hamilton rangemaster final rf stage efficiency

A class C high level modulated RF state typically has an efficiency of 70 - 80%, if I remember correctly.

The efficiency of a Rangemaster can be measured on a test bench by using a dummy antenna and an oscilloscope. A dummy antenna can be a 30 pF mica capacitor connected in series with a 47 ohm carbon composition or carbon film resistor. The 30 pF mica capacitor usually has a +/- 5 % tolerance, and the actual resistance of the nominally 47 ohm resistor can be measured with an accurate DMM. The 30 pF represents a typical capacitance of a 3 m monopole, and the 47 ohms represents a typical ground resistance. Other resistor and capacitor values from 10 to 100 ohms, and from 20 to 40 pF, may also be used to represent different possible antenna installations.

The dummy antenna is connected to the transmitter output in place of the actual antenna and ground. The power output of the transmitter is measured usiing an oscilloscope, with a 10 X probe connected across the resistor in the dummy load. Do not use a 1 X probe, because its capacitance is too high for accurate measurements. Be sure that the oscilloscope is calibrated. Do not use a DMM for any RF voltage measurements. Tune up the transmitter as if a real antenna system were connected, and read the p-p voltage on the oscilloscope. The transmitter power output is P = [(Vp-p)^2]/8R. For example, if the resistor used is 47 ohms, and the voltage reading on the oscilloscope is 3.5 V p-p, the transmitter output power is (3.5^2)/(8)(47) = .0326 W = 32.6 mW. If the transmitter has been set up for 100 mW of input power, this otput power represents an efficiency of 32.6 %.

The efficiency can be estimated using NEC if a field strength reading is available, and the ground resistance is known. Even if a field strength reading is available, the ground resistance of a Part 15 system is practically never known, and has to be simply assumed. By measuring the efficiency directly, like I described above, the transmiter efficiency for several possible ground resistances can be found by repeating the measurements for several different resistors used in the dummy load.

If range/coverage is the goal here, isn't that controlled at least as much by antenna-matching as by output power?

In the case of the Rangemaster, all of the antenna matching circuitry, including the loading coil, is contained inside the transmitter box. The loading coil in the Rangemaster is the secondary winding of the RF stepup transformer connected to the transmitter output stage. So, the output power of the Rangemaster is the power applied by the matching circuit to the antenna system.

The situation is different in the SSTRAN, where the loading coil is recommended to be external to the transmitter box.

Ermi,

Thanks for the excellent tutorial on how to measure the output power of a transmitter! I have used a Tektronix 475 oscilloscope to measure mine in the past.

I do have one comment, and that is that even with a 10x probe, the capacitive load may not be insignificant. I don't have one in front of me, but numbers in the 10-12 pF range come to mind. For best accuracy, this should probably be taken into consideration. A FET probe might have a small enough input capacitance to allow it to be neglected.

Not sure if Ermi's procedure will work with the Rangemaster as it is not designed to look at a 50 ohm load. Why not ask the manufacturer/designer what the efficiency is?

Ermi's calculation does not take into account radiation resistance and inherent ground losses. I would guess the 30% efficiency figure is high. The radiated signal of a short radiator below 50 ohms (5-6 ohms) would have a significant ground loss and would substantially impact efficiency of the final tuning network and radiator. Actual radiation is very likely not much more than several micro-watts. Without knowing the actual radiation resistance, calculating the efficiency is only guess work, even with an O-scope.

A dummy antenna made up of a 30 pF capacitor in series with a 47 ohm resistor is not a 50 ohm load, because the capacitor provides about 3 k ohms of capacitive reactance at the upper end of the AM BCB. Its equivalent parallel network is 3000 ohms of capacitive reactance in parallel with about 200 k oms of resistance. At parallel resonance with the secondary winding of the Rangemaster output RF step-up transformer, it provides a resistive load in the vicinity of 200 k ohms to the Rangemaster output.

I did not take into account the radiation resistance because, in a 3 meter monopole at ground level, it is only about 0.1 ohms, which is very small compared to the ground resistance. In an elevated transmitter, the radiation resistance can be several ohms, depending on the elevation, but it is still usually appreciably lower than the ground resistance.

In my post, I considered only the efficiency of the transmitter, and not the efficiency of the antenna. The 3 meter antenna length limit guarantees poor efficency. For example, if the ground resistance is 47 ohms and the radiation resistance is 0.1 ohms, the antenna efficiency is roughly .1/47 = .00213. If the transmitter has 30% efficiency and therefore applies 30 mW to the antenna, the radiated power is 30(.00213) = .0639 mW = 63.9 uW.

My own 10 X probe is rated for 10 pF, which works out to over 9 k ohms of capacitive reactance at the upper end of the AM BCB. This should not interfere significantly with measuring the voltage across a 47 ohm resistor. I have a FET probe, but I do not use it for the kind of measurements I described.

Hi Ermi,

I was not trying to trash your observation, but simply question how it would apply to a Rangemaster, SSTRAN, or other similar rigs that look into a high impedance.
Happy New Year, John

Gents,

I checked my own 10x probe-- it is an older Tektronix P6122. The rated capacitance is 11 pF. I have to disagree with Ermi on this one point-- I maintain that this capacitance needs to be accounted for in the measurement network, even though it is in parallel with a 47 ohm load. If you model the network, I think you'll agree. You can't just assume that since its reactive impedance is 9,000 ohms, it doesn't matter.

Otherwise, I agree completely with Ermi on his procedure. I don't think it was his intent to account for all of the variables present in a typical Part 15 installation. It's up to you to determine what values you think (or know) are accurate for your particular setup. I would guess that for most of us, the resistive load the transmitter is going to see is not far from 50 ohms, although only a tiny fraction of that is actually radiation resistance. The rest of the resistance is in the ground system. Only in the most ideal installation possible would the ground losses ever start to become similar to, or less than the radiation resistance. I don't know of anyone who has ever set up a Rangemaster over a 120 radial 1/4 wave ground system, but that would sure be interesting to try! Maybe someday somebody will be able to access the ground system of a defunct station and report on their results. That would be fun to read about!

I’ve used this method to test power output successfully in the past. The capacitance of the probe has negligible effect on the tuning because it is shunted by the low resistance load resistor (probe is connected across the load resistor).

As verification l ran a very small spice model consisting of a 1.600 MHz sine wave voltage source feeding the series connection of a 329.822 uH inductor, the 30 pF capacitor and the 47 ohm resistor to ground. With this inductor, the circuit is series-resonant at 1600 kHz as verified by the spice frequency sweep display. My particular scope probes have a capacitance specification of 13 pF, so I added 13 pF across the 47 ohm resistor in the spice model. The resonant frequency changed to 1.60007 MHz.

Resonant frequency without and with the 13 pF probe capacitance:
1.600000 MHz no probe
1.600070 MHz with 13 pF probe

Peak to peak voltage across 47 ohm resistor without and with the 13 pF probe capacitance:
3.50042 Vp-p no probe
3.50089 Vp-p with 13 pF probe