with a good digital camera and I'm able to play the DVD of the trip on the computer using VLC Player, and it is vastly better than any NTSC ever was, so all we are left with is better quality.
Wait, that's a good thing.
As Bruce mentioned, it is possible the component values have changed and this should be checked but I believe there is a fundamental design flaw with using the oscillator as the modulated final stage. Since both the plate and grid voltages are changing with the modulation each has its own effect and together the effect is most likely non-linear performance. It is also very likely that the signal has FM as well as AM present. I don't think a simple tweak will fix this.
Even back in my High School days when I used this transmitter I was aware that the modulation was lacking since I monitored this using a KnightKit oscilloscope but at the time I was not aware of the possible problems and since the audio sounded OK I happily broadcasted without worrying about it.
Neil
I agree with your estimation of the problem based on previous discussions while the group was designing the Big Talker Shortwave Transmitter.
At that time I recall that our amplitude modulation of the final stage was separated from the oscillator by the middle stage "buffer" amp, for the precise reason of keeping the modulation from influencing the purity of the oscillator.
Therefore now, in the spirit of the great Knight Kits, we should design a proper AM tube transmitter.
I'm ready to start at once.
The good old Knight Kit holds a special place of honor for those of us who drooled over the catalog description when we were young. Even after we built it and experienced the hum and marginal audio quality, we were still amazed and fascinated.
The design had its roots going back to the "Phono Oscillators" of the 1930s. Zenith was the most famous, and I think the first.
A "proper" design would include a crystal oscillator, an oscillator buffer stage, and an output stage. Audio would need a preamp stage and a modulator stage for plate modulation. That's 5 tubes (fewer if dual tubes are useable) and a modulation transformer (possibly expensive). A tunable antenna matching network would be a required. The old slug-tuned coils are no longer available, unless you pay big bucks for a custom build. An alternative is to use switchable inductors and a variable capacitor.
The goal would be to at least equal the performance of the present-day "serious" solid state Part 15 AM transmitters.
Crystals for the BC band are all but impossible to find unless you spend big bucks, and a vacuum tube based PLL synthesizer alternative using a higher frequency, cheap crystal is not at all practical. Physical construction is difficult. A sheet metal chassis requires drilling a lot of holes and punching out larger holes for the NOS tube sockets. Point-to-point wiring is a lost art and, even if you like it, it's hard work and often a mess unless you are experienced with pliers, cutters and soldering iron.
I suggest we maintain the Knight Kit's venerable place in history and preserve it as-is for posterity.
It's too true what PhilB says... the Knight Kit is part of a magic time when we "broke in" to radio.
Trying to recreate and build on the original tube experience is like trying to reverse time, a Rod Serling kind of thing.
It's not so bad having solid state perfection thanks to SSTran.com.
Project's over.
Like this one.
I exaggerated the aging component idea,
but it's interesting anyway. Yup, Carl - there
are still a lot of working vintage TVs around.
We have an LPTV NTSC on UHF channel 48 here.
Nobody watches it.
If I had a safe simple AM tube transmitter
to transmit around the house that would
be enough for me.
Yes, it would still be A LOT of work.
You would have to be experienced in
that sort of thing.
Some little company sells a phono
oscillator kit based on the 1937 or
1938 Zenith Phono Oscillator - it is
very simple, but is built on a wooden
board with exposed wiring. Fun for
temporary experimenting, but NOT SAFE
AT ALL. It's modulation is probably a
mixture of AM and FM. I would never
buy it. (Thanks to Neil,
and others, we were able to find that
Part 15 AM started right around 1937 or
1938, just when that oscillator came out.)
As mentioned
on many other boards, the imfamous
Lafayette KT-195 kit (LA-23 assembled)
was seemingly entirely FM modulation on
the AM BCB. How could
that be?? It had several tubes. Weird
ones, two 50BM8s and a 36AM3B. (When
ny 36AM3B rectifier failed, my father went
all over Connecticut to find another, and he
finally did at some hole-in-the-wall electronics
store. This was about 1970.)
There was one circuit around somewhere which
(I think) used a teeny little "can" 1000 khz oscillator
module from some internet parts supplier. This
was connected to one or two vacuum tubes. It
looked like a cool AM BCB transmitter design.
I haven't seen it around the net lately. And
it would only work on 1000 kHz, of course.
My Big Talker transmitter (13.560 MHz Part
15) works very well. You
guys Carl - Neil - MRAM (?) and a few others
put a lot of thought into it, and thanks to
you all, I was able to get it running. That was
a big thrill. The final version was 3 stages -
osc., buffer, and mod/final. After all of that,
it didn't FM anymore - as it did in the beginning.
As for the vacuum tube transmitter idea, you
have to know what you are doing, an you
have to do mechanical work and you have to
love it. And you have to SAFELY work with
high voltage. It's really hard.
Best Wishes,
Bruce