The Amīcaflex 4 transistor reflex radio

[Amie]

Hi Trevor.

The final transistor in that circuit you posted has no bias stabilisation at all.

The two diode reflex detector is an excellent circuit though, the neat AGC trick it does can be seen when you watch the collector voltage rise as the current through the transistor falls on an increase in signal as the set is tuned.... Would make a good signal strength meter circuit....

Even crude bias stabilisation with a resistor to the collector is better than none at all....

Thanks,

Amie.

[Murphyv310]

The output transistor obtains its bias from the 27k resistor between collector and base, agreed no stabilisation, there were many circuits arranged like that in the late 50s and early 60s. A lot of early germanium transistors had substantial leakage and this rather crude method did actually work. Unfortunately these kits were actually quite expensive as such and adding proper biasing would have upped the price and added complexity. In those days kids didn't get expensive gifts unless you were born into a well to do family.

[Amie]

The output transistor obtains its bias from the 27k resistor between collector and base,

Yeah, sorry I must have been looking at a different diagram, Still bad though. A lot of circuits I've seen in old magazines had no biasing at all, I suppose they self biased with the leakage. I have experimented with 2N1302 and GS 109 germanium's but they seemed to be quite well behaved.

I found also that bad biasing causes a transistor amplifier to have non linearity distortion so the gain isn't constant.

In view of the cost of the crystal fuse in the fifties you'd have thought they would have spent a bit more on a few resistors

....

[ppppenguin]

.....but that could be due to rubbish modem chokes with high capacitance. Getting good RFCs these days is like looking for a needle in a haystack.

I dimly remember the choke in my EE8 kit. I don't think it was anything special, just a single section, fat in the middle. While I may still have a few EE8 bits (most likely one of the 150R speakers) I doubt I've got the choke. Putting DC through the speaker was a very clever thing to do.

[Kalee20]

As long as the DC flows in the right direction!

ISTR the better direction was considered to be that which biases the speaker cone slightly outwards, though to me that says the steady field produced by the DC in the coil is acting to demagnetise the field magnet.

DC voltage drop across speaker, and collector-base feedback bias resistor, is rather crude, agree Amie that a couple of extra resistors to safeguard the expensive 3-legged fuse would have been good practice!

[Geordie McBoyne]

Simple single band superhet radio receiver.
The Cossor model 561 "Traveller's Friend".
Only four transistors are employed. TR3 operates as a reflex IF and audio amplifier. 
Information for this set can be found in the 1958/59 Radio and Television servicing book.  The Cossor 561 was introduced September 1957.
From the Radiomuseum:  https://www.radiomuseum.org/r/cossor_tra...d_561.html

Geordie McBoyne.

[Murphyv310]

Hi.
I decided to build up the EE20 circuit onto a Chinese radio kit that was surplus to requirements, as I'd used the IFTs for another project.
Unfortunately I've no suitable high impedance speaker that fits. I also don't have an OPtx to drive the kits speaker successfully. Results are OK, good with an OA79 but less so with a IN60 which gives distortion due to bad clipping on the AF resultant output. Fitting a 1.5m between anode and deck helps but since doing the video I've fitted an OA79 and a reversed diode between the detector Diode anode and deck for a much improved performance.
Just a little bit of fun and reminder of how this little circuit works. Incidentally it works much better than the Sinclair Micromatic I had at one point.
https://youtu.be/e_w2wAHXf7E

[Murphyv310]

Hi Amie.
Sorry for stealing your thread and if the mods think that this should be split then it could be a good idea.
This circuit has just a couple of small mods that have improved things no end. The 15k damping resistor across the RFC is not required and improves output no end especially on weak stations, the addition of the second diode as a doubler improves volume again on weaker stations. The biggest improvement is just a length of wire. It goes from the AF116 collector and is left in close proximity to the aerial coil inducing some feedback, it works best between 1600khz and 750khz then tails off. No adjustment other than spacing is required. Tacking in a big out put tx and speaker gives ample volume to fill a room, much improved quality too.

[Kalee20]

Interesting, Trevor!

Adding the second diode should indeed give an improvement (ISTR you saying your rod pentode receiver didn't show the improvement - I'd be interested in further investigation; what happens if you squirt freezer on the diodes to reduce leakage; etc, if you get it going again). And you found an improvement, isn't it great when results follow theory!

The 'length of wire' laid alongside the aerial coil to produce some feedback: I'm guessing that phasing between tuned winding and base winding is important, else feedback would be in the wrong sense. Again, useful to verify!

[Murphyv310]

Hi Kalee
Indeed when I was working on the rod pentode radio freezing the diodes improved output as it does on a crystal set..
The feedback wire is phase dependent. It has to be placed in parallel to the bar not at 90 degrees and works more reliably at the pick up winding end.