The Amīcaflex 4 transistor reflex radio

[Geordie McBoyne]

from the February 1959 Radio Constructor magazine.  The "Solar Three" a miniature light operated transistor radio.
From the text:
"Apart from its fascinating ability to operate from light rays, the Solar Three adequately meets the normal requirements of a pocket transistor receiver in the personal class. It is completely self-contained, it is housed in a neat plastic case measuring 4 3/8  X  3  X 1 1/4in, and has it's own ferrite frame aerial. As mentioned above, it's output is fed to a miniature hearing aid insert.

The transistors were Ediswan make, TR1 XA103, as specified or it can consist of any similar transistor having an alpha cut-off frequency greater than 2Mc/s.  OC44? TR2 and TR3 may be XB104, OC71? Greater power will be available if an XC101 is employed in the TR3 position. OC72?

Geordie McBoyne.

[Murphyv310]

The Philips EE8/EE20 constructor kit had a 2 or three transistor reflexed radio using AF116 and either one or two AC126 depending if it was phone or speaker, no reaction though.

Single transistor version in photo

[Murphyv310]

The three transistor circuit:

[ppppenguin]

I had an EE8 kit when I was a kid and remember making many things, including radios.

[Geordie McBoyne]

From the June 1961 Radio Constructor magazine another three transistor TRF receiver.  The transistors are the "red spot" type equivalent to the OC71 and AC126. Power came from three midget Mallory cells, each cell providing 1.3volts approximately.  Note there is no load resistor between the detector diode and coupling capacitor.

From the text: On test, the receiver gave excellent results on both Home and Light programs with an aerial 2ft 9in long. The two h.t. cells provided 2.6V approximately, these cells lasting about 10 - 12 weeks using the receiver one hour per day.                                                                                                 
The set  works particularly well well on the St. Pancras to St. Albans train journey.

Geordie McBoyne.

[Amie]

.... Note there is no load resistor between the detector diode and coupling capacitor....
Geordie McBoyne.

Hi.

There are a few circuits that have been published that have a "floating detector". It's something that shouldn't work but often does. I've discussed it with the more knowledgeable, and the consensus amongst those who accept it works is that it's either caused by capacitance across the junction of a germanium diode or it's reverse leakage.

Some designs use an RF transformer to feed a single diode, or a capacitor and choke to ground, for DC reference, but the double diode detector provides its own reference and more output, though the cost of the semiconductors was a big consideration in the 1960s I suppose.

The circuit I've used in this project by using sand transistors rather than soot, means that the diodes also have to be silicon, otherwise the transistor would be biased off.

Thanks.
Amie.

[Murphyv310]

I'm sure I put a post on re the diode and it's explanation when used in the EE8.
I'll repost it in an hour or so.
I must have misposted it on another thread.

[Murphyv310]

Hi.
Heres the write up that I must have posted elsewhere or didn't send

[Kalee20]

It does seem odd to me that the designer didn't complement the single diode, with another, connected so as to form a voltage-doubler / charge-pump / whatever you want to call it.

It would remove the dependency on the non-ideal characteristic of the diode - which is identified in Trevor's attachment as making a big difference to detection efficiency, and which varies significantly from diode to diode, with temperature, etc.

The Amīcaflex does it properly...

[Murphyv310]

Hi Kalee.
Do you remember the two valve rod pentode TRF I made earlier in the year? It also had a germanium Diode detector that had no DC path, it actually worked very well as was, doing it correctly performance massively dropped, I tried all possible combinations of diodes, doublers etc with no success. Sadly due to an error I blew the valves up and subsequently I gave up with it.
Im considering replicating the Philips EE8 circuit and seeing if doing "detection properly" gives an improvement.