6146 AM Modulation

[wc6w]

As part of a curious Amplifier project I'm seeking the proper percentage of Screen Modulation to achieve 100% Plate Modulation of a 6146.  The Screen will be sourced from an independent supply modulated along with the Plate supply. 

 Some associated data points for other Screen Grid tubes:

  The RCA 4CX250B data sheet states 55%.

  The Eimac 4-400A data sheet shows 70%.

  The Eimac Care & Feeding book says a range of 70-100%.

  The Orr Handbook indicates 60% as a starting point.

  And several references suggest that it should be disclosed on the manufacturers data sheet.   Yet, nothing I have seen defines the number for the 6146 ...

  Anyone know the answer? 

73 & Good morning,
   Marv WC6W

http://qsl.net/wc6w/

[KA2DZT]

Marv,

It's a 6146, modulate the screen from a screen dropping resistor off the modulated plate voltage.  Look in the RCA xmitting tube manual for the value resistor, depends on the plate voltage.  If you don't have a tube manual we can can tell you the value.

Fred

[K1JJ]

Marv,

A rough screen "starting point" is probably the best approach. Forget about precise screen % numbers for now and begin with a screen dropping resistor based on recommended  plate modulation screen current..

Once you get the rig running right, then put an audio tone through the system and monitor your THD and IMD on an RF  spectrum analyzer.  A simple $25 SDR band scope will do fine for testing this linearity.  

Then adjust your plate loading, grid drive current, screen current, % of screen modulation and any other available parameters to find the overall sweet spot for cleanliness.   This approach will give you the best modulation and cleanest signal.  Both THD and IMD should be optimized to -30db down or better.  

Each rig, even with the same tube line up, will be different and require slightly different operating parameters for optimization. This is why all the handbooks disagree with each other concerning optimum screen modulation %.  Each situation,  testing method, tank circuit Q's and parts lineup was probably different.  

Adjust everything for highest / cleanest audio peaks,  best THD and IMD and you can't go wrong.  It's always a compromise of playing with parameters until you find that overall sweet spot - especially when adjusting something like % of screen modulation.

T

[AB2EZ]

Marv:

I agree with Tom's comments

Separately, for a tetrode or a pentode (like a 6146A), the plate current is not very sensitive to the plate voltage. The plate current is roughly proportional to the screen voltage.

For that reason, it is the modulation of the screen voltage that is really producing the output RF envelope modulation... even though we all call this "plate modulation"

You should expect to be modulating the screen voltage close to 100% (as would be the case when the screen voltage is obtained from the modulated B+ via a dropping resistor). But, as Tom pointed out... if you can adjust the percentage modulation of the separately modulated screen voltage... you can start with a value of around 70%... and adjust it to produce the best linearity of the RF output envelope v. the audio.

I've done this with a DX-20, and I did find an optimal value... but I don't remember what it was. I believe it was close to 100%. With my Ranger, I sometimes use an external modulator that gives me extremely good fidelity between the modulated RF envelope and the audio input. In that case, I am still obtaining the screen voltage from the stock Ranger plate-to-screen dropping resistor... so the % of screen voltage modulation is essentially the same as the percentage of plate voltage modulation.

The modulation of the plate voltage, in the case of a tetrode or a pentode, primarily serves the purpose of keeping the output efficiency high, not modulating the plate current. I.e. when the peak RF plate current is lower, the peak RF plate voltage swing is lower... and you need less DC/audio frequency plate voltage keep the peak, downward RF plate voltage swing from driving the plate voltage down to 0 volts. By modulating the plate voltage, you keep the output efficiency above 70% for all values of RF output power.

Stu

[N8ETQ]

  Hi,

  A ranger uses 30K. You'll need a clamp..

/Dan

[K1JJ]

BTW,


Dean, WA1KNX talks about how to adjust the ratio between plate and screen modulation to get the best linearity for a 6146B.  He certainly found the sweet spot. It goes to show that linearity can be improved with some work, rather than accepting whatever comes with stock circuits and "normal" parameters.


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

T

[AB2EZ]

Tom

I think the approach that WA1KNX applied in his article is a perfectly good way to make improvements in the linearity of a plate modulated transmitter... adjusting the mix of the total modulated screen-to-cathode voltage that is obtained by connecting the screen to the modulated B+ (via a high-value resistor), and, separately to the un-modulated B+ (via another high-value resistor).

However, it would be (in my view) incorrect to assume that Dean's approach reduced the modulation of the screen voltage to 60%.

One has to keep in mind the "self-modulation" of the screen voltage that occurs when the screen is connected to the B+ (or whatever is used as a fixed screen voltage supply) with a high impedance in series. This can be a high value resistor and/or a suitable choke.

One should also keep in mind that the differential impedance, looking into the screen (i.e. the change in the screen voltage / the associated change in the screen current, when one makes small, incremental changes to the applied screen voltage), is not constant over the range of screen voltages that will applied during modulation. Therefore, the series combination of: a fixed value voltage dropping resistor (for example, 30k ohms in the case of a Ranger) and the impedance looking into the screen (which is in the ballpark of 15k ohms for a 6146, but varies above and below that value, depending upon the actual screen voltage) does not form a constant ratio voltage divider.

When the plate voltage moves up and down, there is a change in the percentage of the electrons ... traveling from the cathode toward the plate... that get captured by the screen. If the plate voltage goes down, a larger percentage of the electrons, that are emitted from the cathode, are captured by the screen. If the impedance from screen-to-screen supply (at audio frequencies) is high, then these electrons "pile up" on the screen... and, therefore, reduce the screen-to-cathode voltage. Likewise, when the plate voltage goes up, a smaller percentage of the electrons that are emitted from the cathode will be captured by the screen. The impedance between the screen and the screen voltage supply will leak off (with an audio frequency time constant) electrons that are already "piled up" on the screen... and the screen-to-cathode voltage will increase (become more positive). This is what is referred to as "self-modulation" of the screen-to-cathode voltage.

For this reason, the screen can be connected (via, for example, a 15k-30k ohm resistor and/or an 8H-12H choke) to a voltage source that is not modulated at all... and, due to self modulation of the screen voltage, one can still achieve full modulation of the RF output envelope, and excellent linearity.

If the screen voltage is being supplied by a low impedance source, e.g. the output of a separately modulated screen voltage supply... then the screen-to-cathode voltage will faithfully follow that source. There will be no self-modulation, because additional electrons that are captured by the screen, due to the modulation of the plate voltage, will immediately be removed by the low impedance screen voltage supply.

This is an important difference between screen voltage modulation using a separate, modulated, low impedance source, v. all of the other methods that incorporate a high impedance (at audio frequencies) between the screen and the source of screen voltage.


With the screen-to-cathode voltage faithfully following the low impedance, modulated screen voltage supply, it will be easy to determine the required percentage of screen modulation, to achieve 100 RF output envelope modulation... and also to determine the associated relationship (i.e. linearity) between: the mix of DC and modulated screen voltage, and the RF output envelope.

Likewise, it would be easy to monitor the screen voltage with an oscilloscope to see what the actual modulation of the screen voltage is... when using any of the methods.

I believe that the required percentage of actual screen-to-cathode voltage modulation, to achieve 100% RF output envelope modulation, will always be close to 100%... regardless of the method used to modulate the screen voltage (including only self modulation).

Of note, with my Ranger, using an external modulator to provide modulated B+, and using the stock 30k ohm screen dropping resistor... the envelope of the Ranger's RF output nearly perfectly tracks the amplitude of the audio input to the external modulator. I observe this with a 2-channel digital oscilloscope. Triggering off of the audio input waveform... it can capture, and hold (until the next triggering event) the audio input waveform, and the RF output waveform. With proper vertical scaling, and a suitable vertical offset, I can superimpose the audio input waveform on the upper edge of the RF envelope. The tracking is nearly perfect.

Stu

[wa3dsp]

When I restored/modified my Ranger I used a mixing circuit as well as a modified clamp.

See  -

http://www.crompton.com/hamradio/JohnsonRanger/RangerRestoration.html

The photos and schematic show the resistive divider I used.

The link to the writeup and formula is on the page but here it is direct -

http://amfone.net/Amforum/index.php?topic=32893.0

[N2DTS]

I always got the best results from a mix of modulated and unmodulated voltage to the dropping resistor when the screen voltage was taken from the high voltage via a screen dropping resistor.
Some rigs did not seem to want to modulate well positive at all stock (32V3) without changes.

With a fixed supply, you can use a choke, and/or a resistor, and set or adjust the voltage and resistor to protect the screen and adjust things for the lowest distortion.
On my big rigs, I use a big pot and a variac so I can play with things while looking at the distortion.