Amplitude Modulation

[ Home ] [ Up ]

Heising Modulation

Viking Valiant Modulation modifications

One of the first things we must understand before discussing characteristics of AM is power measurement. First, there is no such thing as "RMS Power". We find power by multiplying RMS voltage times RMS current, but there is really no such thing as "RMS power". 

We also have equivalent or heating power. This is long term power over a defined period of time. It is the power that does some amount of actual work.

We have average power. Average power is same as equivalent or heating power with an unchanging power level, like a steady unmodulated carrier applied to a constant resistance load. If we close the key on a good stable CW transmitter, we will see the average power displayed on a power meter. It will not be the "RMS power".  It is not the peak power, it is the stable heating power level over some period of time. In this case it would be the time it takes for the envelope meter to reach full scale.

We have peak envelope power. Peak envelope power is the very maximum short term peak of either steady or varying average power levels!

Consider a sine wave with a peak voltage of 100 volts. The RMS voltage is 70.7107 volts, or 100 peak volts. If we placed that voltage across a 50 ohm resistance we would have 70.7107 / 50 = 1.414214 amperes. That would also be 100 watts average power in one complete cycle or any number of equal amplitude cycles that follow. The peak envelope power is also 100 watts.

If we pulsed that power off and on rapidly with a 50% duty cycle the average power would be 50 watts. The peak envelope power would be 100 watts!

RMS is calculated by squaring the value of a function, taking the average (mean value) of the squared function, and finally converting that mean value back by finding the square root of that mean or average. If we had a peak power of 100 watts with a 50% duty cycle the RMS power, if there was such a thing, would be SQRT( (100^2 + 0)/2) = 70.71 watts. We see that 70.71 watts is not the average power, is not the heating or "work" power, and is not the peak power. It isn't anything at all useful! We don't read "RMS power" with any of our instruments.

Characteristics of AM (amplitude modulation)

Let's consider the case of perfect undistorted sine wave modulation of an amplifier stage. The carrier, sidebands, and power levels of the various spectral components making up the signal have a certain ideal relationship. Consider the case below with symmetrical sine wave modulation.

Unmodulated carrier = 100 watts (PEP or) average carrier power.

Average is the same as PEP because, absent amplitude modulation, the carrier level is unchanging over time.

100% steady modulated 100 w carrier = 400 watts PEP or 150 watts average power. Of this 150 watts average power, 100 watts is in the carrier and 25 watts is in each sideband of the AM signal. Hence the carrier is 2/3 of the average power, the sum of power in the two sidebands is 1/3 of total average power under modulation, or 1/6th of the average power in each sideband with 100% modulation.

To 100% sine wave modulate a 100-watt carrier, a modulator must supply 50 watts of audio power. This audio heating or average power adds to the PA RF power for heating (or average power) for an average power of 150 watts. The peak envelope power is four times the carrier power because on audio peaks the carrier doubles in voltage, doubles in current, and quadruples in power.

If we monitor output current we would see a rise to 1.22475 times the steady state current, and also average 1.22475 times the steady state carrier voltage when steady sine wave modulation of 100% is applied.

We will see why these relationships hold true as we explore amplitude modulation. One word of caution, measured values are affected by the type of meter we use and the modulation waveform! Some metering schemes don't fully respond to peaks, and don't fully read the average either. This is a metering problem.

You will also not see the 100% sine wave modulated average power level changes with perfect 100% modulated speech, although PEP will indeed reach four times the carrier on a good sample and hold meter. This is because there is a wider peak-to-average power spread on speech when compared to the peak-to-average power ratio of a steady sine wave modulation. The only good meter to read 100% positive modulation peaks on is a true peak reading meter with adequate peak hold time.

Actually a peak reading meter with adequate hold time is a much more accurate indicator of 100% positive peaks than a spectrum analyzer or oscilloscope.

The best overall modulation percentage indicator would be a specialized device that would sample and hold negative peaks, and sample and hold positive peaks. This would not indicate bandwidth, only percentage of modulation! Actual bandwidth would only be indicated by using a peak holding spectrum analyzer, or in a pinch a very narrow bandwidth tunable receiver with a peak responding AGC and slow decay on a calibrated signal level meter.

Plate Modulation

Plate modulation is the most common method of obtaining amplitude modulation. In order to be plate modulated without distortion, the plate modulated RF power amplifier stage has to maintain a square law power output function with varying plate voltage. If anode voltage increases 50% from modulator power (50% modulation), the RF envelope output power must increase 225% over the carrier value. If the modulator doubles plate voltage (100% modulation), power output should quadruple. If the modulator doubles anode voltage the tube also doubles in anode current, meaning input power on audio peaks is four times the carrier power. Efficiency remains constant at a pretty high value, and so peak RF output power is four times the unmodulated carrier value!

Triodes

As a general rule only plate modulated low-mu or medium-mu triodes provide needed square law power response with anode voltage changes. This occurs only when a triode is operated hard into class C. A triode operating in this manner behaves like a very rapidly off and on switched resistance (switched at the RF excitation rate), and provides a nearly constant load resistance to the much lower frequency plate modulator system. Some of the cleanest, least critical to tune, high level modulated AM transmitters use low mu triodes.

We can see a low-mu triode, as plate voltage increases, has a substantial increase in plate current.  If we pick the correct loadline, the input power will approximately quadruple for every doubling of anode voltage.

For example at A5 we have 1 amperes at 2500 volts, or 2500 watts input power. At A10 we have 1.9 amperes at 5000 volts, or just under 10,000 watts instantaneous plate power.

Remember this is instantaneous power, since the anode is in very short pulses in class C. The average power is MUCH less.

Linearity is not perfect with the 100TL, but if we pick the correct operating loadline, the tube provides very close to square law response. If the modulator doubles anode voltage, peak power would nearly quadruple.

Let's  look at a tetrode.

Tetrodes

A plate modulated tetrode or pentode, without the aid of screen or control grid modulation, will not follow the square law rule. This is because screen grid voltage dominates cathode-to-anode current in a tetrode or pentode.

Let's look at a commonly use amateur tube, the 6146.

Curves A through G represent anode current as anode voltage is varied with constant bias and screen voltage applied. Notice how flat the plate current curves are as anode voltage is varied.

If the modulator doubles anode voltage in an ideal tetrode or pentode amplifier, plate current would not change at all! Most tetrodes are not perfect and will do a little better than this, but will still have considerable distortion if plate modulated.

If we plate modulated a typical pentode or tetrode like the 6146, the system will only achieve 50-60% positive peaks (200 watts PEP for a 100 W carrier) by the time negative peaks reach 100%. For example at -30 volts bias with 200 volts on the screen (curve G), anode current is about 100 mA whether anode voltage 200 volts or 700 volts. If only the anode was modulated, audio would be highly distorted. It would be impossible to obtain 100% positive peaks, and even negative peaks would be grossly distorted.  The easiest way to properly plate modulate a tetrode is to screen modulate at the same time the stage is anode modulated. By applying the correct proportion of modulating voltages to the screen grid and anode, with neither element actually modulating 100%, the system can come very close to producing the desired square law power response. The exact ratio of modulation applied to the screen and anode varies with tuning, loading, grid drive, tube type, and operating voltages. A properly designed plate modulated tetrode is actually not plate modulated, but rather is partially plate and partially screen modulated. We could also modulate the control grid along with the anode, leaving the screen fixed in voltage. Still, the most common plate modulation method of tetrodes, and the method that seems to work adequately, is a combination of screen grid and anode modulation.

There are two ways to modulate the screen and anode. One method "forces" the screen to follow modulated anode voltage by supplying the screen from voltage taken on the PA side of the modulation transformer. This method is shown below:

Components of note are:

C4 (screen bypass) must have high reactance at the highest audio frequency when compared to the parallel combination of R1 and R2.

C6 (screen supply blocking capacitor) must have low reactance at the lowest audio frequency when compared to R2.

R1 determines screen operating current and voltage.

R2 is adjusted in value to provide the best audio linearity at the design value of plate and grid operating currents and voltages. This resistor determines the amount of audio supplied to the screen grid.

A second method is to let the screen self-modulate:

Components of note are:

C4 (screen bypass) must have high reactance at the highest audio frequency when compared to R1.

C6 (screen supply blocking capacitor) must have low reactance at the lowest audio frequency when compared to R1.

R1 determines screen operating current and voltage. This resistor must be very high in value, probably over 5,000 ohms in most cases. If the resistor cannot be made high in value, a series choke that has very high reactance across the audio spectrum must be added in series with R1. R1 is often compromised in value to provide the best audio linearity at the design value of plate and grid operating currents and voltages. This resistor determines the amount of audio supplied to the screen grid.

Self-modulation of the screen works because screen current varies with anode voltage. As the modulator pushes anode voltage higher, screen current decreases. Decreased screen current causes screen voltage to increase. This system normally requires the screen supply voltage to be at least 50% higher than required. For a 6146, the screen source voltage should be at least 300 volts. Screen voltage is often sourced from the anode supply line.

Efficiency Modulation (Grid Modulation and Low Level modulation with Linear Amplifiers)

There are two basic systems that use efficiency modulation, grid modulation and linear amplifiers. When we vary current in a device, we do not have a square law response. With fixed coupling to the load, if we match at full peak power, lower drive levels cause a reduction in efficiency. This is because, unlike with plate modulation, anode supply voltage does not vary. With a conventional deep class C triode and plate modulation, the output device acts presents a constant time-averaged resistance value to the power source. If the modulator doubles anode voltage, anode current also doubles. Since anode voltage and anode current changes in the same proportion, they maintain the same ratio. Anode impedance is E/I, and if the ratio remain constant the anode RF impedance also remains constant. The output device sees the same match (or mismatch) at zero modulation as the output device sees at modulation positive peaks or modulation negative peaks, and with the same conduction angle efficiency remains the same.

Screen modulation, control grid modulation, or linear amplifiers all have a constant anode or collector voltage. This means output device impedance, or E/I of the output device, varies over the audio cycle. The device has highest current on modulation positive peaks, and lowest current on modulation negative peaks. This means the mismatch between the output device and the load varies over the modulation cycle. The normal tuning procedure is to match the output device at maximum positive modulation peak. As the modulation positive peak is reduced the output device has a higher impedance, and this mismatches the device to the tank. The result is a reduction in efficiency as the system moves below the peak positive modulation level, reaching minimum efficiency at maximum negative peak.

The approximate rule with 100% modulation is the device carrier efficiency is half of the device positive peak efficiency. Let's say we have about 70% efficiency at the anode of a tube with 4% tank and other losses for 66% total efficiency. At carrier, plate efficiency will be about 35%. With tank losses of 4% we have an overall efficiency of 31%. This means on carrier 2/3 of the plate input power will be heat, or twice as much heat as carrier power into the tank. With 500 watts PEP output on modulation peaks, the tube anode dissipates about 375 watts of heat.

This is true for screen modulation or linear amplifiers!

Many linear amplifiers with high conduction angles only have about 50% efficiency on peaks, plus the normal procedure is to slightly over-couple. A safe general rule for linear amplifiers is carrier level power dissipation is three times the carrier power.

This means a legal-limit AM linear could have about 1125 watts dissipation during carrier conditions of 375 watts, and on positive modulation peaks output power will be about 1500 watts with 1500 watts of short-term dissipation. This is a reasonable safe estimate.

If a conventional AM linear or screen modulated stage is making more than half the peak efficiency on carrier, odds are very good it has excessive distortion and splatter.

Quality of Audio

Low level modulation often has much less distortion and more fidelity than high level modulation of tetrodes, and more faithfully reproduces the audio input. It is much easier to have low distortion high fidelity audio using low level modulation. To be sure, some of the cleanest AM BC transmitters ever built were low level modulated systems. Unfortunately the low efficiency resulted in high energy consumption, causing most stations to use more energy efficient high level modulation.


The sole shortfall with a linear amp or grid modulation is efficiency. In order to reproduce the input faithfully, the amplifier has to be loaded to handle the PEAK power. This is normally four times the carrier power (or more in some cases). This is because the linear has to be "efficiency modulated".

A safe estimate is 25% carrier efficiency. This means your amp would be making three times the heat as carrier power. An SB220 can safely handle about 500 watts of steady dissipation (inadequate airflow to fully use the tubes) so it is safe at 125 watts carrier when properly tuned.

Very few amps will handle legal limit AM. That would be 375 watts carrier, and a safe estimate would be three times that for dissipation. That would be 1125 watts of heat, which would take a lot of air and a big tube. An 8877 at full rated airflow would work.

Contrary to myth, there is no difference in the sound of any AM transmitter when amplified in a properly tuned and operated linear amplifier. This is because a properly tuned and operated linear, be it a Heath SB220 or anything else, has much less distortion than the typical boat anchor rig. The idea a system has to be plate modulated at the transmitter's output stage for best sound is an absolute myth. The real problem with a linear is NOT the sound. That is an absolute myth. The real problem is heat.

A rig certainly does NOT need to be plate modulated to sound perfect, and as a matter of fact most amateur plate modulated transmitters have terrible distortion as a percentage of modulation. It's just that most people can't actually hear the distortion, they listen to and enjoy the frequency response and might actually "like" a little distortion, and they confuse distortion with good sound.

It's certainly possible to have bad low level modulation, but plate modulating a tetrode guarantees we have to do special tuning and add "circuit tricks" to avoid significant distortion.

HF SSB rigs on AM

Some newer HF transceivers are excellent on AM, with much less distortion and better fidelity than most older amateur AM rigs. The Ten-Tec Orion and the Yaesu FT-1000D are two examples of very good AM rigs.

Peak envelope power, with 100% modulation, is four times carrier power. For 100% modulation in a 100 watt radio, a 100W PEP radio must run 25 watts or less carrier. With a 100-watt radio, peak power on voice peaks should be held to 100 watts or less.

I used to use an IC-751A or IC-706 ICOM on AM. The problem with that ICOM, like with many HF SSB radios, is it uses the ALC to limit power. Turning the output power down or mic control up will not increase the percentage of positive peaks. This is often because the ALC system in most SSB rigs detects power. If we set the carrier power to 25 watts and try to modulate 100% (100 watts), the peaks cause the radio to reduce gain until peaks are back at the carrier power level. The positive peaks stay at 25 watts or so...and the carrier drops to 7.5 watts!

The cure is to run the power level all the way wide open and apply an external stable negative voltage to the external ALC input. Adjust the external negative ALC until the carrier is 20-25 watts, and then the mic control until you have 100 watts on peaks using a good peak reading meter.

An external ALC carrier control can be a 9V battery across a 500k pot in a voltage divider. The positive battery lead goes to ground, and the ALC output comes from the pot wiper. P1 goes to the EXT ALC jack on the radio.

Remember to disconnect the battery when using other modes or when not using the radio!

To use this circuit, run the radio's normal power control wide open. Adjust the pot for 20-25 watts carrier. Adjust the mic control for 100% modulation, or 100 watts PEP (on a 100W radio).

  as of 2/7/2010