RF Voltmeter

Autoranging: Over the years I have owned and used a variety of volt-ohm meters (VOMs), including the famous ‘free’ one from Harbor Freight. Most had separate switch settings for measuring AC volts and DC volts, and separate measurement ranges, such as 0 to 10 volts, 10 to 100 volts, 100 to 1000 volts, etc. All that has gone by the wayside. Now there is one switch setting for volts, or perhaps volts and ohms. It is not necessary to specify AC or DC, or a measurement range. The meter automatically adjusts from millivolts to hundreds of volts, displaying the measurement to 3 or 4 digits precision. Autoranging meters such as my Mustool MDS8207 respond almost instantaneously. Their ranging circuitry is electronic and evidently foolproof, or nearly so.

I wanted to have a go at the auto-ranging concept, but at a simpler or cruder level, so I thought of making an RF voltmeter that would switch ranges automatically, depending on the input level being tested. This page describes the project that resulted. The ‘RF Voltmeter’ is nothing more than an RF rectifier combined with a microcontroller-based DC meter. The boilerplate half-wave RF rectifier sub-circuit can be found on multiple Internet pages. It consists of a Schottky rectifier diode, with minimal filtering / smoothing, as shown in the pictorial diagram above (bottom middle). The DC measurement part of the meter consists of a 3-stage voltage divider for implementing the ranges, and an ESP-12F development board. I had bought a ten-pack of these boards for about $25 (Amazon), without having a project in mind, rather just to see what could be done with them. Previously I had experimented with other NodeMCU boards, as well as with the amazing M5StickC-Plus.

The RF voltmeter does not measure continuously. To trigger a measurement it is necessary to push the (momentary) button. For the display I used a 0.96-inch monochrome OLED, because one was on-hand. There’s not much information to display with a voltmeter, so a small screen suffices. The display remains on for 10 seconds (configurable) and then goes dark. I have learned from my mistake of leaving OLED displays on for a protracted time.

When the button is pressed, A/D values are read ten times (another configurable number) at 1/100 the DC input voltage. If the average reading falls below a specified threshold, the range changes to 1/10 the input voltage and the same test is applied. If the average A/D is still less than the specified threshold the meter switches to 1:1 and measures the rectified raw applied voltage. Thus the meter has a working range of one or two hundred millivolts to about 300 volts DC. However, the rectifier does not have a comparable range. The 1N5711 diode datasheet states that its peak inverse voltage is 70—the specific capacitors used for the prototype build have a somewhat higher voltage rating. In addition to using higher voltage components for the rectifier, it would also be advisable to pay attention to physical layout, if it is desired to increase the RF measurement range.

Two small 3-volt relays, labeled k1 and k2 in the image above, are used to switch between ranges. A couple of LED’s were connected to the unused side of the DPDT relays for debugging, prior to applying voltage to the A/D input of the ESP-12F. The LED’s serve no useful purpose in the finished unit; however, I decided to leave them in place and make a couple of slits in the enclosure front as decoration.

    The ESP-12F is a WiFi board. After assembling and calibrating the meter I added an optional UDP client to the project sketch. There was a practical reason for this, namely to include RF measurement parameters as part of a computer screen display for demonstrating aspects of a different project. WiFi parameters may either be included or omitted when compiling the sketch. If the symbol WIFI_SSID is not defined then the compiled application does not attempt to connect, or to implement the UDP add-on.

    The sketch may be examined or downloaded here. To compile it using the Arduino IDE (v.2), select NodeMCU 1.0 (ESP-12E Module) as the board. Various Internet sources state that the 12F board is a minor revision of 12E, with an improved WiFi antenna being the main or only change. Arduino compile and load functions work the same for both boards. The top section of the sketch documents selectable options, such as whether or not to implement WiFi. A little further down are symbols for UDP IP and port. These should be configured to match the UDP target. Additional operational constants are also configurable, as documented in Sketch comments.

Calibration: There is more than one way to approach calibration. I do not know whether different ESP-12F modules respond more-or-less the same as one another in terms of their A/D conversion. The two that I tested were virtually identical. Both report a raw digital value of 9 or 10 when the analog input pin is grounded. Below about A/D = 12 displayed measurements should not be fully trusted. The A/D ‘trust point’ is a configurable constant named MIN_AD in the sketch.

    I used several different signal sources¹ and measurement methods² to calibrate the prototype unit described and illustrated on this page. A single regression equation sufficed for each of the middle and high ranges. The low range was linear above a certain point. Below that point it was approximately logarithmic. Thus two equations were derived from calibration data for this range. In the sketch, the dividing point is a configurable global constant named NON_LINEAR_REGION, valued at A/D = 80 for the prototype. Below this point a logarithmic fit is used and above it the fit is linear. The high range was calibrated at the DC input point (rectifier disconnected) using a high voltage DC power supply, and was not tested with an RF voltage source.

Although the ordinates are labeled ‘Generator Setting’ both low and medium range RF voltages were also measured using the oscilloscope. As noted above, the rectifier was bypassed when calibrating the high-range, and would not be useable through most of that range unless higher voltage-rated components were substituted. Internally, high-range Vp-p is computed as 2 × the DC regression equation.

Projects Home

1. Two function generators, a QRP transceiver, a couple of batteries and DC power supplies.

2. Oscilloscope for HF and VHF, and bench voltmeter for DC.

Project descriptions on this page are intended for entertainment only. The author makes no claim as to the accuracy or completeness of the information presented. In no event will the author be liable for any damages, lost effort, inability to carry out a similar project, or to reproduce a claimed result, or anything else relating to a decision to use the information on this page.