Germanium Supadetector Radio

[Argus25]

Recently I posted a thread on a radio made with 12V HT valves and an OC16 output transistor.

The radio images attached is another home brew job, again shortwave and medium wave. This one used a standard rotary switch for band change.

Although the chassis is the same size as the valve radio, the construction technique is quite different. This time the chassis is aluminium, not steel. The central part of the chassis was cut out so a pcb could be mounted on top. Stainless steel pressed nuts were used again though and a similar stainless front panel.

This is a hand drawn and home etched pcb (no computer aids for this one).

This technique resulted in a lot of space under the chassis (unlike the valve radio design) so Bulgin C cell holders could be mounted under the chassis to power the radio.

The RF transistors are AF178's and the audio is GET880 pre amp, AC188 driver and OC16 push pull output into a substantial Foster speaker.

In his case the dial assembly is Jackson brand. I added an acrylic light box behind the dial, it should have been deeper though to help diffuse the dial lamps.

The AF178 is an excellent RF transistor, great replacements for AF11x transistors. All the ones in this set are plugged into mil spec sockets with gold plating so they have not had any thermal soldering stressors.

This radio contains an unusual AM detector circuit instead of the usual detector diode, it is called supadetector and it can detect much lower level signals than the germanium diode without square law distortion.

Nearly every transistor in this set is germanium, except for one used as the L/O buffer to drive an external 50R load/cable for a frequency counter, and the silicon transistor in the supadetector circuit.

The supadetector circuit is based on turning the IF output voltage into a current (in the un-bypassed transistor emitter resistance) and driving an inductor which acts as a current source to drive the rectifiers. The fundamental idea for this came from Horowitz and Hill, except they had a different type of current source. Graph attached comparing supadetector with a germanium diode and an ideal detector.

[Al (astral highway)]

Another amazing result, Hugo!

There are many things I love about this, including the ingenious and beautiful way in which you have housed and secured the 3-gang variable capacitor spindle. And again, that OC16 heatsink is a work of art. Did you fashion it yourself and/or is it based on contemporary design?

The circuit theory is unfamiliar, I'll look it up!

[Bazz4CQJ]

Where did you acquire your constructional skills, Hugo? In the days when I worked alongside professional electronics engineers I was always aware that "their's was just a bit better than mine", and sometimes, they were good enough to confirm that for me .

B

[Al (astral highway)]

Originally Posted by Bazz4CQJ

Where did you acquire your constructional skills, Hugo?

I believe Hugo is an eye doctor/surgeon... doesn't get much more intricate than that!

I can see parallels in terms of the fastidious preparation, meticulous understanding and intricacy of some of the designs and projects. Just look at the Pong PCB, to take one example...

[kalee20]

It looks beautiful, I'll bet it works as well as it looks!

Tape and film PCB's are a pain to lay out, and even more of a pain to change - but they sure can look good!

I like the supadetector - constant-current drive into the diode, thus negating the diode drop, followed by current-to-voltage converter (a resistor). But there's an additional diode and a phase-splitting transformer so that at the zero crossing, one diode turns off, the other turns on.

Thanks for sharing this!

[Andrewausfa]

That just looks superb. Did you do the engine turning yourself?

Andrew

[Philips210]

That's outstanding work Argus25, a real joy to look at. You must get a lot of satisfaction building your projects.
It's great to take pride in your work especially when it's to this high standard.
I'm looking forward to seeing more of your completed projects.

Regards
Symon

[Herald1360]

Beautiful but I find the engine turned panel rather distracts the eye making the panel legends difficult to read. Maybe just the lighting?

[Argus25]

Bazz4CQJ, my construction skills are largely self taught from about 5 years of age, though I was fortunate along the way to have some very good mentors. The main ones were my older brother and the late John Stokes in NZ. Most of my early machines/radios & experimental TV's were made from scavenged parts from the local rubbish dump and other places, never getting the result I wanted, but this inspired me to work harder and use better materials and components when later in my life I could finally afford them and make my projects better. I developed an eye for high quality parts.

Andrewausfa, The engine turning on the panel was a part of it I didn't do myself. A fellow at a local machine shop had a tool to do it, which was merely a blunt end of a wooden dowel with abrasive compound in a milling machine bed arrangement. But the panel itself was moved on the bed by hand so if you look you can see many irregularities in the positions of the turning marks. Which turned out to give it a sort of home grown look on a background of a tidy build.

Herald1360, The legends are actually easy to read, part of the problem is the bright flash photo increasing the panel reflections around them and making any defects look more obvious. Under normal lighting it looks a lot better.

One thing I forgot to mention about this radio, it contains two large stud mounted germanium power rectifiers, made around the same time in history as the OC16, these are the OA31. They are on the lelt side hiding under the speaker magnet area. Since this radio runs from an internal 12V battery supply or an external 12V supply I used these rectifiers to mix the supply rails together so I would not need any switching.

I don't have a photo of these diodes on hand. But if you want to see a photo of these amazing metal & glass seal rectifiers from 1958, scroll to about page 6 of this article (where I used some to create a germanium equivalent of a Lucas regulator)

http://nebula.wsimg.com/16f83c1c70ef...&alloworigin=1

[Philips210]

Hi

In pic4, I see that the germanium diodes have coiled leads. I have come across this a few times and would I be correct in thinking the reason for this is to help in heat dissipation when soldering?

Regards
Symon.

[Argus25]

Philips210,

Yes those loops help the heat issue by extending the length of the lead but the diodes still need a clamp on the lead when soldering. Also the loop helps to reduce mechanical stress on the lead to glass junction area, if the leads are bent in and out a bit to fit the diode to the pcb. It is never good to bend a lead too sharply and too close to the edge of a glass signal diode. To create the loop the lead itself was held with round nose pliers to form the loop, so that there were no forces between the lead and the diode glass body doing that.

Al, the heatsink needed in this case to lift the transistors off the pcb so their leads could solder to turret lugs on the pcb, (unlike the one in the other valve radio where it needed to be thermally coupled to the chassis for more cooling as that one the transistor was in class A). The heatsink on this radio is ample for the 3 to 4 watt push pull output stage and barely warms up at normal listening levels.

[Argus25]

I forgot to mention, if you look very closely at the photo of the small supadetector pcb on the initial post, where the transistor leads enter the pcb holes, you will see that it has "plated through holes". I do this on most of my home made pcb's now, with a kit supplied by RS. Also the kit is very useful for repairing pcb's if pads have lifted where the holes are, or simply for strengthening the arrangement, say if you have a to-5 device on a pcb with a heatsink on it, or a wire entering the pcb.