The thought of building a tunable spectrum meter has been muling around in my head for a bit and finally got it built. Basically I thought that I could use a mixer and a local oscillator in combination with the W7ZOI power meter to read the power of a transmitter and its harmonics. I thought about it a bit more and realized that just tuning the LO up and down frequency with a direct conversion type of arrangement would not work. The power meter would detect all frequencies presented to it rather than just the frequency I wanted to measure. After a while I realized that since the mixer would produce the sum and difference of RFin and the LO. If you add 10 MHZ to the signal you want to measure and then use that as the LO then the difference winds up being a fixed frequency of 10 MHz: eg. LO = signal+10 (signal+10)+signal is one mixer output frequency (signal+10)-signal is the other freq which is 10MHz I decided to try a 10 MHz xtal filter as the fixed IF frequency. The plan was to build a 10MHz filter anyway to experiment with Farhan's BITX20 so the work to build the filter wouldn't be wasted if the experiment didnt work. The filter was designed with the AADE filter designer. I spent a day with Chapter 3 of Experimental Methods gathering the crystal parameters of the 10 MHz xtals I had to use as input to the filter designer.. In case anyone is interested in my crystal parameter results, I have a spreadsheet here. Here is the filter (50 ohm in/out 3kHz wide @ 10MHz):![]()
I built a mixer out of a MiniCircuits ADE-1 I added a 3dB PI pad at RFin which is two shunt 300 ohm resistors and a series 18 ohm resistor. I added a 6dB PI pad at IFout to help with matching into the xtal filter this is two shunt 150 ohm resistors and a series 36 ohm resistor. The LOin port should be driven at about +7dBm (about half a volt RMS).
![]()
The Local Oscillator is a homebrew signal generator which is based on the AMQRP DDS-60 and AA0ZZ's PicEl PEGen software. The siggen puts out a constant signal level from about 50kHz to 50MHz and drops about 6dB between 50 and 60 MHz. I had originally been running a DDS30 siggen as my LO but blew that siggen up somehow so that was a clue to me that it was time to upgrade the software and DDS to the DDS60. Now I can measure harmonics up to 50 MHz. The siggen is set to run at +7dBm so that it drives the ADE-1 mixer well.
The attenuator was something that was in my junk box. You can build your own step attenuator. The W7ZOI/W7PUA power meter is based on the Analog Devices AD-8307 logarithmic amp It reads directly in dBm from -70 dBm to over +13 dBm. The article describing the meter is on the cdrom that comes with "Experimental Methods in RF Design" or you can get it from the ARRL website if you are a member. Kanga also has a kit of parts.
I was asked how I made the meter scale. I used Tonne Software's METER program to create the scale. Here is the actual config file that I used for the Meter program I used a RadioShack 0 - 1mA meter 270-1752. The article references a RadioShack 0-15V meter 22-410 which is really a 0-1mA meter in series with a scaling resistor. I think the physical meters are the same size so my meter scale should fit. In any case the config file is here that you can modify to suit.
To use the Manual Spectrum Analyzer, connect your qrp amp or whatever signal you are testing to RF In. Tune the DDS LO to 10 MHz above the oscillator frequency coming out of the RF amp and read the output power of the fundamental on the power meter. (eg. If your amp is running at F=4 MHz then tune the LO to 14 MHz.) Tune the DDS LO to 2*F +10MHz and read the first harmonic power. (eg. for the 4MHZ example, the second harmonic is 8MHz so tune the LO to 18MHz) Tune to DDS LO to the third harmonic + 10 MHz and read. (eg. 3*4 + 10 = 22 MHz) Repeat until out of LO. Note: Keep RFin at a low level, say below -30 dBm to help prevent creation of distortion in the mixer which can mess up your measurements. If necessary an IF amp can be added after the mixer to raise the levels back up to more easily measureable levels.
It worked for me. Let me know if it works for you. Mike KL7R