Revisit AD8307 Power Meter Calibration

This year at our club, annual pig picking BBQ and show-n-tell I had a chance to borrow the kind services of fellow Knightlites member OM John, KU4AF and his brand new and beautiful digital spectrum analyzer.

Connecting the Si570 based oscillator from my Sweeperino   into the Spectrum Analyzer and measuring the power level of the fundamental frequency we measured a -4dB difference in power when compared to measuring the squarewave on the AD8307 based wideband (~500MHz) Power Meter. Using a Low Pass Filter to filter out the odd numbered harmonics I resolved to measure the power in the fundamental on my work bench.

For the experiment I used the well known 10MHz calibration source as developed by K3NHI. This source can be used to establish the -10dBm point on the power meter through the measurement of a DC voltage level at the DC port. This reference is easy to build and is recommended for any homebrew starting out with a basic 50 Ohm bench.

Indeed my result showed a 4.1dB difference in power reading of the fundamental vs the wideband signal.

Herewith below are my Lab Notes of the procedure adopted.

Simple test results of a 7MHz Halfwave Power Output Filter

As shown in the attached schematic and test results. Using my MFJ259B antenna analyzer I was able to obtain a view of the VSWR of this filter over the HF spectrum.

At the very least this illustrated 2 observations to me.

1) A reasonable match to 50 Ohms is obtained over the 40m band.

2) A mismatch exists for frequencies other than the 40m band. Both above and below this band.

The obvious benefit, and the whole point of the filter, is the fact that harmonics are attenuated by the filter. This is all good.

The more interesting Questions to me are :-

1) What is the effect of the reflected harmonic energy when it is absorbed by the Class C Non Linear Power amplifier?

2) Is the Return Loss of the harmonic energy so great (>80 dB) that it is insignificant or does the reflected energy generate significant IMD?

3) Why are diplexers mostly implemented on Mixer outputs but not on output filter circuits?

All good :)

RF Power Meter

Today I finished building my RF Power Meter. This is a design from EMRFD section 7 by W7ZOI. The schematic is self explanatory. It is basically 2 power meters in one. The meters read Peak RF voltage that is then converted to Power using the formula P = V^2/2R.

I have completed the calibration of the bigger input using 15volts DC. I also calibrated it using 10V DC FSD. I can measure accurately a range of +34dBm - +12dBm. The measurements I took are shown in the spreadsheets.

I need to get some bigger power resistors so that I can measure at least 5 watts. Currently even measuring 2.25W generates a lot of heat in the little dummy load.

I performed a sweep using my antenna analyzer of both dummy loads. They are both flat 1:1 SWR well into VHF.

I am looking forward to calibrating the small meter which should allow me to measure down to about -8dBm. Thus I can use it to so some measurement on my receivers.

I have developed 2 calibration charts which show the non linearity of the 1N4148 as can be expected.

I built the meter into a medium quality plastic box that I had lying around.

Refer to the details of the test procedure below:

1	Heading	RF Power Measurements procedure.
2	Label	RF Power Measurements.
3	Date	11/22/2013.
4	Acknowledgements	1.      Author : ZS6RSH. 2.      Reference: EMRFD Section 7 paragraph 7.3.
5	Revision	Rev 1.
6	Revision History	A blog was first written on 11/10/2013. This procedure is being written ‘after the fact’ with the aim of testing the effectiveness of the measurement procedure.
7	Scope	Based on the recommendations in EMRFD it is valid to calibrate my homebrew RF power meter using DC power.  As confirmed in EMRFD, the calibration will be valid through the HF range and into VHF for the specified diode 1N4148. Since the resistors I have used for the dummy load are only 1/4watt rating it is only possible to calibrate between the voltage ranges of 1V-15V at this time. The lower app ~ 1 volt limit is due to the silicon diode becoming non-linear for voltages below that value. Thus the power measurement range will be approximately between +34dBm and +12dBm.
8	History	The accuracy of this DC calibration is primarily dependent on the accuracy of the DC voltage measurements (see UNCERTAINTIES below). Measurements have previously been carried out to understand the accuracy of my two DC voltmeters. The Fluke and Keithley. (Model numbers to be provided). These two meters read the same DC voltages to 2 decimal places over the range 15V – 1V. Refer to xxx For these measurements.
9	Configuration	The homebrew RF Power meter is connected to a variable voltage DC power supply with a variable voltage range of 1VDC – 15VDC and a maximum current capacity of 1Amp. The DC voltage was measured using the Keithly voltmeter connected across the input of the power meter.
10	Test equipment specifications	1.      Keithley model xxx voltmeter. 2.      Homebrew variable voltage current limited, power supply. 3.      Connection leads. Regular leads that came with the voltmeter. 4.      Jumper leads used for the power connections.
11	DUT specifications	Homebrew RF Power meter including dummy load. Power range of 15V FSD, Approx +34dBm - +12dBm.
12	Workbench process	1.      Back off the calibration pot so that the meter cannot be overdriven. 2.      Set the power supply to 15V. 3.      Quickly adjust the cal pot to achieve FSD of 1mA. 4.      Turn off the power supply. 5.      Reduce the power supply voltage so that the meter shows decrements of 1/10 of a milliamp. 6.      For each 1/10 milliamp reduction, quickly record the voltage to 2 decimal places. 7.      Take 10 readings. 8.      Change the FSD to 10Volts. 9.      Take 10 readings as above.
13	Expected Results	The recorded voltages against the ammeter readings should represent the transfer characteristics of a silicon diode of type 1N4148
14	Uncertainties	1.      Dummy load change as a result of dissipation heating. Can be kept to a minimum if the tests are carried out quickly. This variation can be characterized in a separate set of measurements. However the plan is to build a dummy load with QRP power dissipation capabilities in the future. 2.      Specific transfer characteristics of the diode are unknown but will be discovered. 3.      Non linearities in the specific voltmeter readings. Already verified to not be an issue to 2 decimal places. 4.      Variations in ambient temperature during the test period. Not taken into account during this test but could be by recording the temperature for each measurement. 5.      Parallax errors from reading the analog ammeter. Can be read to the nearest 1/100th of a milliamp. 6.      Quick reading of the meters could result in a recording error. 7.      Calibration Graphing errors. However the data was recorded to 2 decimal places. 8.      It is uncertain that the RF Power Meter will record peak RF voltages according to the same transfer characteristics as at DC level. This is assumed to be the case based on the EMRFD reference in section 7 and nothing else at this stage. This consideration is beyond the scope of this set of DC measurements. 9.      RF coupling causing variations in load and measurement characteristics. The SWR was seen to be a flat 1:1 across the HF spectrum using an MFJ259B analyzer. As for 8 above, this consideration is beyond the scope of this set of DC measurements.
15	Preparation	Completed.
16	Perform validation measurements	Completed. The meter calibration pot was set to a minimum to start with to ensure that the meter would not be harmed by overdriving.
17	Perform the full measurement plan	Completed.
18	Observations	No unexpected variations or observations.
19	Change Control	No changes were made to the original plan.
20	Computation	Refer to the attached tables. For each recorded voltage a power value was derived using the formula P=Vpeak^2/2R.  This formula is valid since at RF the meter records peak RF power. The diode rectifies the AC signal and the capacitor charges to the peak value. This power was then converted to dBm and graphed. Thus a major assumption is made here that the power meter will in fact correctly record peak RF values in practice. This validation is beyond the scope of this set of DC measurements.
21	Analysis	The graph of the results shows an expected transfer characteristic in line with a 1N4148 diode over the measurement range.
22	Conclusions	The calibration curves are in line with the expected results and can reasonably be used to explore RF power measurements. Validation of the results, however, is needed in the future against a calibrated RF source.
23	Documentation	Completed

RF Power Meter Schematic