Method for Measuring an IF Filters

by Tony – K1KP

=========Procedure to measure IF filters

Equipment needed:

RF signal generator which covers the filter’s range of frequencies. 50 ohm output impedance.

Step attenuator. This will be your primary signal level standard. Should have at least 70 (preferably 90+) db range and be 50 ohm impedance.

HF general coverage receiver. This must cover the filter’s range of frequencies. It should have a means of turning off the AGC and using manual RF gain control. It will be your primary frequency standard, so a digital display is recommended. It should have an analog S-meter.


(Optional) Audio AC voltmeter (or Scope) This can be used in place of the receiver’s S-meter for better accuracy.


Equipment setup:

The basic setup is to feed the output of the RF generator thru the Step attenuator into the filter. The filter output is fed into the HF receiver. If you are using the Audio voltmeter, connect it to the receiver’s speaker output.

Filter Termination:

For best results, the input and output impedance of the filter should be matched according to the filter specs. This may be difficult as the filter impedance is not always known. In some cases it may be printed right on the filter. Also, some filters are designed to have their inputs and outputs tuned with a small value capacitor. The filter impedances can be matched to 50 ohms by  using a series resistance.

Example: Heathkit filters have Zin = Zout = 2000 ohms. The RF generator has a 50 ohm output impedance. Use 1950 ohms in series with input and output  so the filter ‘sees’ 2K ohms on both ports.

Also, a good ground into and out of the filter is required to get accurate results. A small impedance in the ground can lead to erroneously high stopband ‘blowby’.


In Site Measurements:

 The best results can be had by using the filter in the rig where it will be used (called Rig Under Test or RUT in the following). This is because the termination and grounding issues are taken care of for you. In this case, you can connect the RF generator to the input of the RUT and set it to any frequency the RUT can receive. The RUT will convert the RF generator frequency to the passband range of the filter. Connect the HF receiver used for measurement to a downstream stage to pick off a signal after the filter. Be sure to disable the AGC in the RUT, and set the RF generator output level and RUT RF gain control so the RUT is not overloaded.

Measurement technique:

The HF receiver should be in CW or SSB mode, using as wide a bandwidth as possible. Tune the generator and HF receiver to peak up a signal and find the peak of the filter’s passband. Set the step attenuator to around 70 or 80 dB of attenuation. Adjust the generator output so you have a reading of S9 or so on the receiver.

The procedure for plotting the filter response is to make a series of signal level measurements at the frequencies of interest. These measurements will be relative to a single reference, so you will be able to see the filter’s passband characteristic but not it’s absolute insertion loss.

There are two basic ways to make these relative measurements.

The first is using the S-meter of the HF receiver to indicate a fixed signal level. For this technique, the AGC in the HF receiver is on. At the filter peak, record the S-meter reading. This is your ‘reference level’. Next set the RF generator to the frequency you wish to measure. You can determine exact frequency by zero beating and reading the digital display from the HF receiver. Each time you change the RF generator, retune the HF rig to give the same beatnote (i.e. the HF receiver must track the generator frequency). This also prevents the filter characteristics of the HF receiver from affecting the results – all the measurements are at the same point in the HF receiver’s IF. As you move the RF generator signal away from the peak, the S-meter reading will decrease. Keeping the RF generator fixed at the frequency where you want to make a measurement, decrease the attenuation on the step attenuator until the S-meter comes back up to the recorded reading ‘reference level’. Now you can read the change in filter attenuation by noting the change you made in the step attenuator setting.

The second method is to use an audio AC voltmeter instead of the S-meter to measure the reference level. In this case, turn off the AGC for the HF receiver and use manual RF gain control to make sure the HF receiver is not overloaded. Measure the audio output from the receiver with the AC voltmeter. This is your ‘reference level’. Change the RF generator and HF receiver frequency together. Next adjust the step attenuator to bring the level back up to the reference level. Now read the filter attenuation on the step attenuator as above.

Caveats:

At the extremes of measurement, you need to be sure you are not overloading the HF receiver (or the RUT). Do this by making a 3-6 db decrease in the step attenuator, and checking that signal levels at the measurement point increase appropriately. If they don’t then you may need to readjust the RF generator and start the measurements over with a new ‘reference level’.


Another Way

A very convenient way to measure filters is using a DSP receiver such as the 756 Pro series with a PC equipped with a sound card and audio spectral anaylsis software (shareware). A good program to use is Spectrogram from Visualization Software. The sound card software is used to measure and plot the audio output from the HF receiver. As long as the bandwidth of the HF receiver is wider than the filter you want to test, you can easily sweep the input to the filter while observing the output on the computer screen. In this technique, you need to turn off the AGC in the HF receiver so it doesn’t affect the signal level measured by the sound card.

This technique could also be used to measure the passband  of a RUT directly, as long as the AGC can be turned off.