Push-Pull (full wave) diode detector.

[ukcol]

By the early 1950s domestic AM superhet receiver design had evolved to a point when one manufacturers circuit design was pretty much like another’s. There were some differences of course, particularly in more "up market" sets.

During the years of the development of the superhet there were many more differences between designs, one of which was the use of a full wave diode detector to recover the audio from the I.F.

The only example of this, that I am aware of, is in the 1932 Murphy A8. Unlike the familiar singe diode arrangement, the last I.F transformer has a centre-tapped secondary feeding two diodes (an AC/DD valve).

The point I want to raise is that the advantages of this arrangement seem (to me) a little obscure but perhaps that is why it did not find favour in later designs. The A8 uses a low I.F. frequency so perhaps this simplified the design of the post detector low pass filter? Also being full wave there would be no DC component in the last I.F. transformer’s secondary current. But would these points make any significant difference to performance or cost?

So here are 3 questions -

Have I missed something here?

Do you know of any other domestic AM superhets that use a full wave diode detector?

Is this technique employed in any commercial communications receiver designs?

[Leon Crampin]

Colin, an interesting question. My first reaction is that rectification efficiency has no significance in a detector if the load impedance is high in relation to that of the source - which it nearly always is.

In the case of a power supply, the load current is usually significant and hence the higher efficiency of a full wave configuration is generally used, unless circuit constraints such as those existing in AC/DC sets prevent it.

So, to directly answer your question, my understanding is that because the rectification system in a detector is not supplying a significant amount of current, there is no advantage in a full wave system. As an aside, half wave power supplies were often used in battery sets powered from the mains, signal generators, band III convertors and the like - all of which draw very little current.

I would be similarly unconcerned about DC in the IF transformer magnetic circuit. With low levels of current, relatively few turns on the transformer and a huge air gap in the mag circuit which would not saturate, I can't imagine that this would ever be a problem.

Leon.

[ukcol]

Hi Leon

Thanks for the reply. I'm glad it's not just me.

I've just been looking at the A8 circuit again and there is nothing particularly unusual about the detector load impedance value.

The set does however have an RF amplifier stage AND an extra I.F. stage, so there will be a lot of gain to spare and a larger signal at the detector than would otherwise be the case. Perhaps this has some significance?

Another point that has occurred to me is that in 1932 the superhet was only 4 years old, so perhaps this is an example of cautious over-engineering?

[ukcol]

I'd like to add a further question to those asked in thread #1.

Is the full wave version of the AM diode detector better at handling high modulation levels than the half wave one?

Any more responses to this and the previous questions would be much appreciated.

[quantum]

A 'half wave' diode detector should be just as effective as a 'full wave' one, so long as you get the diode load right. Modern broadcasts with high modulation might mean you have to change the load on some old sets.

I haven't come across p-p diode detection but I have seen detector voltage doubler circuits discussed in an old American magazine. The advantage with this is that you could feed a high slope output valve without an intermediate AF stage. I've tried this circuit myself but I can't say I noticed any improvement...

[PJL]

These are quite early AC valves and the set is going to have an extended bass range so I wonder if it helps reduce hum?...if the set is yours why not try disconnecting one of the diodes?

[Leon Crampin]

Thinking about this some more as I replace the (numerous) Hunts capacitors in my newly acquired VHF64, an interesting circuit configuration which was never tried to my knowledge, came to mind.

Why not have two half wave detectors, one with the diode polarity reversed. If there was no AF pre-amp (not necessary with a highish gain set and high slope output valves) a push pull output stage could be driven directly from the detectors without the need for a phase splitter. I suppose the downside of this arrangement with valve diodes is that one cathode would be "hot" to the IF and consequently its capacitance to the heater would both detune the transformer and may induce hum.

Pity. I might try it with Ge diodes in an idle moment though.

Leon.

[ukcol]

Originally Posted by quantum

........ but I have seen detector voltage doubler circuits discussed in an old American magazine.

That's interesting; I've never come across a detector using a voltage doubler arrangement.

Originally Posted by PJL

...if the set is yours why not try disconnecting one of the diodes?

I wish I did own an A8 but they always seem to attract a higher price than I've prepared or able to pay.

This discussion has arisen out of my interest in circuits in particular, and that interest extends beyond circuits of sets I own and will often include sets I would like to own.

[ukcol]

Leon's idea of using 2 Ge diodes to obtain anti-phase audio signals for a push-pull amplifier sounds a good one. I can't think why it shouldn't work.

[JHGibson]

One very slight advantage of the full wave detector is that the high level ripple in the detector's output is at twice the IF frequency and will not cause instability if coupled back through wiring capacity to an early IF amplifier.
John.

[Herald1360]

Originally Posted by ukcol

That's interesting; I've never come across a detector using a voltage doubler arrangement.

I think it was quite common in sixties transistor reflex (regen) TRF receivers.

Chris

[radioman]

Originally Posted by Leon Crampin

Why not have two half wave detectors, one with the diode polarity reversed. If there was no AF pre-amp (not necessary with a highish gain set and high slope output valves) a push pull output stage could be driven directly from the detectors without the need for a phase splitter.
Leon.

I considered this arrangement a while back but came to the conclusion that unless both sidebands were identical, distortion would occur - so tuning would need to be spot on. This was only a paper exercise though and I didn't actually try it so I suppose I could be wrong !

Andy

[Herald1360]

Originally Posted by Herald1360

I think it was quite common in sixties transistor reflex (regen) TRF receivers.

Chris

Found a circuit example- doesn't use sixties devices, though.