Golborne Vintage Radio

(04-07-2020, 03:40 PM)Amie Wrote: [ -> ]Just one question though, I've seen diodes used as in Trevor's diagram occasionally, But what is its purpose? Is it acting as a clamp?

(There is no DC continuity at its anode so is therefore just floating?)

No, I've never seen a single diode in an official circuit like that, because as you say there is no obvious DC path anywhere on the cathode or anode. Though I can imagine theories to explain why it works, as outlined earlier.

Hi Michael.
Many thanks for the reply. Initially I tried a grid resistor of 1meg and this reduced the output by a fair margin I also tried a small forward bias and again reduced output. So after your suggestions I have now got a 10 meg from G1 to ground and an 18 meg from the diode anode to ground, this has removed the little bit of distortion on audio I have and weaker stations are now at a higher level. I did realise the DC conditions for the diode were wrong. Incidentally although I have lots of germanium diodes I don't have any 1N60's is the 1N60P a good diode as these are available on E-bay?

I also tried another set of Hi-Z phones which have a DC resistance of 2K that gives more volume than the other set that were 5.5k.
Here in Kilmarnock I can receive BBC Ulster at good volume and Spirit Radio more quietly but still with sufficient volume to listen to. I even can pick up Radio Star country but its too weak to really get any entertainment value from.
This of course is with just a ferrite rod aerial.

The modern 1N60 is the absolute best for low level AM detection in terms of sensitivity. Hugely better than OA47 or any signal Schottky. Obviously there is an upper frequency limit and it has more leakage than 1N4148 etc. The current ones don't use the same package or process technology as the 1946 version.
Also works as a pair used in FM detectors, but the PIV can be exceeded. The silicon 1N4148 is fine for most FM detectors that use two diodes.

The lab I had tested many diodes for a beacon RSSI application to set RX agc and TX power on a full duplex radio link.

Brilliant.
Thanks Michael. Will order some today.

You can try replacing the 18M input resistor with an RF choke.
Or a second diode ( anode to ground, cathode to existing cap & anode) and then the only RF and AF load is the grid resistor. Once the input capacitor is charged the only AC or DC load is the grid resistor. Same is true with an RF choke.
The anode load of V1 can be a ferrite ring (or regular inductor at right angles to ferrite) and then a 2nd gang equal to tuning cap, if the inductor equals ferrite rod, will tune. It can be earthed, same shaft as the HT rail is decoupled and a dead short to RF and AF (I hope!).
The modified diode circuit, your own or any of my suggestions, will only load Q of the anode coil and 2nd tuning cap by the load resistance of the grid resistor, especially if the dual diode circuit is used. It's more complicated if a resistor, choke or transformer is used.

Mike Watterson Wrote:

Amie Wrote:Just one question though, I've seen diodes used as in Trevor's diagram occasionally, But what is its purpose? Is it acting as a clamp?

(There is no DC continuity at its anode so is therefore just floating?)

No, I've never seen a single diode in an official circuit like that, because as you say there is no obvious DC path anywhere on the cathode or anode. Though I can imagine theories to explain why it works, as outlined earlier.

.... Just wondering, I know that it does work, after a fashion, I had the experience with the Grundig 80U where the two FM discriminator diodes were missing removing any DC continuity from the AM detector.... Yet it still worked on AM, after a fashion as I said.

Amie.

Oh, and on my suggestion of tuning the anode coil, the capacitor to the diode is in parallel at RF, so needs to be as small as possible, or use a dual gang with that capacitance as padding on the Ferrite rod (smaller tuning range though).

Interestingly I tried a 10 inch speaker and output transformer yesterday and at least two stations were listenable at a pinch.

Amie you are correct about the lack of a DC path on the detector diode. I have also seen some sets with no DC path on some manuals but the set appears to work fine.
In this case the input impedance to the AF amp is extremely high and even adding a 1 meg grid resistor reduced output greatly, only when the values went above 10 meg did I see any improvement.

The audio stage has very good gain indeed and I suspect that the RF stage with the positive feedback is producing very good gain as well, for an extremely simple TRF the results are spectacular, the current drawn from the 9v supply is tiny (I'll measure it today) and the volume is considerably more than I would have imagined. In fact yesterday when driving the speaker my mobile phone microphone was able to pick the audio up just fine.

Mike Watterson Wrote:

Amie Wrote:Just one question though, I've seen diodes used as in Trevor's diagram occasionally, But what is its purpose? Is it acting as a clamp?

(There is no DC continuity at its anode so is therefore just floating?)

No, I've never seen a single diode in an official circuit like that, because as you say there is no obvious DC path anywhere on the cathode or anode. Though I can imagine theories to explain why it works, as outlined earlier.

Agree both!

The diode is working as the detector, but doesn't have an obvious DC path.

Mike W's post #28 (quotes are painful with iPad) advocating resistor on diode input to ground, and diode output to ground, would help. Even better, I'd suggest a second DIODE to ground at the diode input (ie anode of D1), rather than a resistor, connected anode to ground, cathode to D1 anode, because then, with the 270pF capacitor, it would form a voltage doubler or charge pump (choose your name!).

Trevor has built it, and it works. So there MUST be a discharge path, and in fact there is, the leakage in D1. But this is essentially bad design as it relies on a loosely-specified, 'nuisance' characteristic of the diode. Moreover, even for a particular specimen for D1, leakage is not constant - it halves for every 10 deg C drop in temperature. So outside in the winter, I'd expect audio distortion as the leakage in the diode falls to such a low level that the grid voltage of the output valve can no longer follow the modulation envelope.

Absolutely Kalee.
Hit the diode with freezer and hear the results or even solder it when on!!!!
So far using the two very high value resistors do improve matters. Using a 2mf choke on the anode to ground reduces volume and I suspect it's the chokes inherent capacitance that is damping the RF from the first stage. I'll try an old fashioned one tomorrow that will be lower capacitance than the small modern one I tried.

Once I get the 1N60s I'll try the reverse diode to ground etc.