Series Modulation

[kx9dk]

Has anyone built one of these besides WB9ECK?

http://www.amwindow.org/tech/htm/wb9eckseriesmod.htm

Looks like something I'd like to try with my 807 rig.

A little confused over the 6LF6 tubes -- the schematic shows 2 grids and this spec sheet shows 3 grids.

http://www.nj7p.org/Tube4.php?tube=6LF6

Which grid doesn't get used?

Thanks,
Dave KX9DK

[K3ZS]

I might be all wrong on this, but I from what I remember, the grid closest to the plate is the suppressor and it is usually grounded.   I made a series modulator for a DX-20 many long years ago.   It used a single tube in series with the final cathode and a 12AX7 mike amp.    It had excellent sounding audio.  The circuit I used was easy to implement because it required no other voltage except for the filaments.

[steve_qix]

Series modulation works extremely well if set up correctly. All of the class E rigs use series modulation.  Pulse width modulators are series modulators - used in broadcast stations everywhere.  Class H modulators are series modulators.  They are just about the best modulators you can use!

In the specific circuit, there are two R/C time constants that are a little small - at least in my estimation - remembering that I like a flat, no phase shift low end down to single digits!  The 1uf looking into 10k, and the .05uF looking into the 470k resistor. In place of the 1uF cap, I would probably use a 5uF or 10uF capacitor (there is very little voltage, so a 10uF 16V cap should work just fine), and the .05 could be increased to a 1uF.  Yes, it will work just fine as it is - I'm just being a bit of a perfectionist 

It is also easy to implement a negative peak limiter with a series modulator.  I can offer some suggestions if you're interested.  Might as well go all the way 

Also, a loop of negative feedback would probably not be a bad idea.  You could go from the modulator plates (through an appropriate resistor and DC blocking cap) back to the cathode of the first stage if you have sufficient audio voltage at the output of your audio system to tolerate the loss of gain in that first stage.  A small resistor in series with the cathode bypass capacitor at that point would allow you to introduce the negative feedback without giving up all of the gain of the first amplifier.... just some suggestions....

[Opcom]

A nice low power series mod transmitter can be made using one of those old variable regulated high voltage power supplies for the plate voltage. They have plenty of pass tubes. Audio can be injected near the voltage control pot. Some bypass caps have to be removed. The candidate supply ought to be one that does not use a variac, just a pot.

[The Slab Bacon]

A nice low power series mod transmitter can be made using one of those old variable regulated high voltage power supplies for the plate voltage. They have plenty of pass tubes. Audio can be injected near the voltage control pot. Some bypass caps have to be removed. The candidate supply ought to be one that does not use a variac, just a pot.

I have often pondered that or something similar. Hmmmm..............

As far as the series carhode mod rig. To work well you need a mod tube (s) with a low Mu and a low plate resistance, capable of carrying enough current to keep the final working hard. That is most likely the purpose for the 2 parralleled triode-connected sweep tubes. I still think a 6080 / 6AS7 (or a pair) would work even better. It is a low Mu triode with a very low plate resistance. Some years back they were a rather expensive tube, but nowadays not much uses them anymore, and they have become cheap and somewhat plentiful at hamfests.

[WA1GFZ]

A couple 4000 volt IGBTs with a little emitter resistance would work fine.

[IN3IEX]

Be careful with 6080... low max anode voltage. Instead 6LF6s are very good.

My personal opinion is:
http://www.amwindow.org/tech/htm/wb9eckseriesmod.htm   is a very good Hi-Fi design but lot of dissipation.
Traditional transformer plate modulation: most efficient and best solution. Only PWM modulation is more efficient, series class H is more efficient than trasformer modulation if you use not less than three supply voltages.

[R. Fry SWL]

Probably the most efficient AM transmitter in commercial production to date is the "DX" series, invented by Hilmer Swanson of Harris Broadcast Division in Quincy, IL, USA.

This design uses digital techniques to generate the DSB, full-carrier AM waveform used by licensed (analog) AM broadcast stations.  The a-c input to r-f output efficiency of the complete transmitter at 100% sine wave modulation is over 85%, including the a-c power for its internal blowers and all other a-c loads within the transmitter.

Below are some quotes from the product brochure.

• Digital: Harris DX transmitters have Direct Digital Synthesis of
  the RF envelope using true digital modulation, not PDM.
• Power Consumption: DX 25U: 29.1 kW or less typical at 25 kW, 0% modulation;
  43.6 kW or less typical at 25 kW, 100% tone modulation. DX 50:
  58 kW or less typical at 50 kW, 0% modulation; 88 kW or less typical at
  50 kW, 100% tone modulation.

R. Fry

[steve_qix]

Analog series modulation is inherently inefficient, but for low power transmitters run off the AC line it's probably OK.

The modulator tubes (or MOSFETs / IGBTs) can be anything that will otherwise handle the voltage, current, power and have sufficient dissipation.

The required dissipation for the modulator devices can be calculated:  For example, if one wants to run 25 watts input with something like a 6146 in the final, a voltage of 400V at 62.5mA would be reasonable.  The resistance of the RF amplifier is 6400 ohms which is very reasonable for a wide variety of modulator devices.  The load on the modulator IS resistive and constant.

To allow for positive peaks, the total power supply voltage should approach 3 times the carrier DC voltage.  That will allow close to 200% positive peaks.  Ok, for 400V on the 6146 final at carrier, the power supply voltage should be 1200V to allow close to 200% positive modulation.

The power dissipated in the modulator is the voltage drop across the modulator (800V) [that is, 1200V power supply voltage, minus 400V across the RF amplifier at carrier] multiplied by the carrier DC current which is 62.5mA.  That works out to be 50 watts dissipated in the modulator.  The tubes or MOSFETs / IGBTs should have at least a 50 or 60 watt dissipation.

There are a lot of tubes that will work.  A pair of 807s will work nicely in this application.  A pair of 6146Bs would work.  6L6GCs are fine.  8417s - great! 7027As, etc, etc. In addition to audio and typical amateur tubes, all manner of horizontal sweep tubes are out there, and these would certainly work.  Tetrode / pentode connection is fine.  It's a class A amplifier working into a resistive load, so there is nothing particularly unusual about the tubes or solid state devices needed to do the job.

I have built a *lot* of series modulators.  Class A series modulators are inefficient, but they work extremely well and are simple to get working.  Most people try to skimp out on the high voltage, and run less than is needed.  Don't do this.  Run enough voltage to get a lot of positive peaks - 3 times the carrier DC.  The thing will sound great  

[The Slab Bacon]

Be careful with 6080... low max anode voltage. Instead 6LF6s are very good.

Keep in mind that in series cathode modulation the voltage is not that high, it is the current handling capability that is more important. And the linear response created by the low full-on plate resistance.

[steve_qix]

Be careful with 6080... low max anode voltage. Instead 6LF6s are very good.

Keep in mind that in series cathode modulation the voltage is not that high, it is the current handling capability that is more important. And the linear response created by the low full-on plate resistance.

As long as the modulator tubes are sufficient to handle the power, have enough dissipation, and are operated in voltage and current ranges for which they are designed, everything will work properly. 

The class C amplifier, whether the modulation is applied to the negative side (cathode side) or positive side (plate side), is a resistive load, and looks to the modulator just like a resistor.  The voltage and current that the modulator tubes handle in a series modulator are similar to those encountered in a standard transformer coupled modulator, except that the tubes do not have to overcome the (usually lack of) inductance of the modulation transformer.  So, the current handling capacity of the tubes does not have to be unusually high for the given power.

The main difference, as far as the modulator tubes are concerned is the overall power dissipation, which is higher in a series modulator than would otherwise be encountered.