OwenDuffy.net 


Peak amplifier for an RF wattmeter

This article describes a generic peak amplifier for an RF wattmeter, to extend the instrument to measure Peak Envelope Power of SSB telephony waveforms.

Peak Envelope Power

The ITU Radio Regulations define the terms Peak Envelope Power, Mean Power and Carrier Power with regard to a radio transmitter. The terms are defined as:

Note: For use in formulae, the symbol p denotes power expressed in watts and the symbol P denotes power expressed in decibels relative to a reference level.

The advantages of a PEP measurement capability are:

PEP can be measured and displayed on a conventional directional wattmeter extended with a peak hold amplifier. It is a more effective and more accurate method of measuring PEP on speech than using a non storage oscilloscope where the user may miss some peaks.

Design criteria

The design criteria are:

Implementation

Operational amplifiers offer a potential solution to the design criteria. Key factors in selection of a device are the requirements:

Not all op amps satisfy these criteria.

Fig 1: Peak amplifier circuit

Fig 1 shows an amplifier circuit. There are different ways in which this amplifier could be used in an instrument, and they affect the magnitude of the voltage at the amplifier input:

Depending on the application, the amplifier may need to handle a FSD input voltage of upwards of 50mV through to many volts.

An imperfection of real world op-amps is that they have a small input offset voltage, a difference between the inverting and non-inverting inputs to obtain mid-supply output. The input offset voltage varies from part to part, with supply voltage, with input common mode voltage, with temperature etc. One could attempt to provide an external adjustment to null the amplifier offset (and some designers do so), but that doesn't assure long term stability. Further, adjustment of null in the peak hold circuit of stage 1 is a little tricky because of the effect of the diode. there is a better solution, selection of an op-amp with offset and drift specifications more suited to the application.

To ensure that the input offset voltage did not contribute more than about 1% of FSD error, the input offset voltage specified for the op-amp should be less than 1% of the movement FSD voltage.

Two op-amps representative of device families were explored for suitability:

LM358

The LM358 is a low cost dual op-amp designed for low voltage single supply rail operation with a maximum input offset voltage  of 7mV, typically 2mV, and input offset drift of 7μV/°C. Price is about $0.33 for a dual op amp.

The maximum input offset voltage of the LM358 limits its use in this circuit to input FSD of 350mV, or 100mV if one takes a risk on "typical offset" specs rather than "maximum offset". (A quick check of 10 LM358s from the parts bin showed 80% had offset <=1.3mV, and the worst was 2.0mV, well under the maximum offset spec of 7mV.)

The chip, is easy to obtain, cheap, contains the amps for both stages. It is an ideal choice if the input FSD is sufficiently high. If one was designing an instrument from the ground up and had the choice of operating at an input FSD of 0.35V or more, it would be an excellent choice.

Key features of the chip in this application are:

The first stage provides a gain controlled peak amplifier capable of sensing down to ground, with the diode drop eliminated from the output as the diode is within the feedback loop. The output impedance of the first stage is low, allowing rapid charging of C4 capacitor for fast attack. The decay time constant is determined by C4, R3, R5 and R6.

The LM358 has very low temperature compensated input bias current. In this case and due to the resistances used, there is no real benefit in balancing the resistance in the inverting and non-inverting input paths, though this may be wise for some other types of op amp. Values are given for R3 and R4 where input balancing is desired.

If the instrument has multiple ranges, careful attention needs to be given to a value of input resistance that is very close to the resistance of the meter movement. The design tool calculates values for two parallel resistors, R1 and R2, which should allow sufficient accuracy using standard E12 preferred range resistors.

The next stage is a voltage follower with high input impedance (for minimal loading of the capacitor), and a series resistor to drive the meter movement. VR1 is trimmed to obtain the same deflection on a constant carrier as obtained with the amplifier bypassed.

The entire circuit consumes about 5mA.

For battery operation, the regulator should be omitted and a link inserted from its input to output. The circuit should be run from 6V (eg pack of 4 x AA cells) current will be around 2mA (yielding an operating life of about 1500 hours on AA cells).

LTC1050

The LTC1050 is from another generation op-amps. It is a low offset, low drift, chopper stabilised instrumentation amplifier designed for low voltage single supply rail operation with a maximum input offset voltage  of 5μV, typically 0. 5μV, and input offset drift of 0.05μV/°C. These amplifiers are ideally suited to the amplification of low voltage DC signals. Price is around $6 for a single op amp. The ultra low offset and low drift are achieved by an auxiliary amplifier which monitors the offset of the main amplifier and automatically nulls the offset.

The maximum input offset voltage of the LTC1050 limits its use in this circuit to input FSD of 0.25mV, which is much lower than would be experience in real world instruments of this type. (The FSD of a Bird 43 meter movement is one of the lowest at around 45mV.)

The chip, is not as easy to obtain, medium priced, contains only one amp. It is an ideal choice if the input FSD is lower than acceptable using the LM358.

Key features of the chip in this application are:

Fig 2: Lower offset peak amplifier circuit

Fig 2 is a schematic for a lower offset peak amplifier using the LTC1050 in the first stage.

The LTC1050 output is not able to source sufficient current to charge the peak hold capacitor sufficiently quickly, so the design uses a LTC1050 stage with voltage gain to raise the input signal to around 3V peak, followed by an LM358 unity gain / peak hold stage, then another LM358 stage as a voltage follower to drive the meter.

The first stage provides a gain controlled amplifier capable of sensing down to ground, and raising the signal to around 3V, sufficient to mask the input offset voltage of the next stage. The second stage is a unity gain peak hold amplifier with the diode drop eliminated from the output as the diode is within the feedback loop. The output impedance of the second stage is low, allowing rapid charging of C4 for fast attack. The decay time constant is determined by C4 and R8.

The LTC1050 has very low, temperature compensated input bias current. In this case and due to the resistances used, there is no real benefit in balancing the resistance in the inverting and non-inverting input paths, though this may be wise for some other types of op amp. Values are given for R3 and R4 where input balancing is desired.

If the instrument has multiple ranges, careful attention needs to be given to a value of R1 that is very close to the resistance of the meter movement. The design tool calculates values for two parallel resistors, R1 and R2, which should allow sufficient accuracy using standard E12 preferred range resistors.

The next stage is a LM358 voltage follower with high input impedance (for minimal loading of the capacitor), and a series resistor to drive the meter movement. VR1 is trimmed to obtain the same deflection on a constant carrier as obtained with the amplifier bypassed.

The entire circuit consumes about 7mA.

For battery operation, the regulator should be omitted and a link inserted from its input to output. The circuit should be run from 6V (eg pack of 4 x AA cells) current will be around 4mA (yielding an operating life of about 750 hours on AA cells).

Switching

If the amplifier is powered from a shared power supply, attention must be paid to avoiding earth currents flowing on internal paths. This can usually be achieved by single point earthing, connecting the power supply ground to the point where the coax connectors and detector grounds are made. For example, if fitting this amplifier to a Bird 43, connect the shared power supply ground to the sampler assembly, NOT the -ve terminal of the meter movement.

Where battery operation is contemplated, an additional pole set ganged to S1 could be used to switch the DC input line.

Design tool

An Excel spreadsheet containing design formulas is available for download.
Fig 3: Design tool calculations for some example instruments.

Fig 3 shows design values for several proprietary wattmeters. Note the discussion above that R3 and R4 are not really necessary and could be made zero ohms.

Links


© Copyright: Owen Duffy 1995, 2017. All rights reserved. Disclaimer.