Interesting Output Stage?

[pwdrive]

No, I haven't got one but came across this:

http://www.radio-workshop.co.uk/service/...-658rg.pdf

A bit unusual....

Lawrence.

[Murphyv310]

Hi.
I remember this circuit being discussed elsewhere. At first glance it looks odd to say the least but once you grasp what's going on it falls into place.
I've never come across one of these in the flesh and I believe they weren't popular.
Would be interesting to hear if they were any good.

[Mike Watterson]

Saba does this (or very similar) to have push pull mono and single ended stereo on many radio models. The bias on the Saba changes appropriately, so push pull mono is a bit more than twice the power. Only the later versions had even an option for a stereo decoder, the earlier versions would only have used stereo with a separate reel to reel tape or record player. Hence the mono mode was more "important".

Yes it did puzzle me for a moment.

The Saba sounded very good indeed, but the motorised automatic station search was a bit dodgy due to bad contacts on the search button bank. It also had funky adjustable FM IF bandwidth (dealer/factory setting, not for users!)

The other odd output scheme on one Pye and many Grundig battery mains sets is EL42 or similar sharing a transformer (different taps) with a DL94 or DL96, so you get MUCH more volume on mains.

OTH the rather mad Philips LX527AB (various odd features and a later model swaps ONE of two DAF91 for a DAF96 so as to add a DM70!) has two DL94 in push pull, at slightly less than normal 90V HT due to NINE volt LT (to ensure nice bias!). But on AC mains the HT for the DL94 is higher than the rest of the set and is I think about 138V*! It uses more negative bias on mains and close to class B and maybe slightly beyond limits of the valves (but they are Philips valves, right?) (It's certainly loud on mains.)
http://www.radiomuseum.org/r/philips_lx527ab_2.html
It's got a very unusual aerial system too. No need for a SW whip!

[* Edit may be 160V HT, then -12V = 148V !]

[pwdrive]

The origins of that Ferguson amplifier might be in the article linked below, a full description is given, I've vaguely got it but not quite.

Audio mag, Oct. 1958, mag page 19:

http://www.americanradiohistory.com/Arch...58-Oct.pdf

Lawrence.

[Mike Watterson]

Yes, Lawrence, you can see how Saba adapted it to have switchable push pull mono (high power lower distortion as 2nd harmonics cancel and most listening) and single ended stereo (less than 1/2 power per channel, only some records and only if external player available).
I think a Grundig uses a similar circuit to Ferguson. Their Majestic brand in USA was one of the first European made stereo Radios (MPX Radio Stereo came later in Europe than USA).

Thanks for the 1958 link!

[Synchrodyne]

Here are some other references to this and similar kinds of amplifiers, sometimes described as “orthogonal”:



1954 November; Audio magazine; p.36ff; “Simplexing a Standard Amplifier for Two-Channel Operation”.

1958 November; Audio magazine; p.22ff; “For Stereo, The Bi-Ortho Circuit”.

1958 December; Radio-Electronics magazine; p.41ff; “2-Way Stereo Amplifier”.

1959 February; Radio & TV News magazine; p.41ff; “Simplexing for Low-Level Crossover”.

1959 February; Wireless World magazine; p.80; "Single-Amplifier Stereo”.

1959 October; Wireless World magazine, p.445; Brief description of the Emerson 502A record reproducer, which had an orthogonal output.

1960 February; Radio-Electronics magazine; p.44ff; "RCA’s 2-Way Stereo Amplifier”.


The above magazines are all available at:  http://www.americanradiohistory.com.


The technique was also described in “Radio and Television Engineers’ Reference Book”, 4th Edition, 1963.  See p.14-47, which refers to its use in some Murphy stereograms.  The A674SR was one such model.

It appears to have had a small place in the early days of stereo, and in lower level equipment.  One may see how an equipment maker who previously used push-pull outputs in say mid-level mono radiograms might baulk at the idea of having to use a pair of such outputs in similar level stereograms, and would have welcomed the orthogonal approach as a way of retaining some form of “push-pull output” claim without hugely increasing the valve-count.  On the other hand it probably had little appeal to the hi-fi component makers, as it was likely seen as being a “hard sell” into a market of more knowledgeable consumers.

Within a short time though transistors arrived, with fairly early infiltration the mass market equipment.  Given that transistor power amplifiers were mostly output transformerless, there was no place for the orthogonal approach, which used the output transformers for dematrixing.


The “Simplexing for Low-Level Crossover” was an interesting variation.  It referred to what would now be called “bi-amping”, with an electronic crossover ahead of the bass and treble channels.  But instead of separate amplifiers for each of these channels, a single amplifier was used orthogonally.

Cheers,

Steve

[Synchrodyne]

The technique was also described in “Radio and Television Engineers’ Reference Book”, 4th Edition, 1963.  See p.14-47, which refers to its use in some Murphy stereograms.  The A674SR was one such model.

Here is the pertinent page from that book.  It provides a succinct explanation of the technique.

     
 
I wonder if any maker used this circuit with an ultralinear (UL) output.  It should have been possible, as the use of an UL main transformer would not have interfered with the basic process.  Also, UL outputs, both push-pull and single-ended, were sometimes used in the class of equipment in which orthogonal amplifiers were found.
 
Cheers,
 
Steve

[Kalee20]

That's an interesting concept!

If the p-p transformer was UL, then the screen-grid currents would all flow through the difference transformer. So for each transformer, they would contribute to the output. However, if you applied a pure mono signal at the input (both grids driven in anti phase), the thing would look like a UL push-pull amplifier; if you applied the mono signal with both grids driven in-phase, it would look like a SE triode amplifier! All rather horrendous when it comes to trying to optimise the transformers, calculate output power, etc!

[Synchrodyne]

I think though that for the most part, the single-ended component (common-mode or L-R signal) would be of lesser magnitude than the push-pull component (differential of L+R signal), and that the latter would very rarely approach zero.  Then the net effect of the signal-dependent varying load distribution between the push-pull and single-ended transformer primaries would be more akin to that of sliding the UL screen grid tapping points.  Given that the conventional wisdom is that the UL “benefit curve” is reasonably flat over the range from around 20% to around 50% tapping point, in practice the problem may be milder than the extreme-case analysis indicates, particularly if a tapping at the lower end is used.  It falls into something of a grey area, but if we can think of this possibility today, then the chances are that someone thought of it back in the late 1950s and also did some work on it, at least to the go/no go stage.

Another article on orthogonal amplifiers, and one which addresses some of the obvious questions, is:

“Single Push-Pull Stage for Both Stereo Channels”; Radio & Television News for 1959 January, page 48ff, available here:  http://www.americanradiohistory.com/Arch...959-01.pdf.


Cheers,

Steve

[Synchrodyne]

Another possible pathway to applying U-L to an orthogonal amplifier would be to use a separate screen grid primary for the push-pull transformer, and take the centre-tap from this to an appropriate primary tapping (or to a separate winding) on the single-ended transformer. This way distributed loading would be maintained for all input signal conditions. But that is possibly getting into self-defeating complexity.

Cheers,

Steve