Taylor Modulation System ?

[Jim KF2SY]

What exactly was the Taylor super modulation system?  Most of us are familiar
with the various diode "ultra" modulation schemes, but can't find any info on the Taylor system.  Was it an early form of balanced modulator DSB?  I know the late Don Hoisington (k4cjl ?) ran it with great success.

thanks,

Jim

[Todd, KA1KAQ]

Jim -

A quick search turned up this recent post by G:

http://amfone.net/Amforum/index.php?topic=10353.0

That might be a good starting point. It seems this has been covered a few times in recent years, but I may be thinking of the AMRadio email list too.

[k4kyv]

The Taylor modulation scheme is a high efficiency grid modulation scheme, sort of a variation of the Doherty or Terman-Woodward system, but without the quarter wave phasing lines.

You have two tubes in the final.  One is a "carrier" tube and the other is a "peak" tube.  Unlike normal grid modulation where the final must run at about 30% efficiency, the carrier tube is adjusted to run just below the threshold of saturation, at about 60% efficiency.  It is grid modulated the normal way, which means the negative half of the modulation envelope is produced without  distortion.  But the tube is intentionally adjusted to run out of headroom as soon as the positive peak begins. 

That is where the "peak" tube takes over.  It is adjusted to cut-off under unmodulated carrier conditions, but as soon as the positive peak begins, the peak tube takes over where the carrier tube left off, and supplies the additional power beyond carrier power to generate the positive modulation peak. If everything is adjusted just right, the plate current to the carrier tube drops a little under modulation, but the peak tube plate current, which is normally near zero, kicks upwards.

There are two ways to generate the necessary positive peak.  One way is to let the two tubes work into the same tank circuit, but run twice the plate voltage on the peak tube as on the carrier tube.  The other, is to run the same plate voltage on both tubes, but to tap down on the tank coil for the peak tube so that the tank circuit becomes a step-up transformer to supply the positive peak.

The system was described in CQ magazine sometime in the late 50's or early 60's.  I have never actually heard one on the air, but supposedly some hams experimented with it, but that was about the time everyone became obsessed with slopbucket, and it never got much attention.

[Bacon, WA3WDR]

It's odd that the articles about the Taylor system seem to have fallen off the face of the earth.  Evidently it was this sort of Class-B/Class C combination linear amplifier, with a Class B section for carrier level and below, and a sort of Class C section to handle everything above carrier level.  The idea was that it would be significantly more efficient than a regular linear amplifier in AM service.

An easier way to do pretty much the same thing is the Class B Series Modulator.  This has a linear series section running from carrier to zero, and another class B series section going to a higher voltage, and this section kicks in above carrier level to handle the positive peaks.  The RF section can be ordinary Class C, or class E as shown in this system:
http://amfone.net/21stAM/class_b_series.htm

This circuit shows a class B series modulator using transistors and operating at relatively low voltages, but this technique can be done with tubes at higher voltages as well.

[WU2D]

We need to make some new versions of some of these Avant-garde circuits.

What about a linear amplifier based on a Doherty with a combination Class AB Class E arrangement?

Or a complementary Symmetry Class E? That would eliminate two transformers allowing the input and output to be single ended. The only price is finding suitable P-Channel mosfet complements and a bipolar power supply.

Mike WU2D

[Ian VK3KRI]

We need to make some new versions of some of these Avant-garde circuits.

Theres a nice article on the new-fangled  doherty rig for 80M in Feb 1937 QST.  Pretty impressive given Doherty's paper was first presented in 1936.
                                                                 Ian VK3KRI 

[k4kyv]

It's odd that the articles about the Taylor system seem to have fallen off the face of the earth.  Evidently it was this sort of Class-B/Class C combination linear amplifier, with a Class B section for carrier level and below, and a sort of Class C section to handle everything above carrier level.  The idea was that it would be significantly more efficient than a regular linear amplifier in AM service.

It was more like a class-C grid modulated amplifier, control grid modulated, as I recall.  The carrier tube was adjusted just below maximum possible power out, and the peak tube was biased to cutoff at no modulation and during negative peaks.

I suspect there would be some transient distortion no matter how carefully everything was adjusted, and it probably had just enough spurious products to splatter a little.  No doubt much more difficult to tune up than a plate modulated final or conventional leenyar on AM.

Sometimes when I get a little time on my hands, I'll dig through my old CQ's and see if I can find the article.

I did a web search for Taylor modulation, but came up with nothing more than a few references to it, similar to what has been posted in this discussion thread.

[Bacon, WA3WDR]

I also don't know why it was called supermodulation, unless it was a better way to get positive headroom for voice polarity supermodulation.  Compared to a linear amp carrying 150% to 200% modulated AM, the improvement in carrier efficiency is remarkable.

[WU2D]

OK Frank,

I have been playing around in my notebook and have attempted to draw a complementary symmetry AM system. I had a reference on the RF section but had a hard time figuring how to modulate it. Anyway the structure is interesting looking at least.

I know that a circuit like this would have been considered impossible at RF frequencies not that long ago and as you know there is no analog to complementary devices with tubes.

The mosfets are 2 dollar types. I know it has been a hard sell to come up with true complements, but these transistors are close.

Make it real..

Mike WU2D

[Ian VK3KRI]

Sometimes when I get a little time on my hands, I'll dig through my old CQ's and see if I can find the article.

I did a web search for Taylor modulation, but came up with nothing more than a few references to it, similar to what has been posted in this discussion thread.

One of my books has a brief description, but possibly more usefully footnotes to:

Taylor, R. E:  Modulation System US Patent 2,282,347
Villard, O.G:  Overmodulation without Sideband Splatter, Electronics, vol 20, pp09-95, Jan 1947
Villard, O.G:  Supermodulation, an Evaluation and Explanation, QST vol 34, pp 33-40, Dec 1950

                              Ian VK3KRI

[Bacon, WA3WDR]

Thanks Ian!!!

With that info and a little digging, I found a source for a pdf of the patent itself:

Google patent search on 2282347 May 1942 Taylor

[Ian VK3KRI]

Thanks Ian!!!

With that info and a little digging, I found a source for a pdf of the patent itself:

Google patent search on 2282347 May 1942 Taylor

Thats great, because the Villard references are actually for a different form of 'Supermodulation' using plate modulation and a tube to supply RG output on negative modulation peaks below 0Volts.
I found one of the articles in the University library near where I work. Heres a link to it if anyones interested...
http://www.smoke.com.au/~ic/dont_index/SuperMod.pdf

                                                            Ian VK3KRI

[Bacon, WA3WDR]

Thanks again, Ian!  I just saved that pdf in my AM folder.

[WA1GFZ]

The HPSDR guys plan a RF generator board with RF drive and PDM pulse train to generate high efficiency SSB and AM.
I would think many of these old systems don't like to change frequency
or QSY in lid speak.

[Steve - WB3HUZ]

I would think many of these old systems don't like to change frequency
or QSY in lid speak.

Yup. That's why most of these never caught on in any significant way in Amateur Radio.

[Bacon, WA3WDR]

That's why I like the Class-B series modulator and a class C final for a classic analog transmitter design.  It is efficient, and it reduces the 'qsy' problem to the same as any other class C final.

The old analog technology is fascinating and fun, but nowadays digital techniques like PDM, class E and DSP are the state of the art.