Email :
Introduction
Professional test equipment is expensive to buy or hire. But the opportunity is here as below, to emulate a commercial RF millivoltmeter in cheap and simple terms, by its DC output driving a digital multimeter display.
In lieu of access to a radio communications test set, frequency bandwidth response can only be estimated as being good to a few hundred Mhz and an input level range from 1mV to a couple of volts rms. Perhaps a reader giving feedback to CXI could firm up these estimates with proper lab kit.
So the finished unit will do a satisfactory job for hobby use in fault tracing, gain measurement and general radio test work. Don't forget to halve the indicated DMM reading for the true rms input!
Please note this unit is NOT intended to measure RF power, as its sensitive components would be destroyed by the inputting of high RF voltages.
Circuit DescriptionThe complete circuit shown has two sections joined at a possible deliberate linked cable point if required. See the build guide below for details.
A Jfet follower stage having a high input impedance, avoids loading the source circuit under test. Its lowish output impedance connects to an adapted traditional voltage doubler detector commonly used in electronics.
The second section can be used independently to detect and measure signals from a low impedance source such as an RF signal generator, as part of an application process for example, see further below.
Three minor refinements have been incorporated in the second section.
1) Input resistor R3 of 47 ohms has been added, to buffer the indetermined parallel-shunting impedance effect at RF of the circuit, particularly signal clamp diode D2.
2) Conversion from 2 x peak mV to a 2 x rms mV reduction at 71%, are by appropriate calculation of diode load resistors R4/R5/R6 feeding the DMM.
3) To improve detection linearity of low signals, a small additional forward bias via VR1, supplements self-bias of the diodes, by assisting a small signal to be pushed past the non-linear knee portion of diode characteristic.
Build guide at user discretionIn conventional RF style, the unit should be constructed in a small metal or die-cast box with the circuit board supported on spacers from the box floor.
The Jfet input is via a co-ax probe or short croc clip leads through a small grommet hole in one of the box short sides.
By requirement, the whole circuit can be made straightforwardly as shown.
Alternatively an extra break/link facility could be created, by an RF co-ax link through a small grommet hole from the Jfet output to an external free male plug. This plugs into a chassis-mounted female socket for the second section input. That socket is best mounted on the box opposite short side away from the Jfet input leads.
On the prototype unit, the Jfet was not required, but a female BNC socket was used for the second section input, so that a BNC barrel adaptor could be connected to a source signal by the shortest distance possible.
Otherwise, a test cable for inputting an external RF signal to the second section input can be made up if required. This having a plug for the box socket at one end and test connector or clips at the other. Choice of connectors and cable size is at user discretion and parts availability.
A final small grommet hole can accomodate the battery and DMM leads.
AF output is via a 2.5mm or 3.5mm mono jack socket.
Component suggestionsCapacitors are small resin-dipped ceramic, with the exception of C5 which should be a small non-inductive polyester film type.
Resistors are very small 1/8w if available for compactness, or standard 0.6w metal film if construction space permits.
Diodes D1 & D2, are preferably germanium "hot-carrier" Schottky barrier H-P types if available, for the widest possible frequency bandwidth. There is also a newer silicon Schottky 1N5711 equivalent of these, which is purported to be available from Farnell or RS Components and possibly elsewhere.
Alternatively but with restricted bandwidth, germanium diodes such as the preferred OA47 gold bonded and OA90 or OA91 will also suffice. Match two of these with a multimeter for a similar high reverse resistance, so each is of no less than 5 megohms. I use an analogue Avo 8 for this selection test.
Small signal linearityAs explained earlier, this is important and is achieved in this design by feeding a known RF test level to either the Jfet input or second stage input and then increasing VR1 multiturn preset pot from ground, to indicate twice the input rms as DC mV output on the DMM.
Millivoltmeter Calibration
As the zoomable pictures illustrate, an old battery HF signal generator was to hand with a BNC output socket. This supplied a known 50mV rms to a discrete second section via a barrel adaptor. VR1 was increased up from zero, so the output DMM read 100mV DC. (Without bias, the DMM read 72mV DC)
At the circuit Test Point or TP, the forward bias measured 105mV DC. Following this simple calibration procedure, it was felt that no further adjustment was necessary.
To check low levels, 1mV rms input clearly read 2mV DC on the DMM and at high levels, 470mV rms input read 940mV DC on the DMM.
AM modulation was then switched on and a tone was clearly heard from the AF output on a connected bench amplifier.
Application example warts & all for the millivoltmeter and RF signal generator
Bearing in mind that S meter scale calibration isn't an exact science, the known 50mV rms generator output can also be used to spot-check an HF communications receiver or tranceiver S meter. This is done as shown in the block diagram, by tandem connecting via a barrel adaptor a 60dB attenuator and then another barrel adaptor or very short coupling cable to the receiver input.
This little three-resistor pad aid built in a very small die-cast box with suitable rf connectors, reduces 50mV to 50 microvolts as shown.
Calibration alignment can then be made if required to the receiver's S meter drive circuit pots, to set the main S9 scale point and also the remaining intermediate 6dB S points.
This is a slow process and not guaranteed dead accurate due to compromise of impedance matching. However these can be set by adjustment each time of the sig-gen attenuator output, measuring directly to the millivoltmeter as follows:
The upper scale above S9 is cramped, but continuing on,
For S9: 50mV rms & 100mV DC on the DMMFor S8: 25mV rms & 50mV DC on the DMM
For S7: 12.5mV rms & 25mV DC on the DMM
For S6: 6mV rms & 12mV DC on the DMM
For S5: 3mV rms & 6mV DC on the DMM
These sig-gen levels are then transferred in turn to feed the 60dB pad input and the pad output to the receiver input. The meter cal pots are repeat-adjusted until at least the S5 to S9 points on the scale are reasonably accurate.
Signal strength reports which are usually subjective, can finally be given with the confidence of having a meaningful basis to accepted RF standards.
Thus this nice homebrew bit of kit has proved its usefulness for at least one application and any other jobs which may need RF level measuring attention.
ReferencesSemiconductor linearity is often a topic of discussion amongst electronic engineers and the mathematics involved can be very complex. The following pdf by the University of California, explains some of the physics behind the diode:
The following series of articles by Ray Marston may be of interest to some viewers as well:
Bipolar Transistor Cookbook Part 1Bipolar Transistor Cookbook Part 2
Bipolar Transistor Cookbook Part 3
Bipolar Transistor Cookbook Part 4
Bipolar Transistor Cookbook Part 5
Bipolar Transistor Cookbook Part 6
Bipolar Transistor Cookbook Part 7
Bipolar Transistor Cookbook Part 8
Also by same author, Ray Marston the following series of Op-amp Articles may be of interest as well:
Op-Amp Cookbook Part 1Op-Amp Cookbook Part 2
Op-Amp Cookbook Part 3
Op-Amp Cookbook Part 4