High Impedance Buffer and Broadband Amplifier for Digital Freq. Meters

(First published by ) in the Wireless Institute of Australia Amateur Radio magazine, October 1980, last update 14 June 2003.)

AN ARTICLE FOR THE WELL EQUIPPED AMATEUR

With the introduction of synthesized transceivers employing the hetrodyning of several mixer crystals with the VCO output of a PLL system, there has grown the need to measure frequencies at low levels. In the majority of cases, because we are dealing with solid state devices, we have levels that are around the order of 10 dBm or less (0 dBm = 1 mW).

The Impedances around such circuits are not very appropriate for measurement with devices of relatively low impedances, particularly when the circuit impedances can range anywhere between tow hundred and several thousand ohms. Consequently a high gain and high impedance device is required if we are to obtain any measurements and accurate measurements respectively. I am sure that we are all familiar with the operating principle of the GDO, in the same way, loading of any oscillator will cause a resultant shift if frequency.

These two devices, the RF buffer and the broadband amplifier, were primarily designed for the input to the front end of a frequency meter and its prescaler, in particular the EA Digital Frequency Counter. The application was the measurement of a Yaesu FT-901D transceiver because some problems were experienced on the 10m bands.

Those familiar with this transceiver know that the crystals and the VCO cover an approximate frequency range from 15 MHz - 43 MHz. The probe and amplifier were used to obtain measurements over this range with no noticeable shift in the final frequency of the transceiver.

THE HIGH IMPEDANCE BUFFER

Three requirements should be met by the probe:

  1. High input impedance - should be greater than 1 M ohm.
  2. Low input capacitance - typically less than 10 pF.
  3. Wide bandwidth - useful over several octaves.

A JFET was chosen as the active device on the input of the buffer. The JFET was followed by a PNP bipolar tansistor - which is used for impedance transformation. The circuit configuration bootstraps the source resistor to minimize input capacitance.

The FET is a process 50 type with a typical gain of 12 dB at 400 MHz and a noise figure of 4 dB. The quoted input capacitance is 3.5 pf with zero gate to source voltage, although at a Vds of 6.0 volts and a Vgs of -4.0 volts this is significantly improved. A typical device of this process is the MPF102 (although I used a 2N254).

The impedance transforming transistor employed in inverted mode was an AF139, which is a PNP germanium transistor (this was used because if was available in the shack and it has a high Ft. This device is used in TV masthead amplifiers, so it works in the VHF region.) The buffer design is adapted from National Semiconductor's application note (AN32).

The layout is not particularly stringent, although good RF practice should be adopted (keep leads short, particularly around the gate of the JFET.)

The capacitor C1 on the input was included to provide high voltage isolation and should be a good quality high voltage capacitor. If you wish to improve the low frequency you may lower C2 so its impedance is less than 50 ohms at the 3 dB roll off.

Figure 1: High Impedance Buffer

THE BROADBAND AMPLIFIER

National Semiconductor process 43 transistors have been selected because they have a minimum Ft of 600 MHz, some selected devices have Fts within the GHz region. The process 43 transistors are employed in UHF amplifiers and oscillators with collector currents in the range of 1 - 20 mA. Their hfe is between 40 and 200. I chose a 2N3563 as the active device for the amplifier.

THE DC BIAS

The DC bias is important, at high currents we achieve greater bandwidth capabilities and better stabilisation of current gain. Looking at the design curve for Constant Gain Bandwidth it was decided to run the transistor with a current of Ic = 10 mA and a voltage of Vce = 7 Volts as a trade-off in this curve and the supply voltage of 9 Volts (from a No. 216 battery).

Using the following DC network and certain assumptions we will derive the values for the resistors:

  1. Vc = Vcc - Ic Rc ( Ib + Ibias << Ic)
  2. Vb = (R1 + R2) / (R2 Vcc)
  3. Vb = Ve + 0.6 [Vbe 0.6 Volt)
  4. Ve = Ic Re [Ie Ic]

Choosing Ic = 10 mA and Rc = 100 ohms we arrive at R1 = 3.8 kW , R2 = 1 kW and Re = 100 W .

Figure 2: Amplifier DC network

THE RF CONFIGURATION

The key to the bandwidth requirement is to use (RF) negative feedback - which also achieves stabilisation (against positive feedback that can lead to oscillation).

The quoted references in the Ham Radio magazine (now defunct) employ a form of series feedback to achieve gain flatness. The result is constant gain but an unfortunate side-effect is increased input impedance by a factor proportional to the feedback and the beta (hfe) of the transistor. Since beta can be approximated by the following expression hfe ~ Ft / f , where f is the operating frequency, the transistor achieves higher gain at lower frequencies. The other form of negative feedback is shunt feedback. This form lowers the input and output impedance as well as stabilising the current gain of the device.

The overall ultimate design employs the application of both forms of feedback; the design parameters are included below:

The circuit employs a balun to match the transistor's output impedance without loading it too much. It also covers a wide frequency range, however, increasing the number of turns will lower the 3 dB roll-off point.

Figure 3: Amplifier AC network.

The final circuit is a combination of the DC and AC networks. I chose components which resulted in a gain of 19 dB ( Rf / Re=79 ) with Re equal to 4.7 ohms and Rf equal to 510 ohms (5k6 in parallel with 560R) .

The performance of this amplifier was measured using a signal generator and an attenuator driving the amplifier into a resistive load.

Since we lived (at the time) in a fringe area for Channel 6 and Channel 8, Lismore, I was able to use weak TV signals and a colour TV set to perform the gain measurements in the VHF region. The amplifier was preceded by a step attenuator 0 - 30 dB. The attenuator was adjusted for colour dropout with and without the amplifier present. This provided a rough estimate of 6 dB gain at 178 MHz and 3 dB gain at 192 MHz

Figure 4: The Amplifier

ACKNOWLEDGEMENTS

A special thanks to my father, VK2ZAD, for the opportunity to use his reference library and test equipment. Thanks also to Nathan VK2DDT for providing me with the original initiative to build the probe and amplifier.

SPECIFICATIONS

Buffer
Gain ~ 0 dB
Input = 10 M ohm || 4 pF
Output <= 50 ohms.
 
Amplifier
Gain ~ 19 dB
Input ~ 50 ohms
Output ~ 75 ohms
BW ~ 200 kHz - 50 MHz
 

REFERENCES

  1. Wideband IF Autotransformer, John J. Nagle K4KJ, Ham Radio, November 1976, page 10.
  2. Wideband Preamp, Ed Pacyna W1AAZ, Ham Radio, Object 1976, page 61.
  3. General Purpose Wideband RF Amp, Randall Rhea WB4KSS, Ham Radio, April 1975, page 58.
  4. Linear Application Notes, National Semiconductor National Volume 1 AN32, page 7.
  5. Transistors Small Signal Field Effect Power, National Semiconductor.
  6. Solid State Design for the Radio Amateur, ARRL 1977.

 

GLOSSARY