Long ago I realized the integrator with up-down counter. It started counting up at positive going pulse and would "hang" at maximum count till a negative pulse occurred, changing counter state to "down". Clock frequency, manually adjustable, depended on bitrate. Both pulses and clock were displayed on a small screen so setting the proper rate via clock frequency was a breeze. No amount of automatic setting could even come close, copy was 100% even with "bad" signals.
Under proper conditions, noise pulses are much shorter than the morse pulses so have little influence on the count of the morse pulse. They only change the timing of it, which already has been taken account of in a good morse decoding algorithm.
There was a 6U8-based super-regenerative receiver that used a TV horizontal oscillator slug-tuned coil for frequency adjust. That schematic is somewhere in my distant past - maybe I can recover it somehow....
Found CW practice files, various rates, at ARRL.org
That's a good idea, interesting to know how they built the tuned circuit. The last few generations of PAL/NTSC probably used the LOPT (Line Output Transformer) as part of the oscillator?
They also had quartz crystal 3.579545 MHz for NTSC and 4.433618 MHz for PAL. Both systems had line frequencies getting close to 16kHz.
I was able to buy a slug-tuned horizontal coil on Ebay. That was probably at least 10-15 years ago, when TVs still used CRTs. One of these days I'll find the schematic of the receiver that used that sort of coil. I do remember that it was a super-regenerative circuit that used a 6U8.
Last edited:
That one tube VLF receiver (or sound preamplifier) from 1970 is very interesting. Thank you for finding the schematic.
I compare it to our simple OpAmp version as follows:-
Seems that C1 and C2 in parallel for the one tube low band is 4nF same as our 4.1nF resonant capacitor. My knowledge of tubes is very basic but I think it is connected as an inverter, as the grid goes negative it draws less Anode current so Anode voltage goes up. I think the regeneration is negative feedback (or the lack of it) so the first pentode has high voltage gain.
From what you say the slug core in L1 was adjusted to change frequency. L1 must have a lot of turns (the slug core would be a lot smaller than a ferrite rod) typically it could be over 500 turns.
I'm not sure why C3 is there as it likely reduces Q of the resonant circuit, maybe they didn't want any Q? Presumably L2 is there to reduce input from higher frequency bands.
Don't see how the circuit would render the Morse code audible (unless the regeneration did that).
Just goes to show what a tube can do. In the UK the 6U8 (or 6U8A) was an ECF82, often used in TV sets I think first introduced in 1954.
Would be interesting to build this radio for the next SAQ transmission.
The magazine Science and Electronics is an interesting 100 pages! We had Practical Electronics and Wireless World in the UK and it's still interesting to read them.
Last edited:
There is a 60khz LF transmitter (MSF) in UK which outputs the atomic clock time. It is CW machine like SAQ. MSF is not special like the SAQ but they have similar concerns. Data is from carrier on-off. There are custom ferrite core antennas and receivers. Interesting solutions for decoding.
http://www.creative-science.org.uk/MSF1.html
http://www.creative-science.org.uk/MSF1.html
Ah maybe not, well not without modifying it somewhat. I'm not too up on valves but I think the regeneration causes oscillation and to render Morse code audible the oscillation is different from the CW frequency so the resonant circuit is always detuned slightly from the CW frequency.
I'm not too sure about how the pentode is biased but I think Anode current flows through part of L1 so as to bias the Cathode slightly positive of the Grid. Not sure what happens to the small AC component therefore on the Grid. In any event L1 has a DC current flowing through one of it's windings, not a great idea (why not use a low value Cathode resistor for bias). There is no additional high pass filtering.
Can't say I'm a fan of these "clever" but often difficult to understand circuits, think that simplicity is preferable.
To build this I would modify the circuit, remove the regeneration and configure it more like an audio preamp with a fixed gain. Most likely for SAQ replace L1 with the Ferrite Rod. So define the Q, define the Gain, define the filtering and make the output suitable for Laptop PC recording.
Not a fan of the "two in one" valves I would go with two simpler triode valves (tubes) instead of the ECF82 (6U8A) designed specifically for TV sets of the day.
Or being completely honest, think I would use only one triode valve as a front end feeding into an OpAmp for more gain.
Last edited:
Not such a bad idea to have DC current flowing in the tuning inductor, as it is a slug-tuned inductor with a big, big effective air gap...
Yes and the atomic clock based time is encoded as a binary signal that can be read as it is not encoded (other than BCD etc).
Would be quite an achievement to receive and decode this signal from as far away as Canada or the USA.
Last edited:
True not going to saturate the inductor but why do it? A Cathode resistor is simple, accurate and easy to understand. Why mess around with what is your tuned circuit that you rely on to resonate at the right frequency with the Q you want.
I'm just trying to understand the circuit.
From Wikipedia:-
"The signal, also known as the MSF signal (and formerly the Rugby clock), is broadcast at a highly accurate frequency of 60 kHz and can be received throughout the UK, and in much of northern and western Europe. The signal's carrier frequency is maintained at 60 kHz to within 2 parts in 10 to the power of 12, controlled by caesium atomic clocks at the radio station"
"When MSF was first introduced, in 1950, it consisted only of seconds and minute markers, with no coded data. In 1974, a short burst of binary code at 100 bit/s was inserted into the minute marker to specify the time. In 1977, a slow code was introduced, which encoded the time at 1 bit/s over the whole minute (as detailed below). In 1998, the original fast code was removed, leaving the slow code and a simple minute marker"
"The signal, also known as the MSF signal (and formerly the Rugby clock), is broadcast at a highly accurate frequency of 60 kHz and can be received throughout the UK, and in much of northern and western Europe. The signal's carrier frequency is maintained at 60 kHz to within 2 parts in 10 to the power of 12, controlled by caesium atomic clocks at the radio station"
"When MSF was first introduced, in 1950, it consisted only of seconds and minute markers, with no coded data. In 1974, a short burst of binary code at 100 bit/s was inserted into the minute marker to specify the time. In 1977, a slow code was introduced, which encoded the time at 1 bit/s over the whole minute (as detailed below). In 1998, the original fast code was removed, leaving the slow code and a simple minute marker"
Yes that would be a feat, as the USA has WWVB operating at the exact same frequency and with more power. It's even mentioned in that link in post #251
rich -
You are obviously not understanding it, as it works the way it's drawn. A few ma into the inductor is no big deal.
You are obviously not understanding it, as it works the way it's drawn. A few ma into the inductor is no big deal.
WWVB is so strong here in Panama that despite the power company's EMI the signal is visible on the 1.5k BW of the main waterfall spectrum.
Yes quite likely I'm not understanding exactly how it's intended to work - but in a way that's the point. Those kinds of articles were intended for people to build at home and many people who actually build same are kids so it's a good idea to make circuit function clear. Other than my maybe negative comments I think it's a good project.
As for my comment as regards using the resonant circuit at the front end to provide tube Cathode bias well true it's not going to saturate the inductor but why do it? A Cathode resistor is simple, accurate, easy and well defined.
I can go further as regards L1, the use of an inductor with such a small core is a bad idea because of increased turns and therefore reduced Q. Even a Ferrite Rod can be improved upon by using a Ferrite core without any air gap because of less turns and higher Q
Last edited: