German one-valve superhet FM tuner

[regenfreak]

Originally Posted by FERNSEH

From the GFGF article.
The response curve of the FM demodulator.
The choice of the received swing frequency recommended by R.Cantz was 20 to 50Khz.

DFWB.

This is new to me. I got used to seeing S-curves in FM frequency sweep

[regenfreak]

Originally Posted by Synchrodyne

I think grounded grid was probably the modal choice, but neutralized grounded cathode was used by some, also a combination with push-pull feed to grid and cathode, as in the Radford FMT1, which I think used a D&W unit.


Cheers,

In G6Tanuki's link,
http://www.r-type.org/articles/art-011.htm

In the cascode example, the upper triode has grounded grid with a neutralisation capacitor is used not for better stability but to increase signal-to-noise ratio.

[G6Tanuki]

Some associated discussion here: see the "FM in Germany" part about half way down.

https://www.cool386.com/fremodyne/fremo1.html

[regenfreak]

Originally Posted by G6Tanuki

Some associated discussion here: see the "FM in Germany" part about half way down.

https://www.cool386.com/fremodyne/fremo1.html

Thanks I saw that article a long ago when I built my one valve super-regen. I presume the response graph posted by FERNSEH was the response curve of the slope detector. The working principle of quenching etc is anything but simple. The BBC R&D technical report really dimissed the super-regen FM design as "poor" in term of sensitivity and users handling quality.

Here are more examples of German UKW tuners with varieties of designs:


http://www.jogis-roehrenbude.de/Oldi...KW-Vorsatz.htm

[regenfreak]

I have powered it up. It is working. The oscillator only works with at least B+ of 220v drawing total 5ma for both plates.

I have measured the primary inductance of the IF coil to be 9.996 microHenries. I notice that there is a capacitor in series with the bottom of the primary IF to the cathode of the second ECC85 triode. I think it is labelled as C13 (not 100% sure as the scan quality is not good). C13 = 120pF,

Assume the "dynamic" inter-electrode capacitance between the anode and cathode to be Cakd under working conditions, then the overall capacitance C to resonate at 10.7MHz with L = 9.996 microH is C = 22.133p.

1/C = 1/Cakd + 1/C13

Solving Cakd = 27.138pF assuming the series cap is labelled C13.
This ballpark figure ignores the distributed capacitance of IF coil windings which can be significant at VHF. Typically, the value for resonator caps for 10.7MHz IF transformers is about 10pF to 30pF.

This value is compared with the "cold" unloaded Cak = 0.17pF ( or 1.2pF cathode including heater and screen) listed in the ECC85 datasheet. Please correct me if the calculation is wrong.

[Technomaniac]

When I was an apprentice, or maybe even earlier, I built the front end of a "Fremodyne Four" in a spare space on a 5 watt audio amp chassis I had built previously. We had no FM broadcasts out of the capital cities, but when TV arrived, there was one channel right in the middle of the FM broadcast band. It was later moved to a higher frequency. But the tuner worked, though it had 10 % distortion. I didn't have enough knowledge to solve the problem at the time. The "Fremodyne Four" was published as a project by the Australian "Radio Television and Hobbies" or "Electronics Australia" which it became later. A work colleague bought everything I had built during that period, but I have been thinking for a while, of rebuilding the project.

[Boulevardier]

Related question that I've wondered about for years - what is "oscillator squegging"? Can anyone explain it in something like layman's terms?

Mike

[Synchrodyne]

Originally Posted by regenfreak

I have seen triple triode 6GY8, I am yet to see any FM domestic receivers use triple triodes or dual tetrodes...maybe they were used in TV tuners?

The GE T1000 series FM-AM Stereo receivers used as 6EZ8 triple triode as mixer, oscillator and AFC valve, following a 6EW6 pentode RF stage.

The T2000 series of the same era had a 12JN8 triode pentode single-valve front end, with pentode as RF amplifier and triode as autodyne mixer. A solid-state diode was used for AFC.

The RCA RC1210 series FM-AM chassis had a 17C9 double-tetrode single-valve front end.

All of the above had three-gang FM front ends.


Cheers,

[Nuvistor]

Originally Posted by Boulevardier

Related question that I've wondered about for years - what is "oscillator squegging"? Can anyone explain it in something like layman's terms?

Mike

https://en.wiktionary.org/wiki/squegging

[Radio Wrangler]

Oscillators use RF feedback in their operation. The feedback may be obvious on the circuit, but in many oscillators it operates invisibly via internal capacitances of valves and transistors. It is positive feedback and has a side-effect on how the oscillator starts and stops.

Wind up the HT voltage on an oscillator and at some point it will start up. Wind up a bit more HT and yoou may find it starts more quickly, and the output may be a little larger.

Wind down the HT, and the output will diminish, but at some point it suddenly stops. Try increasing the HT from this point, and it does not restart until a noticeably greater HT voltage is reached, the same one we saw in the start-up test. There is hysteresis! There is a range of HT voltage where the oscillator may either be running or not - depending on its pre-existing status.

Circuits with hysteresis can make 'reflex oscillators'

Think of the simple neon bulb oscillators. An HT supply feeds a capacitor via a large-is resistor. There is a neon bulb across the capacitor. The HT supply is greater than the neon bulb's ignition voltage. The capacitor charges up slowly through the resistor, until the voltage is enough and it lights. It now takes current, more current than the resistor supplies to the capacitor becuse we chose the value of resistor to make it so. The capacitor discharges until it gets to the neon bulb's extinction voltage, where the discharge goes out. The resistor current is no longer being upstaged by the now-zero bulb current, so the capacitor starts charging back to the ignition voltage where the cycle repeats.

So an oscillator's start-stop voltages make it a suitable substitute for the neon bulb. For reliable oscillation t theRF frequency, the HT supply to the oscillator has to be stiff enough that the current consumption of the running oscillator cannot pull down the HT supply far enough to stop the RF oscillation.

Squegging problems come from HT supply impedances going higher resistance than intended. Maybe a dropper resistor goes high, or dry-soldered. Battery sets suffer from rising internal resistance in their batteries as the batteries near exhaustion.

So the recipe for squegging or relaxation oscillation is an RC timeconstant, and something exhibiting hysteresis.

It can actually be done deliberately in super-regen receivers. Oscillators become hypersensitive to any help or hindrance at both their start and stop points. Super-regen is a shortening of supersonic regeneration. Supersonic used to mean what we now call ultrasonic, to high pitched to be audible. Regeneration meant positive feedback. So a super regen set would burst into and out of self-oscillation at the received frequency (Your neighbours would hate you!) Audio could be extracted from the current demand on the HT.

David

[Boulevardier]

Thanks Frank & David!

Mike

[FERNSEH]

A dictionary's explanation of the word "Squegging"
squegging.

"An irregular oscillation characterised by short periods of oscillation punctuated by brief periods of quiescence.
Etymology: Uncertain. Possibly from self-quenching."

DFWB.

[regenfreak]

Related question that I've wondered about for years - what is "oscillator squegging"? Can anyone explain it in something like layman's terms?

How to fix squeals and motorboats handbook:

https://canadianvintageradio.com/wp-...Motorboats.pdf

The GE T1000 series FM-AM Stereo receivers used as 6EZ8 triple triode as mixer, oscillator and AFC valve, following a 6EW6 pentode RF stage.

The T2000 series of the same era had a 12JN8 triode pentode single-valve front end, with pentode as RF amplifier and triode as autodyne mixer. A solid-state diode was used for AFC.

The RCA RC1210 series FM-AM chassis had a 17C9 double-tetrode single-valve front end.

All of the above had three-gang FM front ends.

Thanks I seem not be able to find the schematics of any these three examples. Comparison between dual gate mosfet and double-tetrode are often made for their lower inter-electrode capacitance than triodes. I dont know if 17C9 or 9C9 were introduced in the dying days of valves.

Wind up the HT voltage on an oscillator and at some point it will start up. Wind up a bit more HT and yoou may find it starts more quickly, and the output may be a little larger.

In the context of homebrewing VHF oscillator, the ratio L/C seems to be important for oscillation stability but I cant find any literature about it. The Q factor is important but sometimes oscillator refuses to oscillate or become unstable no matter how high Q coil, high B+ used..Obvious problem may be inductive coupling between RF coil and oscillator or not enough turns of feedback in Hartley oscillator. This can be solved by shielding or make the coils at 90 degrees. But once all these factors have been ruled out, there is more beyond the obvious.
The higher L/C ratio, the higher the impedance but I cannot establish the definitive link between high impedance and higher oscillation stability. I think the Q will drop if L/C is large which seems there is a sweet range of values exist between stable and unstable. If L/C is too low, Q is too high, it may go unstable.

hen I was an apprentice, or maybe even earlier, I built the front end of a "Fremodyne Four" in a spare space on a 5 watt audio amp chassis I had built previously. We had no FM broadcasts out of the capital cities, but when TV arrived, there was one channel right in the middle of the FM broadcast band. It was later moved to a higher frequency. But the tuner worked, though it had 10 % distortion. I didn't have enough knowledge to solve the problem at the time. The "Fremodyne Four" was published as a project by the Australian "Radio Television and Hobbies" or "Electronics Australia" which it became later. A work colleague bought everything I had built during that period, but I have been thinking for a while, of rebuilding the project.
Yesterday 11:29 pm

I built Fremodyne with 6C4 and it squeals crazy. Now I only build FM superhets.
This channel has 5 part series on super-regen:
https://www.youtube.com/watch?v=OiTwY5loqsE

[FERNSEH]

From the Radio Museum. The 6C9 and 17C9 double-tetrode and it's application in an FM tuner.
Valve has a special 10 pin base.
https://www.radiomuseum.org/tubes/tube_6c9-10pin.html

DFWB.

[G6Tanuki]

Squegging in the sense of super-regenerative receivers is a form of 'self-quenching'.

The basic mechanism of a superregenerative receiver is that the level of regenerative feedback is increased so signals build up to beyond the point at which the system oscillates; the oscillations are then interrupted in a periodic, controlled fashion to 'quench' them; This quenching can be by way of an externally-generated quench signal [usually applied as DC pulses to the grid] or the superregenerative stage can be 'self-quenching'.
In this mode, the RF oscillations build up to a high amplitude, but are rectified by the diode action between the grid and cathode; the resulting negative DC potential charges the grid-capacitor to a sufficiently high DC value that the valve is cut-off and oscillation is blocked. The DC on the capacitor then leaks away through the grid resistor until the voltage falls to a point where the RF oscillations can start again. But they add negative voltage to the grid, so the oscillation stops.... wash,, rinse, repeat.

You have two time-constants: the RF oscillation time-constant set by the L/C tuned-circuit, and the second time-constant set by the vales of R and C in the grid-circuit. So the 'thing' oscillates at two frequencies; the low-frequency blocking of oscillation by the DC bias gives the quenching needed.

The low-frequency part is the squegging; see also 'blocking oscillator' whih were at one time used extensively in TV timebase circuitry.

[regenfreak]

Originally Posted by FERNSEH

From the Radio Museum. The 6C9 and 17C9 double-tetrode and it's application in an FM tuner.
Valve has a special 10 pin base.
https://www.radiomuseum.org/tubes/tube_6c9-10pin.html

DFWB.

Thanks

I tend to shy away from super-regenerative receivers in favour of more complicated superhet topology. I dont really find it easy to tune and pleasant to listen to. Sensitivity is rather poor. However I heard EF95 or 6AK5 VHF pentode gives the "smoothliest" tuning experience.

[Synchrodyne]

Originally Posted by regenfreak

I dont know if 17C9 or 9C9 were introduced in the dying days of valves.

Around 1961, I think. The attached Sylvania advertisement from 1961 captures the range of single-valve FM front end options then available in the USA.


Cheers,

[regenfreak]

Originally Posted by Synchrodyne

Around 1961, I think. The attached Sylvania advertisement from 1961 captures the range of single-valve FM front end options then available in the USA.


Cheers,

Thanks. Interesting! I have never seen 10-pin valve socket before. I knew about 12 pin compactron.
6kj8 is rather unique and with one triode optimised for Rf and other for oscillator.

[regenfreak]

I forgot that I built this one-valve minimalist FM supehet the year before. It cannot get more simple than this and it was my first working FM superhet with a pentagrid convertor 6SB7Y. I had lots of success building a MW supehet using 6SBY7 for its outstanding DX performance.

It works well with an external antenna. For indoor antenna, I would need a 20dbm LNA pre-amp to pick up all the stations. Obviously it is noisier than miniature triode valves but it happily oscillates all the way to 97.3MHz, super stable ( I use 10.7MHz below)

[Synchrodyne]

Originally Posted by regenfreak

6kj8 is rather unique and with one triode optimised for Rf and other for oscillator.

I am not sure which US valve maker actually developed the 6JK8, but the available evidence points to Sylvania. However, I suspect that the request for this kind of valve may have come from Zenith, who had a history of working with valve makers on new valve types (and later did the same with the IC makers.). Zenith had not long changed its domestic receiver FM front end from the two-valve type (combined AM and FM) to the single-valve type (FM-only), using the 12DT8. It probably wanted to keep to that format with the introduction of stereo, of which it was a major proponent, but also needed more RF gain for a better quieting curve.

Originally Posted by regenfreak

Thanks. Interesting! I have never seen 10-pin valve socket before. I knew about 12 pin compactron.

In the USA there was the decar type, basically a noval with a centre pin. In Europe there was the decal, with the same pin circle as the noval type, but with 10 rather than 9 pins on that circle, with closer spacing.

Originally Posted by regenfreak

I forgot that I built this one-valve minimalist FM supehet the year before. It cannot get more simple than this and it was my first working FM superhet with a pentagrid convertor 6SB7Y. I had lots of success building a MW supehet using 6SBY7 for its outstanding DX performance.

The 6SB7Y was an example of responding to a new development vector by pushing a bit harder up the old vector/blind alley. RCA had positioned the 6BE6 pentagrid as being suitable for FM as well as AM frequency changing, with one valve doing both jobs. By all accounts it did not work very well on FM. So the answer was a better pentagrid, the 6SB7Y. It needed 8 pinouts, so was necessarily an octal, as the noval type had not yet been invented. Meanwhile some setmakers had used the Sylvania 7F8 (loctal) double triode as an FM-AM frequency changer; it was much better at FM. Then GE introduced the 12AT7 double triode as an FM and TV frequency changer, also suitable for AM in FM-AM receivers. This was quickly picked up by Zenith, for example. RCA’s interim response was to issue a noval version of the 6SB7Y, namely the 6BA7, and to suggest that its 6J6 double triode could be used as a combined FM-AM frequency changer (really). Its definitive response was the 6X8 triode pentode in 1951, suitable for AM, FM and AM/FM as well as TV frequency changing. That left the 6BA7 as a pretty good pentagrid for AM work, with a similar noise level to the ECH81. Perhaps though the best heptode for FM work – for anyone inclined that way - was that in the Brimar 12AH8 triode heptode. Brimar seemed to have put some effort into obtaining reasonable FM performance.


Cheers,