RnR Engineering

Broadcast Services and Design

IMPROVED AM MODULATION MONITORING TECHNIQUES

White Paper

 

The classic approach to monitoring AM transmitter modulation is to just attach the mod monitor to the transmitter’s modulation sample connector.  These sample connectors are driven from the transmitter output through a voltage divider and typically provide 1Vrms of carrier signal to drive the mod monitor.  While simple and hassle-free to setup, the results displayed on the monitor often do not accurately depict what is happening in the radiated field from the antenna.  The far field modulation index is sometimes considerably different from what is displayed at the transmitter.

 

To explore these, lets begin with Diagram 1.  Here, a 5 KW transmitter modulated 100% is driving a purely resistive 50 ohm load.  We will model the transmitter as an ideal constant-voltage generator where the transmitter output voltage remains constant regardless of actual load impedance.  In reality, all transmitters have a small internal resistance which causes the transmitter performance to depart slightly from being an ideal constant-voltage generator.   The error in assuming ideal conditions results in only a small error.    

 

                                                   

 

DIAGRAM 1.  TRANSMITTER WITH IDEAL LOAD

 

 

Several things are obvious.  First, a spectral display of the transmitter output shows 500V at the carrier and 250V at each sideband.  Second, a spectral display of the 500:1 divider output shows 1V at the carrier and 0.5V at each sideband.

 

The modulation index at the transmitter output can be easily determined from the 500:1 divider output using the relationship % Mod. =  ( E_L  +  E_U  /  E_C )  X  100 or % Mod. = (0.5V + 0.5V / 1.0 )  X  100 = 100%.

 

The actual modulation index of the radiated envelope can be determined using the carrier and sideband power levels.  Using E²/R, carrier and sideband power levels are calculated as 5000W, 1250W and 1250W respectively.  Using the familiar relationship P_M =  (m²/2)  X  P_C, the far field modulation index is calculated as 100%.

 

So, in the case of an ideal 50 ohm load, modulation index at the transmitter output and the actual radiated envelope modulation index are the same.

 

Now lets examine a real world case with a non-ideal load attached to the transmitter.  In Diagram 2, our assumed ideal transmitter drives an ATU that presents asymmetrical sideband impedances to the transmitter.

 

                                                   

 

DIAGRAM 2.  TRANSMITTER DRIVING LOAD WITH ASYMMETRICAL SIDEBANDS

 

 

Several things are obvious in Diagram 2.  First, since the transmitter is assumed to be ideal, the carrier and sideband voltages are constant and do not change with the asymmetrical load presented to the transmitter (within limits; high VSWR in the load will cause the TX to shut down).   Thus, a spectral display of the transmitter output shows 500V at the carrier and 250V at each sideband, unchanged from Diagram 1.  Second, a spectral display of the 500:1 divider output shows 1V at the carrier and 0.5V at each sideband, unchanged from Diagram 1.

 

The modulation index at the transmitter output can again be easily determined from the 500:1 divider output using the relationship % Mod. =  ( E_L  +  E_U  /  E_C )  X  100 or % Mod. = ( 0.5V + 0.5V / 1.0 ) X 100 = 100%.  Note that even with asymmetrical sideband impedances, the transmitter modulation index remains constant at 100 %.

 

The actual modulation index of the radiated envelope can be determined using the carrier and sideband power levels.  We first calculate the carrier and sideband currents using the voltages and impedances shown in Diagram 2.  These are calculated as 10A, 6.88A and 5.1A.  Then using P = I²R, where R is the real part of impedance, carrier and sideband power levels are calculated as 5000W, 1715W and 1225W respectively.  Using the familiar relationship P_M =  (m²/2)  X  P_C, the radiated envelope modulation index is calculated as 108%. 

 

In these case, despite the engineer’s diligent efforts to maintain the transmitter modulation index at 100%, the true radiated envelope modulation index is 108%.  The envelope in the radiated far field is actually overmodulated!  This station will have problems with on air quality.

 

Many engineers recognize the problem of using the transmitter modulation sample connector output to monitor their transmitter modulation.  Some use a Delta toroid transformer at the input of the ATU, as show in Diagram 3.  The idea is that by monitoring the actual carrier and sideband currents delivered into the load, a better picture of modulation can be obtained.

 

 

DIAGRAM 3.  MODULATION SAMPLE VIA TOROID TRANSFORMER

 

Lets see how this works in our case with asymmetrical sideband impedances.  In Diagram 3, the carrier and sideband currents are 10A, 6.88A and 5.1A as previously calculated.  Based on the toroid conversion factor of 1V/A, the toroid output voltages for the carrier and sidebands are 10V, 6.88V and 5.1V respectively.  Dividing these by 10 for the 20dB attenuator gives carrier and sideband voltages of 1V, 0.69V and 0.51V.

 

The indicated modulation index based on the toroid samples can be easily determined from the toroid voltage samples relationship % Mod. =  ( E_L  +  E_U  /  E_C )  X  100 or       % Mod. = (0.69V + 0.51V /1.0 ) X 100 = 120%.

 

Since the true radiated envelope modulation index is 108% (determined from Diagram 2 above), the indicated value of 120% from the toroid samples is too high, by 12%.  This is because the toroid only samples the current, I, going into the ATU, and not the power, I²R.  If the engineer reduces modulation so his mod monitor indicates 100% from the toroid samples, true radiated envelope modulation index will be only about 90%.  Thus, the station will actually be undermodulated and loudness will be reduced slightly.

 

What is needed is a method of sampling radiated envelope modulation index based on actual carrier and sideband power levels.  One method of doing this is by using a directional coupler inserted just before the ATU input as shown in Diagram 4.

 

 

                

 

DIAGRAM 4.  MODULATION SAMPLE VIA DIRECTIONAL COUPLER

 

The directional coupler samples forward power and provides actual carrier and sideband power levels attenuated by 40 dB.  An additional 24 dB attenuation brings the carrier and sideband sampled power levels down to 0.02W, 0.007W and 0.0049W respectively.  These power levels, when impressed across the 50 ohm input of a mod monitor result in carrier and sideband voltages of 1.0V, 0.59V and 0.49 respectively as shown in Diagram 4.

The radiated field modulation index based on the directional coupler samples can be easily determined from the relationship % Mod. =  ( E_L  +  E_U  /  E_C )  X  100 or % Mod. = ( 0.59V  +  0.49V  /  1.0 )  X  100 = 108 %.

 

It can be seen then that the modulation index indicated by the directional coupler samples agrees exactly with the true radiated modulation index of 108% calculated from actual power levels in Diagram 2 above.  The problem with this method, however, is that directional couplers capable of handling 5 kW at AM frequencies are quite expensive.

 

Perhaps the best method of deriving power samples to drive a mod monitor is to use a loop antenna located some distance from the transmitting antenna.  This is shown in Diagram 5.

 

 

 

DIAGRAM 5.  MODULATION SAMPLE VIA LOOP IN RADIATED FIELD

 

In Diagram 5, we are showing the loop antenna located approximately 100M from the antenna.  There is nothing magic about this distance.  We just want to get far enough away from the antenna to get a true picture of the radiated envelope.

 

At a distance of 100M, the field strength at the carrier and sidebands will be 6.7V/M, 3.9V/M and 3.3V/M.  Assuming an antenna factor of 200 for the loop, the loop voltage output at the carrier and sidebands will be 33.5mV, 19.5mV and 16.5mV.  Multiplying these by the amplifier gain of 30 gives 1.0V, 0.58V and 0.49V.

 

The radiated field modulation index based on the loop output voltages can be determined from the relationship % Mod. =  ( E_L  +  E_U  /  E_C ) X  100 or  % Mod. =   (0.59V  +  0.49V / 1.0 ) X 100 = 108 %.

 

It can be seen then that the modulation index indicated by the loop output voltages agrees exactly with the true radiated modulation index of 108% calculated from actual power levels in Diagram 2 above.  Additionally, a loop antenna has the advantage that it samples the stations signal after it has gone through the ATU, switches and the tower itself.  Any anomalies created through the signal path would be seen in the loop output.

 

The disadvantage to using a loop located 100M out is that the loop signal would have to be brought back to the station through 100M of coax cable.  That is a small price to pay, however, for getting an accurate and true picture of the radiated field modulation index.